INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ENDOTHELIN-1 (ET-1) is a 21-amino acid peptide originally isolated from porcine aortic endothelial cells (42) as a vasoconstricting agent. It has since emerged as one of the main regulators of cardiac function both in physiological conditions and during chronic heart failure, both in humans (40) and in animal models (29, 19). However, its effects on the heart are complex, multiple, and poorly defined. Depending on the species, the stage of ontogenic development and the type of myocyte studied, ET-1 acts as a positive (32, 10) or negative inotropic agent (27). It is reported to accelerate or to fail to accelerate the heart (9, 21) and usually stimulates atrial natriuretic peptide (ANP) secretion by cardiac myocytes (33). ET-1 is also one of the growth factors participating in the hypertrophic response of the heart to changes in its working conditions (31). This diversity of ET-1 actions on the myocardium is largely due to the fact that receptor subtypes are coupled to distinct G protein-linked signal transduction pathways, such as adenylyl cyclase and phospholipase-dependent pathways (30), which regulate cardiac function and phenotype by acting on a number of intracellular targets, including ion channels.

In cardiac myocytes, the L-type Ca²⁺ current (I_Ca) is the main depolarizing current contributing to the plateau-shape of the action potential and governing Ca²⁺ release from the sarcoplasmic reticulum (4). Channels carrying I_Ca are regulated by a number of second messengers, which explains why they mediate the action of a number of hormones and neuromediators on cardiac electrical and contractile properties. This is the case of ET-1, which has been found to modulate I_Ca in cardiac myocytes. Because inhibition and stimulation of the current have both been described (1, 14, 22, 34, 38, 39), it is possible that the effects of ET-1 on I_Ca depend on the species, the stage of ontogenic development, the type of myocyte studied, and the experimental conditions (11).

In human atrial myocytes also, I_Ca is the main current that activates during the plateau phase and that regulates the excitation-contraction process (7). Interestingly, I_Ca regulation by second messengers differs in several respects between human atrial myocytes and myocytes from other tissues or species, as illustrated by the regulation of I_Ca by serotonin (23) and, more recently, by its tonic inhibition by tyrosine kinase (2). In human atrial myocytes, phosphodiesterases (PDE) inhibit I_Ca in the absence of -adrenergic stimulation, indicating that in this tissue there is a basal production and degradation of cAMP (13, 26). All of these aspects of Ca²⁺ channel regulation by second messengers in human atrial myocytes may reflect the functional specialization of these cells, for instance, the constitutive capacity of these cells to secrete ANP. These phenomena may also be due to pathophysiological conditions associated with changes in hormonal regulation of the atrial myocardium, as is probably the case in hemodynamically overloaded atria (16) and chronic atrial fibrillation (35). It is thus important to know how ET-1 regulates the functional properties of human atrial myocardium. It is already known that both ET_A and ET_B receptors are expressed in this tissue (18) and that they are coupled to distinct signaling pathways (36). However, little is known on the cellular targets involved in the action of ET-1 on human atrial myocardium.

The aim of this study was to investigate the regulation of I_Ca by ET-1 in myocytes isolated from surgical samples of the human right atrial appendage, using whole cell configuration of the patch-clamp technique to record ionic currents. We found opposite effects of the peptide on I_Ca depending on the baseline current amplitude.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Clinical data and cardiac myocyte preparation. With approval from our ethics committee, specimens of the right atrial appendage were obtained from 72 patients (5 to 90 yr, mean 56 ± 4 yr) undergoing heart surgery, mainly consisting of coronary bypass surgery (n = 40), mitral (n = 7) or aortic (n = 18) valve repair, and congenital heart defect repair (n = 7). Except for four cases of atrial fibrillation, all the patients were in sinus rhythm. Most patients were on pharmacological treatments that were stopped at least 8 h before surgery (Ca²⁺ channel blockers, -adrenergic antagonists, diuretics, angiotensin-converting enzyme inhibitors, and nitric oxide donors). Myocytes were enzymatically isolated as previously described (2). Briefly, small pieces of atrial appendage were cut up and washed in calcium-free Krebs-Ringer solution containing (in mM) 35 NaCl, 4.75 KCl, 1.19 KH₂PO₄, 16 Na₂HPO₄, 10 HEPES, 10 glucose, 25 NaHCO₃, 134 sucrose, and 30 2,3-butanedione oxime (BDM) (pH 7.4 adjusted with NaOH), gassed with 95% O₂-5% CO₂, and maintained at 37°C. BDM was used transiently to prevent cutting injury (20). Pieces were reincubated in the same solution without BDM and containing bovine serum albumin (5 mg/ml, Hoescht-Behring), 200 IU/ml collagenase (type IV, Sigma; St. Louis, MO), and 6 IU/ml protease (type XXIV, Sigma). After 30 min of digestion, the enzyme solution was replaced by the same solution containing only collagenase (400 IU/ml) for 15 min. Isolated myocytes were suspended in DMEM and incubated at 37°C with continuous gassing with air supplemented with 5% CO₂ for at least 1 h before use.

