This Article Abstract Full Text (PDF) Other Versions of this Article: MCB.00355-07v1 27/19/6669    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Li, D. Articles by Sewer, M. B. Search for Related Content PubMed PubMed Citation Articles by Li, D. Articles by Sewer, M. B.

Cyclic AMP-Stimulated Interaction between Steroidogenic Factor 1 and Diacylglycerol Kinase Facilitates Induction of CYP17

School of Biology and the Parker H. Petit Institute for Bioengineering & Bioscience, Georgia Institute of Technology, Atlanta, Georgia 30332

Received 27 February 2007/ Returned for modification 17 April 2007/ Accepted 20 July 2007

	ABSTRACT

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In the human adrenal cortex, adrenocorticotropin (ACTH) activates CYP17 transcription by promoting the binding of the nuclear receptor steroidogenic factor 1 (SF1) (Ad4BP, NR5A1) to the promoter. We recently found that sphingosine is an antagonist for SF1 and inhibits cyclic AMP (cAMP)-dependent CYP17 gene transcription. The aim of the current study was to identify phospholipids that bind to SF1 and to characterize the mechanism by which ACTH/cAMP regulates the biosynthesis of this molecule(s). Using tandem mass spectrometry, we show that in H295R human adrenocortical cells, SF1 is bound to phosphatidic acid (PA). Activation of the ACTH/cAMP signal transduction cascade rapidly increases nuclear diacylglycerol kinase (DGK) activity and PA production. PA stimulates SF1-dependent transcription of CYP17 reporter plasmids, promotes coactivator recruitment, and induces the mRNA expression of CYP17 and several other steroidogenic genes. Inhibition of DGK activity attenuates the binding of SF1 to the CYP17 promoter, and silencing of DGK- expression inhibits cAMP-dependent CYP17 transcription. LXXLL motifs in DGK- mediate a direct interaction of SF1 with the kinase and may facilitate binding of PA to the receptor. We conclude that ACTH/cAMP stimulates PA production in the nucleus of H295R cells and that this increase in PA concentrations facilitates CYP17 induction.

	INTRODUCTION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In the zonae fasciculata and reticularis of the human adrenal cortex, the biosynthesis of steroid hormones is activated by the peptide hormone adrenocorticotropin (ACTH), which activates a cyclic AMP (cAMP)-dependent signal transduction cascade, resulting in a rapid increase in hormone production and a sustained induction of the expression of steroidogenic genes. We have shown that cAMP-stimulated transcription of one of these steroidogenic genes, CYP17, is mediated by the binding of a complex containing steroidogenic factor 1 (SF1), p54^nrb, and polypyrimidine-tract-binding protein-associated splicing factor to the CYP17 promoter (93). The binding of this complex is concomitant with the cyclical association of several coregulatory proteins, including GCN5, steroid receptor coactivator 1, and ATPase-dependent chromatin remodeling factors (16).

An integral regulator of CYP17 and most other steroidogenic genes in the adrenal cortex and gonads is SF1 (5, 94). SF1 is a nuclear receptor (NR) that plays a key role not only in steroidogenesis but also in endocrine development and sex differentiation (6, 25, 32, 73, 82, 83). Targeted disruption of SF1 in mice resulted in adrenal and gonadal agenesis, absence of the ventromedial hypothalamic nucleus, and impaired expression of pituitary gonadotropins (42, 69, 118). The ability of SF1 to transactivate target genes has been shown to be regulated by phosphorylation (19, 31, 95, 97), sumoylation (12, 54), acetylation (11, 44, 46), and protein-protein interactions (7, 16, 61, 62, 75, 93, 115, 120).

Recently we identified sphingosine (SPH) as an endogenous ligand for SF1 (109). SPH is bound to the receptor under basal conditions, and cAMP treatment decreases the amount of this ligand bound to the receptor. SPH prevents the binding of SF1 to the CYP17 promoter and impairs the recruitment of coregulatory proteins, resulting in decreased cAMP- and SF1-dependent transcription of CYP17 reporter constructs, suppressed CYP17 mRNA expression, and attenuated dehydroepiandrosterone (DHEA) biosynthesis (109). While these studies demonstrated that SPH is an endogenous antagonist for SF1, they did not establish a role for ligand binding in receptor activation.

In addition to our mass spectrometric studies, which identified SPH as an antagonist for SF1, crystallographic studies have demonstrated that phospholipids are present in the ligand binding pocket of the bacterially expressed receptor used for structural analysis (55, 63, 112). Krylova et al. demonstrated that phosphatidylinositol phosphates (PIPs) interact with the ligand binding domain (LBD) of SF1 and that ligand binding is required for maximal activity of the receptor (55), while Li et al. found that the receptor has a large (approximately 1,600 Å) LBD that interacts with phospholipids, such as phosphatidic acid (PA) and phosphatidylethanolamine (PE), that have fatty acyl chains of between 12 and 18 carbons (63). Significantly, the large binding pocket led the authors to speculate that SF1 may readily exchange its ligands to respond to different cellular cues (63). Our studies demonstrating cAMP-stimulated changes in the interaction of SPH with SF1 lend support to the premise that SF1 has multiple ligands and prompted us to identify agonists for the receptor. We show herein that PA is a ligand for SF1 in H295R cells and that the binding of this molecule to the receptor in response to cAMP occurs via the activation of diacylglycerol kinase theta (DGK-).

	MATERIALS AND METHODS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Reagents. Dibutyryl cAMP (Bt₂cAMP) was obtained from Sigma (St. Louis, MO). All phospholipids and SPH were obtained from Avanti Polar Lipids, Inc. (Alabaster, AL). All PA molecular species were dissolved in chloroform, dried under nitrogen gas, and resuspended in 50 mM Tris-HCl, pH 7.4, containing 4 mg/ml fatty acid-free bovine serum albumin (BSA) by ultrasonication and used within 48 h. Cells were treated with lipids by dispersing the solutions into the cell culture medium using a Hamilton syringe. The DGK inhibitor 3-{2-(4-[bis-(4-flurophenyl)methylene]-1-piperidinyl)ethyl}-2,3-dihydro-2-thioxo-4(1H)quinazolinone (R59949) and -amanitin were obtained from EMD Biosciences, Inc. (La Jolla, CA).

Cell culture. H295R adrenocortical cells (85, 99) were generously donated by William E. Rainey (Medical College of Georgia, Augusta, GA) and cultured in Dulbecco's modified Eagle's/F12 medium (Invitrogen, Carlsbad, CA) supplemented with 10% Nu-Serum I (BD Biosciences, Palo Alto, CA), 0.5% ITS Plus (BD Biosciences, Palo Alto, CA), and antibiotics. Jeg3 human choriocarcinoma cells were obtained from Michael R. Waterman (Vanderbilt University School of Medicine, Nashville, TN) and cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and antibiotics.

