2.4. Natriuretic peptides
In mammals, the natriuretic peptide family consists of highly homologous polypeptide cardiac hormones: ANP, BNP and C-type natriuretic peptide (CNP) (for review, see Ruskoaho 1992). ANP and BNP play an important role in cardiovascular homeostasis, functioning as counterregulators of the actions of Ang II (for reviews, see Fyhrquist & Tikkanen 1988, Nakao et al. 1992a, de Bold et al. 1996) (Table 2). Dendroaspis natriuretic peptide (DNP) (Schweitz et al. 1992) and salmon cardiac peptide (sCP) (Tervonen et al. 1998) are new members of the natriuretic peptide family sharing structural, biological and distributional similarities to other natriuretic peptides. Complete disruption of ANP gene results in marked cardiac hypertrophy in mice with a modest increase in blood pressure (John et al. 1995), whereas transgenic mice overexpressing ANP have low heart weight (Klinger et al. 1994). Although BNP gene knockout mice exhibit normal sized hearts, blood pressure and circulating levels of ANP, they develop ventricular fibrosis, suggesting a cardiac myocyte-derived antifibrotic role for BNP (Tamura et al. 2000).
Three distinct receptors exert the effects of natriuretic peptides in mammalian tissues (for review, see Nakao et al. 1992b). Natriuretic peptide receptors-A and -B (NPR-A and NPR-B) are transmembrane single-chain guanylate cyclase-linked polypeptide receptors that stimulate the intracellular cGMP (Chang et al. 1989, Schulz et al. 1989). NPR-C, also known as a clearance receptor, binds natriuretic peptides, whereafter the ligand-receptor complex is internalized and enzymatically degraded (Maack et al. 1987). NPR-A, -B and -C genes are expressed in several tissues, including heart, brain, kidney and adrenal gland (for review, see Yandle 1994). Mice lacking NPR-A develop marked cardiac hypertrophy and chamber dilatation (Knowles et al. 2001).
Cardiac natriuretic peptide gene expression and secretion are activated in response to hypertrophic stimuli (for review, see de Bold et al. 1996). In agreement with these findings, plasma ANP and BNP concentrations are useful cardiac-specific markers (Dagnino et al. 1991). Administration of synthetic BNP (nesiritide), ANP (carpeditide) and DNP as well as vasopeptidase inhibitors (omapatrilat), which are novel molecules inhibiting the activity of NEP and ACE, have beneficial effects on hemodynamic, neurohumoral and renal functions in CHF (Abraham et al. 1998, Troughton et al. 2000b, Lisy et al. 2001, Mizuno et al. 2001b).
Table 2. Comparison of the effects of natriuretic peptides and Ang II
| Biological effects | Natriuretic peptides | Angiotensin II |
|---|---|---|
| Vasoconstriction | ↓ | ↑ |
| Myocyte hypertrophy | ↓ | ↑ |
| Renal sodium secretion | ↑ | ↓ |
| Diuresis | ↑ | ↓ |
| Sympathetic nerve activity | ↓ | ↑ |
| Parasympathetic nerve activity | ↑ | ↓ |
| Aldosterone concentration | ↓ | ↑ |
| Fibrosis | ↓ | ↑ |
| Renin secretion | ↓ | ↓ |
2.4.1. Atrial natriuretic peptide
In 1981, de Bold et al. described the heart as an endocrine organ, since injection of an extract of atrial muscle into rats induced vigorous natriuresis and a fall in arterial pressure, and thereafter ANP was isolated (de Bold et al. 1981, de Bold 1985). ANP is present as a single-copy gene and is organized into three exons separated by two introns (for review, see Nakao et al. 1992a). The ANP gene exhibits a significant homology of the nucleotide sequence among different species (for review, see Rosenzweig & Seidman 1991). The precursor of ANP, preproANP, is converted to proANP, which is a predominant storage form of ANP in specific atrial granules (Vuolteenaho et al. 1985). Next, proANP is cleaved into amino-terminal ANP and biologically active 28-amino acid ANP in mammals (for review, see Nakao et al. 1992a). In humans, transmembrane serine protease, corin, has been shown to convert proANP to ANP (Yan et al. 2000). ANP and other natriuretic peptides share a common structure of a 17-amino acid loop formed by an intrachain disulfide bond between cysteine residues (Misono et al. 1984, for review, see Rosenzweig & Seidman 1991) (Fig. 3).
