Most of the centrally acting antihypertensive drugs, such as clonidine and related imidazoline derivatives, mediate sympathoinhibition, not only via activation of central nervous α₂-adrenoceptors but also via imidazoline I₁-receptors (Bricca et al., 1989; Bousquet, 1997). Imidazoline I₁-receptors are nonadrenergic and noncholinergic neurotransmitter receptors that possess low affinity for norepinephrine and other catecholamines. I₁-receptors are mainly found in the brainstem, adrenal chromaffin cells, and kidneys. In addition, we have recently identified I₁-receptors in heart atria and ventricles and shown that atrial I₁-receptors are up-regulated in rat hypertension and ventricular I₁-receptors are up-regulated in human and hamster heart failure (El-Ayoubi et al., 2002a). In other studies, we demonstrated that acute injections of moxonidine, an imidazoline compound that shows 40 times higher affinity to I₁-receptor versus α₂-adrenoceptors, are associated with enhanced release of atrial natriuretic peptide (Mukaddam-Daher and Gutkowska, 2000), a cardiac hormone involved in pressure and volume homeostasis. Together, these studies led us to suggest that heart I₁-receptors are functional and may be involved in cardiovascular regulation.

Previous binding studies reported [³H]idazoxan binding sites (I₂-receptors) but not I₁-receptors in human atrial appendage; but functionally, these receptors were different from presynaptic imidazoline receptors implicated in inhibition of noradrenaline release. Accordingly, atrial presynaptic imidazoline receptors were considered non-I₁ non-I₂ receptors, and the effects of moxonidine to inhibit noradrenaline release in atrial appendages were attributed to presynaptic α₂-adrenoceptors (Molderings and Gothert, 1999). In contrast, consistent with the presence of I₁-receptors in the heart, Schäfer et al. (2002) have recently shown in isolated perfused rats hearts that moxonidine is able to decrease noradrenaline release independently of α₂-adrenoceptors.

In fact, functional separation between imidazoline I₁-receptors and α₂-adrenoceptors is rather difficult, because these receptors are often colocalized and ligands with affinity to imidazoline I₁-receptors also bind to α₂-adrenoceptors (Bousquet, 1997). However, previous studies indicate that imidazoline receptors and α₂-adrenoceptors are subject to pathophysiological and pharmacological regulation (Yakubu et al., 1990; Ernsberger et al., 1991; Zhu et al., 1997; Ivanov et al., 1998). Therefore, the aim of the present studies was to test regulation of cardiac I₁-receptors versus α₂-adrenoceptors, by showing that I₁-receptors, but not α₂-adrenoceptors are regulated in hypertension and in response to exposure to agonist. Accordingly, studies were performed to 1) demonstrate the presence of α₂-adrenoceptors in the heart and their possible regulation in hypertension, and 2) investigate the effect of chronic in vivo exposure to moxonidine on I₁-receptors and α₂-adrenoceptors in hearts of normotensive rats and spontaneously hypertensive rats (SHR) with established hypertension.

Materials and Methods

Female SHR (12–14 weeks old) with established hypertension and age-matched normotensive Wistar-Kyoto (WKY) and Sprague-Dawley (SD) rats were purchased from Charles River (St. Constant, QC, Canada). Animals were housed in a temperature- and light-controlled room with food and water ad libitum, and maintained for at least 3 days before experimentation. Experiments were performed following the approval of the Bioethics Committee of Centre Hospitalier de L'Université de Montréal, according to the Canadian Guidelines.

Alzet osmotic mini-pumps (2ML1 and 2ML4; Alzet, Cupertino, CA) were implanted subcutaneously in SHR, under isoflurane anesthesia, as we have described previously (Menaouar et al., 2002). These mini-pumps allowed continuous delivery of moxonidine (generous gift from Solvay Pharmaceuticals, Hannover, Germany) or saline vehicle at the rate of 10 μl/h (2ML1), for 1 week, and 2.5 μl/h (2ML4) for 4 weeks. The concentrations of moxonidine were adjusted to allow delivery of 10, 60, and 120 μg/kg/h. The solution of moxonidine was prepared by dissolving the drug in isotonic saline, pH <6.5, and then pH adjusted to 7.0 to 7.4 by NaOH. Rats were sacrificed after 1 and 4 weeks of vehicle and moxonidine treatment, and heart atria and ventricles were separated, snap-frozen in prechilled isopentane, and then stored at –80°C, for receptor analysis by autoradiographic binding, immunoblotting, and RT-PCR.