Immunoprecipitation. For immunoprecipitation assays, H295R cells (4 to 10 150-mm dishes) were transfected with 30 µg of mutant or wild-type pCR3.1-SF1 per dish and treated with 1 mM Bt₂cAMP for 1 h. Nuclear proteins were isolated using the NE-PER nuclear and cytoplasmic extraction reagent (Pierce, Rockford, IL) and incubated with anti-SF1 (Millipore, Billerica, MA) and protein A/G agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA) overnight at 4°C with rotation. Control reactions were carried out by incubating with anti-phospho-cAMP response element binding protein (Millipore) and/or protein A/G and using nuclear extracts isolated from Jeg3 cells. The mixture was then centrifuged, the supernatant removed, and the precipitant subjected to a series of 5-min washes: three times with RIPA buffer [150 mM NaCl, 50 mM Tris-Cl, pH 8.0, 5 mM EDTA, 1% NP-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS], 150 nM aprotinin, 1 mM leupeptin, 1 mM E-64, 500 mM 4-(2-aminoethyl) benzenesulfonylfluoride], and 10 times with phosphate-buffered saline (PBS). The samples were then analyzed by mass spectrometry for detection of phospholipids.

Analysis of phospholipid molecular species. Samples were dissolved in 0.5 ml 9:1 MeOH:H₂O (vol/vol) and analyzed by syringe infusion at a rate of 5 µl/min into an API 4000 quadrupole linear ion trap mass spectrometer. Negative-mode precursor ion scans for m/z 153.1 (typical fragmentation for glycerophospholipids) were performed with the declustering potential set to –100 eV and entrance potential set to –10 eV with a collision energy range of 50 to 70 eV for an m/z range of 300 to 1,000. Species were identified by both m/z matches to a glycerophospholipid database (www.lipidmaps.org) and further class-specific fragmentation (35). PE was identified by positive-mode neutral-loss scans of 141.1 Da. Phosphatidylcholine (PC) was identified by positive-mode precursor ion scans for m/z 184.4. Phosphatidylserine (PS) was identified by neutral-loss scans for 185 Da. Phosphatidylinositol (PI) was identified by negative-mode precursor ion scans for m/z 241. PA was identified by precursor ion scans for m/z 153.1 and absence of signal for that species in the other species-specific scans.

Transient transfection, reporter gene analysis, and mutagenesis. To examine the effect of SPH and PA on SF1-dependent CYP17 reporter gene activity, H295R human adrenocortical cells or Jeg3 human choriocarcinoma cells were subcultured on 24-well plates and transfected with 125 ng pGL3-CYP17-2x57, 25 ng pCMV Tag1-SF1, and 1 ng pRL-TK using Gene Juice (EMD Biosciences, La Jolla, CA). Twenty-four hours after transfection, cells were treated for 6 to 12 h with 1 mM Bt₂cAMP, various PA molecular species (0.1 µM to 10 µM), 5 µM R59949, and/or SPH (01 to 25 µM) and harvested for dual-luciferase assays (Promega, Madison, WI). To determine the effect of mutations in the LBD on SF1-dependent transcriptional activity, cells were subcultured on 12-well plates and 24 h later transfected with pGL3-CYP17 2x57 (93) and wild-type or mutant (H310A, G341K, Y436A, K440E, Y436A/K440E, or G341A/Y436A/K440E) receptor. The effect of DGK overexpression on CYP17 luciferase activity was assessed by transfecting cells with 125 ng pGL3-CYP17-2x57, 1 ng pRL-TK, and 25 ng DGK expression plasmid. The pcDNA3-DGK- expression plasmid was received from James Walsh (Indiana University-Purdue University Indianapolis, Indianapolis, IN), and expression plasmids for DGK- and DGK- were provided by Matthew Topham (University of Utah, Salt Lake City, UT) and Stephen Gee (University of Ottawa, Ottawa, Canada), respectively. Expression plasmids for DGK-1, DGK-2, DGK-1, DGK-2, and DGK- were a gift from Fumio Sakane (Sapporo University School of Medicine, Sapporo, Japan), and pCMV6-DGK- was obtained from Origene (Rockville, MD). SF1 and DGK mutants were prepared using a QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) and were confirmed by sequencing.

Mammalian two-hybrid assay. SF1, GCN5 (pOZ-N-hGCN5; provided by Yoshihiro Nakatani, Dana Farber Cancer Institute, Harvard University School of Medicine, Boston, MA), and human liver receptor homologue 1 (LRH1) (pCI-LRH1; donated by Matthew Redinbo, University of North Carolina, Chapel Hill, NC) were cloned into the MluI and XbaI sites of the pBIND and pACT vectors (Promega, Madison, WI) as previously described (16). Cells were transfected with the pG5 firefly luciferase reporter in combination with the pBIND-GCN5 and pACT-SF1 or pACT-LRH1 vectors to express fusions of GAL4-GCN5 and VP16-SF1/LRH1, respectively, and an SRC1 expression plasmid (pBKCMV-SRC1; generously provided by Bert O'Malley, Baylor University School of Medicine, Houston, TX). Twenty-four hours later, cells were treated with Bt₂cAMP or PA for 16 h and then harvested for analysis of firefly and Renilla luciferase activities.

Real-time RT-PCR. For quantitative reverse transcription-PCR (RT-PCR), total RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA) and amplified using the One-Step SYBR Green RT-PCR kit (Eurogentec, San Diego, CA) and primers (Table 1) targeted at the gene of interest. Primers were designed using Beacon Designer 5.0 software (Bio-Rad, Hercules, CA). CYP17 mRNA expression was quantified using the delta cycle threshold (CT) method normalized to ß-actin content. Real-time RT-PCRs for quantifying the expression of DGK isoforms in H295R cells were also resolved by agarose (2%) gel electrophoresis.

