Transcription factor GATA-4 is activated by phosphorylation of serine 261 via the cAMP/protein kinase a signaling pathway in gonadal cells.

Gonadal gene expression is regulated by pituitary hormones acting through the cAMP/protein kinase A (PKA) signal transduction pathway. The downstream molecular effectors of these signals, however, have yet to be fully understood. We have recently shown that cAMP stimulation of gonadal cells leads to phosphorylation of the transcription factor GATA-4, a key regulator of gonadal gene expression, thus suggesting that this factor might be a novel target for the cAMP/PKA signaling pathway. We now show that the rapid phosphorylation of GATA-4 induced by cAMP in vivo can be blocked by a PKA-specific inhibitor but not by mitogen-activated protein kinase inhibitors, indicating that GATA-4 is predominantly phosphorylated by PKA in response to cAMP in gonadal cells. In addition, using in vitro kinase assays, we show that PKA phosphorylation of GATA-4 occurs predominantly on an evolutionarily conserved serine residue located at position 261. Phosphorylation of GATA-4 Ser261 by PKA enhances its transcriptional activity on different gonadal promoters, an effect that was markedly reduced with a S261A mutant. Moreover, the S261A mutant blunted cAMP-induced promoter activity in gonadal cells. Finally, PKA-dependent phosphorylation of GATA-4 also led to enhanced recruitment of the CREB-binding protein coactivator. This recruitment and transcriptional cooperation were dramatically impaired with the S261A mutant. Thus, our results identify GATA-4 as a novel downstream effector of cAMP/PKA signaling in gonadal cells, where phosphorylation of Ser261 and recruitment of CREB-binding protein likely represent a key mechanism for conveying the cAMP responsiveness of gonadal genes that lack classical cAMP regulatory elements.

pathway, cAMP associates with the regulatory subunits of the protein kinase A (PKA) 1 holoenzyme, allowing dissociation of its catalytic subunits, which then translocate to the nucleus and phosphorylate target proteins (1,2). The best studied target of PKA is the transcription factor cAMP-responsive element-binding protein (CREB), which binds as a dimer to the 8-bp palindromic sequence CRE found in the regulatory region of some cAMP-regulated genes. Phosphorylation of CREB Ser 133 by PKA allows recruitment of the coactivator CREBbinding protein (CBP), which contacts the transcriptional machinery leading to increased gene transcription (3). In the gonads, however, several cAMP-regulated genes, such as steroidogenic acute regulatory protein (Star), steroid 17␣-hydroxylase (Cyp17), P450scc (Cyp11A1), inhibin ␣ (Inha), and aromatase (Cyp19), lack consensus CRE elements. Therefore, transcription factors in addition to CREB must be acting as downstream effectors of cAMP signaling in gonadal cells.
There are several factors that have been shown to be activated by cAMP in gonadal cells. These include stimulatory protein 1 (Sp1), upstream stimulatory factor, estrogen receptor ␣/␤, activating protein 1 (AP-1), CCAAT/enhancer-binding protein (C/EBP) ␤, and Nur77/NR4A1 (4 -8). Although these transcription factors have been shown to contribute to the hormonal regulation of gonadal genes, most are ubiquitously expressed and respond to several stimuli other than cAMP. Moreover, not all cAMP-responsive gonadal promoters lacking CRE elements have binding sites for these factors. In the gonads, the orphan nuclear receptor steroidogenic factor 1 (SF-1/ NR5A1) has been proposed to be a candidate factor downstream of cAMP/PKA. In support of this hypothesis, SF-1binding sites are found in several hormonally regulated genes. Indeed, a role for SF-1 in the cAMP-dependent stimulation of several gonadal genes has been described (9). Although SF-1 can be phosphorylated by PKA in vitro (10), SF-1 is constitutively phosphorylated in vivo by the mitogen-activated protein kinase (MAPK) pathway, and its phosphorylation levels are not affected by cAMP treatment (11). Unfortunately, the absence of adrenals and gonads in SF-1 knockout mice (12)(13)(14) precludes a definitive answer to the role of this factor in these tissues. However, in other steroidogenic tissues normally expressing SF-1, such as the placenta and skin, expression of the cAMPregulated genes Cyp11A1 and Cyp17 was normal in SF-1 Ϫ/Ϫ mice (13,15,16). Therefore, SF-1 might not be the predominant downstream effector of cAMP signaling in gonadal cells.
