Endothelin-1-specific Activation of B-type Natriuretic Peptide Gene via p38 Mitogen-activated Protein Kinase and Nuclear ETS Factors*

Terminally differentiated cardiac myocytes adapt to mechanical and neurohumoral stress via morphological changes of individual cells accompanied by reactivation of fetal pattern of gene expression. Endothelin-1, a powerful paracrine mediator of myocyte growth, induces similar changes in cultured cardiac myocytes as those seen in hypertrophied heart in vivo. By using rat B-type natriuretic peptide promoter, we identified a novel ETS binding sequence, on which nuclear protein binding is activated in endothelin-1-treated cultured cardiac myocytes. This sequence binds ETS-like gene-1 transcription factor and mediates endothelin-1-specific activation of transcription, but not responses to increased calcium signaling via l-type calcium channels, angiotensin II treatment, or mechanical stretch of myocytes. Interestingly, endothelin-1 activated signaling converges via p38 mitogen-activated protein kinase-dependent mechanism on ETS binding site, whereas this element inhibits extracellular signal-regulated kinase activated transcription. In conclusion, given the fundamental role of the interaction of mitogen-activated protein kinases and ETS factors in regulation of eukaryotic cell differentiation, growth, and oncogenesis, these results provide the unique evidence of a endothelin-1- and mitogen-activated protein kinase-regulated ETS factor pathway for cardiac myocytes.

Terminally differentiated cardiac myocytes, due to their inability to divide, adapt to increased mechanical load and the activation of the neurohumoral system by hypertrophy. Initiation of hypertrophic growth is accompanied by a rapid and transient expression of immediate early genes (e.g. c-jun, c-fos, and Egr-1) followed by activation of a pattern of cardiac genes, including atrial and B-type natriuretic peptide (ANP 1

and BNP)
and ␤-myosin heavy chain (␤-MHC), whereas relative expression of a subset of genes is simultaneously decreased (i.e. ␣-myosin heavy chain, ␣-MHC) (1,2). Natriuretic peptides, released into circulation by cardiac myocytes, defend the cardiovascular system against increased hemodynamic load and decrease blood pressure by increasing salt and water excretion and by promoting vasodilation (3). However, when both mechanical and neurohumoral stress of myocytes are sustained, adaptive mechanisms fail leading to maladaptation and, eventually, cardiac dysfunction (i.e. congestive heart failure) (4). Emerging evidence suggests that a number of humoral factors, such as angiotensin II (Ang II) and endothelin-1 (ET-1), participate in this adaptive process and modify the growth of cardiac myocytes (5). Production of ET-1 is increased in humans with cardiac hypertrophy and congestive heart failure, and ET-1 plasma concentrations correlate with the clinical severity of the cardiac failure (6).
There are two distinct G-protein-coupled receptor (GPCR) subtypes for ET-1, ET A and ET B receptors, both of which are widely expressed in a variety of tissues, including myocardium (7). Activation of GPCRs catalyzes exchange of GTP for GDP within the heterotrimeric G-protein complex releasing activated G␣-GTP and G␤␥ subunits, each of which has regulatory functions (8). Binding of ET-1 to its receptors induces G q/11protein-dependent stimulation of phospholipase C, leading to formation of inositol 1,4,5-triphosphate and diacylglycerol (9), which activate the release of calcium from intracellular stores and protein kinase C (10), respectively. Altered calcium handling of myocyte, a feature of human heart failure (11), may result in activation of calcium/calmodulin-dependent kinase (CaMK) and phosphatase (i.e. calcineurin) pathways (12), both of which are implicated in endothelin-1-induced signaling cascades (13). In addition, signaling pathways downstream of ET receptors involve small G-proteins (Ras, RhoA, and Rac) accompanied by activation of mitogen-activated protein kinases (MAPKs) (14).
