Adrenomedullin (AM) and intermedin (IMD; adrenomedulln-2) are vasodilator peptides related to calcitonin gene-related peptide (CGRP). The actions of these peptides are mediated by the calcitonin receptor-like receptor (CLR) in association with one of three receptor activity-modifying proteins. CGRP is selective for CLR/receptor activity modifying protein (RAMP)1, AM for CLR/RAMP2 and -3, and IMD acts at both CGRP and AM receptors. In a model of pressure overload induced by inhibition of nitric-oxide synthase, up-regulation of AM was observed previously in cardiomyocytes demonstrating a hypertrophic phenotype. The current objective was to examine the effects of blood pressure reduction on cardiomyocyte expression of AM and IMD and their receptor components. Nω-nitro-l-arginine methyl ester (l-NAME) (35 mg/kg/day) was administered to rats for 8 weeks, with or without concurrent administration of hydralazine (50 mg/kg/day) and hydrochlorothiazide (7.5 mg/kg/day). In left ventricular cardiomyocytes from l-NAME-treated rats, increases (-fold) in mRNA expression were 1.6 (preproAM), 8.4 (preproIMD), 3.4 (CLR), 4.1 (RAMP1), 2.8 (RAMP2), and 4.4 (RAMP3). Hydralazine/hydrochlorothiazide normalized systolic blood pressure (BP) and abolished mRNA up-regulation of hypertrophic markers sk-α-actin and BNP and of preproAM, CLR, RAMP2, and RAMP3 but did not normalize cardiomyocyte width nor preproIMD or RAMP1 mRNA expression. The robust increase in IMD expression indicates an important role for this peptide in the cardiac pathology of this model but, unlike AM, IMD is not associated with pressure overload upon the myocardium. The concordance of IMD and RAMP1 up-regulation indicates a CGRP-type receptor action; considering also a lack of response to BP reduction, IMD may, like CGRP, have an anti-ischemic function.
Intermedin (IMD; adrenomedullin-2) is a recently discovered 47αα vasodilator peptide (Roh et al., 2004; Taylor et al., 2005) of the calcitionin gene-related peptide (CGRP) family, which also includes adrenomedullin (AM). Increased plasma levels of AM have been detected in hypertension (Sumimoto et al., 1997), myocardial infarction (Nagaya et al., 2000), and heart failure (Willenbrock et al., 1999); levels are greater in hypertensive patients with LVH than those without underlying LVH (Sumimoto et al., 1997). Cardiomyocytes express the precursor peptide and secrete mature AM (Horio et al., 1998). This expression is enhanced in response to mechanical stretch (Tsurada et al., 2000), hypoxic stress (Nishikimi et al., 1998), cytokines (Horio et al., 1998), and angiotensin II (Nishikimi et al., 1998). AM attenuates protein synthesis and expression of ANP in cardiomyocytes (Sato et al., 1997; Tsurada et al., 2000; Bell et al., 2006). Intermedin is expressed, less abundantly than AM, in normal myocardium (Bell et al., 2006); the (patho)physiological significance of intermedin is poorly understood at present, although a cardioprotective effect against ischemia-reperfusion injury has been proposed (Taylor et al., 2005).
The actions of these peptides are mediated by a calcitonin receptor-like receptor (CLR) protein in association with one of three receptor activity-modifying proteins, RAMPS1–3. AM has two specific receptors (AM1 and AM2) formed by CLR combined with RAMP2 or -3, respectively (Hay et al., 2004). AM also has appreciable affinity for a CGRP receptor, composed of CLR and RAMP1. AM1 receptors are highly selective for AM over CGRP and other peptides; AM2 receptors show less selectivity, having considerable affinity for βCGRP (McLatchie, 1998). Intermedin acts nonselectively at all three RAMP-CLR coreceptors (Roh et al., 2004). It is probable that the physiological response of a tissue to each peptide is dependent on the levels of expression of CLR and of the various RAMPs. RAMP2 predominates over RAMP1 in the heart and to a lesser extent in cardiomyocytes; RAMP3 mRNA is much less abundant (Autelitano and Ridings, 2001). A membrane-associated “receptor component protein” (RCP) is also required for activation of signal transduction via the various RAMP-CL receptor complexes (Prado et al., 2001). There is controversy as to whether the orphan receptors RDC-1 (Li et al., 1996) and L1 (Chakravarty et al., 2000) also interact with CGRP and AM; it is now generally considered unlikely that these proteins contribute to peptide binding in the myocardium (Autelitano, 1998; Poyner et al., 2002).