ChIP. Chromatin immunoprecipitation (ChIP) assays were carried out by treating H295R cells with 1 mM Bt₂cAMP or 10 µM di14 PA for 1 h. For temporal ChIP assays (48, 114), H295R cells (subcultured into 100- or 150-mm dishes) were synchronized by treatment with 2.5 µM -amanitin for 2 h. Cells were washed twice with PBS and then treated with 1 mM Bt₂cAMP and/or 5 µM R59949 for time periods ranging from 30 min to 4 h. Cross-linking was performed by incubating the cells in 1% formaldehyde for 10 min at room temperature with gentle shaking, terminated by the addition of glycine (final concentration, 0.125 M) for 5 min, and the cells harvested into RIPA buffer. Lysates were sonicated four times (10 s each) and centrifuged for 15 min at 13,000 x g and 4°C. Fifty microliters supernatant was retained as input. The purified chromatin solutions were precleared with 1 µg rabbit or mouse immunoglobulin G and immunoprecipitated overnight at 4°C on a tube rotator using 5 µg of anti-acetyl histone H4, anti-GCN5, anti-RNA polymerase II (anti-Pol II), anti-SF1 or anti-SRC1, and protein A/G plus (Santa Cruz Biotechnology). All antibodies used in the ChIP assay were obtained from Millipore (Temecula, CA), except for anti-GCN5 (Santa Cruz Biotechnology). The immobilized protein/DNA complexes were subjected to a series of 5-min washes: three times in RIPA buffer, three times in RIPA buffer plus 500 mM NaCl, three times in washing buffer (10 mM Tris-Cl, pH 8, 0.25 M LiCl, 1 mM EDTA, 1 mM EGTA, 1% NP-40, 1% sodium deoxycholate, and protease inhibitors), and three times in Tris-EDTA buffer, pH 8.0. The cross-links were reversed, and DNA was purified by two phenol-chloroform extractions, followed by ethanol precipitation. Real-time PCR was carried out using 4 µl of output, 1 µl of input (diluted 1:4), the iTaq SYBR Green Supermix (Bio-Rad, Hercules, CA), and the following primer pairs: forward (5'-GGC TGG GCT CCA GGA GAA TCT TTC TTC CAC-3') and reverse (5'-CGG CAG GCA AGA TAG ACA GCA GTG GAG TAG-3'), which amplify the region of the CYP17 promoter from position –104 to +43. CT values of the outputs were normalized to input values. PCRs were also resolved on 2% agarose gels.

DGK activity assay. H295R cells were subcultured on 60-mm dishes and treated for 5 min to 1 h with 1 mM Bt₂cAMP. After the desired treatment time, cells were harvested and centrifuged, and nuclear and cytosolic proteins were isolated using a NE-PER Nuclear and Cytoplasmic extraction kit (Pierce Biotechnology, Rockford, IL) as per the manufacturer's instructions. DGK activity was assayed as previously described (8), using octylglucoside-diacylglycerol (OG-DG) mixed micelles that were prepared by mixing 0.25 mM dioleoylglycerol, 55 mM OG, and 1 mM phosphatidylserine in 1 mM diethylenetriamine penta-acetic acid, pH 7.4, vortexing, and sonicating until the suspension appeared clear. Twenty microliters of mixed micelles were added to 70 µl of assay buffer (100 µM diethylenetriamine penta-acetic acid, pH 7.4, 50 mM imidazole-HCl, 50 mM NaCl, 12.5 mM MgCl₂, 1 mM EGTA, 1 mM dithiothreitol, 1 µCi [-³²P]ATP) and 10 µl of cytoplasmic or nuclear extracts. Reactions were initiated by vortexing for 3 s and sonicating for 5 s and then incubated for 30 min at 25°C. The reactions were terminated by the addition of 1 ml chloroform-methanol-1% perchloric acid (1:2:0.75 [vol/vol]), followed by vortexing. One milliliter 1% perchloric acid-chloroform (1:1 [vol/vol]) was added, and the mixture was centrifuged for 5 min at 1,500 x g. The organic phase was washed twice in 1% perchloric acid and then concentrated under a stream of nitrogen and spotted on Silica Gel 60 thin-layer chromatography (TLC) plates (Merck). Plates were developed in chloroform-acetone-methanol-acetic acid-water (10:4:3:2:1 [vol/vol]) and then exposed to a phosphorimager screen. PA was quantified by phosphorimager scanning using a Fluor/Phospho-Imager (Fuji Film) and normalized to the protein concentration in each sample.

Coimmunoprecipitation. H295R cells were transfected with p3xFLAG-CMV-DGK for 48 h and cell lysates isolated and precleared by incubation for 30 min with rabbit immunoglobulin G (Millipore) and protein A/G PLUS-agarose (Santa Cruz Biotechnology). The precleared lysates were immunoprecipitated overnight at 4°C using 5 µg of anti-FLAG M2 (Stratagene) and protein A/G PLUS-agarose. The agarose beads were washed three times with RIPA buffer and three times with PBS and then subjected to SDS-polyacrylamide gel electrophoresis (PAGE) and Western blotting.

In vitro transcription/translation and glutathione S-transferase pulldown. DGKs were expressed in vitro using a TNT T7 Quick Coupled transcription/translation system (Promega, Madison, WI) and [³⁵S]methionine/cysteine (MP Biomedicals, Solon, OH). Immobilized, His-tagged SF1 was expressed in Escherichia coli as previously described (109) and incubated with in vitro-translated DGKs in binding buffer (20 mM HEPES, pH 7.9, 100 mM KCl, 1 mM MgCl₂, 2 mM dithiothreitol, 10% glycerol, 0.05% Triton X-100, and protease inhibitors) for 2 h at 4°C with rotation. The immobilized receptor was then washed three times with binding buffer and then suspended in SDS-PAGE gel running buffer and boiled, and the eluted proteins were analyzed by electrophoresis. The gels were stained with Coomassie and dried, and radiolabeled proteins were detected by phosphorimaging.

Immunostaining and confocal microscopy. Cells were plated on glass coverslips, and then some cells were incubated for 12 h with 1 µM 7-nitrobenz-2-oxa-1,3-diazol-4-yl-PA and treated with 1 mM Bt₂cAMP. After the desired incubation time, coverslips were washed twice with PBS and fixed by incubation in 3.7% formaldehyde for 20 min at room temperature. Coverslips were then permeabilized in 0.2% Triton X-100 for 5 min and blocked for 1 h in 1% BSA. Samples were incubated with anti-DGK- (BD Biosciences), anti-DGK- (Abgent), or anti-SF1 (Millipore) diluted 1:200 in 1% BSA for 1 h, followed by incubation with secondary fluoroscein isothiocyanate- or rhodamine-conjugated antibodies (Santa Cruz Biotechnology) for 1 h. Coverslips were mounted onto slides using Fluoromount G (Southern Biotechnology Associated, Inc., Birmingham, AL), dried for 5 min, and imaged. Confocal images were collected using a laser scanning microscope (LSM 510; Zeiss, Inc., Thornwood, NY) equipped with argon and helium-neon lasers with excitation wavelengths of 488 and 543 nm for fluoroscein isothiocyanate and rhodamine, respectively.