To rule out the influence of blood pressure on receptor regulation, another group of SHR was treated with hydralazine, given at 30 mg/kg/day, in drinking water, for 1 week. The effectiveness of hydralazine was verified by tail cuff measurement of systolic blood pressure before and after 1-week treatment. Then, rats were sacrificed and tissues collected as described above.

Autoradiography. Autoradiography of heart I₁-receptors and α₂-adrenoceptors was performed on frozen heart sections from WKY and SD rats and from saline- and moxonidine-treated SHR, using radiolabeled paraiodoclonidine (¹²⁵I-PIC; 2200 Ci/mmol; PerkinElmer Life and Analytical Sciences, Boston, MA) as we have described previously (El-Ayoubi et al., 2002a). Because ¹²⁵I-PIC binds to both receptor types, autoradiography was performed, separately, in conditions that favor α₂-adrenoceptor binding and in conditions that favor I₁-receptor binding. For α₂-adrenoceptors, the slides were incubated for 1 h at 22°C with 0.5 nM ¹²⁵I-PIC in incubation buffer: 50 mM Tris-HCl (pH 7.7), 5 mM EDTA, 5 mM EGTA, 10 mM MgCl₂, and 50 μM phenylmethylsulfonyl fluoride. Binding was inhibited by increasing concentrations of epinephrine (10^–10–10^–5 M). Binding in the presence of 10^–4 M piperoxan was considered nonspecific. After several washes, the slides were dried, exposed in phosphor-sensitive cassette for 48 h, and then scanned, visualized, and quantified by PhosphorImager (ImageQuant; Amersham Biosciences Inc., Piscataway, NJ).

Autoradiography for I₁-receptors was performed under identical incubation conditions, except for prior incubation of slides with 1 mM phenoxybenzamine and 0.5 mM ethylmaleimide for 35 min at room temperature, to irreversibly inhibit adrenoceptor binding; and by decreasing the concentration of MgCl₂ in the incubation buffer to 0.5 mM, conditions that favor binding to I₁-receptors (Ernsberger et al., 1995). Binding of ¹²⁵I-PIC was competitively inhibited by increasing concentrations (10^–12–10^–5 M) of moxonidine

Membrane Preparation and Immunoblotting. Membranes of ventricular and atrial tissues were prepared in sucrose buffer as described previously (El-Ayoubi et al., 2002a). Protein content was measured spectrophotometrically, using bovine serum albumin as standard.

Immunoblotting was performed (El-Ayoubi et al., 2002a) using 30 μg of denatured protein samples from cardiac tissues and incubation of blots with rat α_2A-, α_2B-, and α_2C-adrenoceptor antiserum (1:500; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), or anti-imidazoline receptor antiserum and nonimmune antiserum (generous gift from S. Regunathan, Department of Psychiatry, University of Mississippi Medical Center, Jackson, MS) diluted 1:1000, or with anti-β-actin (1:500). The blots were incubated with horseradish peroxidase-conjugated anti-rabbit IgG antiserum (1:5000). Immunoreactive bands were visualized by enhanced chemiluminescence detection system (ECL hyperfilm; Amersham Biosciences Inc.), according to the manufacturer's instructions.

Total RNA Extraction and RT-PCR. Total RNA was extracted from the rat heart tissues using TRIzol reagent (Invitrogen, Carlsbad, CA) according to the protocol described by the manufacturer. PCR reactions were performed (Zou and Cowley, 2000) using specific primer pairs for rat α_2A-, α_2B-, and α_2C-receptors, or β-actin (QIAGEN Operon, Alameda, CA). After electrophoresis on agarose gel in the presence of ethidium bromide, fluorescent PCR products were scanned, counted, and analyzed with the ImageQuant software. These data were normalized to the corresponding values of β-actin PCR product in the same samples.

