Increased susceptibility to heart failure in response to volume overload in mice lacking natriuretic peptide receptor-A gene

Abstract

Object: Contribution of the natriuretic peptide system to the development of heart failure (HF) in vivo was examined using mice lacking or having decreased natriuretic peptide receptor-A (NPRA), a guanylyl cyclase-linked receptor.

Methods: Volume-overloaded HF was produced by aortocaval fistula in mice with wild-type (+/+), heterozygous (+/−), and homozygous null mutants (−/−) of the NPRA gene. Severity of HF was assessed 4 weeks after operation on the basis of organ weight, hemodynamics, echocardiographic indices, urinary variables, neurohumoral factors, and myocardial gene expression.

Results: There were no significant differences in lung weight, kidney weight, left ventricular end-diastolic pressure (LVEDP), left ventricular systolic function, or urinary variables among the three sham-operated groups; however, sham-operated (−/−) mice had higher blood pressure and individual cardiac chamber weights than did (+/+) mice. In contrast, (−/−) mice with aortocaval fistula had higher LVEDP, left and right ventricular weights, lung weight, and left ventricular dimension, as well as lower fractional shortening and urinary sodium and cyclic guanosine monophosphate (cGMP) excretion than did (+/+) mice with aortocaval fistula. In addition, ventricular mRNA expression of natriuretic peptides and β-myosin heavy chain increased markedly only in (−/−) mice. Plasma atrial natriuretic peptide, renin, and aldosterone, but not cGMP, showed greater responses to aortocaval fistula in (−/−) mice than in (+/+) mice. Both sham-operated and aortocaval fistula NPRA (+/−) mice almost consistently showed a phenotype intermediate between those of NPRA (−/−) and NPRA (+/+) mice.

Conclusion: These results provide genetic evidence that NPRA signaling protects against HF induced by volume overload in mice.

Atrial natriuretic peptide, Natriuretic peptide receptor-A, Heart failure, Renin–aldosterone system, Natriuresis

1. Introduction

Many neurohumoral factors are involved in the pathophysiology of heart failure (HF). Both vasodilatory factors as well as vasoconstrictory factors are activated in this syndrome [1,2]. The status of HF depends largely on the imbalance between these two systems. Indeed, intravenous administration of natriuretic and vasodilator peptides is effective for the treatment of acute HF [3,4], whereas antagonists of antidiuretic and vasoconstricting hormones are effective for the treatment of chronic HF [5,6]. Despite advances in treatment, mortality remains very high. An improved outcome seems to require a better understanding of the molecular mechanisms responsible for neurohumoral dysregulation in HF.

Recent progress in genetics and molecular biology has culminated in the identification of several candidate genes associated with cardiovascular diseases; however, a few studies have attempted to identify candidate genes associated with HF [7]. Since the natriuretic peptide system is activated in HF and considered to physiologically compensate for HF [8,9], abnormality of the natriuretic peptide receptor-A (NPRA) gene may predispose individuals to HF. A recent study has shown that a deletion mutation in the promoter region of the NPRA gene is associated with left ventricular hypertrophy (LVH) in hypertension [10]. However, whether the mutation of the NPRA gene predisposes to HF remains unknown. To test a hypothesis that the lack or decreased NPRA contributes to the susceptibility to HF, we used a genetic approach in the mouse. We previously showed that ANP-induced cGMP responses depend on the number of NPRA gene [11] and that the lack of NPRA results in LVH [12]. In the present study, we produced a model of HF in mice with wild-type (+/+), heterozygous (+/−), and homozygous null mutants (−/−) of the NPRA gene and evaluated the severity of HF on the basis of organ weight, hemodynamics, urinary variables, echocardiographic indices, neurohumoral factors, and myocardial gene expressions.

2. Methods

The investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the US National Institute of Health and all experiments were carried out in accordance with protocols approved by the Institutional Animal Care and Use Committees of the University of North Carolina.