RNA interference. Cells were subcultured in 12-well plates and 24 h later transfected with 100 nM of small interfering RNA (siRNA) oligonucleotides directed against DGK- using siIMPORTER (Millipore). Cells were transfected for 72 h (DGKs) with siRNA oligonucleotides and either harvested for SDS-PAGE and Western blotting or incubated for an additional 12 h with 1 mM Bt₂cAMP for quantification of CYP17 mRNA expression by real-time RT-PCR.

Western blotting. Cells were harvested into RIPA buffer and lysed by sonication (one 5-s burst) followed by incubation on ice for 30 min. Lysates were centrifuged for 15 min at 4°C and the supernatant collected for analysis by SDS-PAGE. Aliquots of each sample (25 µg of protein) were run on 10% SDS-PAGE gels and transferred to polyvinylidene difluoride membranes (Pall Corporation, Pensacola, FL). Blots were probed with anti-DGK- (BD Biosciences), anti-DGK- (Abgent), and anti-p54^nrb (BD Biosciences), and expression was detected using an ECF Western blotting kit (Amersham Biosciences, Piscataway, NJ) and visualized by scanning blots on a Fluorescence/Phospho-Imager (Fuji Film, Japan).

Statistics. One-way analysis of variance and Tukey's multiple-comparison tests were performed using GraphPad Prism 4.03 software (GraphPad Software, Inc., San Diego, CA).

	RESULTS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Phospholipids are endogenous adrenal SF1 ligands. Using mass spectrometry, we recently demonstrated that SPH binds to SF1 and that this binding to the receptor is modulated by cAMP, where Bt₂cAMP stimulation activated CYP17 mRNA expression by decreasing the amount of SPH bound to the receptor (109). The antagonistic actions of SPH on SF1 function, coupled with the decreased SPH binding to the receptor upon stimulation with Bt₂cAMP, suggested that activation of CYP17, and perhaps other SF1 target genes, requires the exchange of SPH for an activating ligand. Moreover, since both crystallographic analysis (55, 63, 112) and our previous in vitro competition assays (109) revealed that SF1 binds to phospholipids, we performed mass spectrometric analysis of SF1 purified from H295R human adrenocortical cells to determine if phospholipids bind to the endogenous receptor. As shown in Fig. 1, spectral analysis of SF1 immunoprecipitated from control (panel A) or Bt₂cAMP-treated (panel B) cells showed that the endogenous receptor binds to phospholipids. The major peak detected from SF1 isolated from both control and Bt₂cAMP-treated cells was a PA molecule containing acyl chains with a total of 26 carbons and one double bond (m/z = 560.8). Three other PA molecules, 30:0 PA (m/z = 618.8), 35:5 PA (m/z = 678.6), and 40:4 PA (m/z = 736.8), were also bound to the receptor. Additional phospholipid species detected were 20:3 PE (m/z = 502.8) and 34:4 PIP (794.7). Based on our previous studies showing decreased SPH bound to SF1 isolated from Bt₂cAMP-stimulated cells (109), we predicted that the amounts of phospholipids bound to the receptor would increase in response to Bt₂cAMP. Although we did observe an approximately 1.8-fold increase in both 26:1 PA and 20:3 PE and a 2.5-fold increase in the amount of 30:0 PA bound to the receptor purified from Bt₂cAMP-treated cells, the magnitude of these increases, in the absence of internal standards, was not large enough to conclude with confidence that cAMP promoted a statistically significant increase in PA binding to the receptor. We did not observe any Bt₂cAMP-stimulated changes in the amounts of 40:4 PA or 34:4 PIP bound to SF1. In control immunoprecipitations, we also purified the cAMP response element binding protein; however, we were unable to detect phospholipids bound to this transcription factor (data not shown). We also did not detect phospholipids bound to protein A/G in the absence of SF1 (data not shown). Although these data confirm that SF1 binds to phospholipids in human adrenocortical cells, it is possible that more global mass spectrometric analyses for all classes of lipids would reveal other molecules that bind to the receptor. However, based on these mass spectrometric findings, we carried out further studies to determine the functional significance of PA in mediating increased CYP17 gene expression and cortisol biosynthesis in H295R cells.

View larger version (15K): [in this window] [in a new window]   FIG. 1. Identification of endogenous phospholipid ligands for SF1. H295R cells were transfected with pCR3.1-SF1 and treated for 1 h with 1 mM Bt2cAMP. Control (A) or Bt2cAMP-treated (B) lysates were immunoprecipitated with anti-SF1 antibody and phospholipids bound to the receptor extracted and analyzed by mass spectrometry.

View larger version (23K): [in this window] [in a new window]   FIG. 2. Multiple PA molecular species activate SF1-dependent CYP17 transcriptional activity. (A) H295R cells were transfected with pGL3-CYP17-2x57 and then treated with various PA species (0.1 µM to 10 µM) for 16 h. Data graphed are normalized to Renilla activity (pRL-TK) and expressed as n-fold increases of pGL3-CYP17-2x57 activity over pRL-TK activity of the untreated control. Data represent means ± standard errors of means from four separate experiments, each performed in triplicate. (B) Jeg3 cells were transfected with pGL3-CYP17-2x57, pRL-TK, and pCMVTag1-SF1 using Gene Juice. Twenty-four hours after transfection, the cells were treated with 1 mM Bt2cAMP, 10 µM di12 PA, or 10 µM di14 PA for 16 h. Cells were lysed and firefly and Renilla luciferase activities quantified by luminometry. Data graphed represent means ± standard errors of means from two experiments that were each performed in triplicate and are normalized to Renilla activity.

