Role of the Epidermal Growth Factor Receptor in Signaling Strain-dependent Activation of the Brain Natriuretic Peptide Gene*

The epidermal growth factor receptor (EGFR) and ectoshedding of heparin-binding epidermal growth factor (HBEGF), an EGFR ligand, have been linked to the development of cardiac myocyte hypertrophy. However, the precise role that the liganded EGFR plays in the transcriptional activation of the gene program that accompanies hypertrophy remains undefined. Utilizing the human (h) BNP gene as a model of hypertrophy-dependent gene activation, we show that activation of the EGFR plays an important role in mediating mechanical strain-dependent stimulation of the hBNP promoter. Strain promotes endothelin (ET) generation through NAD(P)H oxidase-dependent production of reactive oxygen species. ET in turn induces metalloproteinase-mediated cleavage of pro-HBEGF and ectoshedding of HBEGF, which activates the EGFR and stimulates hBNP promoter activity. HBEGF also stimulates other phenotypic markers of hypertrophy including protein synthesis and sarcomeric assembly. The antioxidant N-acetylcysteine or the NAD(P)H oxidase inhibitor, apocynin, inhibited strain-dependent activation of the ET-1 promoter, HBEGF shedding, and hBNP promoter activation. The metalloproteinase inhibitor, GM-6001, prevented the induction of HBEGF ectoshedding and the hBNP promoter response to strain, suggesting a critical role for the metalloproteinase-dependent cleavage event in signaling the strain response. These findings suggest that metalloproteinase activity as an essential step in this pathway may prove to be a relevant therapeutic target in the management of cardiac hypertrophy.

In response to increased hemodynamic load, cardiac myocytes undergo hypertrophic growth. Initially, this hypertrophy is compensatory, allowing the heart to maintain pump function and cardiac output in the face of the increased work requirement. However, with time, this process becomes maladaptive, leading to cardiac decompensation and, eventually, heart failure (1). In fact, ventricular hypertrophy is an independent risk factor for arrhythmias, sudden death, and heart failure (2,3).
At the cellular level, myocyte hypertrophy is characterized by an increase in cell size and protein synthesis, increased sarcomeric assembly, and changes in gene expression (4). The latter includes sequential activation of immediate early response genes (e.g. protooncogenes such as c-fos and c-jun), a fetal gene program (e.g. atrial natriuretic peptide, skeletal ␣-actin, and ␤-myosin heavy chain), and sarcomeric genes (e.g. cardiac ␣-actin and myosin light chain-2) (4).
Activation of the fetal gene program is one of the most consistent markers of cardiac myocyte hypertrophy. Genes within this group are usually expressed in late embryonic and early neonatal life. In the adult, these genes are quiescent under normal conditions but can be activated following exposure to hypertrophic stimuli such as hemodynamic load. Thus, reactivation of these genes represents a reliable marker for induction of the hypertrophic process, either in vitro (5) or in vivo (6). Brain natriuretic peptide (BNP), 1 although originally isolated from porcine brain (7), is produced predominantly in the heart. As with the fetal genes, BNP gene expression increases dramatically in response to hypertrophic stimuli (8) and plasma BNP levels are used clinically to detect and guide the management of hypertrophy and heart failure in humans (9 -11).