NO generated within normal myocardium by “endothelial-type” nitric-oxide synthase (NOS) plays a pivotal role in reduction of myocardial oxygen consumption and blood pressure (Bayraktutan et al., 1998; Stauss et al., 1999). NO levels are reduced in the plasma of hypertensive patients with LVH (Hua et al., 2001). Chronic administration of the NOS inhibitor Nω-nitro-l-arginine methyl ester (l-NAME) to rats results in hypertension, often (Takaori et al., 1997; Bernatova et al., 1999) accompanied by myocardial remodeling (cardiac hypertrophy and fibrosis), vascular remodeling (medial thickening and perivascular fibrosis), cardiac ischemia and necrosis, and mechanical dysfunction. At cardiomyocyte level, enhancement of protein synthesis and mass is observed in both left and right ventricles together with increased cell width, but phenotypic alterations, manifested as up-regulation of the contractile gene, skeletal α-actin, and cardioendocrine peptides BNP and endothelin-1, accompanied by enhanced expression of the counter-regulatory peptide AM and its receptor components, particularly RAMP3, are found in LV cells only (Bell et al., 2006). It is not known whether intermedin expression within the myocardium is also influenced by NO deficiency. Because l-NAME does not enhance mean pulmonary arterial pressure appreciably or alter pulmonary vascular morphology, pressure overload is unlikely to account for the increased protein mass of RV cardiomyocytes (Hampl et al., 1993). NOS inhibition might be expected to have direct consequences for the myocardium since NO exerts an antigrowth effect on both cardiomyocytes (Matsuoka et al., 1996) and nonmyocytes (Sarkar et al., 1995). The relative contributions of pressure overload, caused by marked hypertension in the systemic vascular bed, and of direct inhibition of NO production locally within the myocardium, to the changes identified in the left ventricle remain unclear. The purpose of the study therefore was to investigate the contribution of pressure overload to these changes by undertaking an intervention study with blood pressure-lowering agents, namely, a direct smooth muscle relaxant, hydralazine (50 mg/kg/day), given concurrently with a thiazide diuretic, hydrocholorithiazide (7.5 mg/kg/day); this combination is considered the best option to directly lower blood pressure compared with other antihypertensive drugs, which stimulate renin secretion (Goto et al., 2000).
Materials and Methods
Experimental Model. The study was performed in accordance with Home Office Guidance on the operation of the Animals (Scientific Procedures) Act 1986, published by Her Majesty's Stationary Office, London. Eight-week-old male Sprague-Dawley rats were assigned to receive either 1) l-NAME (35 mg/kg/day; Alexis Corporation, Laüfelfingen, Switzerland) in drinking water; 2) hydralazine (50 mg/kg/day, H-1753; Sigma-Aldrich, Poole, Dorset, UK) plus hydrochlorothiazide (7.5 mg/kg/day, H-2910; Sigma-Aldrich) in drinking water; 3) l-NAME (35 mg/kg/day) currently with hydralazine (50 mg/kg/day) plus hydrochlorothiazide (7.5 mg/kg/day) in drinking water; or 4) drinking water only (age-matched control) for 8 weeks, and maintained at the Laboratory Service Unit, The Queen's University of Belfast, before sacrifice at 16 weeks of age. Systolic blood pressure was determined by tail cuff sphygmomanometer (Harvard Apparatus Inc., Holliston, MA) (Bell et al., 2004), and the mean of four consecutive blood pressure readings was obtained for each animal at weekly intervals. Body weight was recorded weekly and water consumption daily.