View larger version (26K): [in this window] [in a new window]   FIG. 3. PA increases coactivator binding to SF1 and CYP17 mRNA expression. (A) H295R cells were treated for 16 h with 1 mM Bt2cAMP, 10 µM di12 PA, or 10 µM di14 PA and total RNA extracted for analysis of CYP17 mRNA expression by quantitative RT-PCR. CYP17 expression is normalized to ß-actin content, graphed as n-fold change over level for untreated control, and represents the mean ± standard error of the mean for three separate experiments, each performed in quadruplicate. (B) Total RNA was isolated from H295R cells treated for 16 h with 1 mM Bt2cAMP, 10 µM di12 PA, or 10 µM di14 PA. The mRNA expression of other enzymes in the steroid hormone biosynthetic pathway was determined by quantitative RT-PCR as described in Materials and Methods using PCR primers (Table 1) designed to amplify CYP11A1, CYP11B1/2, CYP21, 3ßHSD, and StAR. Data graphed represent the means ± standard errors of the means for two separate experiments, each performed in triplicate, and steroidogenic gene expression normalized to ß-actin mRNA expression levels. (C) Mammalian two-hybrid experiments were carried out by transfecting H295R cells with pG5-Luc, pACT-GCN5, pBK-SRC1, and pBIND-SF1 or pBIND-LRH1 and then treating with 1 mM Bt2cAMP or 10 µM di14 PA for 16 h. Dual-luciferase assays were carried out on cell lysates and pG5-Luc activity normalized to Renilla activity (pAct-GCN5). Data graphed represent the means ± standard errors of the means for three separate experiments, each performed in quadruplicate. (D) H295R cells were treated for 1 h with 1 mM Bt2cAMP or 10 µM di14 PA and subjected to ChIP as described in Materials and Methods. Purified DNA was amplified by quantitative PCR and the samples resolved on a 2% agarose gel. Results from a representative experiment are shown. Lane 1, control; lane 2, Bt2cAMP; lane 3, di14 PA. (E) Quantification of real-time PCR CT values obtained from ChIP assays. Lysates were immunoprecipitated with antibodies against acetyl histone H4, GCN5, Pol II, SF1, and SRC1. Data graphed in all panels represent the means from two experiments, each performed in duplicate. Outputs are normalized to CT values obtained for 1% input controls, and results are presented as percentages of CT values for untreated control cells.

PA stimulates complex formation. We have recently found that cAMP promotes the assembly of a trimeric complex comprised of SF1, GCN5, and steroid receptor coactivator 1 on the CYP17 promoter (16). Thus, we sought to determine the effect of di14 PA on the formation of this complex by carrying out mammalian two-hybrid assays. H295R cells were transfected with expression plasmids for VP16-SF1 or VP16-LRH1 and GAL4-GCN5 and then treated with Bt₂cAMP and di14 PA for 16 h.

Both Bt₂cAMP and di14 PA (10 µM) significantly stimulated the binding of SF1 to GCN5, and cotransfection of full-length SRC1 potentiated the interaction between SF1 and GCN5 in untreated cells (Fig. 3C). Trimer formation was further stimulated in response to both Bt₂cAMP and di14 PA. Since LRH1 is also a member of the NR5A subfamily of NRs, we also determined the effect of di14 PA on the association of human LRH1 with GCN5. Bt₂cAMP stimulated the binding of LRH1 to GCN5, while di14 PA had no effect on this interaction (Fig. 3C).

PA increases SF1 and RNA Pol II binding to CYP17 promoter. Based on previous studies demonstrating that ACTH and cAMP promote the association of SF1, RNA Pol II, GCN5, and SRC1 with the CYP17 promoter (16), we treated H295R cells with either Bt₂cAMP or di14 PA and assessed both the acetylation state and the occupancy of the promoter by ChIP. As shown in Fig. 3D, exposure of cells to Bt₂cAMP (lane 2) or di14 PA (lane 3) for 1 h promoted the binding of SF1, RNA Pol II, GCN5, and SRC1 to the CYP17 promoter. Moreover, both Bt₂cAMP and PA increased the acetylation of histone H4 at the SF1 binding site, providing further evidence for PA as an endogenous agonist for SF1. Quantification of the changes in promoter occupancy by real-time PCR revealed that both Bt₂cAMP and di14 PA increased the binding of SF1 to the receptor by 5.7- and 4.8-fold, respectively (Fig. 3E). As previously demonstrated with Bt₂cAMP (16), di14 PA also increased the acetylation of histone H4 and stimulated the recruitment of GCN5 and RNA Pol II at the –104/+43 region of the CYP17 promoter. Interestingly, di14 PA was not as effective as Bt₂cAMP in increasing SRC1 binding to the promoter, where Bt₂cAMP increased SRC1 binding by 2.1-fold whereas di14 PA stimulated SRC1 recruitment by only 1.4-fold.

SPH antagonizes PA-stimulated CYP17 reporter gene activity, coactivator recruitment, and mRNA expression. Since SPH is an SF1 antagonist (109), we hypothesized that SPH would inhibit the ability of PA to activate SF1-dependent CYP17 transcriptional activity. To test this hypothesis, the SF1 null, steroidogenic Jeg3 cell line was transfected with pGL3-CYP17-2x57, pRL-TK, and pCMVTag1-SF1 and then treated for 16 h with Bt₂cAMP, di14 PA, and SPH. As shown in Fig. 4A, increasing concentrations (1 to 25 µM) of SPH attenuated the stimulatory effect of di14 PA on CYP17 transcriptional activity. The effects of Bt₂cAMP, PA, and SPH were dependent on the presence of SF1, because the CYP17 reporter plasmid was unresponsive in the absence of the transfected receptor in the Jeg3 cell line (Fig. 2B; also data not shown). Moreover, although transfection of human LRH1 supported Bt₂cAMP-stimulated CYP17 reporter gene activity, PA had no effect on the ability of human LRH1 to activate luciferase expression (data not shown). We also carried out studies to determine the effect of SPH on PA- or Bt₂cAMP-stimulated CYP17 mRNA expression. H295R cells were treated with Bt₂cAMP or di14 PA in the presence or absence of SPH for 12 h, and CYP17 mRNA was quantified by real time RT-PCR. Both Bt₂cAMP and di14 PA increased CYP17 mRNA expression; however, SPH inhibited the induction of CYP17 mRNA expression in response to both of these stimuli (Fig. 4B).

View larger version (12K): [in this window] [in a new window]   FIG. 4. SPH antagonizes PA-stimulated CYP17 luciferase activity and mRNA expression. (A) Jeg3 cells were transfected with pGL3-CYP17-2x57, pCR3.1-SF1, and pRL-TK and treated with 10 µM di14 PA in the presence or absence of SPH (1, 10, or 25 µM) for 16 h. Data graphed are normalized to Renilla activity and are expressed as n-fold increases in pGL3-CYP17-2x57 activity in untreated SF1 null samples. Means ± standard errors of the means for three separate experiments, each performed in triplicate, are graphed. *, P < 0.05, statistically different from untreated SF1-expressing control samples. (B) H295R cells were treated for 12 h with 1 mM Bt2cAMP, 10 µM di14 PA, and 5 µM SPH and total RNA extracted for analysis by real time RT-PCR. CYP17 mRNA expression is normalized to ß-actin mRNA content, and data graphed are expressed as n-fold change from levels for the untreated control group. The asterisk denotes statistical difference from the untreated control group (P < 0.05).

