A direct relationship between plasma aldosterone and cardiac L-type Ca²⁺ current in mice

Abstract

Aldosterone is involved in a variety of pathophysiological processes that ultimately cause cardiovascular diseases. Despite this, the physiological role of aldosterone in heart function remains elusive. We took advantage of transgenic mouse models characterized by a renal salt-losing (SL) or salt-retaining (SR) phenotype, thus exhibiting chronically high or low plasma aldosterone levels, respectively, to investigate the chronic effects of aldosterone in cardiomyocytes devoid of pathology. On a diet containing normal levels of salt, these animals do not develop any evidence of cardiovascular disease. Using the whole cell patch-clamp technique on freshly isolated adult ventricular cardiomyocytes, we observed that the amplitude of L-type Ca²⁺ currents (I_Ca) correlates with plasma aldosterone levels. Larger values of I_Ca are associated with high aldosterone concentrations in SL models, whereas smaller values of I_Ca were observed in the SR model. Neither the time- nor the voltage-dependent properties of I_Ca varied measurably. In parallel, we determined whether modulation of I_Ca by blood concentration of aldosterone has a major physiological impact on the excitation–contraction coupling of the cardiomyocytes. Action potential duration, [Ca²⁺]_i transient amplitude and contraction are increased in the SL model and decreased in the SR model. In conclusion, we demonstrate that the blood concentration of aldosterone exerts chronic regulation of I_Ca in mouse cardiomyocytes. This regulation has important consequences for excitation–contraction coupling and, potentially, for other Ca²⁺-regulated functions in cardiomyocytes.

The classical view that the mineralocorticoid hormone aldosterone maintains Na⁺ and K⁺ balance, blood volume and blood pressure, by controlling transepithelial Na⁺ and K⁺ transport in the kidney, has been recently expanded (Williams & Williams, 2003). The presence of mineralocorticoid receptors (MRs) in a variety of non-epithelial tissues, including brain, heart and blood vessels (Arriza et al. 1987; Lombes et al. 1995), suggested important roles for aldosterone in cardiovascular diseases (Booth et al. 2002; Gomez-Sanchez, 2004), such as cardiac remodelling and fibrosis. Hence, hyperaldosteronism associated with a high salt intake can induce cardiac hypertrophy, with cardiac fibrosis and remodelling, suggesting that aldosterone is an important factor in this pathophysiological process (Brilla et al. 1990; Struthers, 2004). However, a physiological role for aldosterone beyond epithelial transport has been questioned (Funder, 2000). In epithelial tissues, aldosterone selectivity of the otherwise non-selective MRs is achieved by the concomitant expression at high levels of the protective enzyme 11β-hydroxysteroid dehydrogenase (11β-HSD2), which converts glucocorticoids to their 11-keto receptor-inactive derivatives (Edwards et al. 1988; Funder et al. 1988). Non-epithelial tissues appear not to co-express 11β-HSD2 or do so at very low levels (Albiston et al. 1994; Lombes et al. 1995; Qin et al. 2003), so that the unprotected MRs are presumably occupied in the normal state by the much higher levels of circulating glucocorticoids (100- to 1000-fold higher than those of aldosterone). Thus, even though we previously observed an increased L-type Ca²⁺ current after 1–2 days of incubation of rat cardiomyocytes with aldosterone (Benitah & Vassort, 1999; Benitah et al. 2001), the physiological relevance of these observations remains to be established in vivo. We recently documented that cardiac MR overexpression in vivo leads to ion channel remodelling (Ouvrard-Pascaud et al. 2005), but this model did not discriminate an aldosterone effect per se, which is an important issue because modulating the receptor (MR) or the ligand (aldosterone) could lead to different effects.

To assess the role of aldosterone in more physiological conditions, we took advantage of three transgenic mouse models previously described (Hummler et al. 1997; Praderv and et al. 1999a, b; Wang et al. 2004). These models harbour modifications of the amiloride-sensitive epithelial Na⁺ channel (ENaC), thereby presenting a severe salt-losing (SL_s; Hummler et al. 1997; Wang et al. 2004), a mild salt-losing (SL_m; Pradervand et al. 1999a) or a salt-retaining (SR; Pradervand et al. 1999b) phenotype with high or low plasma aldosterone levels, respectively. On a diet containing standard levels of salt, these mice have no signs of cardiovascular disease, exhibiting normal serum corticosterone levels, blood pressure, acid–base electrolyte balance, and serum and urinary electrolyte concentrations. The ventricular myocytes of the different mice models were enzymatically isolated and I_Ca was measured. We found a direct relationship between aldosterone levels and the amplitude of I_Ca. The functional consequences for the cardiomyocytes are important, as shown by the changes that occur in parallel in the action potential (AP) duration, the amplitude of the [Ca²⁺]_i transient and contraction.