Solutions and reagents. The composition of the standard external solutions was as follows (in mM): 136 NaCl, 5.4 KCl, 2 CaCl₂, 10 glucose, 1.06 MgCl₂, 0.33 NaH₂PO₄, and 10 HEPES; pH was adjusted to 7.4 with NaOH. In some experiments, NaCl was replaced by equimolar concentration tetraethylammonium chloride and 4 mM CaCl₂ was used instead of 2 mM CaCl₂. Experiments were carried out at room temperature (22-24°C). The pipette solution consisted of (in mM) 130 CsCl, 2 MgCl₂, 10 HEPES, 15 EGTA, 10 glucose, and 3 MgATP; pH was adjusted to 7.2 with CsOH. In some experiments, 0.42 mM Na₂-GTP was added to the pipette solution (26). Use of a multibarrel system allowed us to exchange the fluid solution bathing the myocyte within 2 s. ET-1 (porcine/human) and ANP (human) were dissolved in 1% acetic acid and stored as stock solutions at 80°C until use. BQ-123 and BQ-788 were dissolved in distilled water and stored as stock solutions at 20°C. Isoproterenol (Iso) was diluted daily in the extracellular perfusion solution. BAY K 8644 and 3-isobutyl-1-methylxanthine (IBMX) were dissolved in ethanol and stored as stock solutions at 20°C. Pertussis toxin (PTX) was diluted with DMEM at a final concentration of 0.1 µg/ml. Incubation with PTX took place at 37°C for 15 h in the incubator. All drugs were purchased from Sigma except for Iso (Sanofi; Winthrop, France).

Current measurements. Macroscopic calcium currents were recorded using the patch-clamp technique in the whole cell configuration. Borosilicate glass pipettes with a tip resistance of 1-2 M were connected to the input stage of a patch-clamp amplifier (Axoclamp 200A, Axon Instruments). Currents filtered at 5 kHz were digitized by a Labmaster (Lab Rac, Scientific Solution) and stored on the hard disk of a personal computer. Data were acquired and analyzed using a program written for our laboratory (Acquis, G. Sadock, CNRS, Gif/Yvettes). Resistance in series, but not the capacitive or leakage current, was compensated for to obtain the fastest capacity transient current. Membrane capacitance was calculated by using the fit of the capacity transient decay. Depolarizing voltage pulses were delivered at 0.2 Hz. The amplitude of the peak I_Ca was calculated as the difference between the amplitude of the inward current at its peak and at the end of the 350-ms test pulses.

Measurement of cAMP concentration. Myocytes were washed twice with 1× PBS and stimulated for 20 min with test compounds. Cells were scraped in 250 µl of 0.01 N HCl and frozen in liquid nitrogen until use. Cell extracts were then thawed and sonicated. The lysates were separated by centrifugation (10,000 g, 10 min), and cAMP was measured in the supernatant using a radioimmunoassay kit (Biotrak, Amersham Pharmacia Biotech).