2.1. Experimental mice

The NPRA (−/−) (n=30), (+/−) (n=28), and (+/+) (n=30) mice used in this study were the N6 or N7 generation of mice backcrossed to C57BL/6, derived from the original mutants as described previously [11]. The extreme cardiac hypertrophy and sudden death occurring in F2 generation males were no longer present in this particular stock of mutants, presumably because of genetic drifts and loss of modifying loci [13,14]. Because blood pressure (BP) and heart weight/body weight (BW) ratios were similar in males and females, both sexes were used in this experiment.

2.2. Creation of aortocaval fistula

Mice underwent aortocaval fistula surgery at 20–40 weeks of age, as previously described [15] with slight modifications adopting the method for producing aortocaval fistula in rat [16,17]. Control animals were operated on in an identical fashion, but a fistula was not established.

2.3. Tail–cuff BP analysis

BP and pulse rate (PR) were measured noninvasively in conscious and restrained mice as described previously [18], before and at 4 weeks after the operation. In brief, BP measurements were made by a single individual between 1 and 5 PM daily on 6 consecutive days with three sessions of 10 measurements (total of 180 measurement). Sessions that gave at least 6 successful BP out of the 10 measurements were included in obtaining the mean BP of individual animals. This excluded about 1% of the sessions.

2.4. Urinalysis

Mice were maintained on a 12-h light/dark cycle and given water and food ad libitum as previously reported [19]. BW, water and food intake, and urinary excretion were measured every 24 h for 3 days. Electrolytes were measured (VT250 Chemical Analyzer) and the values were averaged.

2.5. Echocardiography

Echocardiograms were obtained in gently restrained, conscious mice at 2 and 4 weeks after surgery, as reported previously [20]. Two-dimensional guided M-mode echocardiography was performed with the use of HDI 5000 echocardiograph equipment and a 7.5-MHz transducer [20]. We performed at least five measurements and averaged the values. Left ventricular (LV) end-diastolic diameter (LVDd), LV end-systolic diameter (LVDs), and LV end-diastolic anterior wall thickness (AWT) and posterior wall thickness (PWT) were measured and fractional shortening and relative wall thickness were calculated as follows: fractional shortening=(LVDd−LVDs)/LVDd; relative wall thickness=(2 × PWT)/LVDd.

2.6. Hemodynamic evaluation

Mice were anesthetized with isoflurane and placed in the supine position on warm pads under an operating microscope. A 1.4-French high-fidelity micromanometer catheter (Millar Instruments, Houston, TX) was inserted into the right carotid artery. LV end-diastolic pressure (LVEDP), mean arterial pressure (MAP), and heart rate were recorded under bilateral vagotomy (to limit counter-regulatory autonomic reflexes), as described previously [11]. After hemodynamic parameters were measured, a micromanometer catheter was withdrawn and polyethylene cannula was inserted and 15.5-μm colored microspheres were injected into the aortic arch via the indwelling cannula. At the end of the experiment, individual chambers of the heart, lungs, and kidneys were rapidly excised from each animal, weighed, and frozen in liquid nitrogen.

2.7. Plasma hormone concentrations

Arterial blood samples from mice were rapidly withdrawn as previously described [14,20]. The plasma renin concentration (PRC) was measured by a Gamma coat plasma renin activity kit (Dade Behring, Tokyo, Japan) and determined by the RIA of angiotensin I generated by the incubation of plasma after adding an excess of angiotensinogen provided by bi-nephrectomized rat plasma [21]. The plasma aldosterone concentration was measured by RIA using a SPAC-S aldosterone kit (Daiichi Radioisotope Labs, Tokyo, Japan) [22]. Plasma and urinary cGMP were measured using RIA kits (cyclic GMP assay kit, Yamasa Shoyu, Chiba, Japan) [23]. ANP levels were determined by radioimmunoassay after extraction of plasma using a Sep-Pak C-18 cartridge as previously described [23].

2.8. Quantification of shunt size

We semiquantified shunt severity by injecting 15.5-μm colored microspheres as previously reported method [24]. In brief, blue 15 ± 0.43 μm colored microspheres (Triton Technology) were diluted to a final concentration of 300,000 microspheres/mL in normal saline containing 0.01% Tween 80. After hemodynamic measurement was finished, a micromanometer catheter was withdrawn and polyethylene cannula was inserted. Microspheres (100 μL) was injected into the aortic arch via the polyethylene cannula in the left common carotid artery. After euthanasia, the lungs were removed, weighed, and processed for microsphere recovery per the manufacturer's directions. The mice who had no cardiac hypertrophy or had <5% microsphere recovery in the lungs were excluded from analysis.