The mechanisms connecting ET-1-induced cytosolic signaling to nuclear targets are poorly understood in cardiac myocytes. ET-1 has been reported to activate nuclear factor-B (NFB) transcription factor (15) and, via MAPK-mediated phosphorylation, GATA-4 transcription factor (16). Physical and functional interactions of GATA-4 with its cofactors, nuclear factor of activated T lymphocytes (NFAT) and serum response factor (SRF), are induced in ET-1-treated cardiac myocytes (17,18). However, inhibitory strategies targeted against either GATA-4 protein synthesis or DNA binding affinity have produced different results in ET-1-induced hypertrophic growth of cardiac myocytes, suggesting an increasing complexity of ET-1-dependent pathways (19,20). To elucidate the nuclear mechanisms of ET-1-activated cell signaling, we studied rat BNP (rBNP) as a cardiac myocyte-specific target gene for ET-1 and identified a novel signaling mechanism initiating from cell membrane ET A receptor and leading to p38 MAPK-and ETS factor-dependent induction of the rBNP gene.
Cell Culture and Transfection-Neonatal rat ventricular myocytes were prepared from 2-to 4-day-old Sprague-Dawley rats and transfected with liposome-mediated transfection (20). The Animal Use and Care Committee of the University of Oulu approved the experimental design. Briefly, after preparation, cells were plated at the density of 2 ϫ 10 5 /cm 2 on cell culture plates (Falcon) or on collagen I-coated elastomere plates with flexible bottoms (Bioflex, Flexcell Int.) for stretch experiments and cultured overnight with Dulbecco's modification of Eagle's medium/Ham's F-12 medium (1:1) containing 10% of fetal bovine serum. Transfection of cells (if designated) was performed on the second day in culture. Cells were exposed to 3 l of FuGENE 6 and 1.5 g of DNA (1 g of Luc, 0.5 g of ␤-gal plasmids) per milliliter of complete serum-free medium for 6 h and cultured thereafter in complete serum-free medium. Cells on Bioflex plates were exposed to cyclic mechanical stretch. Frequency of cyclic stretch was 0.5 Hz with pulsation of 10 -25% elongation of cells for 24 h. Cells were stretched by applying a cyclic vacuum suction under Bioflex plates by computercontrolled equipment (FX-3000, Flexcell Int.). In co-transfection experiments, cells were grown on 24-well plates and transfected with 0.45 g of intact Ϫ534BNP-luc or with Ϫ534BNP-luc containing mutation of EBS and 0.05 g of expression plasmid (see "Plasmids and Oligonucleotides") of dnMEK1, dnMKK6, dnJNK1, MEK1, p38␣, JNK1, or pMT2 plasmids with 0.25 g of RSV-␤gal and 1.5 l of FuGENE 6 per milliliter. Empty control plasmid pMT2 was used to equalize the amount of transfected DNA between experimental groups. All experiments were started subsequently on the third day, and cells were harvested according to the experiment during the third or fourth day in culture. Reporter gene activities were measured by using luciferase and ␤-galactosidase (to correct transfection efficiency) assays (Promega) with luminometer. Luciferase activity levels were 9.2 Ϯ 1.0 (average Ϯ S.E.)-fold higher than those of ␤-galactosidase activity. Both luciferase and ␤-galactosidase levels were at the linear range of the assay measurement scale.
Northern Blot Analysis and Radioimmunoassay for BNP-RNA was isolated from cardiac myocytes by the guanidine thiocyanate-CsCl method (21). Northern blots were hybridized with specific cDNA probes for rat BNP, rat ANP, and rat ribosomal 18 S (20) labeled with [ 32 P]dCTP (Amersham Biosciences) with a T7 Quick Prime kit (Amersham Biosciences). To measure the secretion of BNP, samples of incubation medium were subjected to radioimmunoassay of BNP (20).