In the heart, cardiac myocytes are the predominant cells for ANP production (Argentin et al. 1994), although ANP may also be synthesized by fibroblasts (Cameron et al. 2000). Atrium is the most important site for the synthesis of ANP, since the ANP mRNA levels in the atria are 100-fold higher than in the ventricles (for review, see Nakao et al. 1992a). Atrial ANP is secreted in a constitutive and regulative manner (Ogawa et al. 1999), whereas ventricular ANP secretion occurs predominantly in a constitutive pathway (Ruskoaho et al. 1989). Cardiac myocyte stretch is suggested to be a primary stimulus for the release of ANP (Ruskoaho et al. 1986). Also, many neurohumoral secretagogues induce the release of ANP (for review, see de Bold et al. 1996). The major degradation pathways of ANP are receptor-mediated endocytosis and NEP-mediated degradation (for review, see Ruskoaho 1992).
2.4.1.1. ANP gene expression in response to cardiac overload
In experimental animal models, ventricular ANP gene expression and ir-ANP concentrations increase in response to cardiac overload and myocardial hypertrophy (Lee et al. 1988, Marttila et al. 1996). However, whether wall stretch acts directly or via paracrine factors liberated in response to wall distension remains to be clarified (for review, see Ruskoaho et al. 1997). Since the ANP mRNA levels increase 10- to 15-fold in response to hypertrophic stress in the adult ventricle and the basal ANP mRNA levels are relatively low, ANP may be one of the best available molecular markers for cardiac myocyte hypertrophy (Lattion et al. 1986, Kinnunen et al. 1991). Increased plasma and cardiac ANP levels have been measured in patients with CHF related to the severity of cardiac disease (Tikkanen et al. 1985, Yasue et al. 1994, de Boer et al. 2001).
It is probable that neural and endocrine factors are directly involved in stimulating ANP secretion, since ANP release declines to baseline levels despite maintained or repetitive atrial muscle stretch within few minutes both in vivo and in vitro (for review, see de Bold et al. 1996). Indeed, cardiac ANP gene expression and release are stimulated by α-adrenergic agonists, ET-1, glucocorticoids, prostaglandins, thyroid hormone, NO inhibition and Ang II (for reviews, see Ruskoaho 1992, Ruskoaho et al. 1997). The release of ANP is also modulated by β -adrenergic agonists (Agnoletti et al. 1992), acetylcholine (Antunes-Rodrigues et al. 1993), Na+/K+-ATPase inhibitors (Morise et al. 1991) and AVP (Marttila et al. 1996, Magga et al. 1997a).
Stretch does not increase atrial ANP mRNA levels for over a 2-hour period of stimulation in perfused rat heart (Mäntymaa et al. 1993), whereas stretch-induced changes in ANP gene expression occur after 24 h in cultured cardiac myocytes (Gardner et al. 1992, Sadoshima et al. 1992b). Therefore, it seems that more time is required for stretch to induce changes in ANP than BNP gene expression, suggesting that ANP gene expression has characteristics of secondary response genes (Hanford & Glembotski 1996). However, rapid induction of ANP gene expression within five hours by mechanical stretch has been observed in papillary muscle strips (Jarygin et al. 1994).
In rats, both cardiac ANP gene expression and cardiac and plasma ir-ANP concentrations have been shown to increase in response to chronic volume overload induced by deoxycorticosterone acetate (DOCA)–salt treatment and aortocaval shunt, and pressure and volume overload stimulated by binephrectomy (Lattion et al. 1986, Lear & Boer 1995, Yokota et al. 1995). In fact, aortocaval shunt increases left ventricular ANP mRNA levels before the onset of myocardial hypertrophy in rats (Su et al. 1999). In genetically hypertensive animals, SHRs and transgenic rats carrying the mouse Ren-2 renin gene, baseline cardiac ANP synthesis and release are markedly higher than in normotensive controls (Ruskoaho et al. 1989, Marttila et al. 1996). Also, pressure overload induced by aortic coarctation (Rockman et al. 1994, Ogawa et al. 1996) and dogs subjected to rapid ventricular pacing (Perrella et al. 1992) exhibit increased cardiac ANP gene expression in vivo. In response to myocardial infarction-induced cardiac hypertrophy, ANP mRNA levels have been shown to increase both in experimental animal models and in humans (Hama et al. 1995, Omland et al. 1996, Gidh-Jain et al. 1998).