Methods

All experiments were carried out according to the ethical principles laid down by the French Ministry of Agriculture and European Union Council Directives (86/609/EEC) for the care of laboratory animals. All persons who participated in the experiments had the training and authorization to do so (authorization B34-172-16 for animal facility manager).

Transgenic mouse models

Transgenic SL_s mice (Scnn1a^tm1/Scnn1a^tm1TgrαENaC) and wild-type (Scnn1a^+/+) mice were obtained by interbreeding mice heterozygous mutants for the α-ENaC allele (Scnn1a^tm1/+; (Hummler et al. 1996) with or without the transgene TgrαENaC on a NMRI genetic background. SL_m (homozygous mutant, Scnn1b^neo/neo) and wild-type littermates (Scnn1b^+/+) with a partial knockout of the β-ENaC-gene locus were generated as previously described (Pradervand et al. 1999a). Mice were backcrossed onto the C57BL/6J genetic background. Homozygous (m/m) and heterozygous (m/+) mutants at this locus and their littermate controls were obtained by interbreeding heterozygous mutant (m/+) mice. SR mice with the Liddle mutation at position R564 were generated as previously described (Pradervand et al. 1999b). Mice were backcrossed onto the C57BL/6J genetic background and homozygous (L/L) and heterozygous (L/+) mutants for the Liddle mutation and their littermate controls were obtained by crossbreeding heterozygous mutant (+/L) mice. PCR-based genotyping for the gene targeting status was performed by using specific primers, as previously described for the SL_s mice (Hummler et al. 1997), the SL_m mice (Pradervand et al. 1999a) and the SR mice (Pradervand et al. 1999b). Protocols involving animals were reviewed and approved by the state authorities (Service Véterinaire Cantonal, Lausanne, Switzerland). Animals were kept under standard light (12 h:12 h light–dark cycle), were fed diets containing normal levels of salt (3 g kg⁻¹ Na⁺) and had free access to tap water. All animal experiments were performed blinded.

Aldosterone and RU28318 treatment

An osmotic minipump (Alzet model 2004 DURRECT Corp., Cupertino, CA, USA) was implanted for 3 weeks subcutaneously under anaesthesia (sodium pentobarbitone, 30 mg kg⁻¹) to deliver either 50 μg h⁻¹ of RU28318, a MR antagonist (Perrier et al. 2004a) or 50 μg day⁻¹ of d-aldosterone, dissolved in 0.9% NaCl physiological serum. During the period when they were fitted with osmotic minipump, animals (kept together in a cage cleaned twice a week) had free access to food and water, and were checked daily for health problems, pain or distress by trained personnel. No signs of infection, behavioural or physiological changes were observed. Measurement of plasma aldosterone by radioimmunoassay was performed on blood samples obtained by intrathoracic withdrawal after heart excision on animals anaesthetized with an intraperitoneal injection of 50 mg kg⁻¹ sodium pentobarbitone and 1000 IU kg⁻¹ heparin.

Cellular electrophysiology and Ca²⁺ imaging

Experiments were conducted on age- and sex-matched littermates. Animals were treated with heparin (1000 IU kg⁻¹) and anaesthetized with sodium pentobarbitone (50 mg kg⁻¹) administered intraperitoneally. Hearts were removed and cardiac ventricular cardiomyocytes were isolated as previously described (Benitah et al. 2001). Action potentials, Ca²⁺ and outward K⁺ currents were elicited and analysed using solutions and protocols previously described (Benitah et al. 2001; Ouvrard-Pascaud et al. 2005). APs were measured in a standard external solution (mm: NaCl, 140; MgCl₂, 1.1; CaCl₂, 1.8; KCl, 4; glucose, 10; and Hepes, 10, with the pH adjusted to 7.4 with LiOH), while the pipette contained a standard internal solution (mm: KCl, 135; MgCl, 4; EGTA, 5; glucose, 10; Hepes, 10; Na₂ATP, 5; and Na₂-creatine phosphate, 3, with the pH adjusted to 7.2 with LiOH). The outward K⁺ currents were measured with standard internal solution, while the external solution was modified by equimolar replacement of NaCl with 138 mm choline chloride, 1 mm CoCl₂ and 1 mm BaCl₂. I_Ca was recorded with standard solutions with KCl replaced by CsCl.

