| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Testis; |
Department of Cell Biology and Biochemistry,3 Texas Tech University Health Sciences Center, Lubbock, Texas 79430
Department of Biochemistry and Molecular Biophysics,4 Columbia University Medical Center, New York, New York 10032
ABSTRACT
The expression of the steroidogenic acute regulatory protein (STAR) is regulated by PKA in response to trophic hormone stimulation through the second messenger cAMP. However, in steroidogenic cells, the concentrations of hormone necessary to maximally induce cAMP synthesis and PKA activity are often significantly higher than is necessary to achieve maximum steroidogenesis. One general mechanism believed to make PKA signaling more effective is the use of A-kinase anchoring proteins (AKAPs) to recruit PKA to discrete subcellular compartments, which coordinates and focuses PKA action with respect to its substrates. The characterization of AKAP121 has suggested that it enhances the posttranscriptional regulation of STAR by recruiting both Star mRNA and PKA to the mitochondria, thereby permitting more effective translation and phosphorylation of STAR. Testing this hypothesis revealed that cAMP-induced STAR expression and steroidogenesis closely followed AKAP121 abundance when this AKAP was silenced or overexpressed in MA-10 cells but that these changes were effected posttranscriptionally. Moreover, silencing AKAP121 expression in these cells specifically altered the localization of type II PKA regulatory subunit alpha (PKAR2A) at the mitochondria but did not affect its relative expression within the cell. Affinity purification experiments showed that PKAR2A preferentially associated with AKAP121, and cAMP analogs that activate type II PKA induced STAR phosphorylation more efficiently than analogs stimulating type I PKA. This suggests that AKAP121 and PKAR2A serve to enhance steroidogenesis by directing the synthesis and activation of STAR at the mitochondria in response to cAMP.
A-kinase anchoring proteins, cyclic adenosine monophosphate, Leydig cells, signal transduction, steroid hormones
Early studies predicted that in response to challenge with trophic hormone, steroidogenic tissues used activated protein kinase A (PKA) to regulate acute steroidogenesis through phosphorylation events that controlled the delivery of cholesterol to the P450scc enzyme (CYP11A1) [1–3]. Current models hold to this prediction, where steroidogenesis is dependent on the synthesis of the steroidogenic acute regulatory (STAR) protein that functions to mediate the rate-limiting step of steroidogenesis, the transfer of cholesterol from the cytosol and outer mitochondrial membrane to the inner mitochondrial membrane where CYP11A1 resides [4, 5]. Consequentially, the production, activation, and destruction of STAR are tightly controlled in the cell, and a significant body of information currently explains much about how the expression of STAR is regulated [4, 6–12]. From these studies, it has been concluded that STAR is largely responsible for the cAMP-dependent regulation of steroidogenesis. However, it remains unclear how cAMP concentrations that would be predicted to be ineffective at stimulating steroid production can still permit maximal steroidogenesis, a situation that normally occurs in response to trophic hormone stimulation [13–17]. Although many other factors known to affect Star gene transcription may account for increased STAR synthesis, it is thought that compartmentalization of the cAMP-dependent PKA pathway may increase the sensitivity of the steroidogenic machinery to cAMP [3, 18–20].
The characterization of A-kinase anchoring proteins (AKAPs), which tether PKA to discrete subcellular locations through interactions with their regulatory subunits (type I, alpha and beta, or type II, alpha and beta), has shown a prototypical model by which PKA may be organized within the cell, and this functionally related family of proteins has been demonstrated to target PKA to more than a dozen subcellular locations, including the nucleus, centrosomes, plasma membranes, and mitochondria [21–23]. Presently, however, no AKAPs are known to affect STAR-mediated steroidogenesis. The recent discovery of the Akap1 gene that encodes a homologous family of mitochondrial AKAPs being expressed in multiple endocrine tissues has raised the possibility that this novel family of AKAPs is involved in compartmentalizing PKA signaling in steroidogenic cells [24–29]. The mouse PKA anchoring protein AKAP121 is the most prevalent somatic isoform encoded by the Akap1 gene and is of particular interest because it is known to enhance cAMP signaling to the mitochondria as well as to target mRNAs to the mitochondria through its RNA-binding domain [26, 28, 30–34]. We hypothesize that AKAP121 recruits PKA to the mitochondria, permitting more effective translation and phosphorylation of STAR. In the present study, we show that STAR expression and steroidogenesis closely parallel AKAP121 abundance when this AKAP is overexpressed or silenced in MA-10 mouse Leydig tumor cells. In addition, the silencing of AKAP121 appears to diminish the concentration of PKAR2A associated with mitochondria in these cells. Examination of AKAP121 complexes with affinity purification confirms AKAP121 binding to PKAR2A. These data suggest that AKAP121 enhances steroidogenesis by directing the synthesis and activation of STAR at the mitochondria in response to cAMP.
N6,2-dibutyryladenosine-3',5'-cyclic monophosphate, abbreviated (Bt)2cAMP; cAMP; 8-piperidinoadenosine-3',5'-cyclic monophosphate (PIP-cAMP); phorbol 12-myristate 13-acetate (PMA); 22R-hydroxycholesterol (22R); Triton X-100 (TX-100); hexadimethrine bromide (polybrene); protease inhibitor cocktail; Waymouth MB/752 medium (Waymouth); and Dulbecco modified Eagle medium (DMEM) were purchased from Sigma. N6-mono-tert, butylcarbamoyl-adenosine-3',5'-cyclic monophosphate (MBC-cAMP) and 8-(6-aminohexyl) aminoadenosine-3',5'-cyclic monophosphate (AHA-cAMP) were purchased from Biolog (Bremen, Germany). Platinum Pfx DNA polymerase, PCR primers, DH5
host strain E. coli, trypsin-EDTA, tissue culture-grade antibiotics, PBS, OPTI-MEM reduced serum medium, fetal bovine serum (FBS), and horse serum were purchased from Invitrogen (Carlsbad, CA). RNasin, Taq polymerase, avian myeloblastosis virus-RT, restriction endonucleases, and other enzymes were purchased from Promega (Madison, WI).