2.4.1.2. Transcriptional regulation of ANP gene
During embryonic and fetal development, ANP gene is expressed both in the atrial and ventricular but not in the skeletal muscle cells. In the later stage of fetal development, the ANP gene appears to be switched off in the ventricular cells whereas its expression remains high in the atria (Argentin et al. 1994). However, when the ventricles are subjected to hemodynamic overload, ventricular ANP gene expression is reactivated (Lee et al. 1988).
High degree of homology exists between the rat and human ANP genes, suggesting the presence of putative well-conserved regulatory elements (Argentin et al. 1985). By using gene transfer, sequences of 400 bp upstream from the transcriptional initiation site of the human ANP promoter have been reported to be sufficient for the cardiac myocyte-specific gene expression in vitro (Wu et al. 1989) and atrial-specific gene expression in transgenic mice (Field 1988). Furthermore, a 700-bp sequence of the proximal rat ANP promoter is suggested to be sufficient for the stage-specific gene expression during heart differentiation and cardiac muscle-specific gene expression in differentiated atrial and ventricular myocytes (Argentin et al. 1994).
Nkx-2.5 response element (NKE) appears to contribute to the cardiac ANP gene transcription in a chamber- and stage-specific manner in collaboration with other regulatory elements of the rat ANP promoter (Durocher et al. 1996). NKE2, another Nkx-2.5 response element located more distally, is suggested to be required for the high-level activation of the ANP promoter (Shiojima et al. 1999). NKE2 may also be involved in the induction of cardiac ANP gene expression during pathological conditions (Takimoto et al. 2000). Transcription factors Nkx-2.5 and GATA4 have been reported to cooperatively activate ANP and physically interact in vitro and in vivo (Durocher et al. 1997, Lee et al. 1998a). GATA proteins are suggested to play a marked role in the ANP gene expression, since several GATA consensus sites are located between –117 and –3003 bp of the rat ANP promoter (Thuerauf et al. 1994). In this regard, two GATA motifs, residing at –290 and –122 bp of the rat and human ANP promoters, may be important elements for the basal cardiac-specific gene expression of ANP (Grépin et al. 1994). In addition, ANP mRNA levels are elevated chronically in transgenic mice overexpressing GATA4 (Liang et al. 2001b), and GATA4 and SRF synergistically increase the activity of ANP promoter in response to ET-1 administration in vitro (Morin et al. 2001).
The AP-1 binding site in the proximal ANP promoter is reported to be essential for the pressure overload-responsiveness in wall stress-stimulated rat hearts in vitro (Cornelius et al. 1997). In addition to AP-1 site, cAMP response element (CRE) is suggested to mediate the wall stretch-induced increase in the rat ANP promoter activity, since mutations of both AP-1 element and CRE are required to confer complete loss of the inducibility in beating rat hearts in vitro (Cornelius et al. 1997). Furthermore, AP-1-like site is suggested to be an important cis regulatory element in mediating pressure overload-responsiveness of the rat ANP gene in vivo (von Harsdorf et al. 1997). In contrast, there are also studies that cannot define the pressure overload responsive element to locate in the abovementioned region of the rat ANP promoter (Knowlton et al. 1995, Hasegawa et al. 1997). Moreover, components of AP-1 may also inhibit the cardiac gene expression, since cotransfection of c-fos and c-jun expression vectors with ANP reporter construct leads to repression of ANP promoter activity (McBride et al. 1993, McBride & Nemer 1998).
Proximal phenylephrine response element and serum response element (SRE) are suggested to mediate both basal and α1-adrenergic agonist-induced transcriptional activation of the rat ANP gene (Ardati & Nemer 1993, Sprenkle et al. 1995, Morissette et al. 2000). Also, A/T-rich element located distally of the ANP promoter appears to mediate α1-adrenergic inducibility in rat ventricular myocytes (Harris et al. 1997). The ISO-stimulated transcription of the ANP gene is suppressed by dominant negative GATA4 (Morisco et al. 2001), suggesting that GATA4 plays an important role in β -adrenergic stimulation of ANP gene expression in vitro. Furthermore, ET-1-stimulated human ANP promoter activity has been demonstrated to require SRE element (Kovacic et al. 1998). The neuron-restrictive silencer element (NRSE), which is located in the 3" untranslated region of the ANP gene, is involved in the ET-1-induced activation of ANP gene expression. However, NRSE also mediates repression of the ANP gene expression in ventricular myocytes (Kuwahara et al. 2001).