Cardiomyocyte Isolation. Following deep anesthesia of the rats using isoflurane (Abbott Laboratories, Queenborough, Kent, UK), the hearts were rapidly excised, placed in ice-cold saline, and weighed. Excised hearts were cannulated through the ascending aorta, and preparations of cardiomyocytes were isolated from the left and right ventricles, respectively, by enzymatic digestion (collagenase, 0.4 mg/ml; Serva, Heidelberg, Germany) using Langendorff perfusion (Bell et al., 2006). After purification, cells were used immediately for extraction of RNA (RT-PCR protocols) and preparation of membrane protein (immunoblotting protocols); for analysis of cell dimensions, a small amount of each cell preparation was retained and suspended at a concentration of approximately 50,000 viable cardiomyocytes/ml in a “creatinine-carnitine-taurine” medium that consisted of modified glutamine-free medium M-199 supplemented with Earle's salts (Gibco, Paisley, UK), 15 mM HEPES, 5 mM creatinine, 2 mM l-carnitine, 5 mM taurine, 100 μM ascorbic acid, 100 IU/ml penicillin, and 100 μg/ml streptomycin.
Cardiomyocyte Dimensions. Cells were visualized using an inverted phase contrast microscope (Axiovert 10; Carl Zeiss GmbH, Jena, Germany) and displayed on a monitor (Panasonic WV-5410) at a magnification of 988×. Viable cells were selected on the basis of rod-shaped appearance without sarcolemmal blebbing and lack of spontaneous contractile activity in the absence of electrical stimulation. For each heart cell preparation, the widths and lengths (micrometers) were determined for 20 viable cardiomyocytes from each chamber, and a mean value was obtained.
Real-Time RT-PCR. Reported sequences for each gene (Table 1) were used to design, on Primer Express software (PE Applied Biosystems, Foster City, CA), rat-specific primers adapted to RT-PCR conditions, which were synthesized by Invitrogen (Carlsbad, CA). RT-PCR was performed using the following cycle parameters: 10 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C using the ABI Prism sequence detector (PE Applied Biosystems). For each gene, RT-PCR was conducted in duplicate using ABsolute QPCR Sybr Green ROX (ABgene, Epsom, Surrey, UK) in a 2:1 reaction. To ensure the quality of measurements, both negative and positive controls were included in each plate. Analysis was performed using ABI 7000 Prism software. The threshold cycle (Ct) value, at which a statistically significant increase in signal, associated with an exponential growth of PCR product was observed, was used to ascertain expression level. Analysis was performed using ABI 7000 Prism software, and statistical analysis of the RT-PCR results were performed using the Ct value (Ct gene of interest – Ct reporter gene). Relative gene expression was obtained by Ct methods (Ct sample – Ct calibrator) using the control group as a calibrator for comparison of every unknown sample gene expression level. The conversion between Ct and relative gene expression level is n fold induction = 2-Ct relative to GAPDH.
Preparation of Membrane Protein. Viable LV cardiomyocytes were suspended in a 20 mM HEPES-based homogenization buffer, pH 7.4, containing protease inhibitors (Sigma-Aldrich) 0.8 μM aprotinin, 0.1 mM bacitracin, 0.1 mM benzamidizine, 5 mM EDTA, 2 μM leupeptin, and 0.1 mM phenylmethylsulfonyl fluoride and sonicated on ice for 3 bursts of 10 s at 25 W (Vibracell sonicator, Sonics and Materials Inc., Danbury, CT) and then centrifuged at 4°C for 10 min at 1000 rpm (Sigma 3K18; Sigma-Aldrich) to sediment cell nuclei and mitochondrial fractions. The resulting supernatant, containing a crude fraction of plasma membranes, was centrifuged at 12,000 rpm for 55 min at 4°C, and the pellet formed washed with homogenization buffer supplemented with the detergent n-octylglucoside (0.06 mM), sonicated for three bursts of 10 s at 25 W, and incubated on ice for a further period of 15 min. Finally, the suspension was centrifuged at 12,000 rpm for 55 min at 4°C, and the pellet formed was washed with homogenization buffer containing detergent and sonicated for 20 s at 25 W before storage at –70°C pending analysis.