In addition to SF-1 regulatory elements, many cAMP-regulated gonadal promoters also contain GATA motifs for the binding of members of the GATA family of transcription factors. There are six vertebrate GATA factors (named GATA-1 to GATA-6). Four of these are expressed in the mammalian gonads: GATA-1, GATA-2, GATA-4, and GATA-6 (17)(18)(19)(20)(21)(22)(23)(24). In particular, GATA-4 is strongly expressed from the onset of gonadal development and is later found in multiple cell lineages including testicular Sertoli and Leydig cells and granulosa cells of the ovary (17). Because the function of these cells is under hormonal control, GATA-4 may represent a key effector of hormonal signaling in gonadal cells. In agreement with this, GATA-4 has been shown to regulate several cAMP-dependent gonadal promoters including StAR, inhibin ␣, and aromatase PII (25). In addition, we have recently shown that cAMP stimulation of gonadal cells leads to phosphorylation of GATA-4 (26). Although PKA was a likely candidate, the kinase(s) activated by cAMP in gonadal cells and responsible for directly phosphorylating GATA-4 remained unknown.
We now report that GATA-4 is indeed a key effector of cAMP signaling in gonadal cells. Using MA-10 Leydig cells, we show that GATA-4 is phosphorylated in response to cAMP stimulation predominantly through PKA and not the MAPK signal transduction pathway. cAMP-induced phosphorylation occurs on Ser 261 , an evolutionarily conserved residue of the GATA-4 protein. Additionally, we provide a novel mechanism whereby activation of GATA-4 by PKA regulates multiple target genes via an enhanced recruitment of the CBP transcriptional coactivator.
Production of GST-GATA-4 Fusion Proteins-GST-GATA-4 fusion proteins were obtained by cloning various regions of rat GATA-4 in frame with GST using the pGEX2T vector (Amersham Biosciences). Fusion proteins were produced in the Escherichia coli strain BL21 and purified using a glutathione resin (Amersham Biosciences).
Cell Culture and Transfections-CV-1 and L fibroblast cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% newborn calf serum. The MA-10 mouse Leydig tumor cells were grown in Waymouth's medium containing 15% horse serum (30). CV-1 cells were transfected in 24-well plates using the calcium phosphate precipitation method (25). MA-10 cells were seeded in 12-well plates and transfected using the LipofectAMINE reagent (Invitrogen). In brief, 1.5 g of StAR-luciferase reporter and 500 ng of either an empty expression vector or expression vectors for full-length GATA-4 or GATA-4 S261A were transfected with 5 l of LipofectAMINE reagent/well in serum-  and antibiotic-free media. Renilla (phRL-TK) luciferase (Promega, Madison, WI) was used as an internal control. Complete medium was added 6 h after transfection. The next morning, the cells were treated with either vehicle (H 2 O) or 0.5 mM dibutyryl cAMP (db-cAMP; Sigma-Aldrich) for 6 h and then harvested and analyzed using the Dual-Luciferase reporter assay system from Promega. The data reported represent the averages of at least three experiments, each done in duplicate.
Nuclear Extracts and Western Blots-Nuclear extracts were prepared by the procedure described by Schreiber et al. (31). Recombinant GATA proteins were obtained by transfecting L cells with expression vectors for the different GATA-4 proteins. In Western analyses, 10-g aliquots of nuclear extract were separated by SDS-PAGE and transferred to Hybond polyvinylidene difluoride membrane (Amersham Biosciences). Immunodetection of the native GATA-4 protein was achieved using a GATA-4 antiserum (Santa Cruz Biotechnology, Santa Cruz, CA). Hemagglutinin-tagged GATA-4 proteins were detected using an anti-hemagglutinin antibody (BD Biosciences Canada, Mississauga, Canada), and a VECTASTAIN-ABC-Amp™ Western blot detection kit (Vector Laboratories, Burlingham, CA). Duolux™ (Vector Laboratories) was used as chemiluminescent substrate.
