Role of Dosage-Sensitive Sex Reversal, Adrenal Hypoplasia Congenita, Critical Region on the X Chromosome, Gene 1 in Protein Kinase A- and Protein Kinase C-Mediated Regulation of the Steroidogenic Acute Regulatory Protein Expression in Mouse Leydig Tumor Cells: Mechanism of Action

Abstract

Dosage-sensitive sex reversal, adrenal hypoplasia congenita, critical region on the X chromosome, gene 1 (DAX-1) is an orphan nuclear receptor that has been demonstrated to be instrumental to the expression of the steroidogenic acute regulatory (StAR) protein that regulates steroid biosynthesis in steroidogenic cells. However, its mechanism of action remains obscure. The present investigation was aimed at exploring the molecular involvement of DAX-1 in protein kinase A (PKA)- and protein kinase C (PKC)-mediated regulation of StAR expression and its concomitant impact on steroid synthesis using MA-10 mouse Leydig tumor cells. We demonstrate that activation of the PKA and PKC pathways, by a cAMP analog dibutyryl (Bu)₂cAMP [(Bu)₂cAMP] and phorbol 12-myristate 13-acetate (PMA), respectively, markedly decreased DAX-1 expression, an event that was inversely correlated with StAR protein, StAR mRNA, and progesterone levels. Notably, the suppression of DAX-1 requires de novo transcription and translation, suggesting that the effect of DAX-1 in regulating StAR expression is dynamic. Chromatin immunoprecipitation studies revealed the association of DAX-1 with the proximal but not the distal region of the StAR promoter, and both (Bu)₂cAMP and PMA decreased in vivo DAX-1-DNA interactions. EMSA and reporter gene analyses demonstrated the functional integrity of this interaction by showing that DAX-1 binds to a DNA hairpin at position −44/−20 bp of the mouse StAR promoter and that the binding of DAX-1 to this region decreases progesterone synthesis by impairing transcription of the StAR gene. In support of this, targeted silencing of endogenous DAX-1 elevated basal, (Bu)₂cAMP-, and PMA-stimulated StAR expression and progesterone synthesis. Transrepression of the StAR gene by DAX-1 was tightly associated with expression of the nuclear receptors Nur77 and steroidogenic factor-1, demonstrating these factors negatively modulate the steroidogenic response. These findings provide insight into the molecular events by which DAX-1 influences the PKA and PKC signaling pathways involved in the regulation of the StAR protein and steroidogenesis in mouse Leydig tumor cells.

The characterization of protein kinase A- and protein kinase C-mediated steroidogenic acute regulatory (StAR) expression and steroidogenesis suggests that the orphan nuclear receptor DAX-1 is an important regulator of the steroidogenic response in Leydig cells.

Dosage-sensitive sex reversal, adrenal hypoplasia congenita, critical region on the X chromosome, gene 1 (DAX-1; also called Nrob1), an unusual member of the nuclear hormone receptor superfamily of transcription factors, is expressed in the hypothalamic-pituitary-adrenal/gonadal axis and is involved in adrenal and gonadal development, differentiation, and function (1,2,3,4,5). The DAX-1 gene encodes a 470-amino acid protein, whose carboxy-terminal region is homologous to the ligand-binding domain of nuclear receptors, whereas its amino terminus contains a unique 3.5 repeats of a 65- to 67-amino acid sequence that functions as a DNA-binding domain (6,7,8). Mutations in the DAX-1 gene cause adrenal hypoplasia congenita associated with hypogonadotropic hypogonadism, an X-linked inherited disorder characterized by early onset of adrenal insufficiency syndrome (1,2,9). Disruption of DAX-1 in mice results in testicular dysgenesis, which is manifested by decreased testicular size, impaired Sertoli and Leydig cell function, and infertility (10,11). We and others have demonstrated that DAX-1 expression is affected by a plethora of extracellular signals that elevate the expression of the steroidogenic acute regulatory (StAR) protein (12,13,14,15,16,17). However, the precise role of DAX-1 in regulating StAR expression and, thus steroid biosynthesis, is poorly understood.

The 30-kDa mitochondrial StAR protein controls the rate-limiting and regulated step in steroidogenesis, i.e. the mobilization of cholesterol from the outer to the mitochondrial inner membrane in which it is metabolized to pregnenolone by the cytochrome P450 cholesterol side-chain cleavage enzyme (P450scc) (18,19,20,21,22). Regulation of the StAR protein in the adrenal glands and gonads is mediated by multiple signaling pathways, including protein kinase A (PKA) and protein kinase C (PKC), and involves transcriptional and posttranslational activation (21,23,24,25). Transcriptional and/or translational inhibition of StAR expression results in a dramatic decrease in steroid biosynthesis, whereas approximately 10–15% of steroid synthesis appears to be mediated through StAR-independent mechanisms (17,19,23,26). It should be noted, however, that the PKC-dependent induction of steroid synthesis is remarkably low when compared with PKA signaling, although it potentiates gonadotropin and/or cAMP-stimulated steroidogenic responsiveness. We recently demonstrated that the inability of PKC to phosphorylate StAR accounts for the differing steroidogenic responses between PKA and PKC (25,27). Indeed, at least two phosphorylation sites, Ser56/57 and Ser194/195 (in mouse and human, respectively), have been identified in the StAR protein, and mutation of these sites (Ser→Ala) demonstrated that the phosphorylation of Ser194/195 by PKA is necessary to render StAR fully active (28,29,30).

Whereas DAX-1 plays key roles in adrenal and gonadal development and function, it appears to do so by inhibiting the expression of several enzymes and genes that are functionally related to steroidogenesis, including StAR, steroidogenic factor 1 (SF-1), cytochrome P450scc, 3β-hydroxysteroid dehydrogenase, 17α-hydroxylase, Müllerian-inhibiting substance, and high-density lipoprotein (HDL) receptor (6,10,24,31,32,33,34,35). In most of these cases, the inhibitory action of DAX-1 is mediated indirectly rather than by directly binding to these genes. A substantial body of evidence indicates that SF-1 orchestrates the expression of several steroidogenic genes. SF-1 has been demonstrated to colocalize with DAX-1 in the hypothalamus, pituitary gonadotroph cells, adrenal glands, and gonads, and notably SF-1 and DAX-1 are both known to play significant roles in adrenal and gonadal development and function (36,37,38,39,40). Likewise, the nuclear receptor Nur77 (also called nerve growth factor-induced clone B) binds to the nerve growth factor-induced clone B response element (NBRE; 5′-AAAGGTCA-3′), a variant of the nuclear receptor half-site, an element that is similar to, but distinct from, the SF-1 binding motif (5′-CAAGGTCA-3′) and is involved in a large array of biological processes including StAR expression and steroidogenesis (41,42,43,44,45,46). Herein we report that DAX-1 binds, both in vivo and in vitro, to a putative DNA hairpin structure in the mouse StAR promoter. Additionally, we delineate novel mechanisms by which DAX-1 represses both PKA- and PKC-mediated expression of the StAR protein and steroidogenesis as well as serving to antagonize the effects of Nur77 and SF-1 in mouse Leydig tumor cells.