2.4.2. B-type natriuretic peptide
In 1988, Sudoh et al. discovered BNP in porcine brain with biological properties and structural homology similar to ANP (Sudoh et al. 1988, Dagnino et al. 1991). Subsequently, BNP was found to be more abundant in cardiac atria and ventricles than in the central nervous system (Ogawa et al. 1991a, Dagnino et al. 1992). The BNP gene is composed of three exons and two introns (Seilhamer et al. 1989) (Fig. 4). In contrast to ANP, BNP has considerable inter-species diversity of amino acid composition. The posttranslational processing of BNP precursors seems to be different from that of ANP, and the processing sites are not conserved between species, resulting in various lengths of BNP (for review, see Yandle 1994). The predominant circulating forms of BNP are 26, 45 and 32 amino acid peptides in pigs, rats and humans, respectively (for review, see Nakao et al. 1992a). The conversion of BNP precursor to BNP-45 in rats has been reported to demand an endoprotease furin (Sawada et al. 1997).
The major storage form of BNP in the heart is the cleaved mature peptide, although in atrial tissue also prohormones may be stored (for review, see Yandle 1994). BNP is suggested to be released constitutively after secretion (Wei et al. 1993), although storage granules containing both ANP and BNP have been described, demonstrating also a regulatory pathway for BNP secretion (Nakamura et al. 1991, Ogawa et al. 1999). Baseline plasma BNP concentration is approximately 1 fmol/ml in humans, which is one-sixth of the plasma ANP concentration determined simultaneously (Mukoyama et al. 1991). The major stimulus controlling the release of BNP from the atria and ventricles appears to be myocyte stretch (for review, see Ruskoaho et al. 1997). The metabolism of BNP still remains quite an unknown subject, and it is likely that both ANP and BNP share similar metabolic pathways. However, although NEP cleaves ANP mainly at one site, porcine BNP appears to be cleaved at several different sites (Vogt-Schaden et al. 1989). In humans, lower binding activity of BNP than ANP to clearance receptors is suggested to be one reason for the longer plasma half-life of BNP (approximately 22 minutes) compared with ANP (Mukoyama et al. 1991).
2.4.2.1. BNP gene expression in response to cardiac overload
Compared with ANP, plasma concentration of BNP has been shown to be a better marker for impaired left ventricular function and diagnosis of CHF (Kohno et al. 1995, Yamamoto et al. 1996, McDonagh et al. 1998, Dao et al. 2001). In addition, plasma BNP concentrations provide prognostic information in CHF (Omland et al. 1996, Tsutamoto et al. 1997) and guide the treatment of patients with CHF (Troughton et al. 2000a). Plasma BNP concentrations and cardiac BNP mRNA levels have been shown to increase in humans with moderate and severe heart failure and not at early stage of ventricular dysfunction (Wei et al. 1993, de Boer et al. 2001). Also in volume overloaded rats, cardiac BNP gene expression is induced specifically in overt heart failure (Yokota et al. 1995, Langenickel et al. 2000), suggesting a possible role for BNP as a marker of the transition from compensated to severe cardiac dysfunction. In addition, CHF stimulated by rapid ventricular pacing in dogs demonstrates that early left ventricular dysfunction is characterized by selective increase in the atrial BNP gene expression, and overt CHF is provided with additional ventricular BNP gene expression (Luchner et al. 1998).
Cardiac volume overload induced by aortocaval shunt, DOCA-salt treatment and bilateral nephrectomy, which also produces pressure load, increases plasma BNP concentrations and left ventricular BNP mRNA and ir-BNP levels (Lear & Boer 1995, Yokota et al. 1995, Langenickel et al. 2000, Marttila et al. 2001) (Table 3). Also, pressure overload generated by aortic banding in rats stimulates BNP synthesis (Ogawa et al. 1996). In addition to cardiac overload, neural and endocrine factors such as ET-1, Ang II, adrenergic agonists and AVP activate cardiac BNP gene expression and secretion (Bruneau & de Bold 1994, Magga et al. 1994, Bruneau et al. 1996, Liang & Gardner 1998, Magga et al. 1999, Ogawa et al. 1999). In SHRs, ventricular BNP mRNA levels are elevated at the onset of the hypertensive stage and the BNP gene expression correlates with the progression of hypertension (Dagnino et al. 1992). Transgenic mice containing the proximal human BNP promoter exhibit increased promoter activities after two days of acute myocardial infarction in vivo lasting until the end of the four-week experiment. Myocardial infarction also induces endogenous mice BNP mRNA levels in the left ventricles within 48 h (He et al. 2001b). In order to study the BNP gene expression in the absence of ANP, ANP gene knockout mice were produced. Ventricular BNP gene expression increases significantly, but the plasma BNP concentrations remain unaltered in ANP gene disrupted mice (Tse et al. 2001), suggesting the inability of BNP to completely compensate the lack of ANP.