2.9. Quantification of mRNA expression

The gene expression levels of ANP, brain natriuretic peptide (BNP), alpha-myosin heavy chain (MHC), β-MHC, and glyceraldehyde-3-phosphate dehydrogenase in the LV tissue were determined by real-time quantitative reverse-transcription polymerase chain reaction with the use of ABI 7700 and specific primers as reported previously [25].

2.10. Statistical analysis

Data are expressed as means ± S.E.M. Statistical analysis was done by analysis of variance (ANOVA), using Fisher's post hoc test for multiple comparisons. Time-dependent changes were analyzed by MANOVA with repeated measures. Correlation coefficients were calculated using linear regression analysis. Differences were considered statistically significant at P<0.05. Statistical analysis was performed with the use of STATVIEW version 5 (Abacus Concepts, Berkeley, CA) or a JMP software package (SAS Institute, Cary, NC).

3. Results

3.1. Response of organ weight, BP, and PR in mice with each genotype to aortocaval fistula

Physiological profiles of the three genotypes in the sham operation and aortocaval fistula groups are shown in Table 1 and Fig. 1. In the sham operation groups, BW, lung weight/BW, kidney weight/BW, and PR were similar for all three genotypes. However, NPRA (−/−) sham-operated mice had significantly higher BP and weights of individual chambers of the heart than did NPRA (+/−) and (+/+) sham-operated mice. Creation of an aortocaval fistula did not change BW, kidney weight/BW or BP, but significantly increased LV weight/BW and right ventricular weight/BW in all three groups of mice. The average shunt size was not different among the three aortocaval fistula groups ((+/+): 25 ± 7%; (+/−): 27 ± 8%; (−/−): 24 ± 8%; NS). Aortocaval fistula increased lung weight/BW, right atrial weight/BW, and left atrial weight/BW and decreased HR only in NPRA (−/−) mice. Thus, the lack of NPRA not only exacerbated cardiac hypertrophy but also signatures of failing hearts. NPRA (+/−) mice showed intermediate response, but the differences between NPRA (+/−) and NPRA (+/+) mice were not significant.

Fig. 1

Lung weight (W)/BW and LVW/BW in NPRA (−/−), (+/−), and (+/+) mice undergoing sham operation (sham) or induction of heart failure (HF) by aortocaval fistula. (A) Lung W/BW and (B) LVW/BW of sham NPRA (+/+) mice, (+/−) mice, and (−/−) mice, and HF NPRA (+/+) mice, (+/−) mice, and (−/−) mice. **P<0.01 vs. corresponding wild-type control, ^††P<0.01 vs. corresponding heterozygote, ^##P<0.01 vs. genotype-matched sham.

Table 1

Response of BW, cardiac chamber weight, kidney weight, BP, and PR in mice with each genotype to aortocaval fistula

	Genotype (n)	BW, g	RVW/BW, mg/g	LAW/BW, μg/g	RAW/BW, μg/g	Kidney/BW, mg/g	BP, mm Hg	PR, Beats/min
Sham	+/+ (15)	25.8 ± 1.2	0.76 ± 0.04	121 ± 3	183 ± 14	7.11 ± 0.39	112 ± 3	674 ± 8
	+/− (12)	29.3 ± 1.0	0.82 ± 0.04	140 ± 7	176 ± 11	7.27 ± 0.21	128 ± 2*	669 ± 9
	−/− (13)	28.8 ± 1.5	1.22 ± 0.06**,‡	192 ± 18*,†	251 ± 20*,†	7.47 ± 0.19	134 ± 2**,‡	662 ± 10
HF	+/+ (15)	27.6 ± 0.9	0.93 ± 0.05#	160 ± 11	209 ± 9	7.42 ± 0.19	107 ± 3	666 ± 10
	+/− (16)	26.1 ± 1.3	1.02 ± 0.06#	170 ± 12	212 ± 14	7.43 ± 0.30	118 ± 3†	666 ± 10
	−/− (17)	26.7 ± 1.0	1.47 ± 0.08**,‡,##	297 ± 33**,‡,##	391 ± 44**,‡,##	7.97 ± 0.30	141 ± 3**,‡	639 ± 8*,†,#

BW, body weight; RVW, right ventricular weight; LAW, left atrial weight; RAW, right atrial weight; BP, tail–cuff blood pressure; PR, pulse rate; and HF, heart failure.