Materials and Methods

Plasmids, transfections, and luciferase assays

The 5′-flanking −151/−1- and −68/−1-bp regions of the mouse StAR promoter were synthesized using a PCR-based cloning strategy and inserted into the Xho1 and HindIII cloning sites of the pGL3 basic vector (Promega, Madison, WI) that contains firefly luciferase as a reporter gene (29,47). Using the −151/−1 StAR-pGL3 as template, mutation in the DAX-1 binding site was generated using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). The sense strand of the oligonucleotide sequence (mutated bases underlined lowercase letters) used was the following: 5′-GCACAGCCTTgCggccGcAGCATTTAAGGC-3′. Mutations generated in three SF-1 binding sites within the −151/−1-bp region of the mouse StAR promoter, SF-1/1 (−138/−133 bp), SF-1/2 (−46/−40 bp) and SF-1/3 (−102/−96 bp), have been described previously (48). All clones were verified by restriction mapping and confirmed by automated sequencing on a PE Applied Biosystems 310 genetic analyzer (ABI PRISM 310; PerkinElmer, Foster City, CA) at the Texas Tech University Biotechnology Core Facility. The pRL-SV40 plasmid was obtained from Promega. DAX-1, Nur77, and SF-1 expression plasmids have been previously described (24,32,42).

MA-10 mouse Leydig tumor cells (49), a generous gift from Dr. M. Ascoli (Department of Pharmacology, University of Iowa College of Medicine, Iowa City, IA), were maintained in Waymouth MB 752/1 medium supplemented with 15% horse serum and antibiotics (29,47). These cells were transfected using FuGENE HD-transfection reagent (Roche Diagnostics Corp., Indianapolis, IN) under optimized conditions (29,47,48). In brief, the −151/−1 StAR reporters (wild type and mutant) were transfected with different expression plasmids (1:1). The amount of DNA used in transfections was equalized with empty vector. Transfection efficiency was normalized by cotransfecting 10–20 ng pRL-SV40 vector (a plasmid that constitutively expresses renilla luciferase).

Transfection with small interfering RNAs (siRNAs) was performed using X-tremeGENE siRNA transfection reagent (Roche) under optimized conditions (50). Silencer negative control, and the DAX-1/NROB1 siRNAs (no. 1, 5′-CCAACACGACGCAGGAAAU-3′, and no. 2, 5′-GGGACCGUGCUCUUUAACC-3′) were obtained as annealed oligos (Ambion Inc., Austin, TX). The siRNAs were transfected using final concentrations ranging from 50 to 100 nm.

Luciferase activity in the cell lysates was determined by the dual-luciferase reporter assay system (Promega), as described previously (29,47,48). After treatments, cells were washed with 0.01 m PBS, and 300 μl of the reporter lysis buffer were added to the plates. Cellular debris was pelleted by centrifugation at 12,000 × g at 4 C, and the supernatant was measured for relative light units using a TD 20/20 luminometer (Turner Designs, Sunnyvale, CA).

Western blotting

Western blot analyses were carried out using total cellular protein (15,17,30). Briefly, cells were homogenized in lysis buffer, and the supernatant was assayed for total protein. Equal amounts of protein (20–30 μg) were solubilized in sample buffer and loaded onto either 10 or 12% SDS-PAGE (Mini Protean II system, Bio-Rad Laboratories, Inc., Hercules, CA) and electrophoresed as previously described (15,17,30). The proteins were electrophoretically transferred onto Immuno-Blot polyvinyl difluoride membranes (Bio-Rad). Membranes were probed with specific antibodies (Abs) that recognize DAX-1 (3), total StAR (51), phosphorylated StAR (P-StAR) (30), SF-1 (52), Nur77 (Cell Signaling Technology, Inc., Beverly, MA), P450scc (Chemicon International Inc., Temecula, CA), and actin (Santa Cruz Biotechnology, Santa Cruz, CA). After overnight incubation with primary Abs, the membranes were washed and incubated with appropriate secondary Abs for 1 h, washed again, and detection of the different proteins was determined with the chemiluminescence imaging Western Lightning kit (PerkinElmer). Membranes were exposed to x-ray films (Marsh Bio Products Inc., Rochester, NY), and the intensity of immunospecific bands was quantified using a computer-assisted image analyzer (Visage 2000; BioImage, Ann Arbor, MI). Detection of different proteins was assessed using identically processed membranes (in some cases StAR and DAX-1 were detected together); however, where appropriate, the same membranes were also analyzed by stripping and reprobing with additional antibodies.

Quantitative RT-PCR

Total RNA was extracted from different treatment groups using Trizol reagent (GIBCO-BRL, Grand Island, NY). Mouse StAR (18) and DAX-1 (GenBank accession no. NM_007430) were amplified using the following primer pairs: StAR sense, 5′-GACCTTGAAAGGCTCAGGAAGAAC-3′ and StAR antisense, 5′-TAGCTGAAGATGGACAGACTTGC-3′; DAX-1 sense, 5′-GCCGAGGGCCCCCTGGTGGGAC-3′ and DAX-1 antisense, 5′-TCCAGCATCATATCATCCATGCTGAC-3′, as described previously (17,24,32). The variation in RT-PCR efficiency was normalized using an L19 ribosomal protein gene, which was amplified using the following primer pairs: the sense, 5′-GAAATCGCCAATGCCAACTC-3′ and the antisense, 5′-TCTTAGACCTGCGAGCCTCA-3′. Using 2 μg of total RNA, reverse transcription and PCR were run sequentially in the same assay tube containing [α³²P]dCTP in the deoxynucleotide triphosphate mixture (24,32). The cDNAs generated were further amplified by PCR using the primer pairs listed above. The molecular sizes of StAR, DAX-1, and L19 were determined on 1.2% agarose gels, which were then vacuum dried and exposed to x-ray film (Marsh Bio Products) for 1–5 h. The levels of StAR, DAX-1, and L19 signals were quantified using a computer-assisted image analyzer (Visage 2000).