Immunodetection and Quantification. Membrane protein concentration was determined by the method of Lowry. Protein samples were mixed with 4 μl of 1:1 mercaptoethanol/loading buffer [0.5 M Tris (25% v/v), glycerol (20% v/v), 10% sodium dodecyl sulfate (40% v/v), H2O (15% v/v), bromphenol blue (0.005% w/v)] before separation by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (RAMP1 and RAMP2, 20 μg; RAMP3, 80 μg; and CLR, 120 μg of protein per lane) and transfer to polyvinylidene difluoride membrane (0.45 μm; Millipore, Watford, UK). The polyvinylidene difluoride membrane was washed with phosphate-buffered saline (PBS) containing 0.1% (v/v) Tween 20 (Sigma-Aldrich) and blocked overnight in phosphate-buffered saline/0.1% (v/v) Tween 20 solution containing 5% (w/v) Marvel powdered milk (Premier International Foods UK, Lincs, UK). Immunoblotting was performed using primary antibodies directed specifically against rodent CLR (sc-18007 raised in goat; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and rodent RAMP1–3 (sc-11379, sc-11380, and sc-11381 raised in rabbit; Santa Cruz Biotechnology, Inc.) used at a dilution of 1:500 (RAMPs) or 1:200 (CLR). Immunocomplexes were detected using secondary antibodies conjugated to horseradish peroxidase (goat anti-rabbit ab6721 used at a dilution of 1:20,000 (Abcam, Cambridge, UK); donkey anti-goat sc-2020 used at a dilution of 1:40,000 (Santa Cruz Biotechnology, Inc.), and ECL Plus (Amersham Biosciences, Little Chalfont, Buckinghamshire, UK) as substrate and quantified by densitometry (Analytical Imaging System) normalized for protein loading using β-actin (sc-1616 raised in goat; Santa Cruz Biotechnology, Inc.). The specificity of the RAMP and CLR antibodies used to detect the RAMP monomers and CLR protein was confirmed by comparison with immunoblots of membrane samples prepared from Cos7 cells transfected with hRAMP1, hRAMP2, hRAMP3, and hCLR cDNAs (obtained from Dr. David Poyner, School of Life and Health Sciences, Aston University, Birmingham, UK) and rCLR cDNA (obtained from Prof. Walter Born, Laboratory for Orthopaedic Research, University of Zurich, Zurich, Switzerland).
Data Analysis. Data are expressed as means ± S.E. where n denotes number of rats in which systolic blood pressure or heart weight/body weight ratio was measured, or number of heart cell preparations used to analyze mRNA expression, protein levels, or cell dimensions. Statistical analyses were performed by analysis of variance to detect significant differences for between group or within-group effects and post hoc comparisons by Bonferroni or an unpaired Student's t test as appropriate.
Systolic Blood Pressure. Systolic blood pressure was greater (P < 0.05) in rats treated with l-NAME (35 mg/kg/day) at 9 weeks of age onward (i.e., following 1 week of treatment with drug) relative to age-matched control values. The maximum increase was attained at 14 weeks of age and was 80.6 mm Hg greater than the age-matched control value (Fig. 1). Hydralazine (50 mg/kg/day) and HCTZ (7.5 mg/kg/day) together did not influence blood pressure per se but abolished the hypertensive effect of l-NAME.
Heart Weight/Body Weight Ratio. HW/BW ratio tended to increase (P = 0.06) following treatment of 8-week-old animals with 35 mg/kg/day l-NAME for 8 weeks (Fig. 2). This increase (10.0%) was attributed to increased cardiac weight (12.9%; P < 0.05); body weight was not altered. Increases in heart weight and HW/BW ratio (P < 0.05) were abolished by concurrent treatment with hydralazine (50 mg/kg/day) and HCTZ (7.5 mg/kg/day). Treatment with hydralazine and HCTZ also reduced body weight (P < 0.05) by 14.5% in the absence of l-NAME.