BD RetroPack PT67 cells from Clontech (Mountain View, CA) were cultured according to manufacturer guidelines in DMEM supplemented with 10% (v/v) FBS and penicillin/streptomycin (100 U/ml). MA-10 cells, provided by Dr. Mario Ascoli (Department of Pharmacology, University of Iowa College, Iowa City) were maintained in Waymouth medium containing 15% (v/v) heat-inactivated horse serum and 40 µg/ml of gentamicin and cultured as described before [35]. Before conducting experiments, MA-10 cells were washed with PBS and replenished with serum-free Waymouth medium.
The cDNAs for Akap121, Akap84, and the Akap121 mutant deficient for PKA binding (Akap121m) have been described previously [30]. To generate pMSCVpuro-A121, the retroviral expression vector for AKAP121, PCR primers situated in the 5' untranslated region (UTR) and immediately after the termination codon of the mouse cDNA were used to add XhoI and EcoRI restriction endonuclease sites, respectively, to the PCR product after amplification with Pfx DNA polymerase. Digested PCR product was ligated directionally into the pMSCVpuro vector from Clontech. Construction of cDNAs that would express 3xFLAG-tagged AKAP121, AKAP84, and AKAP121m were accomplished similarly with PCR primers that added an in-frame 5' HindIII site and ensured the inclusion of a BamHI site within the 3' UTRs. Digestion with these enzymes permitted directional cloning of the Akap cDNAs into the p3xFLAG-7.1 vector (Sigma) in frame with the repeated FLAG epitope tag to generate p3xF-A121, p3xF-A84, and p3xF-A121m.
Transfections and Transductions
PT67 cells were transfected with the pMSCVpuro retroviral expression vectors with Fugene 6 transfection reagent (Roche Applied Science, Indianapolis, IN) in accordance with the recommendations of the manufacturer. Transfection of the 3xFLAG plasmids into MA-10 cells was accomplished with Fugene HD transfection reagent (Roche).
The production of retroviral particles and the transduction of MA-10 cells were carried out as outlined by Clontech and as previously described [36]. Briefly, we established stably transfected PT67 cell lines expressing either empty pMSCVpuro or pMSCVpuro-A121 and then used sequential rounds of transduction to express the genes in target cells. MA-10 cells to be transduced were seeded 12 h before infection. Medium from PT67 cells expressing either the pMSCVpuro or pMSCVpuro-A121 was added along with polybrene (8 µg/ml) to cultured MA-10 cells for 12 h before being replaced with standard medium. A second 12-h transduction/12-h recovery was administered the following day. Forty-eight hours after the initial transduction, the cells were used for experiments.
The transfection of siRNAs into MA-10 was accomplished with X-tremeGENE siRNA transfection reagent (Roche) according to the manufacturer's directions. All transfections were conducted with a final siRNA concentration of 100 nM. Silencer negative control 1, Silencer glyceraldehyde phosphate dehydrogenase (GAPDH) siRNA, and the Akap1 siRNAs were ordered as annealed oligos from Ambion (Austin, TX). The sequences for the Akap1 siRNAs that follow correspond to the sense strand sequence: Akap1 siRNA #1, 5'-GGUGGAGACACUGAAGGUU-3'; Akap1 siRNA #2, 5'-GGAAGAAUAUAUUGUUGGG-3'.
Total RNA was isolated from cultured cells with TRIzol reagent according to the directions of the supplier. One-step RT-PCR was performed as described previously on total RNA with a single tube assay [37, 38]. The following PCR primers were designed to detect Akap1 mRNA and were used in conjunction with the previously described primers used to detect mRNAs for Star and the ribosomal protein L19 (Rpl19) [39]. (It should be noted that AKAP84 is predicted to be germ cell-specific and that primers were used that would specifically recognize Akap121 mRNA [25].) 1) Akap121 (forward) 5'-GGAGGTGAGGGAGAAGAGGT-3' (bases 685 to 704); 2) Akap121 (reverse) 5'-AGGCAGGGTGGAGATGTAGA-3' (bases 1803 to 1784); 3) Star (forward) 5'-GACCTTGAAAGGCTCAGGAAGAAC-3' (bases –51 to –27); 4) Star (reverse) 5'-TAGCTGAAGATGGACAGACTTGC-3' (bases 931 to 908); 5) Rpl19 (forward) 5'-GAAATCGCCAATGCCAACTC-3' (bases 154 to 173); and 6) Rpl19 (reverse) 5'-TCTTAGACCTGCGAGCCTCA-3' (bases 559 to 540).
PCR products were resolved on 1% agarose gels and visualized either by ethidium bromide staining the wet gels or by vacuum drying the gels and exposing them to autoradiography film (Marsh Bio Products, Inc., Rochester, NY). The signal strengths of the PCR products were determined by densitometry with Scion Image (Scion Corporation, Fredrick, MD).