Chromatin immunoprecipitation (ChIP) assay

ChIP assays were performed using a kit (Upstate/Chemicon, Temecula, CA) as per the manufacturer’s instructions, and carried out under optimized conditions (53). In brief, cells were incubated with formaldehyde (1%) for 10 min at 37 C to cross-link DNA and its associated proteins. Cells were collected, resuspended in lysis buffer, and then sonicated with seven to nine cycles of 10-sec pulses using a Tekmar sonic disruptor (Fisher Scientific, Pittsburgh, PA). The supernatant containing chromatin was cleared with protein A agarose/salmon sperm DNA 50% slurry for 30 min at 4 C with agitation. After centrifugation, the supernatant was immunoprecipitated with IgG, 4–5 μg of Abs to DAX-1 (as above), or phosphorylated cAMP response element-binding protein (P-CREB; Cell Signaling Technology) for 14 h at 4 C, and followed by incubation with protein A agarose/salmon sperm for an additional 1 h. The chromatin-antibody-protein A agarose complexes were subsequently washed with low salt, high salt, LiCl, and Tris/EDTA buffers. Protein-DNA complexes were eluted with freshly made elution buffer (1% sodium dodecyl sulfate, 0.1 m NaHCO₃). NaCl was added (to a final concentration of 20 mm) to the eluates that were then incubated at 65 C for 4 h to reverse the formaldehyde cross-linking. The resulting samples were treated with proteinase K for 1 h at 45 C, and the purified DNA samples were analyzed by a radioactive PCR approach using [α³²P]dCTP in the deoxynucleotide triphosphate mixture (53). PCR was performed with 75–100 ng of DNA and the proximal mouse StAR promoter primers (forward, 5′-CTGGTCCTCCCTTTACACAGTC-3′ and reverse, 5′-GGCGCAGATCCAGTGCGCTGC-3′), spanning bases −170/−149 and −21/−1, respectively (54). A second primer combination [forward, 5′-GTGAGGACAGCTCATACGTGCAC-3′ (bases −3596/−3574) and reverse, 5′-GAACAGGCTTAAGTTAAGACTCC-3′ (bases −3450/−3428)] recognizing the distal region of the mouse StAR promoter was used as a negative control. PCR products were separated on 2% agarose gels. Gels were vacuum dried, exposed to x-ray film (Marsh Bio Products) for 1–5 h, and the resulting signals quantified (Visage 2000).

EMSA

EMSA reactions were performed using nuclear extracts (NE) and in vitro-transcribed/translated DAX-1 (29,47,53). The oligonucleotide probes were engineered and synthesized by heating sense and antisense primers to 65 C for 5 min in annealing buffer [10 mm Tris-HCl, 100 mm NaCl, 1 mm EDTA (pH 7.5)], followed by cooling at room temperature. The sense strands of the oligonucleotide sequences [mutated (Mut) bases in underlined lowercase letters] used were: DAX-1 wild-type (−48/−16 bp), 5′-GGGCACAGCCTTCCACGGGAAGCATTTAAGGCAGCG-3′; DAX-1 Mut, 5′-GGGCACAGCCTTgCggccGcAGCATTTAAGGCAGCG-3′; SF-1 binding site (−48/−35 bp of the mouse StAR promoter), 5′-GGCACAGCCTTCCACG-3′ (48); and consensus NBRE, 5′-GATCTTCGTGCGAAAAGGTCAAGCGCTAG-3′ (55). The specific doubled-stranded oligonucleotides were end labeled with (α³²P)-dCTP (PerkinElmer Life Sciences) using Klenow fill-in reaction and purified using spin columns (29,47). NE (12–15 μg) and in vitro-translated DAX-1 (2–5 μg) were incubated for 15 min at room temperature in a 20-μl reaction buffer [25 mm Tris-HCl, 1 mm EDTA, 4% Ficoll, 10 mm dithiothreitol, 2 μg poly dIdC, 40 ng/μl BSA, and 12 mm MgCl₂ (pH 7.9)] before the addition of a ³²P-labeled probe either alone or in the presence of unlabeled oligonucleotide. When Abs were used, binding reactions were carried out for an additional 45 min on ice before addition of the labeled DNA. The reaction mixture was then subjected to electrophoresis at 200 V for 2 h through a 5% polyacrylamide gel in 0.5× a buffer of 90 mmol/liter Tris-borate, 2 mmol/liter EDTA (pH 8.0). The gels were dried and DNA-protein complexes were visualized with a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).

Statistical analysis

Statistical analysis was performed by ANOVA using Statview (Abacus Concepts Inc., Berkeley, CA) followed by Fisher’s protected least significant differences test. Data represent the mean ± se, and P < 0.05 was considered statistically significant.

Results

Involvement of the PKA and PKC pathways in DAX-1 expression and its relevance to StAR and steroid synthesis

The role of DAX-1 in PKA- and PKC-mediated regulation of StAR expression and progesterone synthesis was investigated. MA-10 cells were treated with dibutyryl cAMP [(Bu)₂cAMP; 0.1 and 0.5 mm] or phorbol 12-myristate 13-acetate (PMA; 1 and 10 nm) for 6 h to activate the PKA and PKC pathways, respectively. As illustrated in Fig 1, the expression of DAX-1 protein was attenuated to approximately 50 to 70% of controls by (Bu)₂cAMP and PMA. Lower doses of (Bu)₂cAMP and PMA, ineffective at inducing StAR protein levels, were also capable of decreasing DAX-1 expression (Fig. 1, A and B). The decrease in DAX-1 levels by (Bu)₂cAMP and PMA was found to correlate with an increase in StAR protein expression. Induction of StAR protein levels in response to (Bu)₂cAMP (0.5 mm) and PMA (10 nm) was 8.6 ± 1.5 and 8.4 ± 1.3-fold over untreated cells, respectively. However, PMA had no effect on P-StAR, an indispensable requirement for optimal steroid biosynthesis (17,25,28,30). Conversely, (Bu)₂cAMP (0.5 mm) strongly increased levels of P-StAR. Accumulation of progesterone in the media was 9.8 ± 1.5 and 423 ± 29-fold with (Bu)₂cAMP (0.1 and 0.5 mm, respectively), and 5.4 ± 0.7-fold with PMA (10 nm) over untreated (1.8 ± 0.2) cells (Fig. 1C and inset). PMA (1 nm) had no apparent effect on progesterone synthesis. These findings show the impact of PKA and PKC activation on directing StAR expression and also the dependence of steroidogenesis on the phosphorylation of StAR by PKA.

Figure 1.

Role of PKA and PKC signaling in DAX-1, StAR, P-StAR, and progesterone levels in MA-10 cells. Cells were treated with two different concentrations of (Bu)₂cAMP (0.1 and 0.5 mm) and PMA (1 and 10 nm), respectively, and subjected to preparation of whole cell lysates for Western blot analysis using 20–30 μg of protein. Representative immunoblots illustrate expression of DAX-1, StAR, and P-StAR in different treatment groups (A). Integrated OD (IOD) values of each band were quantified and compiled data (n = 4) are presented (B). C, Accumulation of progesterone in the media of the same treatment groups was determined and expressed as nanograms per milligram protein. Inset, Control (Con) and PMA (10 nm)-treated progesterone levels. Data represent the mean ± se of four independent experiments. Letters above the bars indicate that these groups differ significantly from each other, at least at P < 0.05.

The inhibition of PKA activity by H89 (0–25 μm) (17,25,30) attenuated (Bu)₂cAMP (0.5 mm)-induced StAR protein and P-StAR levels, which in contrast, significantly reversed (P < 0.01) DAX-1 expression in a dose-dependent manner (Fig. 2A). H89 (25 μm) strongly decreased (≥90%) (Bu)₂cAMP-stimulated StAR, P-StAR, and progesterone synthesis (not illustrated). Similarly, an inhibitor of PKC, GF-109203X (GFX; 0–25 μm) (17,25,30), decreased PMA (10 nm)-treated StAR protein levels, and P-StAR remained undetectable (Fig. 2B). GFX (5 μm) markedly diminished (P < 0.01) StAR protein expression. However, GFX used in conjunction with PMA blocked the decrease in DAX-1 protein levels. These results demonstrate the differential effects of PKA and PKC on the steroidogenic response and its inverse correlation to DAX-1 expression.