Cardiomyocyte Dimensions. The widths of LV and RV cardiomyocytes were increased (P < 0.05) by 16.7 and 20.0%, respectively, following treatment of 8-week-old animals with 35 mg/kg/day l-NAME for 8 weeks (Fig. 3a). These increases were only partly attenuated (P = N.S.) by blood pressure reduction since increases of 10.6 and 10.7% were observed following treatment of 8-week-old animals with 35 mg/kg/day l-NAME for 8 weeks in the presence of hydralazine (50 mg/kg/day) and HCTZ (7.5 mg/kg/day). The lengths of LV and RV cardiomyocytes were not increased significantly (6.1 and 5.4%, respectively) following treatment with l-NAME (Fig. 3b).
Structural and Cardioendocrine Genes Indicative of a Hypertrophic Phenotype. Expression of skeletal α-actin (3.6-fold; Fig. 4, a and b), β-MHC (6.6-fold; Fig. 4, c and d), BNP (3.4-fold; Fig. 5, a and b), ACE (1.8-fold; Fig. 5, c and d), and preproET-1 (1.9-fold; Fig. 5, e and f) mRNAs was increased in LV, but not RV, cardiomyocytes following treatment of 8-week-old animals with 35 mg/kg/day l-NAME for 8 weeks, whereas expression of α-MHC mRNA tended to be reduced (Fig. 4, e and f); expression of c-fos and MLC-2 mRNAs was not altered by treatment in either LV or RV cells (data not shown). Increases observed in LV cardiomyocytes were completely normalized to control values by concurrent administration of blood pressure-lowering agents.
Adrenomedullin, Intermedin, and Their Receptor Component Genes. Expression of preproAM (1.6-fold; Fig. 6a) mRNA was increased in LV, but not RV (Fig. 6b), cardiomyocytes following treatment of 8-week-old animals with 35 mg/kg/day l-NAME for 8 weeks; expression of preproIMD mRNA was increased in both ventricles (8.4- and 8.6-fold; Fig. 6, c and d). Increased expression in LV cardiomyocytes of preproAM mRNA was normalized to control values (P < 0.05) by concurrent administration of blood pressure-lowering agents; although some reduction in preproIMD mRNA expression was evident in RV, expression was not normalized to control values in either ventricle. Expression of CL receptor (3.4-fold; Fig. 7, a and b), RAMP1 (4.1-fold; Fig. 7, c and d), RAMP2 (2.8-fold; Fig. 7, e and f), and RAMP3 (4.4-fold; Fig. 7, g and h) mRNAs was also increased (P < 0.01) in LV, but not RV, by l-NAME treatment, whereas expression of RDC1, L1, and RCP mRNAs was not changed in either ventricle (data not shown). Increased expression of CL receptor and RAMP2 and RAMP3 mRNAs was completely normalized to control values by concurrent administration of blood pressure-lowering agents, whereas that of RAMP1 mRNA was not attenuated.
At protein level (Fig. 8, a–c), treatment with l-NAME increased expression of the RAMP3 monomer (2.5-fold) and to a lesser extent RAMP1 monomer (1.6-fold) and RAMP2 monomer (1.5-fold) within LV cardiomyocytes. Enhanced expression of the RAMP2 and RAMP3 proteins was abolished by concurrent administration of blood pressure-lowering agents, whereas that of RAMP1 was not attenuated. The identity of each RAMP monomer was confirmed by use of molecular weight standards and by comparison with immunoblots of membranes prepared from Cos7 cells transfected with the respective hRAMP cDNAs and from sham-transfected Cos7 cells (Fig. 9, a–c). Although additional bands were observed corresponding to proteins of larger molecular mass (>30 kDa) in both cardiomyocytes and transfected Cos7 cells, these bands were also evident in sham-transfected Cos7 cells, indicating that the antibodies might also recognize discontinuous epitopes present on other unrelated proteins of larger size. All attempts at characterizing these additional bands suggested that they were artifacts and were unlikely to be related to RAMPs or to represent CLR-RAMP complexes or RAMP “homodimers” as described previously by other laboratories (Cueille et al., 2002).