*

P<0.05 vs. corresponding wild-type control.

**

P<0.01 vs. corresponding wild-type control.

‡

P<0.01 vs. corresponding heterozygote.

†

P<0.05 vs. corresponding heterozygote.

#

P<0.05 vs. genotype-matched sham.

##

P<0.01 vs. genotype-matched sham.

3.2. Hemodynamic responses to aortocaval shunt in NPRA (−/−), (+/−), and (+/+) mice

While there were no significant changes in LVEDP in the three sham-operated groups, sham-operated NPRA (−/−) mice had significantly higher MAP than (+/−) or (+/+) mice (Fig. 2). The sham-operated NPRA (+/−) mice also had a slight but significantly higher MAP than NPRA (+/+) sham-operated mice. Creation of an aortocaval fistula did not affect MAP in any group, but significantly increased LVEDP in all three groups. The increase in LVEDP was greater in the NPRA (−/−) mice than in the other two groups. Thus, LV dysfunction was also exacerbated in the NPRA (−/−) mice.

Fig. 2

LVEDP and MAP in NPRA (−/−), (+/−), and (+/+) mice undergoing sham or induction of HF. (A) LVEDP and (B) MAP of sham NPRA (+/+) mice, (+/−) mice, and (−/−) mice, and HF NPRA (+/+) mice, (+/−) mice, and (−/−) mice. *P<0.05 vs. corresponding wild-type control, **P<0.01 vs. corresponding wild-type control, ^††P<0.01 vs. corresponding heterozygote, ^##P<0.01 vs. genotype-matched sham.

3.3. Echocardiographic findings in NPRA (−/−), (+/−), and (+/+) of HF mice

The development of LV dysfunction in the individual animals was monitored using echocardiography (Fig. 3). Before operation, there were no significant differences in LVDd or fractional shortening among the three groups. LVDd increased 2 weeks after operation and further increased 4 weeks after operation in all groups. At 4 weeks, LVDd was significantly greater in NPRA (−/−) mice than in the other two groups. In contrast, fractional shortening decreased 4 weeks after the aortocaval fistula operation only in NPRA (−/−) mice. Consequently, fractional shortening was significantly lower in NPRA (−/−) mice than in the other two groups. Fractional shortening at 2 weeks and 4 weeks after operation significantly correlated with LVDd (2 weeks: r=−0.71, P<0.0001; 4 weeks: r=−0.73, P<0.0001). NPRA (−/−) mice had larger AWT and PWT than did NPRA (+/−) and (+/+) mice before operation (AWT: (+/+): 0.63 ± 0.06; (+/−): 0.78 ± 0.06; (−/−): 0.87 ± 0.08 mm; PWT: (+/+): 0.60 ± 0.03; (+/−): 0.73 ± 0.08; (−/−): 0.83 ± 0.08 mm). Creation of an aortocaval fistula significantly increased AWT and PWT in all three groups of mice at 2 weeks after operation (AWT: (+/+): 0.68 ± 0.06; (+/−): 0.82 ± 0.06; (−/−): 0.90 ± 0.09; PWT: (+/+): 0.65 ± 0.05; (+/−): 0.79 ± 0.09; (−/−): 0.85 ± 0.09) and further increased them at 4 weeks after operation (AWT: (+/+): 0.75 ± 0.07; (+/−): 0.84 ± 0.06; (−/−): 0.98 ± 0.08; PWT: (+/+): 0.69 ± 0.05; (+/−): 0.82 ± 0.07; (−/−): 0.89 ± 0.09). NPRA (−/−) and (+/−) mice had higher relative wall thickness than NPRA (+/+) mice before operation (relative wall thickness: (+/+): 0.39 ± 0.01; (+/−): 0.45 ± 0.03; (−/−): 0.46 ± 0.02), suggesting that NPRA (−/−) and (+/−) mice have concentric hypertrophy. Creation of an aortocaval fistula significantly decreased relative wall thickness in all phenotypes, but NPRA (−/−) and (+/−) mice still had higher relative wall thickness than NPRA (+/+) mice. As a result, LV geometry was transferred from normal pattern to eccentric hypertrophy in NPRA (+/+) mice, whereas LV geometry was transferred from concentric hypertrophy to borderline of concentric and eccentric hypertrophy in NPRA (+/−) and (−/−) mice (data not shown).