For Ca²⁺ imaging, a confocal laser-scanning microscope (Zeiss LSM 510) allowed visualization of cells loaded with the Ca²⁺-indicator dye fluo-3 acetoxymethyl ester (fluo-3 AM, 5 μm, Molecular Probes) as described elsewhere (Gomez et al. 2004). Briefly, fluorescence was excited with an argon ion laser at 488 nm and emission was measured at a wavelength of > 505 nm. During field stimulation (1 Hz), a single myocyte was repetitively scanned along its longitudinal axis at intervals of 1.5 ms for 1.5 s, at least six times. For caffeine experiments, a scan rate of 3 ms for 10 s was used.

Data analysis

All data are shown as means ±s.e.m. collected from at least four different animals for each group. Comparisons are made using Student's unpaired t tests, with P < 0.05 being considered statistically significant.

Results

SL_s mice kept on standard-salt diet present a sixfold elevation of plasma aldosterone levels, compared with age- and sex-matched wild-type (WT) controls (Hummler et al. 1997; Wang et al. 2004). No sign of heart or cellular hypertrophy was denoted, since the heart weight:body weight ratio (heart index) and the membrane capacitance (C_m) were identical to WT (Table 1). However, using the conventional whole cell voltage-clamp technique on isolated ventricular cardiomyocytes, we found that I_Ca exhibited higher densities in SL_s mice (Fig. 1A and B). Neither the time- nor the voltage-dependent properties of the current were changed (Table 2). I_Ca is critical for normal cardiac function, being involved in the generation and conduction of the AP and in contraction (Bers, 2002). The APs were clearly prolonged in SL_s mice (Fig. 1C and D), whereas there was no difference in resting membrane potential (E_R), AP amplitude (AP_amp) or maximum rate of rise of the AP upstroke (dV/dt_max, Table 3). Of note, among K⁺ outward currents that also control the AP duration, we found that the transient one (I_to) was smaller in SL_s mice than in WT mice (at +40 mV, 17.2 ± 1.5 pA pF⁻¹, n = 21, for SL_sversus 21.0 ± 1.1 pA pF⁻¹, n = 26, for WT; P < 0.05), which thereby contributed to lengthening of the AP, whereas the low-concentration 4-aminopyridine-sensitive current (I_Kslow) and the non-inactivating steady-state component (I_SS) were not significantly different in both groups (at +40 mV, the I_Kslow was 7.6 ± 0.7 pA pF⁻¹, n = 22, versus 7.9 ± 0.5 pA pF⁻¹, n = 32; and the I_SS was 5.9 ± 0.2 pA pF⁻¹, n = 23, versus 6.0 ± 0.2 pA pF⁻¹, n = 35, SL_s versus WT, respectively). Concomitant increased I_Ca and decreased I_to is consistent with the Ca²⁺-dependent regulation of I_to expression (Benitah et al. 2001; Perrier et al. 2004b). Furthermore, single cardiomyocytes from SL_s mice displayed larger intracellular Ca²⁺ ([Ca²⁺]_i) transients and contraction (Fig. 1E, F and G). The increase of the [Ca²⁺]_i transient in SL_s mice is not explained by a change in the ability of the sarcoplasmic reticulum to store Ca²⁺, because the amount of Ca²⁺ that can be released by caffeine application is not altered (Fluorescence value [F] normalized to the basal fluorescence [F₀](F/F₀) = 4.3 ± 0.2, n = 30, for WT versus F/F₀= 4.2 ± 0.2, n = 35, for SL_s). Thus, with high circulating plasma aldosterone levels, SL_s mice exhibit an increase of I_Ca, which functionally contributes to AP lengthening and positive cellular inotropy. Three weeks treatment of SL_s mice (n = 4) with a specific MR antagonist, RU28318 (Perrier et al. 2004a), reversed all these effects (Fig. 1), suggesting MR involvement.

Table 1.