Cell Lysis and Mitochondrial Isolation
Whole cell lysates to be used for affinity purification or Western blot analysis were prepared as previously described [40]. Briefly, cells were washed three times with cold PBS, harvested in cold PBS, and then centrifuged and resuspended in NET-2 buffer (50 mM Tris-HCl [pH 7.4], 300 mM NaCl, and 0.05% NP-40) containing 10 µl/ml of protease inhibitor cocktail. Cells were sonicated with a Tekmar TK300 sonic disruptor (Tekmar, Cincinnati, OH) with three 10-sec pulses and then centrifuged at 10 000 x g for 10 min at 4°C. Supernatants were reserved, and protein concentrations were calculated by Bio-Rad Protein microassays (Bio-Rad, Hercules, CA).
The isolation of mitochondria was carried out as previously described with slight modifications [41]. Cells were washed with PBS and collected in TSE buffer (0.25 M sucrose, 10 mM Tris [pH 7.4], and 0.1 mM EDTA) containing 1 µg/ml of protease inhibitor cocktail. The cells were then lysed with a Potter Elvehjem homogenizer by 35 passes at 1000 rpm. The initial lysates were centrifuged twice at 600 x g for 30 min at 4°C, with the supernatants transferred to clean tubes after each step. The resulting supernatants were again transferred to clean tubes and then centrifuged at 12 000 x g for 30 min at 4°C to pellet the mitochondria. Pellets were resuspended in 1 ml of TSE buffer in microfuge tubes and centrifuged at 35 000 x g for 5 min. Resulting mitochondrial pellets were then resuspended in NET-2 buffer, if used for Western blot analysis, or in kinase buffer, if used for in vitro kinase assays. Protein concentrations were determined by microassay as described above.
The 3xFLAG-tagged proteins were expressed and affinity purified from MA-10 whole cell lysates as described by Chu et al. [42] with slight modifications. Cleared cell lysate supernatants were incubated by end-over-end rotation with 10 µl of washed anti-FLAG-M2-agarose beads (Sigma) for 2 h at 4°C. Beads were then gently centrifuged (1000 x g for 5 min at 4°C) and washed five times with NET-2 buffer. Recovered beads (bead-bound fractions) were resuspended in SDS sample buffer (0.1 M Tris-HCl [pH 6.8], 2% SDS, 10% glycerol, 5% 2-mercaptoethanol, and 0.002% bromophenol blue) before analysis by SDS-PAGE and Western blotting.
In Vitro Kinase Assay for STAR Phosphorylation
Mitochondria isolated from cultured MA-10 cells were resuspended in kinase buffer (10 mM morpholinepropanesulfonic acid [pH 7.2], 0.2 mM EDTA, 0.2 mM EGTA, 250 mM sucrose, 15 mM MgCl2, 0.2 mM ATP, 0.2 mM dithiothreitol, 5 mM β-glycerol phosphate, 0.2 mM sodium orthovanadate, and 0.2% TX-100) to give a protein concentration of 3 µg/µl. For each 50-µl reaction, 60 µg of mitochondria was incubated for 30 min at 37°C. For samples treated with recombinant PKA catalytic subunits (Calbiochem, San Diego, CA), we diluted 0.25 µl of the stock catalytic subunit (
2500 U/µl) into 10 µl of kinase buffer containing 100 µg/ml of BSA and used 1 µl of the dilution per reaction.
SDS-PAGE was performed as previously described [43]. Primary antibodies directed against AKAP121 (sc-6439), type I PKA regulatory subunits alpha/beta (PKAR1), PKAR2A (sc-909), type II PKA regulatory subunit beta (PKAR2B) (sc-25424), and alpha-actin (sc-1616), as well as donkey anti-goat immunoglobulin G (IgG) conjugated to horseradish peroxidase (HRP), were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit anti-CYP11A1 antibody was from Chemicon (Temecula, CA). The mouse monoclonal anti-cytochrome oxidase IV (COX4) antibody was purchased from Molecular Probes (Invitrogen). Total STAR was detected as previously described [41], and the rabbit antibody for the detection of phosphorylated STAR (phospho-STAR) [44] was a gift from Dr. Steven King (Scott Department of Urology, Baylor College of Medicine, Houston, TX). Donkey anti-rabbit IgG conjugated to HRP was obtained from Amersham Biosciences (Piscataway, NJ). Anti-mouse IgG conjugated to HRP was purchased from Promega. Chemiluminescent detection was performed with Western Lightning ECL reagent (Perkin-Elmer Life Sciences, Boston, MA).
The synthesis of progesterone from treated MA-10 cells was quantified by examining the recovered medium by RIA as previously described [45]. Triplicate groups of cells were cultured within each experiment, and progesterone concentrations were normalized to total cellular protein and expressed as picograms of progesterone per microgram of protein.
All experiments were repeated independently a minimum of three times. A representative image for each RT-PCR and Western blot experiment is shown, and densitometry data from these analyses were normalized against loading controls and then expressed as the mean fold change relative to the negative controls that were arbitrarily set to 1.0; for STAR and phospho-STAR, Western data were expressed as the fold change relative to (Bt)2cAMP-treated controls. RIA data are expressed as the means ± SEM. When only two groups were being compared, data were analyzed by a Student t-test. Statistical test results were considered significant for P < 0.05 and are denoted within the figures with an asterisk (*). Statistical analysis of multiple groups was modeled by ANOVA, and pairwise comparisons were made by the Fisher protected least significant difference test. Groups denoted with lower case letters within the figures were determined to be statistically different from the other groups for P < 0.05.