Figure 2.

Inhibition of PKA and PKC signaling on DAX-1, StAR, and P-StAR levels in MA-10 cells. Cells were pretreated with increasing doses PKA (H89; 0–25 μm) and PKC (GFX; 0–25 μm) inhibitors for 30 min and then incubated without or with (Bu)₂cAMP (0.5 mm; A) and PMA (10 nm; B) for an additional 6 h. Actin expression was assessed as a loading control (A and B). Representative immunoblots illustrate expression of DAX-1, StAR, and P-StAR using 25–30 μg of total protein. Data are representative of three independent experiments.

The results presented in Fig. 3 show the time kinetics (0–24 h) of (Bu)₂cAMP- (0.5 mm) and PMA (10 nm)-mediated StAR and DAX-1 protein expression and progesterone synthesis in MA-10 cells. (Bu)₂cAMP significantly increased (P < 0.05) StAR protein and progesterone levels by 1 h, elevated them maximally between 4 and 8 h, and both declined thereafter between 12 and 24 h (Fig. 3, A, C, and D). The expression of DAX-1 protein was decreased by 3 h (P < 0.05) and continued to decline with time (up to 24 h). The effects of treatment with PMA on StAR and DAX-1 levels were similar to that seen with (Bu)₂cAMP (Fig 3), with PMA induction being slightly slower than that seen with the cAMP analog, significantly increasing (P < 0.05) StAR and progesterone levels by 2 and 3 h, respectively. Progesterone levels after PMA treatment for 6 h were in the neighborhood of 1–2% of those generated in response to (Bu)₂cAMP. The time-response patterns of StAR and DAX-1 mRNA levels by (Bu)₂cAMP and PMA were qualitatively similar (data not shown, and Refs. 29,30). These findings corroborate the results presented in Figs. 1 and 2 and reinforce the inverse relationship between DAX-1 and StAR in different steroidogenic cells (12,13,14,15,16,17).

Figure 3.

Time-course expression patterns of StAR and DAX-1 proteins and progesterone synthesis in response to (Bu)₂cAMP (0.5 mm) and PMA (10 nm). MA-10 cells were treated for 0–24 h at the indicated doses of (Bu)₂cAMP (A) and PMA (B), and 22–25 μg of total protein were used for Western blot analyses. Representative immunoblots illustrate the expression of StAR and DAX-1 in different treatment groups. Integrated OD (IOD) values for each band were quantified and compiled data from three independent experiments for both proteins are presented (C). Actin expression was assessed as a loading control. Accumulation of progesterone in the media of the same treatment groups was determined and expressed as nanograms per milligram protein, which represent the mean ± se of three independent experiments (D). Asterisks indicate the first time point at which values were significantly different (P < 0.05) from controls.

Repression of DAX-1 by PKA and PKC requires transcription and de novo protein synthesis

To determine the nature of DAX-1 suppression by PKA and PKC signaling, MA-10 cells were treated with (Bu)₂cAMP (0.5 mm) and PMA (10 nm) for 6 h, and transcription and translation were subsequently inhibited by coincubation with actinomycin D (Act; 5 μm) and cycloheximide (Chx; 5 μm), respectively. The results presented in Fig. 4 (A and B) demonstrate that (Bu)₂cAMP and PMA decreased (P < 0.01) the expression of DAX-1 mRNA, responses that were essentially reversed by either Act or Chx. Act and Chx alone had no significant effects on DAX-1 expression. Notably, the decreases in DAX-1 mRNA were found to be associated with increases in progesterone levels in the media (Fig. 4B). These data indicate that attenuation of DAX-1 in response to PKA and PKC requires transcription and translation of a regulatory factor(s) involved in DAX-1 gene transcription or degradation of DAX-1 mRNA in MA-10 mouse Leydig tumor cells.

Figure 4.

Inhibition of DAX-1 expression by PKA and PKC signaling requires transcription and de novo protein synthesis. MA-10 cells were treated without (Con) or with (Bu)₂cAMP (0.5 mm) and PMA (10 nm) in the absence or presence of Act (5 μm) and Chx (5 μm) for 6 h and subjected to DAX-1 mRNA expression by RT-PCR analysis (A). A representative autoradiogram illustrates the expression of DAX-1 and L19 in different treatment groups using 2 μg of total RNA. Integrated OD (IOD) values of each band were quantified and normalized to the corresponding L19 bands (B). Accumulation of progesterone in the media was determined and expressed as nanograms per milligram protein (B). Results represent the mean ± se of three independent experiments. Note the different scales on the graph. Letters above the bars indicate that these groups differ significantly from each other at least at P < 0.05. *, P < 0.05; **, P < 0.01; ***, P < 0.001 vs. control DAX-1 mRNA.

Interaction of DAX-1 with the StAR proximal promoter represses StAR gene expression

Because the repression of DAX-1 is associated with an elevation of StAR expression and steroid synthesis, the mechanism of its regulation by PKA and PKC signaling is of considerable interest. We have previously demonstrated that the proximal −151/−1-bp region of the mouse StAR promoter, which contains the recognition motifs for several trans-regulatory factors, is critical in transcription of the StAR gene (29,47,53,56). ChIP analyses revealed that DAX-1 associates with the proximal StAR promoter (−170/−1 bp) and that both (Bu)₂cAMP (0–1.0 mm, Fig. 5A) and PMA (0–20 nm, Fig. 5B) dose-dependently suppressed protein-DNA interactions. No association of DAX-1 with the distal StAR promoter (−3596/−3428 bp) was observed (Fig. 5C), nor was any DAX-1 binding seen with DNA immunoprecipitated with IgG. As a positive control, we examined the association of P-CREB with the proximal StAR promoter (Fig. 5D), and P-CREB association was found to increase in response to (Bu)₂cAMP (53). These results demonstrate the specificity of the DAX-1-DNA interaction with the proximal StAR promoter and that the PKA and PKC signaling diminish this interaction.

Figure 5.

Effects of PKA and PKC on the association of DAX-1 with the StAR promoter. MA-10 cells were treated without or with increasing doses of (Bu)₂cAMP (0–1.0 mm, A) and PMA (0–20 nm, B) for 6 h, and ChIP assays were carried out as described in Materials and Methods. Cross-linked sheared chromatin obtained from (Bu)₂cAMP and PMA-treated cells was immunoprecipitated (IP) with either IgG or anti-DAX-1 and anti-P-CREB (D) Abs. Recovered chromatin was subjected to PCR analysis using primers (−170/−1 bp, A, B, and D) encompassing the DAX-1 binding site or primers (3596/−3428 bp, C) approximately 3500 bp upstream of the DAX-1 region. Representative autoradiograms illustrate the association of DAX-1 and P-CREB with the StAR promoter in response to (Bu)₂cAMP and PMA. Data shown are representative of three independent experiments.