Data Analysis. The equilibrium dissociation constant (K_d) and maximum binding capacity (B_max) for the ligands used in autoradiography were calculated by the nonlinear method using the Ligand computer program (Elsevier-Biosoft, Cambridge, UK). Densitometric measurements of immunoblots were performed using Scion computer program (National Institutes of Health, Bethesda, MD). Correlation coefficients were calculated from linear regression (Graph-Pad Prism; GraphPad Software, Inc., San Diego, CA). Differences in data obtained from vehicle- or moxonidine-treated rats were compared by nonpaired Student's t test. P < 0.05 was considered significant. All data are expressed as mean ± S.E.M.

Results

Cardiac α₂-Adrenoceptors. Autoradiographic binding of ¹²⁵I-PIC to heart atrial and ventricular sections was inhibited by increasing concentrations of epinephrine. Kinetic parameters obtained from competitive inhibition curves (Table 1) revealed that α₂-adrenoceptor affinity (K_d ≅ 2.5 nM) and B_max in right atria (12.8 ± 0.7 versus 13.4 ± 0.9 fmol/unit area), left atria (12.8 ± 0.4 versus 11.7 ± 0.7 fmol/unit area) and left ventricles (11.7 ± 1.1 versus 12.2 ± 0.5 fmol/unit area) were not altered in SHR compared with WKY rats. Binding to cardiac α₂-adrenoceptors in SHR was also not altered by chronic in vivo moxonidine treatment. B_max remained in 120 μg/kg/h moxonidine-treated SHR right atria at 11.9 ± 0.9 fmol/unit area and K_d at 2.3 ± 0.3 nM. Similarly, kinetic parameters obtained in left atria and left ventricles were not altered in vehicle- or moxonidine-treated SHR (Table 1).

Three α₂-adrenoceptor subtypes were identified in cardiac tissues of SD and SHR, by immunoblotting. Densitometric measurements of the bands corresponding to α_2A-(Fig. 1), α_2B-, and α_2C-adrenoceptors (data not shown) were not significantly different in right atria and left ventricles of SHR versus SD, nor in vehicle- and moxonidine-treated SHR, where variation did not exceed 10%. Furthermore, levels of three subtypes of α₂-adrenoceptor mRNA detected in right and left atria and left ventricles of SHR were also not significantly different among vehicle- or moxonidine-treated groups, where variation did not exceed 10% (Fig. 2).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Representative immunoblot of α₂ -adrenoceptor subtypes (α_2A, α_2B, and α_2C) and β-actin, densitometric and data of α_2A measured in right atria and left ventricles of SD and SHR after treatment with moxonidine (0, 60, and 120 μg/kg/h) for 1 and 4 weeks. Data normalized to corresponding β-actin are presented as percentage of change from SD and vehicle-treated SHR.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

RT-PCR mRNA products of α₂ -adrenoceptors and β-actin in right and left atria and left ventricles of SHR treated with moxonidine (0, 60, and 120 μg/kg/h) for 4 weeks.

Cardiac Imidazoline Receptors. Autoradiography showed that total specific binding of ¹²⁵I-PIC to I₁-receptors was higher in SHR atria (162%) compared with normotensive WKY rats, considered as 100%. Also, total specific binding in atria decreased after treatment with moxonidine at 10, 60, and 120 μg/kg/h for 1 week (Fig. 3).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Top, representative autoradiography of total ¹²⁵I-PIC binding to right atrial tissue sections (after irreversible inhibition of α-adrenoceptor binding) in WKY and SHR after 1-week treatment with moxonidine (0, 10, 60, and 120 μg/kg/h). Bottom, specific binding of ¹²⁵I-PIC to right atrial sections in WKY and SHR after 1-week treatment with moxonidine (0, 10, 60, 120 μg/kg/h). Data are presented as %B/B₀, where B and B₀ represent binding in the absence and presence of increasing concentrations (10^–12–10^–5 M) of inhibiting ligand.