AKAP121 Is Associated with MA-10 Cell Mitochondria
AKAP121 was previously detected in MA-10 cells where its expression was shown to be upregulated in response to hCG treatment [46], but, to our knowledge, the role of AKAP121 in cAMP-induced steroidogenesis has not been studied. MA-10 cells were treated with 0.5 mM (Bt)2cAMP, and the expression of AKAP121 was assessed by Western blot in whole cell lysates, in cytosolic isolates, and in isolated mitochondria for 4 h (Fig. 1). For comparison, STAR abundance in the mitochondria of the same samples was also examined. In this and subsequent experiments, 0.3–0.5 mM (Bt)2cAMP concentrations were used, which induces STAR and steroidogenesis in MA-10 cells to approximately half their maximal capacity. This level of induction was more sensitive to perturbations in either STAR expression or steroid production and permitted easier assessment of the effects of additional treatments on steroidogenesis in these cells. AKAP121 was detected and shown to be highly enriched, if not exclusively localized to the mitochondria. However, AKAP121 concentrations showed little change during the treatment, in contrast to the increase in STAR.
|
Expression of AKAP121 in MA-10 Cells Enhances cAMP-Stimulated STAR Expression and Steroidogenesis
To ascertain whether AKAP121 alters STAR expression or steroidogenesis in response to PKA activation, Akap121 was transduced into MA-10 cells by retroviral-mediated gene transfer. MA-10 cells transduced with either the empty pMSCVpuro virus or the pMSCVpuro-A121 construct were incubated for 6 h in the presence or absence of 0.3 mM (Bt)2cAMP and then harvested to isolate either total RNA or mitochondria. Under these conditions, (Bt)2cAMP produced no significant change in Akap121 mRNA or protein expression; however, MA-10 cells transduced with the AKAP121 virus had more than a 3-fold increase in detectable Akap121 mRNA relative to the control transductions (P < 0.05) and nearly a 4-fold increase in mitochondrial AKAP121 (Fig. 2, A and B; P < 0.05). Star mRNA abundance increased during treatments with (Bt)2cAMP to about 3-fold higher than that in the unstimulated groups, whereas AKAP121 overexpression did not enhance Star mRNA expression in either the presence or absence of (Bt)2cAMP (Fig. 2A). STAR was not readily detected in unstimulated cells, but a comparison of the (Bt)2cAMP-treated groups showed that STAR expression and concentrations of phospho-STAR were doubled when AKAP121 was overexpressed (Fig. 2B; P < 0.03). Concentrations of Rpl19 mRNA (Fig. 2A) and CYP11A1 (Fig. 2B), which served as controls in their respective experiments, did not change as a result of either (Bt)2cAMP treatment or the overexpression of AKAP121. The pattern of steroid production closely mirrored that of STAR and phospho-STAR, with little steroidogenesis detectable in the absence of (Bt)2cAMP (respective progesterone concentrations in the unstimulated control group and AKAP overexpressing groups after 6 h were 2.54 ± 0.1 and 3.09 ± 0.2 pg of progesterone per microgram of protein; Fig. 2C). Stimulation of the transduced cells showed that cells overexpressing AKAP121 produced 610.30 ± 35.2 pg of progesterone per microgram of protein, which was a 3.8-fold increase over the control transductants, which produced 158.58 ± 14.0 2 pg of progesterone per microgram of protein (P < 0.0001).
|
Silencing AKAP121 in MA-10 Cells Reduces cAMP-Induced STAR Expression and Steroidogenesis
The impact of AKAP121 on STAR expression and steroidogenesis in MA-10 cells was gauged with siRNAs specific to the Akap1 gene to reduce the endogenous expression of AKAP121. Transfection conditions for MA-10 cells were optimized with a Gapdh siRNA, which efficiently reduced GAPDH expression by 85% from 24 to 72 h posttransfection and a scrambled negative control siRNA; neither siRNA altered the steroidogenic response of the cells to (Bt)2cAMP stimulation (data not shown). To silence AKAP121, MA-10 cells were transfected with 100 nM negative control siRNA, 100 nM (one of two) different Akap1 siRNAs, or a mixture that contained 50 nM (each) both Akap1 siRNAs. After transfection, all groups were subjected to a 6-h stimulation with 0.3 mM (Bt)2cAMP. Akap121 mRNA concentrations were reduced by 36% and 44% with Akap1 siRNA 1 and Akap1 siRNA 2, respectively (Fig. 3A; P < 0.005). Treatment with both Akap1 siRNAs reduced Akap121 mRNA concentrations by more than 70% (P < 0.0005). Western blot analysis showed that AKAP121 expression closely corresponded to Akap121 mRNA concentrations, falling by nearly 40% when the individual siRNAs were used and by 73% when the siRNA mixture was used (Fig. 3B; P < 0.0001). No statistically significant change in Star mRNA abundance was seen after any of these transfections (Fig. 3A); however, the abundance of both STAR and phospho-STAR closely paralleled AKAP121 expression and was significantly reduced after the silencing of AKAP121 (Fig. 3B). STAR and phospho-STAR were reduced to nearly one-third of control concentrations when AKAP121 was silenced with the pair of siRNAs (P < 0.0001). Detection of CYP11A1 on the same blots confirmed equal loading and also ruled out nonspecific silencing. Paralleling the drop in STAR and phospho-STAR concentrations, progesterone production decreased in groups treated with the Akap1 siRNAs relative to controls (Fig. 3C). Control cells produced 292.75 ± 8.7 pg of progesterone ml–1 µg–1 of protein after 6 h. Cells receiving either of the Akap1 siRNAs produced about half as much progesterone as controls (P < 0.0001), and the cocktail of Akap1 siRNAs reduced steroidogenesis by 75% (P < 0.0001).