Fig. 3

Time course of LV dimension and LV systolic function before and after induction of HF in NPRA (−/−), (+/−), and (+/+) mice. (A) LVDd and (B) fractional shortening (FS) before operation and at 2 weeks and 4 weeks after aortocaval fistula operation. Open circles indicate HF in (−/−) mice; open triangles, HF in (+/−) mice; open squares, HF in (+/+) mice. ^#P<0.05 as compared with the value before operation. *P<0.05 as compared with (+/+) mice, ^†P<0.05 as compared with (+/−) mice.

3.4. Response of cardiac gene expressions to aortocaval shunt in NPRA (−/−), (+/−), and (+/+) mice

We next examined the mRNA expression of genes in LV tissue, which is a characteristic of HF (Fig. 4). In the sham-operated groups, mRNA expressions of ANP and BNP, and β-MHC/alpha-MHC ratio were higher and the mRNA expression of NPRA was lower in NPRA (−/−) mice than in the other two groups. The mRNA level of NPRA was dose-dependent, as expected. Creation of an aortocaval fistula markedly increased mRNA expressions of ANP and BNP and β-MHC/alpha-MHC ratio in NPRA (−/−) mice. These increases were significantly greater in NPRA (−/−) mice than in the other two groups. Creation of an aortocaval fistula did not change the mRNA expression of NPRA in any group. Thus, NPRA signaling pathway inhibits the cardiac fetal gene expression both in the basal and HF condition.

Fig. 4

Myocardial gene expression levels in NPRA (−/−), (+/−), and (+/+) mice undergoing sham or induction of HF. (A) Gene expression of ANP, (B) gene expression of BNP, (C) gene expression of β-MHC/alpha-MHC ratio, and (D) gene expression of NPRA in sham NPRA (+/+) mice, (+/−) mice, and (−/−) mice, and HF NPRA (+/+) mice, (+/−) mice, and (−/−) mice. *P<0.05 vs. corresponding wild-type control, **P<0.01 vs. corresponding wild-type control, ^†P<0.05 vs. corresponding heterozygote, ^††P<0.01 vs. corresponding heterozygote, ^##P<0.01 vs. genotype-matched sham.

3.5. Response of urinary parameters to aortocaval shunt in NPRA (−/−), (+/−), and (+/+) mice

To examine the importance of NPRA signaling on the body fluid retention in HF, we analyzed urinary parameters. Urinary variables in the three genotypes of mice 4 weeks after sham operation or aortocaval fistula creation are shown in Fig. 5. Urinary sodium excretion (U_NaV), urine volume (UV), and urinary chloride excretion (U_ClV) were similar for all three genotypes in sham-operated mice, whereas urinary cGMP excretion (U_cGMPV) was significantly lower in NPRA (−/−) mice than in NPRA (+/+) mice. UV and U_ClV after aortocaval fistula did not differ significantly among the three genotypes. However, creation of an aortocaval fistula reduced U_NaV only in NPRA (−/−) mice. U_NaV was therefore significantly lower in NPRA (−/−) mice than in the other two groups. In contrast, aortocaval fistula significantly increased U_cGMPV both in NPRA (+/+) and NPRA (+/−) mice, but not in NPRA (−/−) mice. Consequently, U_cGMPV was significantly lower in NPRA (−/−) mice than in NPRA (+/+) and NPRA (+/−) mice. Thus, lack of communication between the heart and kidney by NPRA signaling pathway appears to contribute to the more severe body fluid retention in NPRA (−/−) HF mice.