At protein level (Fig. 8d), treatment with l-NAME tended to increase expression of CLR (1.4-fold) within LV cardiomyocytes, but increases were not statistically significant. The identity of the CLR protein (∼42 kDa) was confirmed by comparison with immunoblots of membranes prepared from Cos7 cells transfected with hCLR or rCLR cDNAs and from sham-transfected Cos7 cells (Fig. 9d). Although an additional band was observed corresponding to a protein of larger molecular mass (∼50 kDa) in both transfected Cos7 cells and, less abundantly, cardiomyocytes, this band was also detected in sham-transfected Cos7 cells, indicating that the antibody used might also recognize a discontinuous epitope present on another unrelated protein (∼50 kDa).
Anatomical hypertrophy was evident following treatment with l-NAME (35 mg/kg/day) as indicated by elevated HW/BW. However, there were notable dose-dependent observations in relation to previous findings (Bell et al., 2006) since heart weight was elevated (in contrast to the lack of increase at a lower dose of 20 mg/kg/day), whereas body weight was maintained (in contrast to marked reduction, combined with increased morbidity and mortality, observed at 50 mg/kg/day). A dose intermediate between 20 and 50 mg/kg/day would seem to be well tolerated and optimal for establishing a robust and stable model of cardiac remodeling in long-term NO deficiency.
Changes at organ level reflect a combination of cardiomyocyte hypertrophy and necrosis, vascular remodeling, proliferation of nonmyocytes, and deposition of extracellular matrix protein. Marked attenuation of increased cardiac mass by treatment with hydralazine together with hydrochlorothiazide indicates that many of these processes are likely to be initiated in response to pressure loading of the heart. Increased width of both LV and RV cardiomyocytes is consistent with previous reports of enhanced protein synthesis and mass of cardiomyocytes from both chambers (Bell et al., 2006), even though the right ventricle is not subjected to the same pressure loading as the left since long-term administration of l-NAME, although causing remarkable hypertension in the systemic vascular bed, does not enhance mean pulmonary arterial pressure or alter pulmonary vascular morphology (Hampl et al., 1993). Incomplete attenuation of l-NAME induced increases in cell width by blood pressure reduction indicates a pressure-independent component to the pathogenesis of cardiomyocyte hypertrophy in this model, which is likely to reflect NO deficiency within the myocardium since NO itself is known to exert direct inhibitory effects on cell growth and proliferation (Sarkar et al., 1995; Matsuoka et al., 1996; Bartunek et al., 2000). Decreased local NO production might also compromise blood flow within the myocardium, thereby promoting ischemia and contributing indirectly to cardiomyocyte hypertrophy and necrosis (De Oliveira et al., 1999).
Phenotypic alterations manifested in LV, but not RV, cardiomyocytes, namely, enhanced expression of the structural proteins β-MHC and sk-α-actin and activation of cardioendocrine pathways, characterized by increased expression of BNP, ACE, and ET-1 mRNAs, were ameliorated by blood pressure reduction, indicating that these mechanisms are recruited exclusively in response to mechanical stress upon cardiomyocytes because of pressure loading caused by profound hypertension within the systemic vasculature. Recruitment of cardioendocrine signaling pathways might be expected to exert a modifying effect upon systemic vascular tone, affecting preload and afterload, in addition to influencing the coronary circulation and elaboration of cardiomyocyte hypertrophy directly.