Closer analysis of the proximal region in the mouse StAR promoter predicted a DNA hairpin structure (Fig. 6A) at position −44/−20 bp (containing a stem of five nucleotides and a loop of 15 nucleotides; evaluated using Mfold software, (57)], and thus, the role of this hairpin on DAX-1 binding was analyzed. MA-10 cells transfected with the −151/−1 StAR reporter segment demonstrated 3.4 ± 0.5- and 1.8 ± 0.3-fold increases in luciferase activity by (Bu)₂cAMP (0.5 mm) and PMA (10 nm), respectively, over untreated cells (Fig. 6B). Deletion of the 5′-flanking region to −68/−1 bp (which still contains the hairpin loop), diminished (P < 0.05) basal and consequently (Bu)₂cAMP and PMA mediated StAR reporter activity without affecting fold responsiveness, when compared with the −151 construct. Overexpression of DAX-1, in the presence of the −151/−1 bp reporter, decreased basal promoter activity approximately 50% but did not affect (Bu)₂cAMP- and PMA-mediated fold activity (Fig. 6C). The −151/−1 bp construct was chosen based on higher reporter responses in comparison with the −68/−1-bp StAR segment. The introduction of mutations in the hairpin loop resulted in a StAR promoter that was no longer inhibited by DAX-1 but rather was modestly increased (∼15%) in basal and (Bu)₂cAMP-/PMA-mediated StAR gene expression (Fig. 6C). These findings point to the functional importance of the −44/−20 bp motif in recruiting DAX-1.

Figure 6.

Identification of a DNA hairpin structure in the mouse StAR promoter and the effect of DAX-1 on PKA- and PKC-mediated StAR promoter activity. A, Presence of a putative DAX-1 binding motif in the mouse StAR gene. This motif is located between −44 and −20 bp and is composed of five nucleotides stem and 15 nucleotides loop. B, MA-10 cells were transfected with different StAR reporter plasmids (−151/−1 and −68/−1 bp) in the presence of 20 ng of pRL-SV40 vector (renilla luciferase for determining transfection efficiency). Schematic representations of different StAR reporters (−151 and −68 bp) are shown (bottom panel). pGL3 basic (pGL3) was used as a control. Cells were also transfected with empty vector (pcDNA3) or DAX-1 expression plasmid, within the context of the −151/−1 wild-type and mutant StAR reporter segments as indicated, in the presence of pRL-SV40 (C). After 36 h of transfection, cells were incubated for a further 6 h in the absence (basal) or presence of either (Bu)₂cAMP (0.5 mm) or PMA (10 nm), and luciferase activity in the cell lysates was determined and expressed as relative light units (RLU, luciferase/renilla). Relative fold activity is illustrated above bar diagrams. Data represent the mean ± se of four independent experiments. Letters above the bars indicate that these groups differ significantly from each other at least at P < 0.05.

To obtain more insight into these mechanisms, the ability of the predicted hairpin structure to bind DAX-1 was studied using EMSA. As illustrated in Fig. 7A, a ³²P-labeled probe (−48/−16 bp) encompassing the hairpin loop binds to a protein in MA-10 NE, and this DNA-protein complex was nearly abolished with cold competitor (Fig. 7A; compare lanes 2 and 3). NE obtained from (Bu)₂cAMP- (lane 4) and PMA (Fig. 7A, lane 5)-treated MA-10 cells contained a significantly diminished (P < 0.01) DNA-protein complex. Importantly, protein-DNA binding was reduced when the probe was mutated such that the loop would be disrupted (Fig. 7A, lanes 6–8), suggesting the importance of this loop in DAX-1 binding. Figure 7B (lanes 1–5) demonstrated that the binding of MA-10 NE to the −48/−16 bp was inhibited with the minor groove binder, distamycin A, in a concentration-dependent manner (0–100 μm), suggesting this compound affects DNA conformation (6,58). The hairpin loop was also able to bind in vitro-translated DAX-1 (Fig. 7C, lanes 1–5), and the resulting protein-DNA complex was further enhanced with an increasing amount of DAX-1 (Fig. 7C; compare lanes 1 and 2). The binding of in vitro-translated DAX-1 to the −48/−16 was also strongly inhibited by DAX-1 Ab and distamycin A (Fig. 7C; compare lanes 2, 4, and 5). These data demonstrate that DAX-1 specifically binds to the DNA hairpin structure and represses StAR gene transcription.

Figure 7.

The binding of the hairpin region in the StAR promoter to endogenous DAX-1 in MA-10 NE and in vitro-translated DAX-1 using EMSAs. MA-10 NE (12–15 μg) obtained from control (Con, lanes 2, 3, and 6), 0.5 mm (Bu)₂cAMP (lanes 4 and 7), and 10 nm PMA (lanes 5 and 8)-treated cells were incubated with the ³²P-labeled wild-type (WT) and Mut DAX-1 probes (−48/−16 bp) encompassing the hairpin loop (A) as described in Materials and Methods. The specificity of DAX-1 binding (−48/−16 bp) in untreated MA-10 NE (lanes 1–6) was also assessed with increasing amounts (1–100 μm) of distamycin A (B). Binding of in vitro-translated DAX-1 protein (lanes 1–5) was evaluated with the −48/−16 probe (C). Protein-DNA binding was augmented with an increasing amount of DAX-1 (lanes 1–5). Binding was challenged with cold competitor (lane 3), DAX-1 Ab (lane 4), and distamycin A (lane 5). Cold competitor (−48/−16 bp) was used at 100-fold molar excess. Migration of free probes is shown in A. These experiments were repeated three times, and representative phosphor images from each group are illustrated.

Silencing of DAX-1 increases StAR expression and steroidogenesis

Given the complex interplay between PKA/PKC signaling and DAX-1 function, the suppressive effects of DAX-1 on StAR expression and progesterone synthesis were studied by silencing DAX-1 in MA-10 cells. Cells transfected with 100 nm of either of two DAX-1-specific siRNAs (see Materials and Methods) diminished DAX-1 expression (not illustrated). However, cells transfected with 50 nm of each siRNA (100 nm total) drastically decreased (80–90%) DAX-1 protein and DAX-1 mRNA levels (Fig. 8). As shown in Western (Fig. 8A) and RT-PCR (Fig. 8B) analyses, (Bu)₂cAMP (0.5 mm) and PMA (10 nm) treatments decreased (P < 0.01) DAX-1 expression in cells transfected with a negative control siRNA. Interestingly, the silencing of DAX-1 augmented basal and (Bu)₂cAMP/PMA mediated StAR expression and progesterone synthesis between 1.6- and 3.8-fold (Fig. 8C), demonstrating further that DAX-1 decreases the steroidogenic response.

Figure 8.