EMSA-Nuclear extracts from cardiac myocytes (22) were prepared and protein concentration from each sample was colorimetrically determined (Bio-Rad Laboratories). Double-stranded oligonucleotide probes were sticky end-labeled with [ 32 P]dCTP by Klenow enzyme (22). For each reaction mixture (20), 6 g of nuclear protein and 2 g of poly(dI-dC) were used in a buffer containing 10 mM HEPES (pH 7.9), 1 mM MgCl 2 , 50 mM KCl, 1 mM dithiothreitol, 1 mM EDTA, 10% glycerol, 0.025% Nonidet P-40, 0.25 mM phenylmethylsulfonyl fluoride, and 1 g/ml each leupeptin, pepstatin, and aprotinin. Reaction mixtures were incubated with a labeled probe for 20 min followed by gel-electrophoresis (250 V, for 1.5-1.75 h normally and for supershift 2.5 h) on 5% polyacrylamide gel. Subsequently, gels were dried and exposed in a PhosphorImager screen and analyzed with ImageQuaNT (Amersham Biosciences). To confirm DNA sequence specificity of the protein-DNA complex formation, competition experiments with 10, 50, and 100 molar excesses of unlabeled oligonucleotides with intact or mutated binding sites were performed. In addition, DNA binding activity assays of Octamer-1 (Oct-1) transcription factor were run in parallel to control the specificity of the altered DNA binding activity (16). For competition and supershift experiments appropriate competitor oligonucleotides or antibodies were added to reaction mixture 20 min before labeled probe.
Immunoblot-Western blot analysis was done using standard reagents and protocol (16). Briefly, 20 g of nuclear proteins was used for SDS-PAGE, followed by Western blotting utilizing antibodies against phospho-Elk-1 (Ser-383) and Elk-1 (binds both phosphorylated and dephosphorylated Elk-1). Finally, blots were developed with goat anti-Rabbit IgG-HRP and enhanced luminescence reagents followed by exposure of film for 1-3 min (Hyperfilm ECL, Amersham Biosciences).
Statistical Analysis-Results are expressed as mean Ϯ S.E. To determine the statistical significance between two groups and for analysis of multiple groups, the Student's t test and one-way analysis of variance followed by the least significant difference post hoc test were used, respectively. Differences at the 95% level were considered statistically significant.

ET-1 Activates BNP Gene Transcription via a Novel ETS
Binding Sequence-Cultured neonatal ventricular myocytes serve as a useful experimental model for the study of myocyte growth and the intracellular mechanisms involved. Primary neonatal ventricular myocytes (referred to hereafter as cardiac myocytes) used in this study respond to a variety of stimuli by the activation of hypertrophic growth program virtually identical to that of developing neonatal and the pathologic adult myocardium (5). ET-1 activated cardiac genes associated with hypertrophic growth program, such as ANP and BNP (Fig. 1A), ofwhichBNPsynthesiswasaugmentedthroughanET A receptordependent mechanism (Fig. 1B) (23). In addition, cardiac myocytes subjected to ET-1 develop hypertrophy, including activation of protein synthesis (16,20) and increased cell size accompanied by reorganization of sarcomeres (Fig. 1C).
In the context of the rBNP promoter, previous studies have indicated that the proximal 5Ј-flanking region of 535 bp is sufficient to confer the induction by GPCR agonists (20,24).
The proximal Ϫ534 bp of BNP promoter contain multiple potential cis-acting elements (25) such as GATA elements. Proximal GATA elements participate in the activation of rBNP gene transcription by ␣-adrenergic agonist phenylephrine (24) but not by ET-1 (20). Therefore, we utilized a computer-based search for novel cis-acting elements of the rat BNP gene and identified an ETS binding sequence (EBS) and AP-1 binding sequence at Ϫ498 and Ϫ373 bp of the rBNP promoter, respectively. To characterize the role of these elements, we introduced site-directed mutations to EBS and AP-1 binding sequences and to Ϫ90/Ϫ81-bp tandem GATA elements of the rBNP promoter ( Fig. 2A). Endothelin-1 activated proximal Ϫ534-bp rBNP promoter dose-dependently via ET A receptor, and the induction was inhibited by mutation of EBS (Fig. 2, B and C). The combination of mutations of AP-1 and/or GATA elements with the mutation of EBS did not further decrease the inhibition seen with rBNP promoter containing mutated EBS (data not shown). Importantly, the induction of rBNP transcription in myocytes by mechanical stretch or Ang II was unaffected by the mutation of EBS (Fig. 2D). To investigate this ET-1 responsive element further, we studied the complex formation of cardiac nuclear proteins with EBS of rBNP promoter.