|
To ensure that the use of Akap1 siRNAs was not affecting other mitochondrial enzymes involved in steroidogenesis, MA-10 cells were transfected with either the negative siRNA control or the cocktail of Akap1 siRNAs and then treated with either 0.3 mM (Bt)2cAMP or 25 µM 22R for 6 h. The hydroxylated cholesterol derivate 22R is water soluble, serves as a membrane-permeable substrate for CYP11A1 in the mitochondria, and does not require STAR to traverse the mitochondrial membranes and intermembrane space. Treating MA-10 cells with 22R saturates the steroidogenic machinery with a substrate that is rapidly converted to a steroid hormone unless other blocks to steroidogenesis are in place. STAR and AKAP121 expression were not significantly affected by 22R and responded to the siRNAs and (Bt)2cAMP as before, but steroidogenesis induced after 22R treatment was not affected after silencing AKAP121, indicating that the steroidogenic machinery was intact (Fig. 4).
|
Silencing AKAP121 Alters PKAR2A Localization
To evaluate what effects silencing AKAP121 might have on the compartmentalization of PKA, we profiled the abundance of PKA regulatory subunits in MA-10 whole cell lysates (Fig. 5, left panels) and in isolated mitochondria (Fig. 5, right panels) by Western blot analysis. MA-10 cells were transfected with the mixture of Akap1 siRNAs or with control siRNA, as used in Figure 3, and then each group was treated with or without (Bt)2cAMP as before. As shown above, the pair of Akap1 siRNAs strongly attenuated AKAP121 abundance and (Bt)2cAMP-stimulated STAR expression (Fig. 5). Both type I (PKAR1) and type II (PKAR2) PKA regulatory subunits were observed in whole cell lysates, but under the conditions tested, neither (Bt)2cAMP nor the Akap1 siRNAs caused a significant change in the overall abundance of any of the PKA regulatory subunits (Fig. 5, left panels). Importantly, though, the abundance of mitochondrially localized PKAR2A was reduced by 51% in groups where AKAP121 had been silenced (Fig. 5, right panel; P < 0.0005). Other than PKAR2A, the concentrations of the PKA regulatory subunits in the mitochondria were unaffected by Akap1 siRNAs. Similarly, (Bt)2cAMP did not affect the mitochondrial concentrations of any of the PKA regulatory subunits. Both the alpha and beta isoforms of PKAR1 could be detected when independent antibodies specific to each isoform were used, but their patterns of expression and localization were unchanged by the Akap1 siRNAs and (Bt)2cAMP treatment (data not shown). For simplicity, concentrations of PKAR1 are shown with an antibody that does not discriminate between the PKAR1 alpha and beta isoforms. Actin and COX4 served as loading markers for the whole cell lysates and mitochondrial fractions, respectively.
|
PKAR2A Associates with AKAP121 in MA-10 Cells
AKAP family members bind PKAR1 and/or PKAR2 with different efficiencies [27, 47–49]. To ascertain which isoforms preferentially associate with AKAP121 in MA-10 cells, we used affinity purification techniques to isolate epitope-tagged Akap1 gene products. Three different AKAP1 variants were N-terminally tagged with the 3xFLAG epitope (Fig. 6A) and transiently expressed in MA-10 cells. Summarily, 3xF-A121 includes the wild-type protein endogenously expressed in MA-10 cells. The sequence for 3xF-A84 derives from a splice variant of the Akap1 gene thought to be germ cell-specific [25], and it contains the same PKA-binding domain as AKAP121 but lacks the C-terminal region, including the KH and Tudor domains that mediate RNA binding and self-association [29, 50]. The 3xF-A121m construct was generated from a previously described mutant form of AKAP121, in which two leucine-to-proline substitutions in the PKA-binding domain prevent proper association with the regulatory subunits of PKA [30]. Lysates from cells transfected with either the empty 3xFLAG vector or the tagged Akap1 cDNAs were affinity purified and probed with an anti-FLAG antibody. The FLAG-tagged AKAPs were readily detected in the lysates and bead-bound fractions but not in the wash fractions, whereas no FLAG reactivity was observed in the cells transfected with an empty vector (Fig. 6B, top panel). PKAR1 was detected as before, with an antibody that recognized both the alpha and beta isoforms, but it was not observed copurifying with any of the tagged AKAPs (Fig. 6B, middle panel, bound fractions). PKAR2A was copurified with both 3xF-A84 and 3xF-A121 but not with 3xF-A121m or in the control transfected group, showing that this association was mediated through the PKA-binding domain common to the wild-type AKAPs (Fig. 6B, bottom panel, bound fractions). An immunoreactive band with an apparent molecular mass of 32 kDa was consistently seen in the lysates probed with the PKAR2A antibody; however, this protein did not persist in the affinity-purified complex. (Notably, this background band is reported by Santa Cruz for this product; we observed it consistently in whole cell lysates and, to a lesser degree, in isolated mitochondria.) PKAR2B was also examined by this technique, and it did not appear to copurify with any of the AKAPs (data not shown).
|
Type II PKA Agonists Stimulate STAR Phosphorylation
The activation of protein kinase C by PMA was previously shown to promote the synthesis but not the phosphorylation of new STAR in MA-10 cells [44]. It was further demonstrated that 50 µM (Bt)2cAMP, a concentration normally ineffective at inducing Star gene transcription and steroidogenesis, was capable of activating the PKA necessary to phosphorylate STAR generated in response to PMA. To discriminate whether certain isoforms of PKA are more effective in phosphorylating STAR, we extended these conditions with pairs of cAMP analogs that can preferentially serve as ligands for either PKAR1 or PKAR2 [19, 51–53].