Morphometric variables observed in the different groups

	Heart index (mg g−1)	Cm (pF)
SLs model
WT	6.4 ± 0.3 (na= 6)	165.7 ± 5.2 (nc= 72)
SLs	6.4 ± 0.3 (na= 5)	173.4 ± 5.3 (nc= 47)
SLs+ RU28318	6.3 ± 0.3 (na= 4)	171.8 ± 7.5 (nc= 27)
SLm model
+/+	8.2 ± 0.4 (na= 4)	149.7 ± 10.3 (nc= 29)
m/+	8.1 ± 0.5 (na= 5)	144.6 ± 8.3 (nc= 28)
m/m	7.9 ± 0.3 (na= 7)	144.0 ± 5.7 (nc= 44)
SR model
+/+	8.3 ± 0.3 (na= 5)	154.4 ± 6.7 (nc= 37)
L/+	8.1 ± 0.4 (na= 4)	153.4 ± 5.9 (nc= 33)
L/L	8.2 ± 0.3 (na= 5)	148.2 ± 8.3 (nc= 35)
Aldosterone treatment
Control	7.4 ± 0.2 (na= 4)	155.6 ± 5.4 (nc= 29)
+ Aldosterone	7.7 ± 0.5 (na= 4)	160.0 ± 7.2 (nc= 22)

Heart index, heart weight:body weight ratio; C_m, cellular membrane capacitance; n_a, number of animals; and n_c, number of cells.

Figure 1. Chronic high aldosterone levels in SL_s mice are associated with increased cardiac Ca²⁺ current.

Representative current traces (A) and averaged current–voltage (I_Ca–V) relationships of I_Ca densities (B) revealed significant increases of I_Ca in SL_s (red, 26 cells) compared with wild-type cardiomyocytes (WT, blue, 39 cells), whereas MR antagonist (RU28318) treatment (+RU, green, 27 cells) reversed the I_Ca increases. Superimposed traces of ventricular APs (C) and histograms of mean AP area (D) show a significant lengthening of AP duration in SL_s compared to WT mice; this effect reversed with RU28318 treatment. Also shown are representative line-scan images of fluo-3 fluorescence recorded in a WT (E, top), an SL_s (E, middle) and an RU28318-treated SL_s cardiomyocyte (E, bottom), and bar graphs corresponding to the pooled peak amplitude [Ca²⁺]_i transients (F) and cell shortening (G). The number of cells analysed (n) is indicated. Error bars are s.e.m.*P < 0.05, **P < 0.005, ***P < 0.0005 SL_sversus WT and #P < 0.05, ##P < 0.005, ###P < 0.0005 +RU versus SL_s.

Table 2.

The time- and voltage-dependent properties of I_Ca in the different groups

	ICa (pA pF−1)	V50 (mV)	k (mV)	Tpeak (ms)	τfast (ms)	τslow (ms)
SLs model
WT	—	−27.4 ± 0.4 (n = 35)	5.4 ± 0.3 (n = 35)	5.3 ± 0.1 (n = 39)	9.6 ± 0.6 (n = 39)	70.9 ± 1.1 (n = 39)
SLs	—	−26.0 ± 0.3 (n = 24)	5.1 ± 0.3 (n = 24)	5.2 ± 0.2 (n = 26)	8.7 ± 0.3 (n = 26)	68.2 ± 1.3 (n = 26)
SLs+ RU28318	—	−26.6 ± 0.2 (n = 19)	5.2 ± 0.2 (n = 19)	5.0 ± 0.2 (n = 27)	8.5 ± 0.4 (n = 27)	69.8 ± 1.0 (n = 27)
SLm model
+/+	—	−29.4 ± 0.6 (n = 21)	5.7 ± 0.6 (n = 21)	5.8 ± 0.2 (n = 29)	10.2 ± 0.6 (n = 29)	74.1 ± 1.6 (n = 29)
m/+	—	−27.9 ± 0.6 (n = 21)	5.9 ± 0.5 (n = 21)	5.9 ± 0.2 (n = 28)	10.1 ± 0.6 (n = 28)	70.5 ± 1.1 (n = 28)
m/m	—	−28.3 ± 0.5 (n = 29)	5.6 ± 0.4 (n = 29)	5.9 ± 0.1 (n = 44)	10.0 ± 0.3 (n = 44)	70.5 ± 1.3 (n = 44)
SR model
+/+	—	−28.9 ± 0.6 (n = 13)	6.2 ± 0.5 (n = 13)	5.6 ± 0.1 (n = 24)	10.2 ± 0.5 (n = 24)	77.6 ± 2.8 (n = 24)
L/+	—	−28.3 ± 0.5 (n = 13)	5.9 ± 0.4 (n = 13)	5.9 ± 0.2 (n = 19)	10.0 ± 0.7 (n = 19)	75.5 ± 1.7 (n = 19)
L/L	—	−29.9 ± 0.7 (n = 17)	6.2 ± 0.6 (n = 17)	5.6 ± 0.2 (n = 19)	10.5 ± 0.6 (n = 19)	72.6 ± 2.6 (n = 19)
Aldosterone treatment
Control	−7.2 ± 0.4 (n = 29)	−30.1 ± 0.3 (n = 28)	6.2 ± 0.5 (n = 28)	6.1 ± 0.2 (n = 29)	10.9 ± 0.5 (n = 29)	65.8 ± 1.2 (n = 29)
+ Aldosterone	−9.0 ± 0.5 (n = 19)*	−29.9 ± 0.4 (n = 18)	5.6 ± 0.3 (n = 18)	5.6 ± 0.3 (n = 19)	10.7 ± 0.6 (n = 19)	67.2 ± 2.7 (n = 19)