Silencing of DAX-1 and its relevance to StAR expression and progesterone synthesis. MA-10 cells were transfected with either a negative control siRNA at 100 nm (Con siRNA) or a mixture of two DAX-1-specific siRNAs at 50 nm (100 nm total, DAX-1 siRNA), as described in Materials and Methods. After 36 h of transfection, cells were treated without (Con) or with (Bu)₂cAMP (0.5 mm) and PMA (10 nm) for an additional 6 h and subjected to Western (DAX-1, StAR, and actin) and RT-PCR (StAR, DAX-1, and L19) analyses. Representative immunoblots and autoradiograms illustrate the expression of DAX-1, StAR, and actin (A) and StAR, DAX-1, and L19 (B) in different treatment groups, respectively. Actin and L19 were used as controls in Western and RT-PCR analyses, respectively. Similar results were obtained from three different experiments. C, Accumulation of progesterone in the media was determined and expressed as nanograms per milligram protein, and the data shown represent the mean ± se of three independent experiments. Letters above the bars indicate that these groups differ significantly from each other, at least at P < 0.05.

Role of Nur77 and SF-1 on StAR gene expression and their consequences with DAX-1

To better understand the involvement of DAX-1 in the transcriptional repression of the StAR gene, the roles of Nur77 and SF-1 with respect to PKA and PKC signaling were studied. Earlier studies have demonstrated that SF-1 can bind to the −46/−40 bp of the StAR promoter (this sequence lies within the hairpin) (37,48,59), and this motif is also a target for Nur77 (46). Thus, it was of interest whether the hairpin structure could recognize Nur77 and SF-1. It can be seen in Fig. 9A that the binding of MA-10 NE to the −48/−16 bp region could be markedly inhibited with unlabeled probe (Fig. 9A, lane 3) and that a DAX-1 Ab (Fig. 9A, lane 4) also strongly decreased complex formation. Additionally, protein-DNA binding was attenuated (P < 0.05) by the presence of unlabeled consensus NBRE sequence (Fig. 9A; compare lanes 2 and 5). The binding was also competed with an unlabeled SF-1/2 consensus binding site sequence (SF-1/2 BS; −48/−35 bp, Fig. 9A; compare lanes 2 and 6) that has previously been demonstrated to bind SF-1 in MA-10 NE (48). More specifically, the binding of MA-10 NE with the −48/−35 bp sequence (SF-1/2 site) was shown to be effectively inhibited by its consensus sequence or by a SF-1 Ab. In the present study, partial competition of the protein-DNA complex by unlabeled SF-1 and Nur77 binding sequences suggests that DAX-1 predominantly binds to the −48/−16-bp region. CREB Ab that recognizes the CRE family proteins had no effect on the protein-DNA complex, indicating CREB does not bind to this region and confirms previous findings (supplemental Fig. 1, published as supplemental data on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org) (29). Importantly, the protein-DNA complex could not be competed by unlabeled probe bearing mutations localized to the hairpin loop but that left the SF-1/Nur77 binding element intact (Fig. 9A, lane 7). These results indicate that the DNA hairpin structure can bind both Nur77 and SF-1 and that mutation of the predicted loop appears to decrease the binding of these factors within this region.

Figure 9.

Assessment of Nur77 and SF-1 binding to the DAX-1 hairpin structure and roles of these factors in StAR promoter responsiveness. A, MA-10 NE (15 μg) was used to determine protein binding to the ³²P-labeled DAX-1 (−48/−16 bp) motif (lanes 2–7). Binding of untreated MA-10 NE (Con, lane 2) to the labeled probe (−48/−16) was challenged with its unlabeled (−48/−16 bp) sequence (lane 3), DAX-1 Ab (lane 4), consensus NBRE (Con NBRE, lane 5), a confirmed SF-1 BS (lane 6), and with a mutant DAX-1 sequence (−48/−16 Mut, lane 7). MA-10 cells were transfected with empty vector (pcDNA3), Nur77, and SF-1 expression plasmids, within the context of one of the −151/−1 StAR reporter segments: either wild-type (−151 WT) or a reporter containing mutations in one of the SF-1 binding sites (SF-1/1M, SF-1/2M, SF-1/3M), in the presence of 15–20 ng of pRL-SV40 (B). After 48 h of transfection, cells were collected and luciferase activity in the cell lysates determined by [relative light units (RLU), luciferase/renilla] and presented in terms of percent activity. Data represent the mean ± se of four independent experiments. Letters above the bars indicate that these groups differ significantly from each other, at least at P < 0.05.

The influence of Nur77 and SF-1 was further investigated by determining StAR promoter responsiveness within the context of the −151/−1 StAR reporter segment, which is known to contain three SF-1 recognition motifs (46,48). MA-10 cells transfected with either the Nur77 or SF-1 expression plasmids, in the presence of the −151/−1 StAR reporter segment, demonstrated approximately 2-fold increases in StAR reporter activity when compared with controls (Fig. 9B). Specific mutations targeting each of the SF-1 recognition motifs decreased (P < 0.05) basal as well as Nur77 and SF-1 induced StAR promoter activity with the order of importance being SF-1/3>SF-1/1>SF-1/2. Notably, transactivation of the StAR gene by Nur77 was lacking when the SF-1/3 mutation was used. These data demonstrate that both Nur77 and SF-1 can target three SF-1 binding sites present within the −151/−1 bp region of the StAR promoter and positively influence transcription of the StAR gene (46,48).

In additional studies, MA-10 cells transfected with the Nur77 expression plasmid in the presence of the −151/−1 StAR reporter segment increased (Bu)₂cAMP- (1.9 ± 0.4-fold) and PMA (1.6 ± 0.2-fold)-mediated StAR reporter activity relative to the responses seen in mock-transfected cells (Fig. 10A). However, cells coexpressing Nur77 and DAX-1 demonstrated decreased (P < 0.05) basal and (Bu)₂cAMP-/PMA-mediated StAR reporter activity when compared with Nur77-induced responses. SF-1 expression also augmented both basal and (Bu)₂cAMP-/PMA-stimulated StAR promoter responsiveness, and the magnitude of the responses over mock-transfected cells were 2.9 ± 0.7- and 2.1 ± 0.4-fold by (Bu)₂cAMP and PMA, respectively. SF-1 mediated StAR gene transcription was also diminished in cells coexpressing SF-1 and DAX-1 (Fig. 10A). These data demonstrate that DAX-1 represses transactivation potential of Nur77 and SF-1 in regulating StAR gene transcription.

Figure 10.

The effects of the expression of Nur77, SF-1, and DAX-1, independently or in combination on StAR promoter activity; the effects of DAX-1, (Bu)₂cAMP, and PMA on Nur77, SF-1; and P450scc expression in MA-10 cells. A, Using the −151/−1-bp StAR reporter segment, cells were transfected with empty vector (pcDNA3), Nur77, SF-1, DAX-1, or a combination thereof as indicated in the presence of 15–20 ng of pRL-SV40. After 36 h of transfection, cells were incubated for an additional 6 h in the absence (basal) or presence of (Bu)₂cAMP (0.5 mm) and PMA (10 nm), and luciferase activity in the cell lysates was determined [relative light units (RLU), luciferase/renilla] and expressed in terms of fold activity. Data represent the mean ± se of three to five independent experiments. Cells were also transfected with either empty vector (pcDNA3) or DAX-1 expression plasmids and subjected to immunoblotting for Nur77, SF-1, and P450scc expression using 20–25 μg of total cellular protein (B). Cells were treated with (Bu)₂cAMP (0.5 mm) and PMA (10 nm) for 6 h, and the expression of Nur77, SF-1, and P450scc proteins was determined by immunoblotting. Actin expression demonstrates equal loading. Representative immunoblots illustrate the expression of Nur77, SF-1, and P450scc proteins in different treatment groups. Integrated OD (IOD) values of each band were quantified and compiled data from three independent experiments are presented (lower panel). Letters above the bars indicate that these groups differ significantly from each other, at least at P < 0.05. *, P < 0.05; **, P < 0.01; ***, P < 0.001 represent significant differences in comparison with respective controls.