Analysis of in Vivo GATA-4 Phosphorylation-MA-10 cells were seeded in 6-well plates using complete media. The next day, the cells were washed and maintained in phosphate-free Dulbecco's modified Eagle's medium for 30 min. [ 32 P]Orthophosphate (0.5 mCi/ml) was added, and the cells were incubated for an additional 2 h. The cells were then treated with either vehicle or different signaling pathways inhibitors (50 M PD90059 (ERK1/2 pathway), 10 M SB203580 (p38 MAPK pathway), and 10 M H89 (PKA pathway) from Calbiochem, San Diego, CA) for 30 min. The cells were finally stimulated for 2 h by adding either vehicle (H 2 O) or 0.5 mM db-cAMP. The 2-h cAMP treatment period is within the range of times that others have reported for the induction of GATA phosphorylation in various systems using either in vivo phosphorylation or increased GATA binding as functional endpoints (32)(33)(34)(35). After stimulation, the cells were washed and lysed in 300 l of radioimmunoprecipitation buffer (10 mM sodium phosphate, pH 7.2, 1% Igepal (Sigma-Aldrich), 0.1% SDS, 1% sodium deoxycholate, 2 mM EDTA, 150 mM NaCl, 50 mM sodium fluoride, 0.2 mM sodium vanadate, 1 mM phenylmethylsulfonyl fluoride, and 20 g/ml of the protease inhibitors aprotinin, leupeptin, and pepstatin (Sigma-Aldrich Canada)). Radiolabeled GATA-4 was purified by immunoprecipitation with 2 l of GATA-4 antiserum (Santa Cruz) for 2 h at 4°C. A 20-l aliquot of protein G-Sepharose (Amersham Biosciences) was then added, and incubation was continued for an additional 2 h at 4°C. Immunocomplexes were washed five times in radioimmunoprecipitation buffer, resuspended in 1ϫ Laemmli loading buffer, heated at 95°C for 5 min, separated by 10% SDS-PAGE, and transferred to polyvinylidene difluoride membrane. Labeled GATA-4 protein was visualized by autoradiography.
In Vitro Phosphorylation Assay-In vitro kinase assays were performed using 1 g of the different GST-GATA-4 proteins and 10 units of PKA catalytic subunit for 30 min at 30°C (New England Biolabs, Mississauga, Canada) in a total volume of 40 l containing 1 l (10 Ci) of [␥-32 P]ATP (3000 Ci/mmol) in the presence of 50 M of cold ATP. The reactions were terminated by adding 1ϫ Laemmli loading buffer and heating the mixture for 5 min at 95°C. After SDS-PAGE, the labeled proteins were transferred to polyvinylidene difluoride membrane and visualized by autoradiography. The total protein was visualized by Coomassie staining of the gel.
Pull-down Assay-In vitro pull-down assays were performed as previously described (26). A portion of the CBP protein (amino acids 1200 -1900) was 35 S-labeled using the TNT in vitro transcription/translation kit from Promega. The GST-G4 255-440 fusion protein (500 ng) was in vitro phosphorylated as described above but with unlabeled ATP. Unincorporated ATP was eliminated by one wash in 500 l of binding buffer without bovine serum albumin (26) before incubation with 35 Slabeled CBP.

GATA-4 Is
Phosphorylated through the cAMP/PKA Pathway in Vivo-Recent results obtained in our laboratory using an antibody against a phosphoserine residue indicated that the GATA-4 transcription factor was phosphorylated in response to cAMP stimulation of MA-10 cells (26). Although our results suggested that PKA might be responsible for GATA-4 phosphorylation, the exact signaling pathway and kinase involved had not been identified. To further analyze the phosphorylation status of GATA-4 and identify the signaling pathway involved, [ 32 P]orthophosphate metabolic labeling studies were performed in quiescent and db-cAMP-stimulated MA-10 Leydig cells; the experimental protocol is depicted in Fig. 1A. As shown in Fig. 1B, GATA-4 is constitutively phosphorylated in these cells in the absence of any stimulation. GATA-4 phosphorylation levels, however, were significantly increased (more than 3.5-fold) following db-cAMP treatment (Fig. 1B, top panel). This increase occurred without a change in total GATA-4 protein levels (Fig. 1B, lower panel), which is consistent with previous studies showing that hormonal stimulation of MA-10 cells has no effect on GATA-4 mRNA and protein levels (7,26,36,37). Because GATA-4 has recently been shown to be phosphorylated at amino acid Ser 105 through the p38 and ERK1/2 MAPK pathways in cardiac cells (33,35) and because cAMP is also known to activate the MAPK pathways (6), we used different inhibitors to identify which signaling pathway is responsible for GATA-4 phosphorylation in gonadal cells. As shown in Fig. 1B, cAMP-induced GATA-4 phosphorylation was completely abrogated by H89, a PKA inhibitor, but not by SB203580 or PD90059, which are inhibitors of the p38 and ERK1/2 MAPK pathways, respectively. The