V₅₀, potential of half-maximal inactivation; k, slope factor of steady-state inactivation; T_peak, activation kinetics expressed as the time from the onset of +10 mV voltage step to peak current amplitude; τ_fast and τ_slow, fast and slow components, respectively, of decay phase of I_Ca at +10 mV; and I_Ca, Ca²⁺ current density at +10 mV. n, number of cells;

^*P < 0.05.

Table 3.

Action potential characteristics measured in the different groups

	ER (mV)	APamp (mV)	dV/dtmax (V s−1)	AP area (mV s−1)
SLs model
WT (n = 41)	−80.9 ± 0.4	116.3 ± 1.0	172.1 ± 4.9	—
SLs (n = 35)	−80.8 ± 0.4	117.9 ± 1.4	171.9 ± 5.9	—
SLs+ RU28318 (n = 20)	−80.5 ± 0.5	117.7 ± 1.3	170.6 ± 7.4	—
SLm model
+/+ (n = 25)	−80.1 ± 0.4	120.2 ± 1.5	169.3 ± 4.9	—
m/+ (n = 47)	−80.2 ± 0.3	118.8 ± 1.2	172.4 ± 6.3	—
m/m (n = 64)	−79.8 ± 0.4	118.9 ± 0.8	168.9 ± 4.7	—
SR model
+/+ (n = 29)	−78.7 ± 0.4	115.2 ± 1.3	167.2 ± 5.3	—
L/+ (n = 23)	−79.1 ± 0.7	117.3 ± 1.9	169.7 ± 6.3	—
L/L (n = 28)	−79.4 ± 0.4	116.6 ± 1.1	167.9 ± 6.4	—
Aldosterone treatment
Control (n = 31)	−79.7 ± 0.4	116.5 ± 1.6	169.9 ± 8.0	2.1 ± 0.09
+ Aldosterone (n = 25)	−79.3 ± 0.5	116.9 ± 1.4	168.8 ± 6.9	2.7 ± 0.19*

E_R, resting membrane potential; AP_amp, AP amplitude; dV/dt_max, maximum rate of rise of AP upstroke; and AP area evaluated by integration of voltage variation over its duration. n, number of cells;

^*P < 0.05.

To investigate whether or not increased I_Ca is a general feature during a chronic, lifelong exposure to elevated plasma aldosterone levels, we examined mice exhibiting a milder salt-losing phenotype (SL_m). With normal salt intake, SL_m mice exhibit a mild increased plasma aldosterone level compared with the sex-matched littermate WT mice (+/+; 1.4- and 1.8-fold, in heterozygous (m/+) and homozygous (m/m) mice, respectively; Pradervand et al. 1999a). No significant changes in either heart index or C_m were observed (Table 1). Cells from the m/+ and m/m mice showed a gradual increase in I_Ca compared to those from +/+ mice (Fig. 2A), which appeared without alterations of the time- or the voltage-dependent properties of the current (Table 2). The increase of I_Ca amplitude contributed to the progressive lengthening of the AP (Fig. 2B). No changes in E_R, AP amplitude or dV/dt_max were found (Table 3). There was also a gradual increase in the [Ca²⁺]_i transient (Fig. 2C) and in the cell shortening (Fig. 2D), which was significant in m/m mice. Sarcoplasmic reticulum Ca²⁺ load was unchanged (F/F₀= 3.8 ± 0.4, n = 24, for +/+; F/F₀= 3.8 ± 0.2, n = 19, for m/+; and F/F₀= 3.9 ± 0.3, n = 36, for m/m mice). Altogether, these results are consistent with our previous finding of an in vitro aldosterone-induced increase of I_Ca in isolated adult rat cardiomyocytes (Benitah & Vassort, 1999; Benitah et al. 2001).

Figure 2. Mild increased aldosterone levels in SL_m mice contribute to a gradual increase in I_Ca.