View larger version (10K): [in this window] [in a new window]   FIG. 5. PA stimulates DHEA release from H295R cells. Cells were treated for 24 h with 1 mM Bt2cAMP or 10 µM di14 PA and the amount of hormone released into the medium quantified by enzyme-linked immune assay. DHEA concentrations are normalized to the protein content of each sample. The asterisk denotes statistical difference from the untreated control group (P < 0.05).

View larger version (22K): [in this window] [in a new window]   FIG. 6. Mutagenesis of amino acids in the LBD decreases the transactivation potential of SF1. (A) Cells were transfected with pGL3-CYP17-2x57, pRL-TK, and wild-type or mutant SF1 expression plasmids (cloned into the pCMVTag1 vector) and treated with 10 µM di14 PA for 16 h. Luciferase activity in cell lysates was analyzed by luminometry. Data graphed are normalized to Renilla activity and expressed as n-fold increases in pGL3-CYP17-2x57 activity in untreated SF1 null samples. Data shown are the means ± standard errors of the means for at least three separate experiments, each performed in triplicate. (B) H295R cells were transfected with 100 ng of expression plasmid for wild-type or mutant FLAG-tagged receptor for 48 h and Western blotting performed on lysates. (C) Cells were transfected with 25 µg pCMVTag1-SF1 (left) or pCMVTag1-SF1-G341E (right). Forty-eight hours after transfection, lysates were isolated and immunoprecipitated with anti-FLAG antibody and phospholipids bound to the receptor extracted and analyzed by mass spectrometry.

View larger version (28K): [in this window] [in a new window]   FIG. 7. DGK activity is required for cAMP-stimulated CYP17 transcription. (A) H295R cells were transiently transfected with pGL3-CYP17-2x57, pRL-TK, and pCMVTag1-SF1 and then treated with Bt2cAMP (1 mM) in the presence or absence of R59949 (5 µM) for 16 h. Data are normalized to the luciferase activity of the Renilla gene and expressed as the n-fold increase over luciferase activity in the untreated control cells. Data shown are the means ± standard errors of the means for four separate experiments, each performed in triplicate. *, statistically different from untreated control; P < 0.05. (B) H295R cells were transfected with pGL3-CYP17-2x57 and pRL-TK. Twenty-four hours later, cells were pretreated for 30 min with 5 µM R59949, followed by treatment with 1 mM Bt2cAMP, 10 µM di14 PA, 10 µM di14 PC, 10 µM di14 PE, or 10 µM di8 PI(3,4,5)P3 for 16 h. Luciferase activity was quantified in the isolated cell lysates by luminometry, where data are normalized to the luciferase activity of the Renilla gene and expressed as the n-fold increase over luciferase activity in the untreated control cells. Data shown are the means ± standard errors of the means for two separate experiments, each performed in triplicate. (C) Cells were treated for 12 h with 1 mM Bt2cAMP and/or 5 µM R59949 and total RNA extracted for quantification of CYP17 mRNA expression by real-time RT-PCR. CYP17 mRNA expression is normalized to ß-actin mRNA content, and data graphed are expressed as n-fold changes from results for the untreated control group, where an asterisk denotes statistical difference from the control group. (D) -Amanitin-synchronized cells were treated with Bt2cAMP and/or R59949 for 30 min to 4 h and cross-linked in 1% formaldehyde and the sheared chromatin immunoprecipitated with an antibody against SF1. DNA bound to the immobilized receptor was purified and SF1 binding to the CYP17 promoter (–104/+43) quantified by real-time PCR as described in Materials and Methods. Data graphed represent means from two experiments, each performed in duplicate. Outputs are normalized to CT values at each time point.

cAMP stimulates nuclear DGK catalytic activity. Many studies have demonstrated roles for nuclear DGK activity in varied cellular processes. For example, DGK- is activated by nerve growth factor in PC12 cells (101) and thrombin in IIC9 fibroblasts (8, 107), DGK- promotes myogenesis in C2C12 cells (23), and DGK- plays a role in regulating the cell cycle in CHO-K cells (72). Interestingly, both DGK- (102) and DGK- (23) are localized in the speckle domains of the nucleus. Nuclear speckles are punctate structures that are enriched in pre-mRNA splicing factors and have been shown to sequester posttranslationally modified SF1 (11, 12). We postulated that cAMP-stimulated PA production and binding to SF1 occurs at these sites of transcriptional inactivity. Thus, we determined the effect of activation of the ACTH/cAMP signaling pathway on nuclear DGK activity. H295R cells were treated for 5 to 120 min with Bt₂cAMP and the isolated nuclear and cytosolic fractions incubated with radiolabeled phosphate and diacylglycerol using an OG-DG mixed micelle assay (8). Radiolabeled PA was resolved by TLC and quantified by phosphorimager scanning and densitometric analysis. As shown in Fig. 8A, Bt₂cAMP rapidly increased DGK activity and the amount of generated PA in the nucleus. This increase in PA peaked at 5 min, remained elevated for 60 min, and then returned to basal levels at the 120-min time point (Fig. 8B). No significant changes were detected in the amount of PA produced in the cytoplasm (data not shown).

View larger version (22K): [in this window] [in a new window]   FIG. 8. cAMP increases DGK activity, and SF1 synergizes with DGK to stimulate CYP17 transcriptional activity. (A and B) H295R cells were treated for 5 min to 2 h with 1 mM Bt2cAMP and nuclear extracts purified for analysis of DGK activity using the OG-DG mixed-micelle assay (see Materials and Methods). Radiolabeled PA produced was resolved by TLC (representative plate shown in panel A) and quantified by phosphorimager scanning and densitometry (panel B). (C) Expression of DGK isoforms in H295R cells was determined by real-time RT-PCR of control RNA (50 ng) and the PCRs resolved on a 2% agarose gel. (D) Cells were transfected with pGL3-CYP17-2x57, pRL-TK, pCR3.1-SF1, and DGK expression plasmids and then treated for 16 h with 1 mM Bt2cAMP. Data graphed are normalized to Renilla activity (pRL-TK) and expressed as n-fold increases of pGL3-CYP17-2x57 activity over pRL-TK activity of the untreated control group. Data shown represent the means ± standard errors of the means from at least three separate experiments, each performed in triplicate.