Competitive inhibition curves were plotted from values obtained from normotensive and hypertensive vehicle- and moxonidine-treated rats and presented as percentage of B/B₀, where B and B₀ represent, respectively, binding with and without moxonidine (Fig. 3). Kinetic parameters calculated from these curves using the Ligand computer program revealed that 1-week treatment dose dependently decreased B_max in SHR right and left atria. At the lowest dose of 10 μg of moxonidine, B_max decreased from 40.0 ± 2.9 to 18.2 ± 0.4 fmol/unit area (p < 0.01) in right atria and from 27.7 ± 2.8 to 12.3 ± 0.6 fmol/unit area (p < 0.04) in left atria. The doses of 60 and 120 μg moxonidine decreased B_max in rat right and left atria to values not significantly different from two normotensive controls (Table 2). Four-week treatment did not have additional effects, so that at 120 μg of moxonidine, B_max in right atria represented 9.0 ± 0.3 fmol/unit area. Moxonidine treatment did not affect B_max and K_d of I₁-receptors in right and left ventricles of moxonidine- and vehicle-treated SHR (Table 2).

The presence of three immunoreactive imidazoline receptor protein bands was shown in cardiac tissues by immunoblotting. The apparent molecular masses of these proteins were around 160, 85, and 29/30 kDa. Densitometric measurements of bands corresponding to the 160-kDa band was only slightly increased in atria of SHR, and almost not detected in normotensive SD rats and in moxonidine-treated SHR for 1 or 4 weeks. On the other hand, the density of bands corresponding to 85-kDa protein increased significantly (p < 0.05) in right atria of SHR to 134.6 ± 3.3% compared with normotensive control (considered as 100%).

Figure 4 shows that, compared with vehicle-treated SHR (considered as 100%), chronic moxonidine treatment resulted in a significant (p < 0.01) decrease in the intensity of the bands corresponding to 85-kDa proteins to represent 83.4 ± 1.9, 59.9 ± 2.7, and 51.8 ± 3.0% in 10, 60, and 120 μg moxonidine-treated SHR for 1 week, respectively. Treatment with moxonidine at 60 and 120 μg/kg/h for 4 weeks, resulted in a mild additional decrease in the intensity of the 85-kDa bands to 51.1 ± 2.2 and 46.8 ± 3.3%, respectively (Fig. 4). A modest increase in the intensity of ∼29-kDa band (data not shown) was observed in right atria of SHR (107 ± 2%) that decreased to 94 ± 1, 87 ± 3, and 83 ± 5% after 1 week of moxonidine at 10, 60, and 120 μg, respectively. In left ventricles, chronic treatment of SHR with moxonidine (120 μg) for 1 and 4 weeks did not alter the intensity of the bands corresponding to 85 kDa, but slightly decreased the 29-kDa band to 89 ± 3%, and after 4 weeks to 89 ± 1%.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Representative immunoblot and densitometric measurement of the 85-kDa imidazoline receptor protein in right atria of SHR after treatment with moxonidine (0, 10, 60, and 120 μg/kg/h) for 1 week and 60 and 120 μg/kg/h for 4 weeks. Data normalized to corresponding β-actin are presented as percentage of change from vehicle-treated SHR (considered as 100%). *, p < 0.01 versus vehicle-treated SHR.

Compared with corresponding WKY (considered as 100%), the percentage of increase in SHR right atrial B_max correlated with the percentage of increase in the density of the 85-kDa band (R² = 0.7744; p < 0.03), but not with the 29-kDa protein band. Moxonidine treatment resulted in a dose-dependent decrease in B_max and in the 85-kDa band compared with corresponding saline-vehicle treated SHR (considered as 100%). Figure 5 shows that the percentage of decrease in B_max correlated with the percentage of decrease in the 85-kDa band (R² = 0.5700; p < 0.006), but not with the 29-kDa band (R² = 0.1754; N.S.), suggesting that the 85-kDa protein may represent imidazoline I₁-receptors in the heart. A weak, but significant correlation was found between the 85- and the 29-kDa protein band (R² = 0.2717; p < 0.03), in moxonidine-treated SHR, implying that the two receptor proteins arise from the same gene.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Top, correlation between percentage of decrease in moxonidine-treated SHR (vehicle-treated SHR considered as 100%) right atrial B_max obtained from competitive binding assays versus percent decrease in the density of the 29-kDa (dotted line) and 85-kDa (solid line) bands obtained by immunoblotting. Bottom, correlation between percentage of change in the density of right atrial 85-versus 29-kDa bands obtained by immunoblotting.