Comparison of pooled maximal conductance of I_Ca (G_ICamax, A), AP area (B), peak amplitude of [Ca²⁺]_i transients (C) and cell shortening (D) in n cardiomyocytes of littermate control (+/+), homozygous (m/m) and heterozygous (m/+) mice upon normal salt diet. Values are means ±s.e.m.*P < 0.05, **P < 0.005.

To test whether changes in I_Ca correlated with low aldosterone levels, we examined a salt-retaining (SR) mouse model (L/L, L/+, mouse model for the Liddle's syndrome with salt-sensitive hypertension; Pradervand et al. 1999b). In homozygous L/L mutant mice, on a normal salt diet, plasma aldosterone levels were threefold lower, whereas heterozygous L/+ mice showed intermediate levels of aldosterone (Pradervand et al. 1999b). Heart index and C_m were similar in all groups (Table 1). In contrast to the two other SL models, a decrease in I_Ca (Fig. 3A and B) and a shortening of AP duration (Fig. 3C and D) are obvious in L/L mice, whereas intermediate values were found in L/+ mice. Neither the time- nor the voltage-dependent properties of I_Ca varied measurably (Table 2), suggesting a genuine decrease in the number of Ca²⁺ channels at the surface membrane, since the surface area remained constant. Consistent with this, β-adrenergic stimulation of I_Ca was similar in +/+, L/+ and L/L animals (at 0 mV, 1 μmol l⁻¹ isoprenaline increased I_Ca by 143 ± 3%, n = 11, in +/+; 140 ± 4%, n = 9, in L/+; and 140 ± 7%, n = 6, in L/L mice). It is worth noting that in both the L/L and L/+ mice, K⁺ outward currents did not significantly change compared with +/+ mice (at +40 mV, for +/+, L/+ and L/L mice, respectively, I_to was 20.3 ± 1.9 pA pF⁻¹, n = 12; 20.9 ± 1.5 pA pF⁻¹, n = 12; and 21.1 ± 1.8 pA pF⁻¹, n = 16; I_Kslow was 4.5 ± 0.4 pA pF⁻¹, n = 12; 4.2 ± 0.6 pA pF⁻¹, n = 14; and 5.2 ± 0.7 pA pF⁻¹, n = 16; and I_SS was 6.3 ± 0.3 pA pF⁻¹, n = 13; 5.9 ± 0.3 pA pF⁻¹, n = 15; and 6.4 ± 0.3 pA pF⁻¹, n = 16).

Figure 3. Lifelong exposure to low aldosterone levels in SR mice is combined with a decrease in I_Ca.

Superimposed sample I_Ca traces recorded at +10 mV (A) and averaged I–V relationships (B) comparing current densities in control (+/+, open circles; 24 cells), heterozygous (L/+, grey circles; 18 cells) and homozygous littermate mice (L/L, filled circles; 19 cells). Representative AP recordings (C) and pooled AP area (D), revealing a gradual shortening of AP for L/+ and L/L mice compared with +/+ littermates. Error bars indicate s.e.m., n the number of experiments and *P < 0.05, **P < 0.005.

Taken together, our results show that, in animal models with no sign of cardiovascular pathology, plasma aldosterone levels are correlated in vivo with a titration of cardiac I_Ca. In Fig. 4, the percentage change in pooled maximal conductance of I_Ca (G_ICamax, relative to corresponding WTs) of the different models used in this study is plotted as a function of relative plasma levels of aldosterone. Remarkably, we found a direct linear relationship (r = 0.99, P < 10⁻³) between the logarithm of plasma aldosterone and G_ICamax.

Figure 4. Direct relationship between aldosterone levels and cardiac I_Ca.

Plot of changes, expressed as a percentage relative to corresponding controls, in pooled maximal conductance of I_Ca(G_ICamax)versus plasma aldosterone level (log scale) of studied models. The dotted line is a linear fit to transgenic model data points weighted by s.e.m. Data from wild-type mice treated 3 weeks with aldosterone (○; +Aldo) via osmotic minipumps fit remarkably well to this relationship.