PreproIMD mRNA was expressed under normal conditions in both LV and RV cardiomyocytes, less abundantly than preproAM mRNA. However, expression of preproIMD mRNA was increased markedly in both ventricles in response to NO deficiency following administration of l-NAME, and in contrast to preproAM mRNA, which was elevated within LV cardiomyocytes only, was not normalized by blood pressure reduction. These data indicate divergence in regard to mechanisms of recruitment of these cardioendocrine peptides within the cardiomyocyte. Activation of AM within the LV is initiated primarily by systemic hypertension; indeed, increased plasma levels of AM have been detected in hypertension (Sumimoto et al., 1997), and release of the peptide from cardiomyocytes has been demonstrated in vitro in response to mechanical stretch (Tsurada et al., 2000). Plasma levels have been found to correlate better with LV mass than with blood pressure; augmented release of AM may therefore represent a counter-regulatory response to LV hypertrophy rather than hypertension per se (Sumimoto et al., 1997). In contrast, synthesis of intermedin is largely attributable to a pressure-independent mechanism; demonstration of a cardioprotective effect of intermedin against ischemia-reperfusion injury (Taylor et al., 2005) would support the hypothesis that this peptide is released from both ventricles in response to hypoxia. Indeed, cardiac ischemia and necrosis are often reported in this model of NO deficiency (Takaori et al., 1997; Bernatova et al., 1999).
In agreement with previous data obtained following administration of a lower dose (20 mg/kg/day) (Bell et al., 2006), RAMP2, RAMP3, and CLR mRNAs were all increased in LV but not RV cardiomyocytes by l-NAME (35 mg/kg/day); these further increases in expression correlated with additional increases in systolic blood pressure at the higher dose. This pattern of mRNA expression, together with the corresponding changes manifested at protein level, was completely normalized by blood pressure reduction. Since RAMP2 and RAMP3 both constitute AM receptors when in combination with CL receptor protein (Hay et al., 2004), increased abundance of these receptor complexes would be expected to act in tandem with increased expression of AM itself to augment the influence of this counter-regulatory signaling cascade in regulating myocardial contractility (Ikenouchi et al., 1997; Ihara et al., 2000) and opposing the pathogenesis of cardiomyocyte hypertrophy (Sato et al., 1997; Tsurada et al., 2000; Bell et al., 2006). RAMP3, normally expressed much less abundantly than RAMP2 within the cardiomyocyte (Autelitano and Ridings, 2001), was up-regulated to the greater extent. Such changes in relative abundance of these RAMPs have implications for affinity and selectivity of mediator-receptor interactions, since AM2 receptors are less selective for AM over other peptides of this family than AM1 receptors (McLatchie, 1998).
In contrast, LV cardiomyocyte expression of RAMP1 mRNA, which was only slightly elevated following administration of 20 mg/kg/day l-NAME despite a notable increase of systolic BP (Bell et al., 2006), was markedly elevated, to a similar extent as that of RAMP3 mRNA at the higher dose used in the present study, and this was not normalized by BP reduction. These data indicate that myocardial expression of the RAMP1 gene is largely independent of the regulatory influence of pressure loading. RAMP1, when in combination with the CL receptor protein, constitutes a CGRP-responsive receptor complex (Hay et al., 2004) at which AM also has moderate affinity (McLatchie, 1998). The vasorelaxant peptide CGRP, found predominantly as a neurotransmitter localized in the heart within the sensory innervation comprising capsaicin-sensitive A(δ) and C-fiber afferent nerves (Mulderry et al., 1985), often in association with the coronary vasculature from which nerve fibers extend into the myocardium, forming a nerve plexus that is particularly well developed within papillary muscles, exerts multiple effects on the cardiomyocyte in vitro, including positive inotropy (Bell and McDermott, 1994), initiation of a hypertrophic phenotype (Bell et al., 1995; Bell et al., 1997), and direct cardioprotective effects against ischemic injury (Li et al., 1996). Neuronal release of CGRP is enhanced in response to myocardial ischemia (Franco-Cereceda et al., 1987). Enhanced expression of the RAMP1 protein would increase the probability of interaction of this neuronally released CGRP with its receptor, thereby augmenting cellular responsiveness. In addition to possible beneficial effects on coronary hemodynamics, localized release of CGRP could also serve to counter the cardiac ischemia and necrosis associated with NO deficiency, and also to promote compensatory hypertrophy and improve contractility of surviving myocardium. It is interesting to note in this regard that cardiomyocyte hypertrophy is also initiated in vitro in response to oxidative stress (Sabri et al., 1998; Siwik et al., 1999).