The epidermal growth factor receptor (EGFR) belongs to the ErbB family of receptor tyrosine kinases (12). There are four members of the ErbB family: EGFR/ErbB1/HER1, ErbB2/ HER2/neu, ErbB3/HER3, and ErbB4/HER4. Ligands of the epidermal growth factor (EGF) family bind to the ErbB receptors resulting in homodimerization or heterodimerization and subsequent autophosphorylation of specific tyrosine residues in the cytoplasmic domains of the receptors. This leads to the binding of adapter proteins and subsequent activation of downstream signaling pathways that define receptor biological activity. In vertebrates, these ligands exhibit some specificity. Major ligands for the EGFR include EGF, transforming growth factor-␣, amphiregulin, and heparin-binding EGF-like growth factor (HBEGF) (13). HBEGF has attracted particular interest as a signaling intermediate in G-protein-coupled receptor signal transduction (14,15). The generation of HBEGF results from metalloproteinase-dependent cleavage of HBEGF from its membrane anchor (ectoshedding). This cleavage event is linked through a mechanism that is only partially understood to Gprotein-coupled receptor occupancy (14). Soluble HBEGF is then free to associate with EGFR and promote local autocrine/ paracrine effects in the same or in neighboring cells. HBEGF transactivates the EGFR in response to G-protein-coupled ligands such as ET-1 (14,15). Asakura et al. (16) have shown that inhibiting ectoshedding of HBEGF prevents ET-1-dependent stimulation of myocyte hypertrophy in vitro and the development of load-dependent cardiac hypertrophy in vivo. HBEGF also plays a critical role in normal heart development. Iwamoto et al. (17) recently reported that HBEGF-null mice develop severe heart failure associated with dilated ventricular chambers, diminished cardiac function, and grossly enlarged cardiac valves.
Neonatal rat cardiac myocytes have provided a useful model to investigate myocyte hypertrophy that develops following exposure to pharmacological (e.g. ␣-adrenergic agonists, endothelin) (18,19) or mechanical (e.g. cyclic strain) (20) stimuli in vitro. These myocytes undergo changes in cell size, architecture, protein synthesis, and gene expression that parallel those seen in hypertrophied myocardium in vivo. As a hypertrophic stimulus, mechanical strain stimulates induction of immediate early genes (20 -25), fetal genes (5,20,22), and sarcomeric genes (26,27) in cardiac myocytes. We have previously reported that applying mechanical strain to neonatal rat cardiac myocytes activates the human BNP gene promoter (5,28,29). This activation involves a direct as well as an autocrine/paracrine component (28,30) and signals through a combination of extracellular signal-regulated kinase (ERK) and p38 mitogenactivated protein kinase (p38 mitogen-activated protein kinase) dependent events (30).
The EGFR has been implicated in the development of myocyte hypertrophy in vitro and the development of load-dependent cardiac hypertrophy in vivo (16,31). However, the precise role that the EGFR plays in activation of the hypertrophic gene program remains undefined. We have investigated the involvement of liganded EGFR in the process linking mechanical strain to stimulation of the human (h)BNP gene promoter. We have established that this linkage exists and that it requires the participation of both reactive oxygen species and endothelin at different points in the signaling cascade.
Cell Culture and Application of Mechanical Stretch-Ventricular myocytes were prepared from 1-day-old neonatal Sprague-Dawley rats by alternate cycles of 0.05% trypsin digestion and mechanical disruption as described previously (32). Cells were cultured on collagen Icoated Flex plates (Flexcell International Corp., McKeesport, PA) in Dulbecco's modified Eagle's medium (H21) containing 10% enriched calf serum (Gemini Bioproducts, Woodland, CA), 2 mM glutamine, 10 units/ml penicillin, and 100 units/ml streptomycin for 24 h post-isolation. A glass-cloning cylinder (1 cm in diameter) was placed in the center of each well to preclude cell attachment, thereby positioning the majority of adherent cells on the outer 75% of the culture surface where distension is maximal (33). Following intervention (transfection and/or pharmacological treatment), cells were subjected to cyclical stretch (60 Hz) on the FX3000 unit (Flexcell International Corp.) at a level of distension sufficient to promote a calculated increment in surface area of ϳ20% at the point of maximal distension on the culture surface (33).
Protein Synthesis-Protein synthesis was assessed by measuring [ 3 H]leucine incorporation in cultured ventricular myocytes. Freshly isolated cells were cultured in 24-well plates for 24 h and then changed to serum-free medium and subjected to treatment as indicated in Fig.  5A for 48 h. Cells were pulsed with [ 3 H]leucine in leucine-free medium for the final 4 h of the treatment period, washed three times with ice-cold PBS, and extracted with 10% trichloroacetic acid at 4°C for 30 min. Cell residues were rinsed in 95% ethanol, solubilized in 0.25 N NaOH for 2 h, and neutralized with 2.5 mM HCl in 1 mM Tris-HCl (pH 7.5). Radioactivity was measured in a liquid scintillation counter.