DAX-1 overexpression in MA-10 cells was also capable of decreasing (P < 0.05) Nur77 and SF-1 protein levels between 45 and 70% when compared with their respective mock-transfected controls (Fig. 10B). Similarly, DAX-1 decreased expression of the P450scc protein expression by about 25% (Fig. 10B and supplemental Fig. 2) (10,14). Cells treated with (Bu)₂cAMP (0.5 mm) and PMA (10 nm) demonstrated 4.2 ± 1.3- and 2.9 ± 0.7-fold increases in Nur77 protein expression over untreated cells, respectively, and these levels were decreased (P < 0.05) after DAX-1 expression. Conversely, neither (Bu)₂cAMP nor PMA was able to alter SF-1 protein levels, suggesting these analogs do not affect SF-1 protein synthesis (60,61). Whereas (Bu)₂cAMP had no effect on P450scc protein levels, PMA slightly elevated P450scc expression (Fig. 10B). These findings demonstrate that the suppression of StAR gene expression by DAX-1 occurs, at least in part, due to a transrepression mechanism initiated as a consequence of DAX-1 simultaneously inhibiting the expression of both Nur77 and SF-1 in MA-10 mouse Leydig tumor cells.

Discussion

Regulation of the StAR protein and steroid biosynthesis in steroidogenic tissues has been demonstrated to be mediated through multiple signaling pathways and transcription factors that generate both positive and negative effects on the steroidogenic response (6,12,21,27,56,62,63,64,65). Whereas many well-described signals up-regulate the steroidogenic response, the roles of inhibitory events in StAR expression and steroidogenesis are not as well understood. The nuclear receptor DAX-1 is endowed with transcriptional repressor activity that is invariably abolished in individuals harboring DAX-1 mutations, suggesting that the impairment of its transrepressor function is linked to the pathogenesis of adrenal hypoplasia congenita (4,7). DAX-1 is an atypical nuclear receptor that lacks a zinc-finger DNA binding motif; instead, it harbors an amino-terminal Gly- and Ala-rich repetitive region that serves as a DNA binding domain (6,8,31,66). Nevertheless, studies have demonstrated that DAX-1 does not necessarily require binding to DNA to exhibit its repressive effects but can inhibit the transcription of several genes through discrete mechanisms (31,33,34,35,67).

The present findings outline the events by which the activation of the PKA [by (Bu)₂cAMP] and PKC (by PMA) signaling pathways can down-regulate DAX-1 and concomitantly up-regulate StAR expression in a time-dependent fashion. The down-regulation of DAX-1 by PKA and PKC signaling requires de novo transcription and translation of an unknown gene or set of genes, suggesting that the degradation of DAX-1 is dependent on the synthesis of a protein induced by either of these pathways. Consequently, DAX-1 serves as a potent repressor of steroidogenesis, the effects of which are relieved by activation of either PKA or PKC. It should be noted that in MA-10 cells, the steroid production elicited by PKC alone is compromised by the obligatory need for PKA to phosphorylate and activate StAR (31,33,34,35,67). As such, the effects of DAX-1 are more clearly seen at the level of StAR protein expression rather than at the level of steroid biosynthesis. Interestingly, the repression of DAX-1 by (Bu)₂cAMP and PMA was found to be essentially reversed by the pathway specific inhibitors H89 and GFX, respectively, which, in contrast, attenuated StAR expression and steroid synthesis. Therefore, it is conceivable that a balance between the repressor function of DAX-1 and the inducer functions of PKA and PKC may allow for a finer-tuning of the steroidogenic machinery in mouse Leydig cells. One possible source of contention in evaluating these data, however, is that the selected doses of H89 and GFX used (5–25 μm) here to inhibit the PKA and PKC pathways have also been shown to affect several other kinases, albeit in different systems, that are under the influence of these same signaling pathways (68,69,70,71). Studies have documented that 5- to 10-μm concentrations of H89 and GFX can be ineffective in inhibiting the PKA and PKC pathways (72,73,74); however, these inhibitors have been used at 40–50 μm in different cellular models (73,75,76,77), suggesting that differing cellular levels of PKA and PKC, and other variations can give rise to the need to use varying amounts of these inhibitors to achieve similar effects. Whereas the results from this (the present data, and Refs. 17,25,26,30,78) and other (79,80,81) laboratories have demonstrated that relevant concentrations (10–50 μm) of H89 and GFX are required for the inhibition of PKA and PKC in Leydig cells, it is not currently known whether these inhibitors affect other kinases in MA-10 mouse Leydig tumor cells.

An intriguing aspect of the present findings is the identification of a novel DNA hairpin structure at position −44/−20 bp in the mouse StAR promoter that appears to be responsible for the recruitment of DAX-1. The functional relevance of this hairpin was verified by three independent approaches demonstrating that DAX-1 binding to this site effectively represses steroid synthesis by impairing transcription of the StAR gene. First, in vivo ChIP data showed a clear association of DAX-1 with the StAR proximal promoter, and both (Bu)₂cAMP and PMA progressively diminished this interaction. Second, reporter gene analyses suggested that DAX-1 binds to a putative hairpin loop in the StAR promoter and suppresses StAR gene expression, whereas disruption of the loop abrogated the inhibitory function of DAX-1. Third, EMSAs demonstrated that the binding of a protein present in MA-10 NE to the −48/−16 bp site mirrored the complex formation seen using in vitro-translated DAX-1 and that this DNA-protein complex was specifically affected by cold competitors, a disruption of the hairpin loop, by distamycin A or a DAX-1 Ab. Inhibition of protein-DNA binding by distamycin A indicated that DAX-1 interacts with the minor groove of the DNA helix, a characteristic of high-mobility group box proteins, and suggests that this drug alters the secondary structure to which DAX-1 appears to bind (6,58). Previous work demonstrated that DAX-1 binding to the DNA hairpin requires an intact stem loop and that the binding was lacking in the absence of a loop. Furthermore, the binding of DAX-1 to DNA stem loops appears to be stronger with higher stem lengths than with short stems and is also dependent on the nucleotide compositions in the loop (6,66). In this study we demonstrated that a predicted hairpin structure at −44/−20 containing five paired stem nucleotides and 15 loop nucleotides binds DAX-1 and that the disruption of the hairpin loop blocks the ability of DAX-1 to bind and prevents the repression that DAX-1 imparts on the promoter. However, we cannot exclude the possibility that mutations may affect DAX-1 binding through a mechanism independent of the disruption of the hairpin loop. Whereas these results are in agreement with earlier studies demonstrating that DAX-1 binds to a DNA hairpin at −61/−27 bp in the human StAR promoter and decreases StAR expression and steroidogenesis (6,66), more conclusive demonstration of a hairpin loop structure in the mouse StAR promoter will require additional investigation. Two stable hairpins that bind DAX-1 have already been shown to be present in the mouse DAX-1 promoter (6). In contrast, suppression of the HDL receptor gene transcription by DAX-1 does not require DAX-1 binding to a DNA hairpin (35). It has been reported that the ligand-binding domain of DAX-1 can bind to mRNAs that in turn modulate the activity of the DAX-1 amino-terminal repeats (82). Therefore, it appears unlikely that a single mechanism is used by DAX-1 as it exerts its transcriptional repression on the different enzymes and genes involved in steroidogenesis.