SF1 binds to DGK- and DGK- in vitro and in vivo. To further explore the mechanism by which SF1 and DGKs synergistically stimulate CYP17 reporter gene activity (Fig. 8D), we first carried out in silico analysis of the primary sequences of DGKs for the presence of NR boxes or LXXLL motifs. These leucine-rich motifs regulate transcriptional activation by enabling coregulatory proteins to bind to NRs (17, 34, 74, 77). Several putative LXXLL motifs are present in all of the DGK isoforms that are expressed in H295R cells, except DGK- (Fig. 9A). DGK-, DGK-, DGK-, and DGK- contain one putative SF1-interacting motif, while DGK-, DGK-, and DGK- contain two putative binding sites. To determine if SF1 interacts with DGKs in the H295R adrenocortical cell line, we performed immunoprecipitation assays and found that the receptor also binds to both DGK- and DGK- and this binding is dependent on the expression of DGK- or DGK-, because suppressing the translation of these kinases substantially decreases the amount of receptor immunoprecipitated from the cells (Fig. 9B). We also performed glutathione S-transferase pulldown assays using in vitro-translated, radiolabeled DGKs and bacterially expressed receptor to determine if SF1 binds to DGK- or DGK-. As shown in Fig. 9C, SF1 bound to both DGK- and DGK-. This binding was dependent on functional LXXLL motifs in DGK-, because mutation of these sites decreased interaction with SF1 (Fig. 9C). To assess the effect of mutating the NR boxes on CYP17 transcriptional activity, we transfected H295R cells with pGL3-CYP17-2x57, pCM VTag1-SF1, and wild-type or NR box mutant DGK- and DGK-. Mutation of both LXXLL motifs in DGK- attenuated the synergistic activation of CYP17 reporter gene activity (Fig. 9D), further supporting a role for direct interaction between DGK- and the receptor for SF1-dependent CYP17 transcription. Mutation of the lone NR box in DGK- also decreased the ability of SF1 and DGK- to synergistically stimulate an effect on luciferase activity but to a lesser extent than was seen with the DGK- mutants.

View larger version (21K): [in this window] [in a new window]   FIG. 9. DGK binds to SF1. (A) Alignment of LXXLL motifs in human DGK isoforms. (B) Cells were transfected with expression plasmids containing FLAG-tagged DGK- or DGK- and pCDNA3.1-SF1 (Myc-tagged) and treated for 1 h with 1 mM Bt2cAMP. Lysates were immunoprecipitated with anti-FLAG antibody and protein A/G Sepharose. SDS-PAGE and Western blotting with an anti-Myc antibody were used to detect SF1 bound to the immunoprecipitated kinases. Some cells were first transfected with siRNA oligonucleotides against DGK or DGK for 72 h and treated for 1 h with 1 mM Bt2cAMP, and then cell lysates were isolated and immunoprecipitated using either anti-DGK- or anti-DGK- antibodies. Purified proteins were subjected to SDS-PAGE and Western blotting with an anti-SF1 antibody. (C) Immobilized SF1 was incubated with in vitro-translated, radiolabeled wild-type or mutant DGK- or DGK-, and the reaction mixtures were washed and subjected to SDS-PAGE. Acrylamide gels were stained, dried, and exposed to a phosphorimager screen. Binding was detected by scanning phosphorimager screens using a Fuji Fluor/Phospho-Imager. (D) H295R cells were transfected with pGL3-CYP17-2x57, pRL-TK, and wild-type or LXXLL mutant DGK- or DGK- expression plasmids. Transfected cells were treated with 1 mM Bt2cAMP for 16 h and the luciferase activity quantified by luminometry.

View larger version (31K): [in this window] [in a new window]   FIG. 10. SF1 colocalizes with DGK and PA in the nuclei of H295R cells. Cells were plated onto glass coverslips, fixed, permeabilized, and incubated with anti-SF1 and anti-DGK- (A) or anti-SF1 and DGK- (B) for 1 h. Coverslips were washed and incubated with anti-fluorescein isothiocyanate and antirhodamine, and immunofluorescence was detected by confocal microscopy. (C) Cells were plated on glass coverslips, incubated with NBD-PA for 12 h, fixed, permeabilized, and incubated with an antibody against SF1. Washed coverslips were incubated with antirhodamine and subjected to confocal microscopy.

Suppression of DGK- attenuates cAMP-stimulated CYP17 mRNA expression.

View larger version (21K): [in this window] [in a new window]   FIG. 11. cAMP-dependent CYP17 transcription requires DGK-. (A) Cells were transfected with siRNA oligonucleotides directed against DGK- or DGK- for 72 h and then treated for 12 h with 1 mM Bt2cAMP. Total RNA was extracted, and real time RT-PCR was performed. CYP17 mRNA expression was normalized to ß-actin content. Values graphed represent the means ± standard errors of the means for three separate experiments, each performed in quadruplicate. *, P < 0.05; statistically different from untreated control. (B) Lysates from cells transfected with siRNA oligonucleotides against DGK-, DGK-, or a scrambled sequence were isolated and analyzed by SDS-PAGE and Western blotting using anti-DGK-, anti-DGK, or anti-p54nrb, as described in Materials and Methods.

	DISCUSSION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

We recently demonstrated that SPH is a ligand for SF1 in the human adrenal cortex (109). SPH antagonizes the receptor and prevents the initiation of CYP17 transcription by impairing coactivator binding to the receptor (109). Our studies presented herein reveal that PA is another ligand for SF1. Mass spectrometric analysis of the receptor that was purified from H295R cells identified several phospholipids bound to SF1, with a 26:1 PA molecule being the most abundant (Fig. 1). Although we have identified phospholipids bound to SF1 purified from H295R cells and our functional data suggest that PA may be the endogenous ligand for the receptor, our analysis of only a small subset of possible ligands, i.e., phospholipids, precludes concluding that PA is the predominant, physiologically relevant ligand for SF1. It is likely that performing unbiased mass spectrometric analysis for all classes of lipids and for other molecules, such as carbohydrates, may identify other molecules that specifically bind to the receptor. In order to comprehensively analyze all classes of molecules that interact with SF1 in different cell types and under different physiological conditions, we are currently optimizing conditions for liquid chromatography tandem mass spectrometry. These mass spectrometric studies coupled with thorough analyses of the functional significance of any identified molecules are likely to reveal the molecule(s) that serves as a ligand for SF1.