Immunostaining-Ventricular myocytes were cultured in 4-chamber slides for 24 h, changed to serum-free medium, and then subjected to the indicated treatments for 48 h. Cells were then washed with PBS and fixed with 3.7% paraformaldehyde at room temperature for 20 min followed by PBS containing 0.2% Triton X-100 for 2 min. Slides were blocked with PBS containing 0.2% bovine serum albumin and 0.1 g/ml normal horse IgG for 1 h and incubated with mouse anti-rat sarcomeric ␣-actinin antibody (EA-53) in the above blocking solution at room temperature for 1 h. Cells were washed three times with PBS and incubated with Texas Red-conjugated horse anti-mouse secondary antibody (Vector Laboratories) at room temperature for 1 h. After three consecutive washes, slides were mounted with VectorShield mounting medium and viewed using fluorescence microscopy. Cell surface areas of individual cells were quantified using OpenLab (Improvision) planimetry software from two-dimensional images.
Wild-type and Mutant Plasmids-The construction of Ϫ1595 and Ϫ904 human hBNP-luciferase has been described previously (34). To generate Ϫ1315 rat ET-1-luciferase, oligonucleotide primers with sequences 5Ј-GTCAATGTGCTTTATGTGTG-3Ј (sense) and 5Ј-TCACCG-GAGCGCAAAGCGTC-3Ј (antisense) were used to amplify a genomic fragment extending from Ϫ1315 to ϩ100 in the human ET gene (35). KpnI (sense) and BglII (antisense) restriction sites were added to the respective 5Ј borders to facilitate subcloning. PCR products of the expected size were cut with KpnI and BglII and subcloned into KpnI/ BglI-cut pGL3, a promoter-less luciferase reporter plasmid (Promega, Madison, WI).
Transfection and Luciferase Assay-18 -24 h following isolation, cardiac myocytes (10 6 ) were transiently co-transfected with 0.5 g of the indicated BNP luciferase reporters and 0.05 g of actin-␤-galactosidase using Lipofectin reagent (Invitrogen) under conditions recommended by the manufacturer. 18 -24 h post-transfection, cells were changed to serum-free medium and maintained for 24 h prior to experimentation. At the end of the appropriate intervention, cells were washed twice with PBS and lysed with lysis buffer (Promega). Luciferase activity was measured using the luciferase assay system (Promega). ␤-Galactosidase activity was assayed using the Galactolight Plus chemiluminescence assay (Tropix, Bedford, MA). For each well, luciferase levels were normalized to ␤-galactosidase activity.
Measurement of HBEGF-Cultured myocytes were deprived of serum for 24 h and then cultured in phenol red-free Dulbecco's modified Eagle's medium H21. Following the experimental interventions indicated, medium was collected and concentrated using a Centricon centrifugal filter device (YM-3, Millipore). Samples were centrifuged at 6500 ϫ g for at least 3 h until volume was reduced from 4 to 0.5 ml. Concentrates were incubated with 2.5 g of anti-HBEGF antibody and protein A-agarose overnight at 4°C. Immunoprecipitates were washed three times, boiled, and electrophoresed on 15% SDS-polyacrylamide gels. Using conventional Western immunoblotting techniques as described above, HBEGF levels were assessed using goat polyclonal anti-HBEGF IgG and donkey anti-goat IgG.
Statistical Analysis-Data were analyzed by one-way ANOVA using Student-Newman-Keuls or Bonferroni Multiple Comparison post-hoc tests as applicable.