Treatment of SHR with hydralazine for 1 week resulted in blood pressure reduction from 193 ± 8to135 ± 5mmHg(p < 0.02), whereas blood pressure remained in control rats at 186 ± 11 mm Hg. However, hydralazine treatment did not alter imidazoline receptor proteins measured by immunoblotting.

Discussion

The major findings of this study are as follows: 1) First time localization of α₂-adrenoceptors in heart atria and ventricles; and 2) demonstration that heart imidazoline I₁-receptors but not α₂-adrenoceptors are regulated in SHR, and in response to chronic in vivo exposure to a selective imidazoline receptor agonist, suggesting that heart I₁-receptors are subject to regulation. In addition, 3) the parallel change in receptor B_max and the 85-kDa imidazoline receptor protein, suggest that this protein may represent cardiac I₁-receptors.

Pharmacological and molecular cloning studies have revealed three α₂-adrenoceptor subtypes: α_2A (α_2D in rats), α_2B, and α_2C (Link et al., 1996; Altman et al., 1999). In the human heart, mRNA for all three α₂-adrenoceptors subtypes have been detected by PCR (Brodde and Michel, 1999). However, probably due to very low expression relative to α₁- and β₁-adrenoceptors, previous studies have not been successful in demonstrating α₂-adrenoceptors in the heart at the protein level through radioligand binding studies (Brodde and Michel, 1999). In the present study, demonstration of α₂-adrenoceptors in the heart was achieved by multiple approaches. Quantitative receptor autoradiography was used in conditions where binding of ¹²⁵I-PIC to adrenoceptors versus I₁-receptors was optimized by using high MgCl₂ (10 mM) concentration in the incubation buffer (Ernsberger et al., 1995). Furthermore, because radioligands cannot fully discriminate between α₂-adrenoceptor subtypes, further identification was obtained by immunoblots and RT-PCR, using specific rabbit polyclonal antibodies and primers for each subtype (Zou and Cowley, 2000). However, cellular localization of these receptors needs further experiments, because receptor subtypes were detected in whole cardiac tissue, which involves several cell types, including fibroblasts and myocytes, myocardial blood vessels, nerve terminals and intracardiac neurons (Armour, 1999).

The α_2A, abundant in the central nervous system, mainly in brain stem, is directly involved in regulating sympathetic outflow and seems to be the major presynaptic autoinhibitory receptor subtype (Altman et al., 1999). The α_2B is more abundant in arterial vascular smooth muscle cells and mostly responsible for vasoconstriction, and it is responsive to altered salt handling. The function of α_2C is not yet clear, but it may be the presynaptic autoreceptor in human atria (Rump et al., 1995; Hein, 2001).

The physiological significance of α₂-adrenoceptors in various heart chambers is beyond the scope of the present study. This study, however, provides strong evidence that α₂-adrenoceptors are present, albeit at low levels, in the rat heart atria and ventricles, at the levels of synthesis, protein expression, and binding activity, but these receptors seem not to be regulated by moxonidine, a selective agonist of imidazoline I₁-receptors. Because brain α₂-adrenoceptors have been shown to be selectively down-regulated in response to α₂-adrenoceptor agonists (Yakubu et al., 1990), and kidney imidazoline receptors to be down-regulated in response to imidazoline receptor agonists (Hamilton et al., 1993), the present findings imply that α₂-adrenoceptors in the heart interact weakly with moxonidine.