Three analogs of cAMP were used in these experiments, in addition to (Bt)2cAMP. PIP-cAMP preferentially binds site A of PKAR1 and site B of PKAR2 with similar affinities, and although not efficient at activating PKA alone, this analog synergizes effectively with complementary analogs that can bind to the opposite site on either isoform of PKA [54, 55]. Either AHA-cAMP, which highly prefers site B of PKAR1, or MBC-cAMP, which binds to site A of PKAR2, was used in combination with PIP-cAMP to preferentially activate its respective PKA isoform [55, 56]. The type I (0.5 µM AHA-cAMP and 100 µM PIP-cAMP) and type II (0.5 µM MBC-cAMP and 100 µM PIP-cAMP) analog pairs were intentionally used at low concentrations that were incapable of independently inducing STAR expression or eliciting a steroidogenic response in MA-10 cells after 6 h (Fig. 7, lanes 4 and 6). Notably, these same concentrations were both capable of modestly activating the Star promoter (data not shown). In comparison, MA-10 cells treated with 10 nM PMA or 0.3 mM (Bt)2cAMP produced STAR, but PMA-induced STAR was not phosphorylated, as seen in whole cell lysates from MA-10 (Fig. 7A). When used in combination with PMA, the type I pair did not enhance STAR expression, phosphorylation, or activity beyond what was seen with PMA alone (Fig. 7, lanes 3 and 5). The type II pair resulted in the phosphorylation of PMA-induced STAR to concentrations comparable with those seen in the group treated with 0.3 mM (Bt)2cAMP (Fig. 7A) and stimulated about half as much progesterone (Fig. 7B). Notably, the steroidogenic response from cells treated with PMA and the type II analog pair was 6.2 times greater than when PMA was used with the type I pair (P < 0.001).
|
Given that PKAR1 was readily detected at the mitochondria in MA-10 cells (Fig. 5), it was surprising to observe that it was unable to phosphorylate STAR (Fig. 7). To ensure that type I PKA at the mitochondria was indeed activated by its agonists, the experiment shown in Figure 7 was repeated, but STAR phosphorylation was accomplished in vitro. MA-10 cells were grown in two groups treated with or without 10 nM PMA for 6 h to induce STAR. After this treatment, the mitochondria were isolated from each group and immediately used for in vitro kinase assays with cAMP, recombinant PKA catalytic subunits, or the pairs of PKA-preferring analogs. Unphosphorylated STAR was detected in the mitochondria from PMA-treated cells, but neither recombinant PKA catalytic subunits (which do not need cAMP for activity) nor 100 µM cAMP added in vitro resulted in detectable phospho-STAR, indicating that the PKA in these assays did not have access to the STAR that had been internalized to the inner mitochondrial membrane (Fig. 8A). The inclusion of TX-100 in the kinase buffer, which solubilized the mitochondria, resulted in the phosphorylation of STAR by both the recombinant kinase and endogenous PKA activated by cAMP (Fig. 8A). With TX-100 included in the kinase buffer, we tested the ability of the type I and type II analog pairs to activate mitochondrial PKA in vitro, with the same concentrations that had been applied to the cells in culture as shown in Figure 7. Interestingly, under these conditions, both pairs of analogs were effective at activating mitochondrial PKA, as judged by the concentrations of phospho-STAR in comparison with each other or with samples treated with either cAMP or recombinant PKA (Fig. 8B).
|
The data given in the present study provide the first direct evidence that an AKAP is involved in the posttranscriptional regulation of STAR-mediated steroidogenesis and, specifically, that the accumulation of STAR through PKA-dependent pathways is enhanced by AKAP121. We currently hypothesize that PKA has at least three roles in regulating STAR expression: first, PKA serves to control the Star promoter [4]; second, PKA enhances the translation of Star mRNA [57, 58]; and finally, PKA promotes STAR phosphorylation and activation [7, 58, 59]. Based on our present data, it is likely that PKA in complex with AKAP121 directs one or both of these latter two events. This level of regulation presumably reflects the importance of fine-tuning the concentrations of STAR-mediated steroidogenesis, and the impact of PKA in moderating both the synthesis and activity of STAR underscores the critical role that cAMP plays in this process. Previous studies have already defined how Star gene transcription is acutely regulated by PKA through numerous transcription factors, and the treatment of steroidogenic tissues with cAMP or one of its analogs quickly activates the Star promoter [4, 43, 60]. Although significantly less is known regarding how PKA is involved in the posttranscriptional regulation of STAR [8, 57, 61], the manifold examples that have already been brought to light in other systems have strong implications for the involvement of AKAPs, and particularly for AKAP121 [24, 30–32, 62].