RESULTS
Application of cyclic mechanical strain to cultured neonatal myocytes for 5 min activated the EGFR (Fig. 1). Western blots of EGFR immunoprecipitates demonstrated a strain-dependent increase in the phosphotyrosine content of a 170-kDa protein that co-migrated with the EGFR itself (Fig. 1A). Pooled data from three independent experiments indicated that mechanical strain increased phosphorylation of EGFR by 287 Ϯ 13% compared with static control levels (p Ͻ 0.05) (Fig. 1B), whereas there was no change in absolute levels of EGFR protein (Fig. 1C). The ability of the EGFR inhibitor, AG-1478, to abrogate straindependent phosphorylation of the EGFR (Fig. 1, A and B) suggests that strain stimulates the intrinsic kinase activity of EGFR, leading to autophosphorylation of the receptor.
In some systems, activation of the EGFR has been shown to result from cleavage of HBEGF from its membrane anchor (ectoshedding) and subsequent occupancy of the receptor by this ligand (14). We asked whether cyclic strain increased HBEGF ectoshedding in our myocyte cultures. As shown in Fig.  2A, strain effected a time-dependent increase in HBEGF levels in the culture medium. The increase was detectable at 5 min, peaked after 15 min at 477 Ϯ 87% of static control levels (p Ͻ 0.01), and began to normalize after 30 min of mechanical strain.
A component of the strain effect on the hBNP promoter involves the local generation of the vasoactive peptide ET (28,30), which can account for 50 -85% of the strain response. 2 ET-1 has been shown previously to activate the EGFR (16). Therefore, we asked whether ET-1 would promote HBEGF FIG. 1. Effect of mechanical strain on EGFR phosphorylation. Cardiac myocytes were deprived of serum for 24 h, pretreated with the EGFR inhibitor, and AG-1478 (250 nM) for 1 h and then subjected to a static or strain environment for 5 min. Panels A and B, cells were lysed, immunoprecipitated (i.p.) with an anti-EGFR antibody, and assayed for phosphorylation by Western blotting using an anti-phosphotyrosine antibody. Panels A and C, membranes were then stripped, and EGFR levels were determined by immunoblotting with an anti-EGFR antibody. Data are expressed as means Ϯ S.E., n ϭ 3. **, p Ͻ 0.01 versus static control.
ectoshedding from cardiac myocytes in our cultures. Treating cardiac myocytes with ET-1 did in fact increase HBEGF secretion into the medium to 500 Ϯ 128% of control levels (p Ͻ 0.05) (Fig.  2B), raising the possibility that ET plays a role upstream of HBEGF secretion in the strain-dependent signaling pathway.
We next asked whether HBEGF can activate the hBNP promoter directly. As shown in Fig. 3, both EGF (Fig. 3A) and HBEGF (Fig. 3B) significantly increased hBNP promoter activity, implying that HBEGF-dependent activation of EGFR plays a role in mediating strain-dependent activation of the hBNP promoter. To address this question more directly, we examined the effect of the EGFR inhibitor AG-1478 on straindependent hBNP promoter activation. The strain increased hBNP promoter activity to 369 Ϯ 58% versus the static control (p Ͻ 0.001). This increase was completely abolished by AG-1478 (Fig. 4), demonstrating a requirement for activated EGFR in promoter induction. In contrast, strain-dependent hBNP promoter activity remained intact in the presence of the platelet-derived growth factor receptor kinase inhibitor, AG-1296 (Fig. 4), implying specificity in the signaling mechanism linking strain to the BNP transcriptional response.
HBEGF also increased protein synthesis assessed as [ 3 H]leucine incorporation into acid-insoluble protein, cell size, and sarcomeric assembly in these myocyte cultures, demon-strating global activation of the hypertrophic program in these cultures (Fig. 5).