inhibitory effect of H89 was not simply due to a blockade of basal GATA-4 phosphorylation because H89 treatment in the absence of cAMP only resulted in a slight decrease in phosphorylation levels (Fig. 1B, right panel). Thus, PKA appears to be the predomi- To map the relevant phosphoacceptor site, we next examined the GATA-4 primary amino acid sequence for potential PKA phosphorylation sites. One putative PKA consensus phosphorylation site that matches the high affinity RRXS motif is present at amino acid position 261 (Ser 261 ) between the two zinc fingers of the DNA-binding domain ( Fig. 2A). This region is perfectly conserved between the rat, mouse, human, bovine, rabbit, frog, fish, and chick GATA-4 proteins ( Fig. 2A). To characterize the ability of this site to undergo PKA-mediated phosphorylation, in vitro kinase assays were performed using various GST-GATA-4 fusion proteins (Fig. 2B). PKA efficiently phosphorylated a GATA-4 C-terminal fusion protein (GST-G4 255-440) containing the consensus PKA motif, whereas Nterminal constructs lacking this region (GST-G4 1-200 and GST-G4 1-260) could not be phosphorylated by PKA. Therefore, the PKA target site(s) on GATA-4 are located between amino acids 261 and 440. Consistent with this, mutation of Ser 261 (GST-G4 255-440 S261A) or a further N-terminal deletion that removes Ser 261 (GST-G4 302-440) severely impaired PKA-mediated phosphorylation (Fig. 2B, lower panel). As revealed by total protein staining (Fig. 2B, lower panel), the absence of phosphorylation observed with GST fusion proteins lacking Ser 261 was not due to low amounts of protein. mone-induced gonadal gene expression, then phosphorylation of GATA-4 might enhance its transactivation properties on target gonadal promoters. Indeed, we have recently reported that PKA activates GATA-4 transcriptional activity on the murine StAR promoter (26). As shown in Fig. 3, we now report that in addition to StAR, PKA enhances the GATA-4-dependent activation of several gonadal promoters including 17␣hydroxylase, aromatase, and inhibin ␣. The stimulatory effect of PKA most likely occurs through phosphorylation of GATA-4 because PKA alone (Fig. 3, white bars) had no effect on promoter activity. This stimulation was also strictly dependent on GATA-4 binding to DNA because deletion or mutation of the GATA elements in the respective target promoters prevented transactivation by GATA-4 and enhancement by PKA (data not shown). Therefore, PKA-mediated phosphorylation of GATA-4 activates its transcriptional activity on a variety of GATA-dependent promoters that are active in gonadal cells.
PKA-dependent Activation of GATA-4 Requires Ser 261 -To establish the importance of Ser 261 for basal and PKA-stimulated GATA-4 transcriptional activities, the full-length wildtype (FL) or mutated (FL S261A) GATA-4 proteins were used in cotransfection experiments (Fig. 4A). Consistent with its role in PKA-mediated phosphorylation (Fig. 2), mutation of GATA-4 Ser 261 (FL S261A) markedly reduced its activation by PKA without affecting basal transactivation (Fig. 4A). This decrease was not due to differences in expression, nuclear localization, or stability of the two different GATA-4 constructs because both the wild-type and S261A GATA-4 proteins were expressed at similar levels (Fig. 4B). The residual enhancement of GATA-4 S261A by PKA is likely attributable to low affinity phosphorylation site(s) located in the C-terminal region of the GATA-4 protein (Fig. 2B).
As previously mentioned, the GATA-4 protein contains another phosphorylation site at Ser 105 , which is a target of the p38 and ERK1/2 MAPK in cardiac cells (33,35). Therefore, it remained possible that the PKA enhancement of GATA-4 transcriptional activity could be indirectly mediated through the MAPK pathway. Although a constitutively active form of MAPK/ERK kinase 1 (MEK1 CA) could enhance full-length GATA-4 transcriptional activity (Fig. 4C), this enhancement was lost when the N-terminal domain of GATA-4 containing the MAPK target Ser 105 was removed (⌬N 2 ). In contrast, PKA could still enhance the activity of this same construct. This differential effect was not due to differences in the amount of GATA-4 protein present because both the full-length (FL) and ⌬N 2 GATA-4 constructs were expressed at similar levels (Fig.  4D). Moreover, the enhancing effects of PKA and MEK1 were not due to increases in GATA-4 expression because the total GATA-4 protein levels were not altered by either kinase (Fig.  4D). Therefore, GATA-4 appears to contain at least two independent phosphorylation sites of which Ser 261 is indispensable for the maximal transcriptional enhancement by PKA.