ET-1 Activates Specific Complex Formation of Cardiac Nuclear Proteins with EBS-Previously, it has been demonstrated that exposure of cardiac myocytes to hypertrophic agonists selectively activate transcription factors followed by specific DNA binding affinity and eventually activation of transcription (16,26). To analyze whether there is specific binding activity on EBS of rBNP promoter by cardiac nuclear proteins, we utilized EMSA, and, when designated, performed the protein-DNA binding reactions in the presence a 100-molar excess of unlabeled competitor oligonucleotides (Fig. 3A). We found that three specific shifts (complexes C1-3) were constantly formed, of which C1 and C2 migrated closely to each other separately from a faster migrating complex C3. Competitor oligonucleotides with mutated nucleotides from Ϫ497 bp (mut 2) to Ϫ491 bp (mut 8) failed to compete against intact EBS and were found essential for formation of complexes C1-3 (i.e. 5Ј-cCCG-GAAGtg-3Ј), suggesting similarity of binding proteins in these complexes. Interestingly, a labeled high affinity ETS binding sequence (SC1) (27) formed complexes similar to C1-3 (Fig.  3A). The core sequence for ETS factor binding of SC1 (Table I) is similar to EBS of rBNP promoter. When SC1 oligonucleotide was used as a unlabeled competitor against nuclear protein binding on EBS of rBNP promoter, SC1 dose-dependently inhibited protein binding on EBS, and the competition was abolished by single nucleotide mutation of SC1 (i.e. 5Ј-CCGaAAGTG-3Ј) (Fig. 3B). An NFB binding oligonucleotide (Table I), as an another potential GGA sequence binding element, was not able to compete against binding on EBS of the rBNP promoter (Fig. 3B). Moreover, with EBS of rBNP, nuclear proteins extracted from adult rat heart (left ventricle) formed identical complexes C1-3 in EMSA (data not shown), suggesting that ETS factors binding on EBS of rBNP are present in adult heart.
In view of specific complex formation of cardiac nuclear proteins with EBS of BNP promoter, we analyzed whether ET-1 treatment of cardiac myocytes had any effect on the nuclear protein binding activity on EBS (Fig. 4, A-D). Initially, a rapid and transient increase in formation of complex C2 was detected at 5 and 15 min. Later, a sustained increase of EBS binding activity was observed at 3 h for complexes C1 and C3, which remained elevated at 24 h. In agreement with our previous studies (16), all the nuclear extracts were tested in parallel for Oct-1 binding activity, which was not altered by ET-1 suggesting a specific activation of EBS binding (data not shown). Taken together, ET-1 treatment of cardiac myocytes initiated a biphasic activation of protein binding on the EBS of rBNP, and these specific complexes closely resembled complex formation of nuclear proteins with a high affinity ETS binding sequence.
Elk-1 Binds on EBS of rBNP Promoter-Functional ETS factor binding regulatory cis sequences have been characterized extensively in both viral and eukaryotic genes, but there are exceptionally few reports of the cardiac genes regulated by ETS factors (28). Gupta et al. (29) have reported a repressor role for 3Ј-flanking EBS resulting in myocyte-restricted expression of ␣-MHC gene, whereas other studies available have described an interaction of serum response element (SRE) with an EBS in the activation of c-fos gene in cardiac myocytes (30,31). EBS of rBNP differs significantly from those reported previously: (i) Ϫ2.2 kb of the 5Ј-flanking region of rat BNP gene contains no potential SRE for TCF (ternary complex factor)-SRF synergy (18) and (ii) EBS functions as an activator element of rBNP transcription exclusively by ET-1. To identify the transcription factors binding on EBS, we used specific antibodies added to binding reaction prior radioactively labeled EBS and subsequent EMSA. We utilized antibodies for NFB (p50 and p65), AP-1 complex (c-fos, c-jun, jun-B, and jun-D), Nkx-2.5, SRF, ETS-1/-2 (DNA binding domain), Fli-1 (a member of ETS transcription factor family), and activated Ser-383-phosphorylated Elk-1 (Phospho-Elk-1). We found that ET-1-activated C2 complex was supershifted by antibody raised against Phospho-Elk-1 (Fig. 5A). In contrast, none of these antibodies had any effect on GATA-4 binding activity on the Ϫ90/Ϫ81-bp rBNP GATA binding site in EMSA run in parallel (data not shown).