We next explored the role of endogenous ET in signaling the strain-dependent increase in HBEGF in the myocyte cultures. We have reported previously that ϳ60% of the BNP promoter response to strain is dependent upon local generation of ET in the myocyte cultures (28). Our subsequent studies have indicated that the ET contribution to the strain response can be as high as 85% in some cultures. 2 The results presented in Fig. 6A indicate a dominant role for ET in generating the response to strain in these cultures. Strain led to an ϳ2.3-fold increment in BNP promoter activity. BQ 610, a type A ET receptor antagonist, effected a reduction in basal promoter activity (ϳ55% inhibition) and nearly completely blocked the response to strain. The strain led to an ϳ9-fold increase in HBEGF levels in this series of experiments (Fig. 6B). Interestingly, although BQ 610 had little effect on basal HBEGF levels, it virtually eliminated the induction by mechanical strain. Furthermore, the induction of BNP promoter by ET-1 (ϳ13-fold in this study) was nearly completely blocked by the EGFR inhibitor AG-1478 (Fig. 6C). Collectively, these data demonstrate that endogenous ET plays a major role in contributing to the strain-dependent induction of the hBNP promoter and that it does so by increasing the generation of HBEGF.
If strain-dependent activation of the hBNP promoter requires activation of the EGFR, inhibition of metalloproteinasemediated cleavage of membrane-bound pro-HBEGF might be predicted to abrogate the promoter induction. Pretreatment of cardiac myocytes with the metalloproteinase inhibitor GM-6001 prevented strain-dependent secretion of HBEGF (Fig. 7A) and attenuated hBNP promoter activation in response to strain (Fig. 7B), suggesting that metalloproteinase activity is required for the latter response.
Mechanical strain generates reactive oxygen species (ROS), in particular superoxide, that appear to be linked to increased protein synthesis (36) in cardiac myocytes. This led us to investigate the role of ROS as mediators of the strain-dependent induction of the hBNP promoter. The antioxidant, N-acetylcysteine, reduced both basal and strain-dependent HBEGF secretion in the myocyte cultures (Fig. 8A), suggesting that strain induces HBEGF ectoshedding through the generation of reactive oxygen species. As expected, N-acetylcysteine attenuated the ability of cyclic strain to activate the hBNP promoter (Fig. 8B), although it had no effect on the ability of ET-1 to do so (Fig. 8C). This is noteworthy because, as noted above, a major component of the response to strain requires generation of ET (28,30). These data suggest that ROS signaling is positioned upstream of ET in the signaling cascade linking strain to increased hBNP promoter activity. The ability of strain to activate the endothelin promoter is in fact abolished by N-acetylcysteine (Fig. 8D), a finding that supports this hypothesis.
Recent reports have implicated NAD(P)H oxidase as one source of reactive oxygen species that promote EGFR ligand release (37,38). Therefore, we questioned whether NAD(P)H oxidase might be involved in signaling the strain response in cardiac myocytes. Apocynin is an NAD(P)H oxidase inhibitor that prevents the binding of the cytosolic subunits of the NAD(P)H oxidase to its membrane-bound subunits, thereby preventing oxidase activation and subsequent generation of superoxide (39). Apocynin inhibited both basal and strain-dependent HBEGF secretion from cardiac myocytes (Fig. 9A) and blocked strain-dependent hBNP promoter activity (Fig. 9B). As with N-acetylcysteine, apocynin had no effect on the ability of ET-1 to stimulate the hBNP promoter (Fig. 9C); however, straindependent activation of the endothelin promoter was abolished by apocynin (Fig. 9D). This finding is consistent with the concept that NAD(P)H oxidase activity and ROS-dependent signaling lie upstream of ET production, which in turn is positioned upstream from hBNP promoter activation. DISCUSSION Cardiac myocytes undergo hypertrophy in response to hemodynamic load. This process is characterized by an increase in cell size and protein synthesis, increased sarcomeric assembly, and changes in gene expression (4) including activation of immediate early response, fetal, and sarcomeric genes (4). As with the fetal gene program, BNP gene expression increases dramatically in response to hypertrophic stimuli (8). Here, we demonstrate that induction of the BNP gene in the neonatal rat cardiac myocyte requires activation of the EGFR. These findings are compatible with the model summarized in Fig. 10 in which mechanical strain through the activation of NAD(P)H oxidase increases ROS generation, which then leads to increased production of ET. ET induces metalloproteinase-mediated cleavage of pro-HBEGF and ectoshedding of HBEGF. HBEGF activates the EGFR, which in turn leads to activation of the hBNP promoter.