In independent series of experiments, we examined the effects of chronic infusion of aldosterone, via osmotic minipumps, on I_Ca and AP in wild-type mice. Control experiments were conducted in unimplanted animals, since in pilot experiments we did not observe AP or I_Ca alterations in rats implanted with osmotic minipumps filled with this physiological serum compared to unimplanted control animals (data not shown). Chronic aldosterone infusion significantly (P < 0.05) increased the plasma aldosterone concentration above control levels (1177.5 ± 80.4 versus 274.5 ± 94.6 pg ml⁻¹, in aldosterone-treated (n = 4) versus sex-matched littermate wild-type mice without implanted minipumps (n = 4)). Heart index and C_m were not different (Table 1). Freshly isolated ventricular myocytes exhibit increased I_Ca amplitude (Table 2) and AP duration (Table 3) in aldosterone-treated mice compared to control littermates. When relative G_ICamaxversus plasma aldosterone concentration was plotted in Fig. 4, the values fitted nicely to the data obtained from transgenic mice. As we previously observed in isolated rat ventricular myocytes (Benitah & Vassort, 1999), it is worth noting that short-term exposure to 100 nm aldosterone has no effect on I_Ca in mice ventricular myocytes (data not shown), which suggests that the observed effects reflect a long-term modulation of the L-type Ca²⁺ channel.

Discussion

The present study provides, for the first time, in vivo evidence that lifelong exposure to aldosterone exerts a physiological regulation of the L-type Ca²⁺ current. This regulation, shown here in adult mouse ventricular myocytes, reflects changes in the density of Ca²⁺ channels, rather than a direct modulation of their activity, and might be involved in long-term adaptation of the cardiomyocytes to circulating aldosterone. The physiological consequences on cardiac output are obvious, since Ca²⁺ controls electrical activity and inotropy.

In comparison to its well-documented effects on pathology (Booth et al. 2002; White, 2003; Struthers, 2004), there is only little information available about the physiological role of aldosterone in the heart. For example, it has been shown that aldosterone has rapid inotropic effects in the isolated perfused rat heart (Barbato et al. 2004). However, the relevance of these effects in vivo remain uncertain and are poorly documented (White, 2003). Most of the studies examining the in vivo effects of aldosterone on cardiac function have been performed using high dietary salt intake, raising the question of whether the observed actions are due to aldosterone per se or not (Wang et al. 2004). We have previously described a specific and genomic regulation of I_Ca by aldosterone occurring in vitro (Benitah & Vassort, 1999; Benitah et al. 2001; Lalevée et al. 2005), and an increase in I_Ca in transgenic mice overexpressing MRs in cardiomyocytes in vivo (Ouvrard-Pascaud et al. 2005). We now demonstrate that plasma concentrations of the MR natural ligand, namely aldosterone, correlate with cardiac I_Ca amplitude as determined in cardiomyocytes isolated from animals exhibiting different blood levels of aldosterone (Fig. 4). Increased circulating aldosterone concentrations induce larger values of I_Ca, whereas lower values of I_Ca are observed with reduced plasma aldosterone levels. Systemic effects of aldosterone are unlikely to account for these effects. Indeed, animals (SL and SR models) had no clinical signs of pathology with normal dietary salt intake. Notably, blood pressure and plasma electrolytes were normal (Hummler et al. 1997; Praderv et al. 1999a, b; Wang et al. 2004). In addition, there was no sign of hypertrophy at the cellular level, suggesting that the increase in I_Ca does not occur as a consequence of morphological, structural and size changes of the cardiomyocytes.

The modulation of I_Ca amplitude by aldosterone described here is consistent with a genuine long-term physiological regulation of Ca²⁺ channels. The absence of alteration in kinetics and voltage-dependent properties of I_Ca suggests that the changes reflect changes of Ca²⁺ channel expression rather than channel modulations. Consistently, β-adrenergic stimulation of I_Ca was similar in control and experimental animals, at least in the Liddle model. Moreover, aldosterone had no effect when applied acutely at high concentrations. As expected, we showed that the effect of aldosterone on Ca²⁺ entry has a major impact on inotropy by regulating the AP duration, [Ca²⁺]_i transient and contraction. With significant variations of the [Ca²⁺]-transient amplitude one would expect an effect on Ca²⁺-dependent inactivation of the current. However, in the voltage-clamp experiments a high concentration of EGTA (5 mm) in the pipette filling solution buffered Ca²⁺, preventing the [Ca²⁺]_i transient feedback effect on I_Ca. Thus no difference in Ca²⁺-dependent inactivation (τ_fast) was observed. Na⁺ transport regulation by aldosterone in the heart (Mihailidou et al. 2000) could also contribute to inotropy by secondarily increasing Ca²⁺ influx through Na⁺–Ca²⁺ exchange (Bers, 2002). The caffeine-evoked [Ca²⁺]_i transient declined at a similar rate in SL_s (1.8 ± 0.2 s, n = 31) compared to WT (1.8 ± 0.3 s, n = 18). In the SL_m model, the inactivation kinetics of caffeine-evoked [Ca²⁺]_i transients were also unchanged (1.6 ± 0.3 s, n = 16 for +/+; 1.6 ± 0.2 s n = 10 for m/+; and 1.6 ± 0.2 s, n = 14 for m/m mice). These results suggest that Na⁺–Ca²⁺ exchange activity is not modified. However, the increased Ca²⁺ influx during the AP is associated with a greater Na⁺ influx through the Na⁺–Ca²⁺ exchanger, which might therefore contribute to an increase in AP duration. The acid–base balance is also modulated in cardiac cells by aldosterone following an increase in the Na⁺/H⁺ antiport activity (Korichneva et al. 1995), an effect that might lengthen the AP through an inhibition of most cardiac ionic conductances (including I_Ca) and affect Ca²⁺ homeostasis (Carmeliet, 1999). Even if we cannot definitely exclude such an effect, the fact that AP prolongation is associated with increased I_Ca in SL mice (Figs 1 and 2) tends to rule out this possibility. In addition, aldosterone does possess autonomic effects, potentiating the cardiac action of catecholamines (Wang, 1994; Yee & Struthers, 1998), which might increase I_Ca (Pignier et al. 2000). However, activation of cardiac myocyte α₁-adrenergic receptors by catecholamines induces hypertrophic growth (Simpson et al. 1991), which was not the case in our models.