One of the earliest indicators for the involvement of AKAPs in steroidogenesis was the realization that in many steroidogenic cell types, the concentrations of trophic hormone necessary to maximally induce cAMP synthesis are many times higher than is necessary to achieve maximum steroidogenesis [13, 14, 16]. It was quickly resolved that small increases in cAMP after trophic hormone stimulation led to the occupation of high-affinity or "preferred" cAMP receptor proteins, which are sufficient to permit maximal steroidogenesis [15, 17, 20]. However, the mechanisms by which maximal STAR synthesis and steroidogenesis are achieved in the context of submaximal cAMP concentrations are only now being discerned [44, 63]. It has been suggested that trophic hormone binding elicits only modest adenyl cyclase activity, with relatively few molecules of cAMP occupying key sites and permitting the PKA-dependent processes of steroidogenesis to proceed [17, 18, 44, 64–66]. The emergence of AKAPs as a means to localize PKA within the cell provides an attractive mechanism for defining the specificity of these "preferred receptors," by means of limiting the access of PKA to specific substrates, and for enhancing their effectiveness, by concentrating the enzyme to discrete locations within the cell [22, 23]. We propose that PKA bound to AKAP121 and the outer mitochondrial membrane represents such a "preferred receptor." By channeling cAMP signaling to mitochondrial substrates, including STAR, AKAP121 enhances the effectiveness of low cAMP concentrations in steroidogenesis. Our current data support this concept by demonstrating that the PKA-dependent induction of STAR is greatly enhanced by the presence of AKAP121 in Leydig cells. The biological consequences of the presence or absence of AKAP121 were directly observed, because silencing its expression reduced (Bt)2cAMP-stimulated steroidogenesis in Leydig cells, whereas overexpressing AKAP121 greatly enhanced their steroidogenic output. This becomes more interesting in light of the recent observation that Akap1 gene expression is upregulated in response to hCG in MA-10 cells [46]. Notably, (Bt)2cAMP did not regulate AKAP121 expression under the conditions we tested, placing both AKAP121 expression and cAMP production downstream of LH signaling. Thus, it seems plausible that AKAP121 serves as a hormonally regulated means of enhancing cAMP signaling to the mitochondria. These data mesh well with previous studies that have established a growing precedent for AKAP involvement in steroidogenesis [24, 46, 52, 67, 68]. Earlier experiments in granulosa cells demonstrated that PKA is redistributed as a result of the induction of specific AKAPs in response to FSH [24], and more recent data suggest that AKAP121 is one of the many AKAPs involved [34]. As preantral rat follicles differentiate before ovulation, the response of granulosa cells to cAMP is observed to change, partly due to the increased expression and repartitioning of type II PKA, a change that is coincident with STAR induction [52, 67, 68]. It seems likely that the ability of AKAP121 to enhance STAR-mediated steroidogenesis plays a role in granulosa cell differentiation. Adding another layer to this complexity, studies in rat Leydig cells demonstrated an interaction between the PKAR1A anchoring protein PAP7 and the peripheral-type benzodiazepine receptor, a mitochondrial membrane protein also required for cholesterol transfer into the mitochondria during acute steroidogenesis [69]. Although it appears certain that AKAPs are regulating STAR expression and steroidogenesis, exactly which specific AKAPs are involved and how they do so are only now being determined.
We find that AKAP121 enhancement of STAR expression is entirely posttranscriptional, as evidenced by our observation that STAR expression and phosphorylation, but not Star mRNA abundance, were correlated with the presence of AKAP121. Although the mechanistic basis of this regulation is not entirely proven, the expression of AKAP121 appeared to modulate the abundance of STAR under conditions in which Star mRNA concentrations were fixed. One hypothesis initially suggested by Li et al. [46] is that AKAP121 binds and recruits Star mRNA to the outer mitochondrial membrane, which, in turn, permits more efficient expression of STAR. AKAP121 is known to act in this fashion with at least two other mitochondrially targeted mRNAs: those for manganese superoxide dismutase (Sod2) and the F2 subunit of the ATP synthase F0 complex (ATP5J2) [31]. AKAP121 and its human homolog, AKAP149, are the only presently known AKAPs that possess an RNA-binding KH domain. Thus far, mRNAs known to interact with these AKAPs have been shown to do so through unique double-hairpin structures in their 3' UTR [31, 32]. Analysis of the 3' UTR of STAR mRNA in the same manner as Sod2 and ATP5J2 transcripts by Mfold software predicts two stem-loop structures bearing close resemblance to those that bind AKAP121 [70, 71], but whether these structures promote the binding of Star mRNA to AKAP121 remains to be proven. These findings are made more interesting by the recent discovery that alternative splicing of the Star mRNA 3' UTR is directly influenced by PKA activation, which, in turn, leads to altered rates of degradation for the different Star transcripts [72]. The data presented do not directly test whether Star mRNA can bind AKAP121 in Leydig cells; however, several preliminary experiments in vitro and in vivo suggest that that they do interact, supporting such a mechanism and suggesting a novel means by which STAR function is tied to the regulation of its synthesis.