HBEGF was originally identified as a mitogenic factor that acts through the EGFR to promote cell division in fibroblasts and vascular smooth muscle cells (40). More recently, the process of HBEGF shedding has received considerable attention as an intermediate in signaling the events, leading to cardiac hypertrophy (41). Expression of the HBEGF gene is enhanced in hypertrophied left ventricles of spontaneously hypertensive rats (42) and in the ventricles of rats following myocardial infarction (43). It is also increased in neonatal or adult rat cardiac myocytes induced to undergo hypertrophy with an ␣-adrenergic agonist (44). Exogenous HBEGF has been shown to increase protein content in adult rat cardiac myocytes (44) and to increase protein synthesis, cell size, and sarcomeric assembly in cultured neonatal rat myocytes (see Fig. 5), all consistent with the induction of hypertrophy in these cells.
Independent studies have established a link between metalloproteinase-dependent processing of pro-HBEGF to HBEGF and the development of hypertrophy. Asakura et al. (16) recently reported that inhibiting ectoshedding of HBEGF prevents ET-dependent stimulation of myocyte hypertrophy in vitro and the development of load-dependent cardiac hypertrophy in vivo. Angiotensin II-induced EGFR activation and cardiac hypertrophy were also abrogated by metalloproteinase inhibition using either a pharmacological inhibitor or expression of a dominant negative mutant of metalloproteinase 12 (ADAM12) (16). Here, we show that inhibiting metalloproteinase activity using GM-2001 blocks both secretion of HBEGF into the extracellular milieu (Fig. 7A) and strain-dependent activation of the hBNP promoter in neonatal rat cardiac myocyte cultures (Fig. 7B). These data suggest that metalloproteinase-dependent cleavage of pro-HBEGF is required for straindependent activation of the BNP promoter and imply that selective metalloproteinase inhibitors may prove useful as therapeutic agents capable of interfering with the signaling circuitry leading to cardiac hypertrophy and subsequent cardiac dysfunction. This possibility is particularly attractive in that these metalloproteinases are located in the cell membrane and are accessible from the extracellular compartment.
A variety of metalloproteinases are activated by free radicals (45). For example, matrix metalloproteinase-2 is induced by mechanical strain in vascular smooth muscle cells through ROS formation (46). Generation of the EGFR ligand, amphiregulin, by ADAM17 was suppressed by antioxidants such as N-acetylcysteine or dimethylthiourea (37). Predictably, activation of the EGFR is also redox-sensitive. Autophosphorylation of the EGFR is stimulated by ROS (37,47,48), and in coronary arteries, stretch-enhanced EGFR phosphorylation operates through a ROS-dependent mechanism (38). In our studies, N-acetylcysteine reduced both basal and strain-dependent HBEGF secretion in the myocyte cultures (Fig. 8A), which in turn attenuated the ability of strain to activate the hBNP promoter (Fig. 8B). Recent reports have implicated NAD(P)H oxidase as a potential generator of superoxide, which in turn promotes EGFR ligand release (37,38). In coronary arteries, stretch-dependent EGFR phosphorylation is prevented by inhibition of NAD(P)H oxidase (38). In cardiac myocytes, NAD(P)H oxidase-dependent ROS generation has been implicated as playing a role in both ␣ 1 -adrenoreceptor-dependent (49) and angiotensin II-dependent (50) hypertrophy. In the studies presented here, suppression of NAD(P)H oxidase activity attenuated strain-dependent processing of HBEGF (Fig. 9A) as well as hBNP promoter activity in cultured cardiac myocytes (Fig. 9B). Taken together with the studies described above, this finding suggests that EGFR activation is an important intermediary step in the pathway linking strain-dependent ROS generation to myocyte hypertrophy and the changes in gene transcription that accompany it.