Reporter gene assays using PA molecules that contain various acyl chain lengths and degrees of unsaturation demonstrated that the di14 PA species was the most efficacious activator of SF1 of those tested; however, both the di12 and di18:0 PA species also significantly stimulated luciferase activity. Except for the distearoyl PA species, most PA molecules with one or more 18-carbon acyl chains did not significantly increase CYP17 transcriptional activity (Fig. 2A). These data are in agreement with the findings of Li et al. demonstrating that phospholipids comprised of two C₁₂-C₁₆ acyl chains are optimal for stabilizing SF1 in an active conformation (63). Our findings also indicate that although di14 PA had the greatest stimulatory effect on CYP17 reporter gene activity, other PA molecular species can activate the receptor, which provides evidence that the ligand binding pocket of SF1 is able to accommodate differing acyl chain lengths. It is equally possible that the ligand binding pocket can accommodate other classes of lipids. Further, mass spectrometric analysis of SF1 purified from H295R cells also detected short-chain PE species (Fig. 1), and crystallographic studies by others have identified various phospholipids, such as PIP (55) and PE (63), bound to the receptor. However, neither short-chained PE molecules (di12, di14, or di16) nor di8 PI(3,4,5)P₃ was able to stimulate SF1-dependent CYP17 reporter gene activity (data not shown). Collectively, these studies support a role for dynamic ligand binding and dissociation from the cavernous 1,600-ų ligand binding pocket in response to ACTH stimulation.

Reporter gene assays determined that di14 PA was the most efficacious inducer of CYP17 transcriptional activity (Fig. 2A). However, mass spectrometric analysis of phospholipids bound to SF1 purified from H295R cells revealed that the major phospholipid bound to the receptor was a 26-carbon PA species with one double bond, suggesting that the preferred endogenous PA ligand may be a PA molecule that is structurally dissimilar to the di14 PA. The fact that diverse PA species elicited differing magnitudes of activation (Fig. 2A) may indicate that some of these molecular species are partial agonists. We predict that when bound to some PA species and possibly other classes of lipids, SF1 may adopt conformations intermediate between inactive and fully active conformations, with suboptimal coactivator binding. An alignment of SF1 and the peroxisome proliferator-activated receptor (suggests that loss of PA responsiveness in the SF1 inner ligand binding pocket mutant H310A (Fig. 6A) may be functionally analogous to the selective loss of binding and responsiveness to the full agonist rosiglitazone but not a partial agonist that was observed with the peroxisome proliferator-activated receptor H323A mutant (106). Thus, we predict that SF1 H310A may be useful in screens for partial agonists; however, further biophysical studies using phospholipids comprised of acyl chains of various lengths are under way to further characterize the structural and biochemical parameters required for activation and repression of SF1-mediated transcription.

Although the data presented herein, as well as the findings of others (55, 63, 112), provide compelling evidence for an integral role of ligand binding in regulating the ability of SF1 to activate target gene transcription, our studies (Fig. 3A) also suggest that ligand binding is not sufficient for maximal activation of the receptor. It is likely that in addition to increasing the production of PA in H295R cells, activation of the cAMP signaling pathway also stimulates recruitment of coregulatory proteins in a ligand-independent manner. Moreover, since several studies have established the role of posttranslational modifications (11, 12, 19, 31, 44, 46, 54, 60, 95) in controlling receptor function, it is probable that cAMP maximally activates target gene transcription by integrating multiple regulatory mechanisms. As previously discussed, acetylation (11, 44, 46), sumoylation (12, 54, 60), and phosphorylation (19, 31, 95) have all been shown to regulate the ability of SF1 to induce target gene transcription; thus, we propose that ACTH/cAMP activates CYP17 transcription by promoting a series of changes, including posttranslational modification, coregulatory protein recruitment, and ligand binding, all of which are necessary for optimal CYP17 gene expression. We have previously found that phosphatase activity is required for cAMP-dependent CYP17 expression (95, 96), and it is established that phosphorylation of SF1 on S-203 is required for coactivator recruitment (31). Thus, we conclude that the ACTH/cAMP signaling cascade orchestrates a dynamic, coordinate series of events, all of which facilitate steroidogenic gene transcription and subsequent hormone biosynthesis. The order of these interdependent events remains to be fully characterized.

The identification of a molecule that activates SF1 suggests that the receptor exchanges ligands in response to extracellular signaling. We have shown that ACTH/cAMP stimulates the dissociation of SPH from the ligand binding pocket (109), thereby allowing acutely synthesized PA to bind to the receptor as SPH exits the ligand binding pocket. The flexibility of the entryway to the ligand binding cavity (63) supports a role for the exchange of SPH for PA during the switch from an inactive to a canonical active conformation. Previous findings that SPH inhibits coactivator binding and CYP17 transcriptional initiation (109) contrast the increased binding of GCN5 to the CYP17 promoter in response to PA observed in the current studies (Fig. 3D) and indicate that allosteric rearrangement in response to ACTH/cAMP-induced ligand production may facilitate coregulator binding and/or exchange. Indeed, we have preliminary evidence that a series of posttranslational modifications enables the structural changes in the LBD of the receptor that occur during ligand exchange (E. B. Dammer et al., unpublished observations). However, further in vitro and in vivo biophysical analysis of the ligand binding event is necessary to identify the factors that contribute to receptor activation.

We demonstrated that DGK- catalyzes the reaction that yields the ligand for SF1 in response to activation of the cAMP signaling pathway. DGK- is a member of a growing family of kinases, all of which phosphorylate diacylglycerol to produce PA. To date, 10 isoforms have been identified, several of which, including DGK-, are localized in the nucleus (22, 104, 105, 108, 110, 113). DGK- contains a proline-rich region and three cysteine-rich domains in the N terminus, centrally localized Ras-associating and pleckstrin homology domains and a C-terminal catalytic domain (90). While the functions of many of the other domains in DGK- are unclear, the catalytic activity of the enzyme has been shown to depend on all of its domains (65). It has been postulated that the cysteine-rich domains of the enzyme are required both for correct folding of the protein and for substrate presentation (65). Moreover, DGK- activity is modulated by -thrombin and nerve growth factor in the nucleus of IIC9 fibroblasts (8) and PC12 adrenal cells (101), respectively, and by Rho A (37) and norepinephrine (111) in the cytoplasm.

Although the precise mechanism by which PA is delivered to SF1 is unclear, we demonstrate that DGK- and SF1 interact (Fig. 9B and C), and thus, we postulate that the receptor may translocate to distinct regions of the nucleus for ligand binding. DGK- is found in nuclear speckles along with hyperphosphorylated RNA polymerase II and the splicing factor SC-35 in various cell types, including PC12, HeLa, and MCF-7 (102). Nuclear speckles are punctate subcompartments enriched in splicing factors and are located in the interchromatin regions of the nucleoplasm (33, 57). These structures are dynamic in their size, shape, and number, and it is thought that this plasticity is controlled in a cell type-specific manner according to the levels of gene expression and in response to metabolic and environmental signals that influence the available pools of active splicing and transcription factors (57). Moreover, since several kinases have been localized to nuclear speckl