Fig. 5

Urinary parameters in NPRA (−/−), (+/−), and (+/+) mice undergoing sham or induction of HF. (A) UNaV, (B) UcGMPV, (C) UV, and (D) UClV of sham NPRA (+/+) mice, (+/−) mice, and (−/−) mice, and HF NPRA (+/+) mice, (+/−) mice, and (−/−) mice. *P<0.05 vs. corresponding wild-type control, **P<0.01 vs. corresponding wild-type control, ^†P<0.05 vs. corresponding heterozygote, ^††P<0.01 vs. corresponding heterozygote, ^#P<0.05 vs. genotype-matched sham, ^##P<0.01 vs. genotype-matched sham.

3.6. Response of neurohumoral factors to aortocaval shunt in NPRA (−/−), (+/−), and (+/+) mice

In order to examine the mechanism of exacerbation of HF in NPRA (−/−) mice, we next analyzed neurohumoral factors in three genotypes of sham-operated and aortocaval fistula mice (Fig. 6). In sham-operated mice, NPRA (−/−) had higher ANP and lower PRC, aldosterone, and cGMP levels than NPRA (+/+). There were no differences in plasma ANP, cGMP or PRC between NPRA (+/−) and (+/+) mice, although NPRA (+/−) had lower aldosterone levels than NPRA (+/+) mice.

Fig. 6

Plasma neurohumoral factor levels in NPRA (−/−), (+/−), and (+/+) mice undergoing sham or induction of HF. (A) Plasma ANP, (B) plasma cGMP, (C) PRC, and (D) plasma aldosterone of sham NPRA (+/+) mice, (+/−) mice, and (−/−) mice, and HF NPRA (+/+) mice, (+/−) mice, and (−/−) mice. *P<0.05 vs. corresponding wild-type control, **P<0.01 vs. corresponding wild-type control, ^†P<0.05 vs. corresponding heterozygote, ^††P<0.01 vs. corresponding heterozygote, ^##P<0.01 vs. genotype-matched sham.

Creation of aortocaval fistula significantly increased plasma ANP and PRC in all three genotypes. In aortocaval fistula mice, NPRA (−/−) still had higher plasma ANP levels than the other two groups. In contrast, NPRA (−/−) mice exhibited exaggerated increase of PRC of the in response to volume overload and, although their PRC levels were still slightly lower than those in other two groups, the difference was no longer significant. Aortocaval fistula increased plasma cGMP levels in NPRA (+/−) and NPRA (+/+) mice, but not in NPRA (−/−), and the cGMP levels in aortocaval fistula NPRA (−/−) mice remained lower than the other two groups. In contrast, creation of an aortocaval fistula increased plasma aldosterone only in NPRA (−/−). Consequently, there was no difference in plasma aldosterone levels between NPRA (−/−) and NPRA (+/+) mice with aortocaval fistula. Plasma levels of cGMP and renin in the NPRA (+/−) mice were intermediate between NPRA (−/−) and NPRA (+/+) mice, but plasma aldosterone levels in (+/−) mice were significantly lower than both (−/−) and (+/+) mice. Thus, a failure of controlling the renin–angiotensin–aldosterone activation due to a lack of NPRA–cGMP signaling pathway may be contributing to the exacerbation of HF in NPRA (−/−) mice.

4. Discussion

In the present study, sham-operated NPRA (−/−) mice were characterized by higher BP, left and right ventricular weights, and left and right atrial weights as compared with wild type mice. These results are consistent with our previous findings showing that the NPRA system is involved in the regulation of BP systemically and the inhibition of cardiac hypertrophy locally [11,13]. We also showed that sham-operated NPRA (−/−) mice had a higher plasma ANP level and lower cGMP, PRC, and aldosterone levels than NPRA (+/+) mice. Furthermore, sham-operated NPRA (−/−) mice had higher mRNA expressions of ANP and BNP and a higher β-MHC/alpha-MHC ratio than NPRA (+/+) mice. In contrast, there were no differences in urinary variables between these two groups. These findings suggest that BP in NPRA (−/−) mice appears to be increased to maintain urinary volume and sodium homeostasis via the mechanism of pressure-natriuresis, accompanied by inhibition of the renin–angiotensin–aldosterone system.