Expression of L1, which is normally found at very low levels within cardiomyocytes (Autelitano, 1998), of RDC-1 (Li et al., 1996), expressed abundantly in the myocardium, and of the RCP protein, which moderates the efficacy of coupling of the RAMP/CLR complex to G protein activation (Prado et al., 2001), was not influenced by l-NAME administration, indicating that altered abundance of these proteins within the cardiomyocyte is unlikely to participate in the pathophysiological changes induced by NO deficiency. This supports the conclusion of Poyner et al. (2002) that LI and RDC-1 do not represent receptors for the CGRP/AM family, but rather are orphan GPCRs for which the true ligands remain to be identified.
Intermedin interacts nonselectively both at AM receptors, formed by the association of RAMP2 or -3 with CLR, and at CGRP receptors formed by the association of RAMP1 with CL (Poyner et al., 2002; Roh et al., 2004). Although enhanced expression of preproIMD mRNA is not accompanied by enhanced expression of receptor component genes in RV, in contrast to LV, some increase in activity of intermedin signaling pathways might also be anticipated as a consequence of increased peptide levels. Although the functional significance of this peptide is poorly understood at present, the robust increase in intermedin expression indicates an important role in the cardiac pathology resulting from NO deficiency, but, unlike AM, intermedin is probably not associated primarily with pressure overload upon the myocardium. The concordance of intermedin and RAMP1 up-regulation in LV supports a CGRP-type receptor action; considering also the relative unresponsiveness to BP reduction, intermedin may, like CGRP, serve primarily an anti-ischemic, cardioprotective function, although an additional attenuating influence for this peptide on cardiomyocyte hypertrophy, mediated by stimulation of the AM2 receptor, cannot be discounted, particularly since RAMP3 expression is also enhanced. Expression of intermedin and RAMP1 mRNAs is likely to be a dynamic process, tightly regulated in response to prevailing levels of myocardial ischemia, and extent of ongoing cell necrosis and apoptosis, and requirement for remodeling of adjacent healthy myocardium.
We thank Dr. Maria Gerova (Slovak Academy of Sciences) for helpful discussions regarding the l-NAME model and Dr. David Poyner and Prof. Walter Born for generous provision of CLR and RAMP cDNAs to enable us to confirm the identity of the RAMP monomers and CLR protein.
- Received July 18, 2005.
- Accepted November 30, 2005.
This study was funded by a Grant FS: 2001030 (to D.B. and B.J.M.) from the British Heart Foundation.
ABBREVIATIONS: IMD, intermedin; CGRP, calcitionin gene-related peptide; AM, adrenomedullin; LVH, left ventricular hypertrophy; CLR, calcitonin receptor-like receptor; RAMP, receptor activity modifying protein; RCP, receptor component protein; NOS nitric-oxide synthase; l-NAME, Nω-nitro-l-arginine methyl ester; LV, left ventricular; RV, right ventricular; RT-PCR, reverse transcription-polymerase chain reaction; Ct, threshold cycle; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HW, heart weight; BW, body weight; hRAMP, human receptor activity modifying protein; MHC, myosin heavy chain; ACE, angiotensin-converting enzyme; ET, endothelin; CL, calcitonin receptor-like; BNP, brain natriuretic peptide; HCTZ, hydrochlorothiazide; hCLR, human calcitonin receptor-like receptor; rCLR, rat calcitonin receptor-like receptor; bp, base pair(s).
- The American Society for Pharmacology and Experimental Therapeutics