Most importantly, the present study confirms our previous finding that I₁-receptors are present in the heart and extend to demonstrate that up-regulated atrial I₁-receptors in SHR (El-Ayoubi et al., 2002a) are normalized by chronic in vivo exposure to I₁-receptor agonist. These receptors seem to be unrelated to imidazoline I₂-receptors, previously identified in the heart (Molderings and Gothert, 1999), because, by definition, the ligands used in the present study (^125I-PIC and moxonidine) show very low affinity to I₂-receptors (Bousquet, 1997).

Immunoblotting of heart membranes showed multiple molecular mass imidazoline receptor proteins similar to those so far described in brain and heart (El-Ayoubi et al., 2002a). Levels of 85-kDa proteins were increased in atria of untreated SHR compared with normotensive rats. Chronic moxonidine treatment, for short and long duration, was associated with decreased density of the 85-kDa bands in SHR atria. It is interesting to note that atrial 85-kDa but not the 29-kDa imidazoline receptor proteins vary in parallel to values of B_max for I₁-sites determined by ¹²⁵I-PIC binding. This correlation leads us to propose that the 85-kDa protein may represent I₁-receptors in the heart, as has been suggested by Ivanov et al. (1998). Furthermore, the positive correlation between changes in the two receptor proteins implies that they may arise from the same gene.

Regulation of imidazoline receptors has been previously reported in other tissues and under different physiological and pharmacological manipulations, usually in a manner distinct from α₂-adrenoceptors. Chronic treatment with the prototypic antidepressant imipramine, down-regulates I₁-receptors in rat brainstem, without affecting α₂-adrenoceptors (Zhu et al., 1997). Also, renal I₁-receptors are up-regulated by subpressor doses of angiotensin II infusion (Ernsberger et al., 1991), and in kidneys of SHR (El-Ayoubi et al., 2002b), whereas α₂-adrenoceptors are either unchanged or decreased in 1K1C rat kidneys (Li et al., 1994).

In the present study, imidazoline receptors were not different in hearts of two normotensive strains, WKY and SD, but up-regulated in SHR hearts, and then normalized and down-regulated after rat treatment with moxonidine, for short or long duration. The mechanisms involved in receptor up-regulation may include sympathetic overactivity, elevated blood pressure, increased cardiac mass, and activated intracardiac neurohormones, such as angiotensin II and norepinephrine. Treatment with moxonidine inhibits or counteracts these effects and eventually may indirectly lead to down-regulation of its receptor. We have previously shown that moxonidine dose dependently reduced blood pressure in SHR (Menaouar et al., 2002). However, the dose of 10 μg of moxonidine, which had no effect on blood pressure in those rats (Menaouar et al., 2002), significantly reduced I₁-receptor protein and B_max. Furthermore, treatment of SHR with hydralazine, a vasodilator antihypertensive compound that reduced blood pressure to a similar magnitude achieved by 120 μg of moxonidine, had no effect on imidazoline receptor proteins. Further studies are needed to clarify the mechanisms involved, but the present results argue against receptor down-regulation occurring in response to reduction in blood pressure per se, and in favor of a direct effect of the ligand on the receptor. Other investigators demonstrated that stimulation of I₁-receptor with moxonidine leads to activation of PC-PLC and generation of diacylglycerol, which activates several isoforms of protein kinase C (Ernsberger, 1999). Protein kinase C results in functional desensitization of the I₁-receptor through phosphorylation of serine and threonine residues in the receptor intracellular loop (Eason and Liggett, 1996).

In conclusion, this study demonstrates that heart I₁-receptors but not α₂-adrenoceptors are up-regulated in SHR and normalized by chronic antihypertensive treatment with moxonidine. Cardiac I₁-receptor normalization occurred after 1 week of treatment, the time point when moxonidine resulted in reversal of left ventricular hypertrophy in these rats (Menaouar et al., 2002). Also, the presence of I₁-receptors in atria, tissues known to secrete or respond to natriuretic peptides, atrial natriuretic peptide and brain natriuretic peptide, suggest a functional relationship between the two systems. Therefore, heart I₁-receptors are subject to regulation by the cardiovascular environment. Future antihypertensive treatment with imidazoline drugs should consider the heart as a major target organ.