Given the cumulative evidence demonstrating that Elk-1 transcription factor is regulated via site-specific phosphorylation (32), we further investigated activation of Elk-1 by using Western blot analysis. The site-specific phosphorylation (Ser-383) of Elk-1 transiently increased at 5 and 15 min in cardiac myocytes treated with ET-1 (Fig. 5B). Therefore, Elk-1 phosphorylation is temporally related to increased formation of complex C2 in EMSA (Fig. 4B). In addition, antibodies raised against NFATc3, NFATc4, GATA-4, STAT-1 and -3 and com-

FIG. 3. Characterization of ؊498 BNP ETS binding sequence (EBS).
A, competition analysis of ETS binding sequence of rat BNP promoter and comparison with SC1 probe. B, competition analysis for nuclear protein binding on ETS binding sequence of rBNP promoter with NFB binding site and SC1 oligonucleotides. EMSA using double-stranded oligonucleotide corresponding to ETS binding sequence at Ϫ498 bp of rBNP promoter ((Ϫ498) BNP) and oligonucleotide corresponding to a high affinity ETS binding site (SC1) as a radioactively labeled probes was carried out with 100-molar excess of unlabeled competitor oligonucleotides (for sequence see Table I). For the binding reaction, 6 g of a mixed nuclear protein extract from untreated and ET-1 (100 nM, 15 min and 24 h)-stimulated cardiac myocytes was used. Binding reactions with EMSA were repeated separately four times from different nuclear extracts, and the three specific complexes invariably formed are indicated as C1, C2, and C3.
petitor oligonucleotides of AP-1, GATA-4 binding sequences (Ϫ373 bp and Ϫ90 bp/Ϫ81 bp of rBNP) and Nkx-2.5 binding element (at Ϫ240 bp of rat ANP gene) or shift-inducible element failed to have any effect on complex formation of cardiac nuclear proteins with EBS (data not shown). In conclusion, ET-1 induces phosphorylation and binding of Elk-1 on EBS of rBNP (complex C2) and likely activates additional ETS related proteins whose identity we were unable to identify by using specific antibodies against ETS related or other transcription factors.
ET-1 Activates rBNP Transcription via EBS by p38 MAPKdependent Mechanism-Hypertrophic stimuli, including GPCR agonists (e.g. ET-1, phenylephrine, and Ang II) and mechanical stress activate MAPKs in cardiac myocytes (14,16,33). Elk-1 transcription factor is a known downstream target for MAPKs (34,35), and thus we investigated their potential role in activation rBNP transcription via EBS by using expression plasmids for each MAPK pathway. A minimal amount of expression plasmids (0.05 g) was selected to prevent quenching and to confirm that the effects of overexpression were specific for each MAPK cascade. Dominant negative (dn) kinases of MEK1 (upstream of ERK), MKK6 (upstream of p38), and c-Jun N-terminal kinase-1 (JNK) were utilized to inhibit specifically each pathway. Interestingly, only the inhibition of p38 MAPK-dependent mechanism attenuated the activation of rBNP transcription by ET-1 (Fig. 6A). Vice versa, the activation of p38 MAPK resulted in EBS-dependent activation of rBNP promoter activity (Fig. 6B). Contrastingly, activation of ERK signaling by forced expression of MEK1 activated rBNP transcription only when signaling via EBS was blocked by mutation (Fig. 6B). Furthermore, pharmacological blockade of p38 MAPK pathway by p38 MAPK inhibitor (SB203580, 20 M) prevented ET-1 (100 nM, 15 min)-induced Ser-383 phosphorylation of Elk-1 (data not shown).