In the present study we applied a model of volume overloaded HF in NPRA (−/−), (+/−), and (+/+) mice by creating an aortocaval fistulae which has been reported to produce overt congestive HF with impaired hemodynamics in mice [15]. We found that aortocaval fistulae in NPRA (+/+) mice induced characteristic features of HF, including increased left and right ventricular weights, LVEDP, plasma ANP, cGMP, PRC, LV dimension, and mRNA expression of ANP as compared with sham-operated mice. Importantly, we also found that aortocaval fistulae in NPRA (−/−) mice further increased left and right ventricular and atrial weights, accompanied by marked increases in expression of the marker genes such as ANP, BNP, and β-MHC in the left ventricle. Thus the volume overload worsened LV systolic function and further increased LVEDP in NPRA (−/−) mice as compared with the NPRA (+/−) or with NPRA (+/+) mice. The marked exacerbation of HF in the NPRA (−/−), despite that these mice have all other compensatory mechanisms intact, demonstrates that the natriuretic peptide system directly protects against progression of HF and LV dysfunction. A recent study suggests that mice with cardiomyocyte-restricted inactivation of the NPRA gene have cardiac dysfunction [26], suggesting a protective role of cardiac NPRA in the maintenance of cardiac function. However, since exaggerated HF in NPRA (−/−) mice with aortocaval shunt was associated with not only cardiac hypertrophy, but also increased plasma volume, and a loss of the vasodilatory effect of natriuretic peptides, the observed decrease in LV systolic function might be a consequence of complex orchestration of natriuretic peptide signaling in the heart, kidney and vasculature. Indeed, previous studies indicate no direct effect of ANP on cardiac function [27]. Further studies are thus needed to clarify the direct local effect of NPRA on cardiac function.

Our study also showed that U_NaV was lower in NPRA (−/−) aortocaval fistula mice than in the NPRA (+/−) or (+/+) aortocaval fistula mice, 4 weeks after operation. The lower U_NaV was accompanied by decreased U_cGMPV. Previous studies showed that despite similar UV and U_NaV at baseline in NPRA (−/−) and (+/+) mice [28,29], ANP infusion increases UV, U_NaV, and U_cGMPV only in NPRA (+/+) mice, but not in NPRA (−/−) mice [28]. They also showed that plasma volume expansion increases U_NaV, UV, and U_cGMPV in NPRA (+/+) mice, but not in NPRA (−/−) mice, suggesting that volume-induced release of ANP has an important role in diuresis and natriuresis [28]. Indeed, we and other investigators previously reported that a natriuretic peptide receptor antagonist, HS-142-1, significantly decreases U_NaV in animal model of HF without changing hemodynamics, suggesting that natriuretic peptide system exclusively compensates for HF through natriuretic action [30,31]. Taken together, these results suggest that communication between the heart and kidney via the NPRA signaling pathway is a compensatory mechanism for HF.

The synthesis and release of renin in the kidney are controlled by physiological determinants such as salt intake, arterial pressure, and sympathetic nerve activity [32]. ANP is also known to suppress renin secretion [2] and aldosterone synthesis [33]. However, PRC and plasma aldosterone levels were lower in sham-operated NPRA (−/−) mice than in NPRA (+/+) mice. This is most likely because of the elevated arterial pressure in NPRA (−/−) mice which leads to inhibition of renin synthesis and release from kidney juxtaglomerular cells [34]. Interestingly, NPRA (−/−) mice had greater increases in PRC and aldosterone after aortocaval fistula than did NPRA (+/+) mice, without change in mean arterial BP. We suggest that the lack of inhibitory effects on renin and aldosterone secretions through NPRA signaling pathway may underlie the greater response of PRC and aldosterone to aortocaval fistula in the NPRA (−/−) mice. The greater response of the renin–aldosterone system caused by a lack of feedback via the NPRA signaling pathway, in turn, is likely to be a contributing factor for the progression of HF in these mice. This explanation concurs with prior observations by others that activation of the renin–angiotensin–aldosterone system is intimately involved in the development of HF in the aortocaval fistula model [35]. It is also supported by a previous study by Wada et al., who demonstrated that administration of natriuretic peptide antagonist, HS-142-1, increases plasma renin and aldosterone levels in an animal model of HF [31]. Conversely, however, we note that a possibility remains that the worsened heart failure in the NPRA (−/−) mice may be contributing to the increased plasma renin concentration and aldosterone levels. Further studies are necessary for elucidating the cause–effect relationships.