In addition to the long-term stimulation of BNP during hemodynamic stress, BNP gene expression increases as quickly as the expression of proto-oncogenes in response to hypertrophic stimuli in vitro (Mäntymaa et al. 1993, Nakagawa et al. 1995) and in vivo well before the development of LVH (Magga et al. 1994). Moreover, BNP mRNA levels are elevated rapidly in response to PE and phorbol ester stimulation, and transcript stabilization by these agents further increases the half-life of BNP mRNA (LaPointe & Sitkins 1993, Hanford et al. 1994, Hanford & Glembotski 1996). However, although mechanical strain increases BNP gene expression at the transcriptional level, it may not stabilize the BNP transcripts (Liang et al. 1997).
Table 3. Left ventricular BNP synthesis in experimental animal models
| Animals | Stimulus | LVH | ΔRR | Time/Age | LV BNP mRNA | LV ir-BNP | Reference |
|---|---|---|---|---|---|---|---|
| AT1A KO mice | CoA | ↔ | ↑ | 30 min | ↑ | − | Harada et al. 1998 |
| WKY rats | AVP/PE | ↔ | ↑ | 1 h | ↑ | ↑ | Magga et al. 1994 |
| SHRs | AVP/PE | ↑ | ↑ | 1 h | ↑ | ↑ | Magga et al. 1994 |
| TGR(mREN-2) | AVP | ↑ | ↑ | 2 h | ↔ | ↔ | Marttila et al. 1996 |
| Wistar rats | AMI | − | ↓ | 4 h | ↑ | ↑ | Hama et al. 1995 |
| SD rats | Nephr. | ↔ | ↑ | 1 d | ↑ | ↑ | Marttila et al. 2001 |
| Wistar rats | Ao-shunt | ↑ | ↓ | 3 d | ↔ | − | Langenickel et al. 2000 |
| SD rats | DOCA | ↑ | ↔ | 1 w | ↔ | ↔ | Yokota et al. 1995 |
| SD rats | DOCA | ↑ | ↑ | 5 w | ↑ | ↑ | Yokota et al. 1995 |
| SHRs | - | ↔ | ↑ | 5 w | ↑ | − | Dagnino et al. 1992 |
| SD rats | CoA | ↑ | ↑ | 6 w | ↑ | ↑ | Ogawa et al. 1996 |
| SHRSPs | - | ↑ | ↑ | 12 w | ↑ | ↑ | Ogawa et al. 1991a |
| TGR(mREN-2) | - | ↑ | ↑ | 12 w | ↔ | ↑ | Marttila et al. 1996 |
| ANP KO mice | - | ↑ | − | 12 w | ↑ | ↔ | Tse et al. 2001 |
| SHRs | - | ↑ | ↑ | 10-22 m | ↑ | ↑ | Magga et al. 1994 |
| Ao-shunt, aortocaval shunt; AMI, acute myocardial infarction; ANP-/- mouse, ANP gene knockout mouse; AT1A, angiotensin II type 1A receptor; AVP, arginine8-vasopressin; CoA, aortic coarctation; DOCA, deoxycorticosterone acetate; ir, immunoreactive; KO, knockout; LV, left ventricular; LVH, left ventricular hypertrophy; Nephr., bilateral nephrectomy; PE, phenylephrine; ΔRR, change in blood pressure; SHRSP, spontaneously hypertensive rat-stroke prone; TGR(mREN-2), transgenic rats carrying the mouse Ren-2 renin gene; WKY, Wistar-Kyoto. | |||||||
2.4.2.2. Transcriptional regulation of BNP gene
The rat BNP promoter possesses multiple potential GATA and AP-1 binding sites in the 5" flanking sequence, which are suggested to be important elements for BNP regulation (Grépin et al. 1994, Thuerauf et al. 1994). The sequences and positions of the GATA and AP-1 motifs are highly conserved in the rat, dog and human BNP genes (Grépin et al. 1994) (Fig. 5). In rat and human BNP gene, GATA site at -30 bp likely serves as a binding sequence for TATA-protein (Seilhamer et al. 1989, Thuerauf et al. 1994). By using gene transfection into the rat cardiac myocytes in vivo, the proximal -114 bp containing GATA and AP-1-like elements have been shown to be sufficient for basal ventricular-specific expression of the rat BNP gene (Marttila et al. 2001). Also in vitro, the determinant for cardiac specificity has been suggested to locate within the proximal -114 bp, since the deletion of sequences between -2200 and -114 bp does not affect the high-level activity of the rat BNP promoter (Grépin et al. 1994). By transfecting the rat BNP-luciferase