Because the GATA-4 Ser 261 mutation blunted PKA-mediated transcriptional enhancement in heterologous cells, we next examined the effect of this mutation on cAMP-induced StAR promoter activity in MA-10 cells (Fig. 5). Stimulation of MA-10 cells with db-cAMP activated the StAR promoter about 6-fold; cAMP induction was not affected by exogenously expressed GATA-4 (Fig. 5, compare the gray and stippled bars). This induction, however, was blunted by overexpression of the nonphosphorylatable GATA-4 S261A mutant (Fig. 5, solid bar), which competes with endogenous GATA-4 protein for DNA binding. Thus, these results suggest that phosphorylation of GATA-4 Ser 261 is essential to obtain maximal activation of GATA-and cAMP-dependent gonadal promoters in response to hormonal stimulation.

Cooperation between GATA-4 and CBP on Gonadal Promoters
Is PKA-dependent-The CBP transcriptional coactivator is an important regulator of gene expression through its ability to interact and cooperate with several transcription factors such as CREB in response to cAMP/PKA signaling (3). Because p300/CBP has also been reported to physically interact in vitro with the DBD of GATA-4 (38), we tested whether GATA-4 could cooperate with CBP on a series of gonadal promoters and whether this cooperation required protein phosphorylation by PKA (Fig. 6). No significant cooperation between GATA-4 and CBP was observed in the absence of PKA on the StAR, 17␣hydroxylase, aromatase, and inhibin ␣ promoters (Fig. 6A, open  bars). In the presence of PKA, however, strong synergy was observed on all natural promoters tested (Fig. 6A, solid bars). In addition, the PKA-dependent cooperation between GATA-4/ CBP was also observed on a synthetic reporter (Fig. 6B, left  panel) consisting of a single copy of the consensus GATA element from the proximal StAR promoter fused to the unresponsive minimal Mü llerian inhibiting substance promoter (Fig. 6B,  right panel). Once again, this effect was independent of changes in total GATA-4 protein levels (Fig. 6C). Thus, GATA-4 binding to DNA is necessary and sufficient for the PKA-dependent GATA-4/CBP transcriptional cooperation.
Next, the domain of GATA-4 required for the PKA-dependent synergy with CBP was mapped (Fig. 7). The GATA-4 protein contains two independent activation domains (ADs) that flank its DBD. Deletion of the GATA-4 C-terminal AD (⌬C 1 ) did not impair synergism. However, deletion of both GATA-4 ADs (⌬N 1 C 1 ), leaving only the DBD, severely blunted cooperation with CBP. These results suggest that maximal cooperation with CBP requires at least one functional AD and an intact GATA-4 DBD, the latter being involved in DNA binding and direct physical interaction with p300/CBP (38). In agreement with these requirements, a heterologous AD from the viral protein VP16 fused to the GATA-4 DBD (VP16-⌬N 1 C 1 ) could restore PKA-dependent cooperation with CBP (Fig. 7). Because PKA-mediated phosphorylation of Ser 261 is important to activate the GATA-4 transcription factor, we next tested whether phosphorylation of GATA-4 Ser 261 is also required for transcriptional synergism with CBP. As shown in Fig. 7, mutation of GATA-4 Ser 261 into alanine markedly impaired transcriptional synergism with CBP (50-fold activation compared with 125-fold for wild-type GATA-4). Thus, phosphorylation of GATA-4 Ser 261 is likely an important mechanism for the re- cruitment of the CBP coactivator to cAMP-dependent promoters containing GATA regulatory elements.