Activation of the BNP promoter by mechanical strain involves a direct effect of the stimulus on the cardiac myocyte and an indirect autocrine/paracrine effect that involves the sequential generation of angiotensin II and ET-1 (28). By inference, activation of the ET promoter by mechanical strain (Figs. 8 and 9) probably reflects induction of the autocrine/paracrine component of the response. More interestingly, NAD(P)H activation and ROS generation appear to lie upstream of ET-1 gene expression in the strain-dependent pathway, because blocking NAD(P)H activity or scavenging free radicals with N-acetylcysteine ablated strain-dependent induction of the ET-1 gene but had no effect on the ability of exogenous ET-1 to activate the BNP promoter. Thus, the generation of ET would appear to be the critical ROS-sensitive step in this cascade. However, it is worth noting that the ET antagonist BQ610 blocked strain-dependent HBEGF generation within minutes (Fig. 6B)  ulus application. Although ET transcription may be activated within minutes following application of the strain stimulus, it seems unlikely that this could result in significant new ET synthesis within this time frame. These findings may indicate that ROS are involved in promoting both release of pre-formed ET as well as the subsequent activation of ET gene expression (Figs. 8D and 9D) without impacting on ET-dependent signal transduction (Figs. 8C and 9C). Alternatively, it may be that ET and ROS function in parallel rather than in series in signaling the early response to strain.
The magnitude of the contribution of ET to the response to strain is larger than we have reported previously (28). The reason(s) behind this discrepancy is unclear. We transfected cells in the previous study by electroporation and in the present study using Lipofectin reagent; however, a limited number of transfections carried out with electroporation demonstrated a similar level of ET dependence (data not shown), indicating that this is not technique-related. In any case, although the link between ET (and the autocrine-paracrine component of the strain response) and HBEGF generation can clearly be made in these studies, the relatively small contribution of the "direct" (i.e. ET-independent) effect to the response to strain in the current studies precludes assignment of the HBEGF/EGFR as a major effector for this pathway.
We have previously reported that a number of nuclear receptor agonists including the peroxisome proliferator activator receptors (PPAR), subtypes ␣ and ␥, vitamin D receptor, and retinoic acid/retinoid receptors attenuate ET-1-dependent hypertrophy and hBNP promoter activity in neonatal cardiac myocytes (51,52). Moreover, pioglitazone, a PPAR␥ agonist, has been shown to abrogate left ventricular remodeling and myocyte hypertrophy following experimental myocardial infarction (53). Interestingly, ligand activation of PPAR␥ inhibits HBEGF promoter activity in rat intestinal cells, possibly by suppressing the transcriptional activities of AP-1 and Ets (54). PPAR␥ ligands also reduce levels and activities of various metalloproteinases including metalloproteinases 1, 2, 3, 7, 9, and 13 (53, 55-61). Vitamin D receptor ligands modulate EGFR levels and/or EGFR activation in keratinocytes (62), breast cancer cells (63), human colon cancer cells (64), squa-mous carcinoma cells (65), and osteoblast-like cells (66), whereas retinoid X receptor ligands have been shown to downmodulate EGFR levels (67) and EGFR signaling (i.e. Stat3 activation) in squamous carcinoma cells (68). It is tempting to speculate that PPAR␥-, retinoid X receptor-, and/or vitamin D receptor-dependent inhibition of hypertrophy might involve the blockade of metalloproteinase-dependent HBEGF ectoshedding or reduction in availability of the pro-HBEGF substrate in cardiac myocytes.
In summary, the activation of the EGFR plays an important role in mediating mechanical strain-dependent activation of the BNP promoter. This pathway sequentially involves oxidant signaling, ET-1, and ectoshedding of HBEGF. Metalloproteinase activity as a critical intermediate in this pathway may represent a logical target for therapeutic interventions designed to control the development of cardiac hypertrophy and the sequelae that accompany it. Strain activates NAD(P)H oxidase, which generates ROS. This promotes production of ET-1, which stimulates metalloproteinase-mediated cleavage of pro-HBEGF to HBEGF, activates the EGFR and subsequently leads to an increase in hBNP promoter activity.