vectors into the cardiac myocyte cultures, deletion of the proximal AP-1-like motif decreases BNP promoter activity fourfold and deletion of two GATA motifs at -90 bp causes another fourfold reduction of BNP promoter activity (Grépin et al. 1994). Moreover, co-transfection of the rat -116 bp BNP-luciferase reporter with GATA4 or GATA6 results in 4-fold activation of the BNP promoter, and the response is potentiated by PE in cardiac myocytes (Liang et al. 2001b). Furthermore, when cardiac myocytes are transfected with rat BNP-luciferase construct, mutation of two GATA sites at -90 bp decreases GATA-mediated induction by 50 % in response to GATA4 overexpression in vitro (Thuerauf et al. 1994). Similarly, dominant-negative GATA4-engrailed fusion construct inhibits GATA4 or GATA6-induced BNP gene transactivation, suggesting the involvement of GATA factors in regulating BNP promoter activity (Liang et al. 2001b). Moreover, BNP mRNA levels are elevated in transgenic mice overexpressing GATA4 (Liang et al. 2001b). In rats, mutation of the AP-1-like element has no significant effect, whereas mutation of two GATA elements at -90 bp decreases left ventricular BNP promoter activity to 57 % of the intact -114 BNP construct activity in response to one-day nephrectomy in vivo (Marttila et al. 2001).
The rat BNP gene is 65 % homologous with the human BNP gene (LaPointe et al. 1996). Human BNP promoter consisting of sequences between -1818 and +100 bp has been shown to be more active in cardiac myocytes than in fibroblasts in vitro. This sequence consists of positive and negative regulatory elements that contribute to the BNP gene expression in cardiac myocytes (LaPointe et al. 1996). The analysis of the proximal human BNP promoter has revealed that rather than a single element, several tandemly arranged cis elements are responsible for the myocyte-specific promoter activity (LaPointe et al. 1996). In vivo, the human BNP promoter containing sequences from –408 to +100 bp has been shown to confer the cardiac-specific BNP gene expression (LaPointe et al. 1996, He et al. 2001b).
In contrast to rat BNP promoter, the gene construct containing human BNP proximal promoter (-111 to -40 bp) is inactive when transfected into cardiac myocytes, suggesting that GATA and AP-1 elements may not be critical determinants of human BNP gene activity (LaPointe et al. 1996). However, the GATA element at –85 bp of the human BNP promoter and unidentified more distal cis elements have been shown to be targets for ET-1 (He & LaPointe 2001a). Furthermore, three NF-AT binding sites have been demonstrated in the human BNP promoter, and NF-AT3 and GATA4 transactivate human BNP promoter synergistically in neonatal rat cardiac myocytes (Molkentin et al. 1998). M-CAT binding factors, also known as transcription enhancer factor-1-like proteins, contribute to basal human BNP promoter activity by binding to M-CAT-like elements. The proximal M-CAT element mediates the human BNP promoter activation stimulated by ISO and cAMP (He et al. 2000a). In rat BNP promoter, M-CAT may also mediate transcriptional stimulation in response to α1-adrenergic treatment (Thuerauf & Glembotski 1997). Furthermore, mechanical strain and ET-1 have been shown to enhance human BNP promoter activity through p38 MAPK, which operates via the NF-κ B element with three shear stress response element-like structures of the human BNP gene in neonatal rat ventricular myocytes (Liang & Gardner 1999, Liang et al. 2000b).