AKAP121 may also direct STAR-mediated steroidogenesis by tethering PKA to the mitochondria, because the effective translation of STAR and its posttranslational activation are both dependent on PKA activity. In steroidogenic cells, STAR is rapidly mobilized to or translated at the outer mitochondrial membrane, the first location where the protein is readily observed [73]. Because the half-life of STAR increases from minutes to hours after its translocation into the mitochondria, it is apparent that the spatiotemporal localization and activation of STAR are rigidly controlled in the cell [74, 75]. In adrenocortical cell lines, Star mRNA being actively translated by polyribosomes fails to generate normal STAR concentrations at the mitochondria when PKA activity is reduced or inhibited, and similar studies designed to characterize the effects of the phosphorylation of STAR have shown that PKA can enhance STAR expression posttranscriptionally [59, 61]. This has led to the hypothesis that PKA increases STAR expression by enhancing a cotranslational import mechanism, and we predict that this occurs in concert with AKAP121 [46, 73]. PKA may also be involved in repressing the degradation of STAR as well as in enhancing the ability of STAR to transfer cholesterol into the mitochondria [7, 58, 59, 76]. Such a mechanism for PKA cannot be excluded because STAR itself is a substrate for PKA, and the specific phosphorylation of STAR on at least one of two consensus PKA-phosphorylation sites (Ser56/57 and Ser194/195 in mouse/human) is necessary to render it fully active [7, 8]. Our data indicate that AKAP121 preferentially associates with PKAR2A in MA-10 cells, because this subunit is readily copurified with AKAP84 and AKAP121. Notably, the association of PKAR2A with the mitochondria in these cells was dependent on the presence of AKAP121. Viewed in conjunction with our overexpression and silencing data suggests that AKAP121 enhances the PKA-dependent posttranscriptional processing of STAR by recruiting type II PKA to the mitochondria. This result stands in contrast to earlier reports that suggested that type I PKA relayed cAMP signaling to the steroidogenic machinery in Leydig cells [19]. To address this issue specifically, we examined how type I and type II PKA affected STAR phosphorylation. Although both types of PKA are present in MA-10 mitochondria and both are capable of phosphorylating STAR in vitro, type II PKA was far more effective at phosphorylating STAR in vivo. Consequently, under conditions in which STAR availability was not rate limiting, the activation of type II PKA permitted more steroid hormone synthesis than the stimulation of type I PKA. Notably, the phosphorylation of STAR in vitro was accomplished only when detergent was used to solubilize the mitochondria, suggesting that STAR that has been translocated to the inner mitochondrial membrane can no longer be accessed by PKA. We propose that PKA compartmentalization at the mitochondria provides type II PKA with better access to STAR as it is brought into the mitochondria and that AKAP121 is responsible for approximating the kinase to its substrate. Importantly, this does not eliminate the critical role for type I PKA in regulating steroidogenesis. Although type II PKA agonists more readily activated STAR and permitted greater steroidogenic output, they did not maximally induce steroidogenesis. We predict that AKAPs are likely to orchestrate both type I- and type II-specific events controlling steroidogenesis, not only with regard to regulating STAR expression but also to controlling the traffic of cholesterol within the cell. Supporting this view, two laboratories have reported that peptides that disrupt either type I or type II PKA from docking with their cognate AKAPs have reduced hormone-stimulated steroidogenesis [77, 78].
Our results indicate that AKAP121 can regulate the posttranscriptional synthesis of STAR, in turn serving to tune the capacity of steroidogenic cells to cAMP signaling. However, several aspects regarding how AKAP121 actually performs this role now await further investigation. As already mentioned, the capacity for AKAP121 to bind Star mRNA remains to be conclusively demonstrated. In addition, the recent characterization of Akap1 knockout mice showed no immediately discernible aberrations in steroidogenesis, although normal ovulation was disrupted in the females [34]. Given the complex means of regulating the Star gene and the importance of continued steroidogenesis, it seems likely that other mechanisms are in place to ensure successful cholesterol transfer in these animals, and as such, these animals may present an excellent model for examining the role of AKAP121 in steroidogenesis. To date, the best understood mechanisms for regulating the synthesis of STAR are those governing Star gene transcriptions, where PKA and other signaling pathways are known to rapidly activate and shut down the STAR promoter in response to different stimuli. In the present study, we have demonstrated that the posttranscriptional regulation of STAR by type II PKA through AKAP121 can greatly modulate steroidogenesis in Leydig cells and suggest a novel mechanism for regulating the synthesis of STAR.
FOOTNOTES
1Supported by funds from the National Institutes of Health (NIH) grant HD-17481 and grant B1-0028 from the Robert A. Welch Foundation (to D.M.S.). J.J. was supported by the NIH training grant DK-007328. 
Correspondence: 2FAX: 806 743 2990; e-mail: doug.stocco{at}ttuhsc.edu
Received: 22 August 2007.
First decision: 20 September 2007.
Accepted: 31 October 2007.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  P. R. Manna, M. T. Dyson, and D. M. Stocco Regulation of the steroidogenic acute regulatory protein gene expression: present and future perspectives Mol. Hum. Reprod., June 1, 2009; 15(6): 321 - 333. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Duan, N. Cherradi, J.-J. Feige, and C. Jefcoate cAMP-Dependent Posttranscriptional Regulation of Steroidogenic Acute Regulatory (STAR) Protein by the Zinc Finger Protein ZFP36L1/TIS11b Mol. Endocrinol., April 1, 2009; 23(4): 497 - 509. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. R. Manna, M. T. Dyson, Y. Jo, and D. M. Stocco 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 Endocrinology, January 1, 2009; 150(1): 187 - 199. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. R Manna and D. M Stocco The role of JUN in the regulation of PRKCC-mediated STAR expression and steroidogenesis in mouse Leydig cells J. Mol. Endocrinol., November 1, 2008; 41(5): 329 - 341. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Jana, X. Yin, R. B Schiffer, J.-J. Chen, A. K Pandey, D. M Stocco, P. Grammas, and X. Wang Chrysin, a natural flavonoid enhances steroidogenesis and steroidogenic acute regulatory protein gene expression in mouse Leydig cells J. Endocrinol., May 1, 2008; 197(2): 315 - 323. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