In addition to MAPK cascades, ET-1 is a known activator of Ca 2ϩ -dependent signaling such as CaMK and calcineurin in cardiac myocytes (13). Therefore, we further examined whether activation of Ca 2ϩ -dependent pathways induced rBNP transcription similarly via the p38 MAPK-and EBS-dependent mechanism as ET-1. Cells were exposed to dihydropyridine agonist Bay K8644 to activate Ca 2ϩ influx via L-type Ca 2ϩ channels (36). Bay K8644 induced rBNP transcription independently of EBS, and the induction was equally inhibited by dominant negative expression plasmids of each MAPK cascade (Fig. 6C). Similar results were obtained with physiological activation of L-type Ca 2ϩ channels using ␤ 1 -adrenergic receptor stimulation by isoproterenol (data not shown). Conclusively, these results suggest that ET-1 activates rBNP gene transcription via EBS in a p38 MAPK-dependent mechanism, whereas induction of BNP gene by robust activation of Ca 2ϩ -dependent signaling likely requires activity of all three MAPK cascades in parallel. In contrast, intact EBS of rBNP promoter inhibits the activation of transcription by the ERK pathway alone. DISCUSSION The founding member of the ETS family in mammalians, ETS-1, was originally discovered due to homology to v-ets, an oncogene of avian E26 ("E-Twenty-Six") retrovirus genome (37). ETS factors share a winged helix-turn-helix domain for DNA binding, and they bind to 5Ј-GGA(T/A)-3Ј ETS binding sequence (EBS) (27,38). ETS factors have been implicated in transforming and tumor-associated products as well as cell cycle and apoptosis regulator genes, and they regulate hematopoiesis and the development of central nervous system, bone and cartilage, mammary glands, and vasculature (28, 39).

FIG. 5. Analysis of EBS binding proteins.
A, Elk-1 binds on EBS of rBNP and is a major protein in C2 complex. EMSA using doublestranded oligonucleotide corresponding to EBS of rBNP ((Ϫ498) BNP) as a radioactively labeled probe was used with 6 g of nuclear proteins from cardiac myocytes treated with ET-1 (100 nM, 15 min). For supershift reaction, 2.5 l of specific antibodies raised against NFB (p50 and p65), the members of AP-1 complex, Nkx-2.5, SRF, ETS-1/-2 (DNA binding domain), Fli-1, or Ser-383-phosphorylated Elk-1 (P-Elk-1) were added to the binding reaction. Supershifted C2 complex is indicated (SS). B, ET-1 increases phosphorylation of Elk-1. Nuclear proteins from control and ET-1-treated (100 nM) cardiac myocytes were subjected to SDS-PAGE. An antibody raised against Ser-383-phosphorylated Elk-1 was used to reveal the amount of activated Elk-1 protein (p-Elk-1) and total amount of Elk-1 (Elk-1) was quantified by immunoblotting filters with another antibody raised against Elk-1 that binds both phosphorylated and dephosphorylated Elk-1 proteins. A representative Western blot of four to five separate experiments for each time point is shown. Therefore, the identification of target genes for ETS signaling has emerged to a major area of research interest.
Although there are multiple ETS factors in the heart (40), their functional role has not been revealed. Activation of c-fos promoter via Elk-1 together with SRF composes a regulatory mechanism of c-fos gene present in most eukaryotic cell types, including cardiac myocytes (30,31). In addition, ETS-domain transcription factors act through EBS containing no adjacent SREs, likely in complex with alternative cofactors, but only a single report is available of such a regulatory EBS acting on a gene expressed in myocytes (␣-MHC) (29). Yet, multiple genes activated during hypertrophic growth of cardiac myocytes contain potential EBS in their 5Ј-flanking regulatory region, such as genes for ␤-MHC and ANP (29). In the present study, we identified a novel EBS at Ϫ498 bp of rBNP promoter, constituting the rBNP gene as a first cardiac target for ETS signaling that is activated during hypertrophic growth.