Interpretation of the differential effect of NPRA genotype on the susceptibility in HF is complicated, because there were baseline differences in cardiac chamber weights and BP among the three genotypes, and because previous epidemiological studies have shown that hypertension and LVH is a risk factor for HF. In the current study, we observed that aortocaval fistula caused hypertrophy and increased ANP levels in all three genotype, but that it caused pulmonary congestion only in NPRA (−/−). It is likely that the higher basal BP and the presence of LVH contribute to the increased susceptibility to HF in these mice. However, Arnal et al. [36] previously reported that the aortocaval fistula did not increase the frequency of HF in spontaneously hypertensive rats with LVH compared to that in normotensive Wistar rats, and that aortocaval fistula increased the heart weight/body weight ratio more in normotensive rats than in hypertensive rats. Their data thus suggested that heart weight/body weight increase or susceptibility of HF in aortocaval fistula model is independent of BP and LVH of animals. Recently, Mori et al. [36] produced aortocaval fistula in ANP (−/−) mice on a low salt diet in an attempt to eliminate the effect of BP and compared left ventricular weight and echocardiographic findings with those in the ANP (+/+) mice. They showed that ANP (−/−) aortocaval fistula mice exhibited higher heart weight and wall thickness than ANP (+/+) aortocaval fistula mice despite that their BP was the same, indicating again that ANP has an antihypertrophic effect independent of BP. These studies indicate that high BP and the presence of LVH do not seem to greatly contribute to the susceptibility in HF induced by aortocaval fistula, and that the lack of NPRA–cGMP signaling, but not high BP or LVH, plays a direct role in the pathogenesis of HF in the NPRA (−/−) mice.

Finally, both sham-operated and aortocaval fistula NPRA (+/−) mice almost consistently showed a phenotype intermediate between those of NPRA (−/−) and NPRA (+/+) mice; however, NPRA (+/−) mice appeared to show a phenotype a little closer to NPRA (+/+) mice rather than to NPRA (−/−) mice. We previously reported that ANP-induced cGMP response is directly correlated with the number of functional NPRA genes [37]. Indeed, in the present study, UcGMPV and plasma cGMP levels depended on the number of NPRA genes. Thus, the effect of number of NPRA genes on cGMP production appears to be dose-dependent, whereas that of cGMP on phenotype may not necessarily be dose-dependent. There may be some threshold to cGMP-mediated effects on phenotype.

Clinically, heart failure is a pathologic process characterized by a decline of heart contractility. Our study clearly has demonstrated that NPRA signaling pathway attenuates the progression of HF induced by volume overload and predicts that mutations in the NPRA gene would likely predispose humans to HF. Indeed, a recent study has shown that a deletion mutation in the promoter region of the NPRA gene with a reduced receptor activity is associated with an increased susceptibility LVH in patients with hypertension [10]. Therefore, if the reduced NPRA gene activities are identified in patients with valvular heart disease, ischemic heart disease, or chronic renal failure, intensive medical attentions need to be paid towards the prevention of HF.

Acknowledgments

This work was supported in part by an NIH grant, HL49277 and HL62845, by the Scientific Research Grant-in-Aid 14570692 from the Ministry of Education, Culture, Sports, Science and Technology, Japan, and by the Science Research Promotion Fund from the Promotion and Mutual Aid for Private Schools of Japan.

We thank Ms. Jeniffer Wilder and Longuan Xu for their technical assistance. We thank Dr. Nobuo Shirahashi for useful advice on statistical analysis. We also thank Dr. Kathleen Caron and Dr. Masao Kakoki for useful advice and discussion.