PKA-mediated Phosphorylation of GATA-4 Ser 261 Enhances the Physical Interaction with CBP-The fact that transcriptional cooperation between GATA-4 and CBP requires PKAmediated phosphorylation of GATA-4 Ser 261 suggests that protein phosphorylation by PKA might also enhance the interaction between the two proteins. To test this possibility, in vitro interaction experiments were performed using GST-GATA-4 fusion proteins and a fragment of the CBP protein (amino acids 1200 -1900) previously shown to interact with GATA-4 (38). As previously reported by Dai and Markham (38), GATA-4 and CBP do interact in the absence of PKA phosphorylation (Fig. 8, lane 2). Consistent with the transcriptional cooperation data (Fig. 7), PKA-mediated phosphorylation of the wild-type GATA-4 fusion protein (GST-G4 255-440 WT) enhanced the interaction with CBP (Fig. 8, lane 3). In contrast, no increase was observed with the GATA-4 S261A mutant that cannot be phosphorylated by PKA (Fig. 8, lane 5). DISCUSSION In gonadal cells, it is well established that most gonadadotropin-regulated events, such as increased steroidogenic gene expression in Leydig cells, are mediated through the cAMP/ PKA signaling pathway (39). In several tissues, including the gonads, the classic downstream effector of cAMP/PKA signaling is the phospho-CREB/CBP complex, which specifically binds to CRE regulatory elements leading to stimulation of target promoters (1,2). Several cAMP-regulated gonadal genes, however, lack consensus CRE elements, which suggests that transcription factors other than CREB are responsible for conveying the cAMP responsiveness of those genes. Based on its expression in multiple cell types of the testis and ovary and on the fact that it can activate several gonadal promoters (40), GATA-4, much like CREB, is likely a common regulatory factor that coordinates the cAMP responsiveness of multiple gonadal genes. Indeed, our present data strongly support a role for GATA-4 as a novel downstream effector of hormonal signaling in gonadal cells.
GATA-4: a Novel Mediator of cAMP Signaling in Gonadal Cells-[ 32 P]Orthophosphate metabolic labeling of MA-10 cells confirmed that endogenous GATA-4 protein is directly phosphorylated by PKA in response to cAMP stimulation (Fig. 1). The consensus PKA phosphorylation site that we have identified (GATA-4 Ser 261 ) is perfectly conserved across several species. This strong conservation suggests a critical role for this site that has been preserved during evolution. In addition to GATA-4, all other vertebrate GATA proteins as well as GATA factors from Drosophila (dGATA-A) and sea urchin (SpGATAc) contain PKA phosphorylation sites. This suggests that PKAmediated phosphorylation of GATA factors might have a broader role in the control of cAMP/PKA-dependent processes in different tissues and organisms. For example, adrenal gland function is tightly controlled by adrenocorticotropic hormone through the cAMP/PKA pathway, which regulates expression of several steroidogenic enzyme encoding genes, such as Star and Cyp17 that lack CRE regulatory elements. Although GATA-4 is not present in the adult adrenal, GATA-6, whose transcriptional activity is also stimulated by PKA on the StAR promoter (26), is abundantly expressed (41,42).
Several GATA factors have already been implicated as downstream effectors of signaling pathways in other systems. For example, phosphorylation of GATA-3 has been shown to be involved in the regulation of interleukin-5 gene expression in T helper 2 cells (32). In hematopoietic cells, interleukin-3-dependent phosphorylation of GATA-2 stimulates proliferation of hematopoietic progenitors (43), and phosphorylation of GATA-1 regulates erythroid maturation of murine erythroleukemia cells (44). GATA-4 is also phosphorylated in response to hypertrophic stimuli in the heart (33)(34)(35)(45)(46)(47)(48). Although some of those cases involved cAMP production, GATA phosphorylation was not directly mediated by PKA but rather involved activation of the MAP kinases p38 and ERK1/2. This can be achieved by cross-talk between the cAMP and MAPK pathways, which involves a new class of cAMP-binding proteins termed Epac (exchange protein activated by cAMP) that bind to and activate Ras-related small GTPases, thus providing entry into the ERK1/2 or p38 MAPK pathways (49,50). In the heart, MAPK-mediated phosphorylation of GATA-4 occurs on Ser 105 , which is a consensus phosphorylation site for MAPK (33,35). Although GATA-4 could be potentially phosphorylated by MAPK on Ser 105 by cAMP stimulation of MA-10 Leydig cells, cAMP-induced phosphorylation was not blocked by MAPK inhibitors (Fig. 1). This suggests that cross-talk between cAMP and the MAPK pathway is not responsible for cAMP-induced phosphorylation of GATA-4 in these cells. Rather, GATA-4 is predominantly phosphorylated on Ser 261 , which lies within a consensus motif (RRXS) for PKA phosphorylation. To our knowledge, this is the first demonstration of cell signaling that involves direct phosphorylation of a GATA factor by PKA. The fact that GATA-4 is preferentially phosphorylated by PKA on Ser 261 in MA-10 cells does not formally exclude a role for Ser 105 phosphorylation. In fact, gonadal cells do respond to certain stimuli that activate the cAMP pathway but without the involvement of PKA (6). In these cases, activation of MAPK appears to play a predominant role (6). Therefore, divergent signaling pathways and kinases might converge on GATA-4 to regulate gonadal gene expression and function. Interestingly, such a mechanism has recently been described for the yeast GATA factor Gln3, which acts as a final effector of two key nutrient and sensing pathways via differential phosphorylation by two different kinases (51).