Importantly, treatment of myocytes with ET-1 activated specific complex formation between nuclear proteins and EBS of BNP. Similar activation of binding affinity on rBNP gene has been reported of GATA-4 transcription factor (16). Moreover, it is evident that activation of rBNP gene by forced expression of p38 MAPK alone requires both intact EBS and GATA elements (16) at the proximal promoter region. However, ERK is also capable of phosphorylating GATA-4 in a manner similar to p38 MAPK (16), but it appears that activation of rBNP promoter by the ERK pathway is inhibited by signaling via EBS (schematic presentation in Fig. 7). This model of nuclear regulation may target parallel MAPK pathways in a stimulus-dependent manner. Our hypothesis is supported by previous studies indicating that activation of rat BNP promoter during myocyte stretch appears to require intact GATA binding sequence at Ϫ90/Ϫ81 bp (25) but not EBS at Ϫ498 bp of rBNP. Taken together with the fact that ETS factors have been shown to regulate tissuerestricted gene expression by utilizing both gene activation and repression mechanisms (41,42), this study provides unique evidence for the role of ETS transcription factors as a gatekeeper of MAPK-mediated cardiac gene activation.
Elk-1, identified in protein complex binding on EBS of rBNP, is directly regulated via site-specific phosphorylation by MAPK pathways (34,35) and conversely inactivated by dephosphorylation by phosphatases, including calcineurin (43). Originally, studies with T cells demonstrated that calcineurin dephosphorylates NFAT leading to nuclear translocation (44), but Molkentin et al. (45) described a role for the calcineurin-and NFAT-dependent signaling pathway also in myocyte hypertrophy. In the present study, robust activation of calcium signaling via L-type calcium channels resulted in the activation of BNP promoter independently of EBS, and all three MAPK cascades were equally required. Therefore, increased calcium signaling, mechanical stretch, and Ang II activate rBNP promoter via mechanisms that appear to uncouple the requirement of EBS of BNP gene. Calcineurin has been shown to direct signaling through JNK and ERK cascades (46,47) and to inactivate signaling via Elk-1 through dephosphorylation (32). In addition, accumulation of intracellular calcium may lead to autoinhibition of DNA binding activity of ETS factors via sitespecific CaMK-mediated phosphorylation (48). In conclusion, recent studies (49, 50) combined with our results suggest an intimate relationship between MAPKs and Ca 2ϩ -dependent pathways in the development of cardiac hypertrophy.
It is noteworthy that pathophysiological effects of ET-1 extend beyond myocardium to the regulation of proliferation (51) and contraction (52) of vascular smooth muscle cells, and the most notable effects are seen in coronary arteries (53). Therefore, it is intriguing to speculate that the ET-1-activated nuclear pathway described in the present study may not be limited to the regulation of cardiac gene expression but might extend to involve additional ETS factors expressed in vasculature, such as ETS-1, the expression of which is activated by ET-1 in cultured vascular smooth muscle cells (54) and in human endothelial cells during tumor angiogenesis (55). Interestingly, ET-1 increases vascular endothelial growth factor production in ovarian cancer cells via ET A receptor (56), whereas ETS factors activate genes regulating angiogenesis, such as vascular endothelial growth factor receptor R1 (57) and -2 (58). Furthermore, ET-1 is an endogenously produced growth factor implicated in tumor cell proliferation (59), and ET A receptor antagonism may prove useful treatment strategy in malignancies where this protein is overexpressed (60). Indeed, the ET-1-activated and MAPK-regulated ETS factor signaling provides an attractive target for the ongoing study of cellular growth, proliferation, and oncogenesis.
In summary, our results demonstrate that ET-1 activates rBNP gene via mechanisms distinct from those of increased calcium influx via L-type calcium channels, Ang II, or mechanical stretch of myocytes. Importantly, ET-1 induced binding of ETS factors (Elk-1) on EBS of rBNP promoter, and this element was required for the activation of transcription by ET-1. The p38 MAPK-dependent mechanism mediated activation of rBNP transcription via EBS, whereas activation of rBNP promoter via the MEK1/ERK pathway was inhibited by intact EBS. In conclusion, given the fundamental role of the interaction of MAPKs and ETS factors in the regulation of eukaryotic cells, these results provide the unique evidence of an ET-1-specific and MAPK-regulated ETS factor pathway for cardiac myocytes.