Another interesting finding in this study is the involvement of Nur77 in StAR expression and the observation that DAX-1 negatively modulates Nur77-induced transactivation of the StAR gene. The overexpression of Nur77 enhanced (Bu)₂cAMP- and PMA-mediated StAR gene expression, and these increases were diminished by DAX-1. Like StAR, the expression of Nur77 is increased by both (Bu)₂cAMP and PMA and down-regulated by DAX-1, suggesting that the induction of Nur77 and the attenuation of DAX-1 are crucial for steroidogenesis in Leydig cells. Previous studies have shown that DAX-1 interacts with Nur77 and inhibits its transcriptional effects by competing with Nur77 binding to the steroid receptor coactivator-1 (83). Nur77 has been reported to trans-activate the genes for 17α-hydroxylase, 21-hydroxylase, and 20α-hydroxysteroid dehydrogenase using cell-specific signal transduction pathways and/or combinations of cofactors, i.e. steroid receptor coactivator, CREB binding protein (CBP), p300/CBP-associated factor, and vitamin D receptor-associated protein-205 (43,44,84). Studies have demonstrated that the induction of CREB by (Bu)₂cAMP is correlated with its increased presence on chromatin and the recruitment of the coregulator CBP/p300 to the StAR promoter (53,64,85). Hence, it is likely that the PKA- and PKC-mediated induction of Nur77 results in the recruitment of CBP/p300 to function in regulating StAR transcription. This hypothesis agrees with earlier studies demonstrating that (Bu)₂cAMP and human chorionic gonadotropin can induce Nur77 expression in Leydig and adrenal cell lines (42). Notably, in K28 mouse Leydig cells, Nur77 expression correlates with increased steroid synthesis but not StAR expression (42). In contrast, we observed that the increased levels of Nurr77 that occur in conjunction with the activation of either PKA or PKC were tightly associated with increased StAR gene expression, an observation that is consistent with findings that were published during the preparation of this manuscript (46). This seeming contradiction could be due to differences in cell signaling specificity, experimental conditions, and/or the presence of unknown factor(s) involved in StAR expression. Our current data suggest that Nur77 is capable of binding to three SF-1 binding sites at −138/−133 bp (SF-1/1), −46/−40 bp (SF-1/2), and −102/−96 bp (SF-1/3) within the −151/−1 bp region of the mouse StAR promoter and that the mutation of these SF-1 binding sites represses Nur77 mediated transactivation of the StAR gene (46). Previously, it has been demonstrated that the −46/−40 bp region (SF-1/2) of the mouse StAR promoter binds SF-1 and that protein-DNA binding was markedly decreased with its unlabeled sequence or SF-1 Ab (48). Consequently, we predict that the DAX-1 stem loop containing the SF-1/2 motif can bind to both Nur77 and SF-1 because their binding was competed for with consensus NBRE and SF-1 binding (−46/−40 bp) sequences, respectively, but was not affected after the disruption of the hairpin loop. Further studies to elucidate these mechanisms are currently underway and will lead to a better understanding of the involvement of Nur77 in regulating StAR gene transcription.

The physiological roles and tissue distribution of DAX-1 are strikingly similar to those of the nuclear receptor SF-1, and this factor is known to play pivotal roles in the cAMP-regulated expression of several steroidogenic enzymes and genes, including StAR (33,36,37,56,60). In the present study, DAX-1 was demonstrated to inhibit SF-1-induced transactivation of the StAR gene by PKA and PKC, an observation in agreement with the repression of SF-1 mRNA and steroidogenesis by DAX-1 seen in previous studies (24,32). However, neither (Bu)₂cAMP nor PMA was able to alter SF-1 protein levels, suggesting that these agents influence posttranslational modification of SF-1. Indeed, forskolin has been shown to enhance the phosphorylation of SF-1 leading to an increase in SF-1 binding to the StAR proximal promoter and a concomitant increase in StAR gene transcription (60). In agreement with our previous findings (48), we now provide evidence that the −48/−16 promoter region encompassing the DNA hairpin binds to SF-1 and that the disruption of the hairpin loop prevents formation of the protein-DNA complex. Furthermore, (Bu)₂cAMP and PMA were able to partially rescue the transactivation potentials of Nur77 and SF-1 in DAX-1-mediated repression of the StAR gene (3). DAX-1 has been shown to repress both Müllerian-inhibiting substance and the HDL receptor gene transcription by affecting the synergism between SF-1 and GATA-4 (34) and SF-1 and sterol regulatory element-binding protein-1a (35), respectively. Whether DAX-1 binding to the DNA hairpin affecting the neighboring GATA-4 binding site (−66/−61 bp) (86) requires further investigation. Nevertheless, it is tempting to speculate that combinatorial effects of other factors, which interact/cooperate with SF-1 (for example, GATA-4, CCAAT/enhancer-binding protein-β, CREB/cAMP response element modulator, specificity protein-1) in transcriptional regulation of the StAR gene, may also be negatively influenced by DAX-1.

The expression of DAX-1 strongly suppresses StAR transcription, demonstrating one of the important and effective means by which DAX-1 inhibits steroid biosynthesis, and these events are influenced by Nur77 and SF-1. In this study we were able to corroborate previous findings (10,14), demonstrating that in MA-10 cells overexpressing DAX-1 the expression of P450scc is inhibited. DAX-1-deficient male mice up-regulate aromatase expression, an event associated with Leydig cell hyperplasia and degeneration of germinal epithelium, leading to infertility (11). DAX-1 is expressed in Sertoli cells, regulated developmentally and is involved in the development of spermatogenic cells (3,40). Therefore, in addition to the physiological importance of DAX-1 in the development and function of the adrenal glands, it also appears to play key roles in controlling testicular function.

In summary, these findings provide insight into the mechanisms by which DAX-1 represses StAR expression and steroidogenesis and extend our understanding of the interplay between PKA/PKC signaling and the repressional events governing the StAR promoter in MA-10 mouse Leydig tumor cells. DAX-1 binds to a DNA hairpin structure in the mouse StAR promoter and effectively suppresses steroid synthesis by inhibiting transcription of the StAR gene. Furthermore, the down-regulation of DAX-1 increases the expression of Nur77 and SF-1, two transcription factors that can activate the StAR promoter. Thus, removing the potent repression elicited by DAX-1 is a key event in regulating StAR-mediated steroidogenesis.