Mechanism of Phospho-GATA-4 Action in Gonadal Cells-In addition to stimulating GATA-4 transcriptional activity on several target promoters, phosphorylation of GATA-4 also modulates its ability to cooperate with certain transcription factors. For example, GATA-4 can synergize with SF-1 in the absence of PKA (27). GATA-4/SF-1 synergism was nonetheless enhanced by PKA (data not shown). PKA action is also essential for transcriptional synergism between GATA-4 and other transcription factors such as C/EBP␤ (26) and CBP (this study). Thus, modulation of the intrinsic transcriptional properties of GATA-4 and its ability to cooperate with other transcription In vitro pull-down assays were performed using 500 ng of bacterially produced phosphorylated (ϩPKA) or unphosphorylated (ϪPKA) GST-GATA-4 fusion proteins and an in vitro translated 35 S-labeled fragment of the CBP protein (amino acids 1200 -1900). After extensive washes, the bound proteins were separated on a 10% SDS-PAGE gel and visualized by autoradiography. Input corresponds to 10% of the total 35 S-CBP used in each assay.
factors are two mechanisms (which are not mutually exclusive) whereby gonadal gene expression might be regulated by the PKA-mediated phosphorylation of GATA-4. Another potential mechanism involves protein-protein interactions with coactivators. Although GATA-4 can interact with the coactivator CBP in the absence of any posttranslational modification, PKAmediated phosphorylation of GATA-4 Ser 261 significantly enhanced this interaction (Fig. 8). Moreover, on target gonadal promoters, this enhanced phospho-GATA-4 Ser 261 /CBP interaction was critical for maximal GATA-4/CBP transcriptional synergism. Interestingly, these results are reminiscent of the interaction between CBP and phospho-CREB Ser 133 . In the classical cAMP/PKA pathway, cAMP elicits a 3-4-fold increase in CREB phosphorylation levels (52). This is very similar to our observed increase in GATA-4 phosphorylation levels after cAMP stimulation of MA-10 Leydig cells. Thus, it appears that phosphorylation of GATA-4, much like CREB, allows for a stronger interaction with CBP, ultimately leading to increased gene transcription. Importantly, this novel mechanism provides new insights into our understanding of the hormone-dependent regulation of genes that lack classical cAMP regulatory elements.
Because the GATA-4 protein contains two distinct phosphoacceptor sites (Ser 105 and Ser 261 ), differential phosphorylation in response to various signals might constitute another important regulatory mechanism. Like GATA-4, CREB can also be phosphorylated on different serine residues by different kinases depending on the stimuli. For example, phosphorylation of CREB Ser 142 by the casein kinase II and Ca 2ϩ -calmodulin kinase II prevents CBP recruitment and target gene activation (3). This allows for signal discrimination through CREB leading to differential output. Although no data are currently available concerning coactivator recruitment upon phosphorylation of GATA-4 Ser 105 in response to a stress signaling pathway, it is still tempting to speculate that a similar mechanism could also exist for GATA-4. In this way, phosphorylation of a common transcriptional regulator by different kinases in response to stress or cAMP signals would translate into differential gene expression via modulation of coactivator recruitment. Through this mechanism, GATA-4 could represent the cornerstone of a large transcriptional complex (involving other transcription factors such as SF-1 and C/EBP␤, and the coactivator CBP/p300) that is required for the activation of different sets of genes in response to hormonal and stress signaling in the gonads and other endocrine tissues.