Our present results provide new insights into a controversial issue. In the face of a 100-fold higher concentration of glucocorticoids in circulating plasma, how might aldosterone access and activate MRs? Indeed, MRs have similar affinity for glucocorticoids and aldosterone (Arriza et al. 1987; Funder et al. 1988). This question has been resolved by the finding of the enzyme 11β-HSD2, which catalyses, with its cosubstrate NAD, the interconversion of the active glucocorticoid corticosterone to its MR-inactive 11-keto congener 11-dehydrocorticosterone in the rodent (or cortisol to cortisone in man). This mechanism allows aldosterone selectivity in epithelial and vascular smooth muscle cells (Funder et al. 1988; Funder, 2004). However, cardiac expression of 11β-HSD2 has been reported to be extremely low (Edwards et al. 1988; Funder et al. 1988; Qin et al. 2003; Funder, 2004). The suggestion has therefore emerged that the cardiac MRs are, under ordinary physiological circumstances, occupied by glucocorticoids, impeding aldosterone action in cardiomyocytes (Funder, 2000 2004; Qin et al. 2003). However, we show here that plasma aldosterone levels determine I_Ca amplitude. This modulation might involve classical MR activation, as suggested by RU28318 experiments in SL_s mice (Fig. 1), which also tend to rule out a non-specific effect on cardiomyocyte metabolic state (Qin et al. 2003). More importantly, the lower plasma aldosterone concentrations in the SR model are associated with decreased I_Ca density, suggesting that MRs are not completely occupied by gluocorticoids. This conclusion is consistent with the idea that the specificity of aldosterone versus glucocorticoid action is not entirely pre-receptor and that post-11βH-HSD2 events are also important (Farman & Rafestin-Oblin, 2001).

Although the myocardium is not considered as a classical target organ for mineralocorticoid action, deregulation of I_Ca modulation by aldosterone might be involved in pathological states involving the cardiovascular system. For example, the aldosterone system plays an important role in the arrhythmogenic ionic remodelling, notably involving I_Ca increase, that develops before cellular hypertrophy (Perrier et al. 2004a). Two clinical trials indicate that MR blockade, in addition to standard therapy, is beneficial in hypertrophic heart diseases and reduces the rate of sudden death. The RALES study demonstrated a decrease in mortality with spironolactone treatment in patients with severe heart failure (Pitt et al. 1999), while the EPHESUS study showed a benefit of eplerenone, a second generation MR antagonist, in patients with acute myocardial infarction and left ventricular systolic dysfunction (Pitt et al. 2003a). Half of the benefit in both RALES and EPHESUS was attributed to a decrease in sudden cardiac death. Independent of blood pressure alterations, aldosterone, because of its ability to stimulate fibrosis (Lijnen & Petrov, 2003; Pitt et al. 2003b), is an important determinant of human left ventricular hypertrophy, which is associated with cellular electrophysiological alterations (Perrier et al. 2004a). Hence, the aldosterone system might be an important contributor to altered electrical activity of the myocardium through Ca²⁺ signalling modifications (Takeda et al. 2002; Perrier et al. 2004a).