Statistical analysis. Values are expressed as means ± SE. Differences between values were tested for statistical significance by using Student's paired and unpaired t-tests. Fisher exact test was used to compare the distribution of the effect of ET-1 on I_Ca in the various conditions tested (see Fig. 7). P values <0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Heterogeneous effects of ET-1 on I_Ca in human atrial myocytes. Figure 1 shows examples of the effects of external application of 10 nM ET-1 on I_Ca elicited by 350-ms depolarizing test pulses from 60 to 0 mV. No sodium current was recorded, as illustrated in Fig. 1, which also shows that a sizable current was only observed at potentials more positive than 50 mV; there was no evidence for a T-type Ca²⁺ current. In 68 of 154 myocytes studied, ET-1 decreased I_Ca (33 ± 2%, n = 68, P < 0.0001), whereas in the remaining myocytes it increased the current (99 ± 7%, n = 60, P < 0.001) (Fig. 1, A and B) or had no effect (i.e., the ET-1 effect did not exceed the classic rundown of I_Ca or the run-up phenomenon sometimes observed at the beginning of the current recording) (n = 26). The inhibitory effect of ET-1 on I_Ca was associated with a 10-mV shift of the current-voltage relationship toward positive potentials (Fig. 1C), whereas the stimulatory effect shifted the current-voltage relationship leftward (Fig. 1D). Neither the inhibitory nor the stimulatory effect of ET-1 was associated with a change in the apparent reversal potential of the current, ruling out major changes in its ionic selectivity. To further eliminate the possibility that part of the ET-1 effects was due to modulation of a contaminating sodium current, in some experiments NaCl was replaced by an equimolar concentration of tetraethylammonium chloride in the external solution. In these conditions, ET-1 still stimulated or inhibited I_Ca depending on the baseline current amplitude (not shown). Plots of the percent change in I_Ca in the presence of ET-1 against the current density measured before peptide application (Fig. 2A) showed that the stimulatory effect was observed in cells characterized by an I_Ca with a relatively low density (2.3 ± 0.1 pA/pF, n = 60) and was never observed for a current density of more than 5 pA/pF. The inhibitory effect predominated in cells with relatively high current densities (5.8 ± 0.3 pA/pF, n = 68) (Fig. 2B). ET-1 had no effect in 26 myocytes whatever the baseline current density. There was no relationship between the cell size and the response of I_Ca to ET-1 (Fig. 2C). Similar heterogeneous effects of ET-1 were observed using an external solution containing 2 or 4 mM Ca²⁺ (n = 8). As shown in Table 1, there was no clear relationship between the type of ET-1 effect on I_Ca and the patient's clinical history, including -adrenergic antagonists treatment. This probably reflects the fact that currents of small or high density were not associated with a given clinical parameter, in keeping with published observations that a number of factors can alter I_Ca including hemodynamic overload of the atria, which is associated with a number of cardiopathies (16, 35).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Opposite effects of endothelin-1 (ET-1) on L-type Ca2+ current (ICa). A and B: current traces elicited by a 10-mV incremental test pulse protocol from a holding potential of 60 mV in control conditions (control) and at the steady-state effect of 10 nM ET-1 in two distinct myocytes (numbers indicate the membrane potential at which current was recorded). A: ICa was inhibited by ET-1 (membrane capacitance 109 pF). B: peptide stimulated the current (membrane capacitance 80 pF). C and D: current density-voltage relationships of ICa in myocytes with currents stimulated (C) or inhibited (D) by ET-1. Each point is the average current (mean ± SE) density in 10 cells. *P < 0.05 and **P < 0.01 compared with control.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Type of effect of ET-1 on ICa depends on the basal current density. A: percent change in ICa on ET-1 application plotted against current density before peptide application (155 myocytes from 35 patients). B: current density in myocytes in which ICa was stimulated (n = 60 myocytes) or inhibited (n = 68 myocytes) by ET-1. Values are means ± SE. ****P < 0.0001. C: normalized distribution of the stimulatory and inhibitory effects of ET-1 plotted against membrane capacitance.

ET-1 inhibits I_Ca prestimulated by Iso. The aim of the following experiments was to determine the nature of the relationship between the I_Ca amplitude and its modulation by ET-1. First, we tested the effect of ET-1 in myocytes pretreated with the ₁-adrenergic agonist Iso (1 µM). As illustrated in Fig. 3, A and B, when I_Ca was stimulated by Iso (150 ± 13%, n = 7, P < 0.01) ET-1 always decreased I_Ca by 32 ± 7% (5.9 ± 0.7 vs. 9.1 ± 1.1 pA/pF, n = 7, P < 0.05). When the current was stimulated by a mechanism not involving ₁-adrenergic regulation, such as that induced by the dihydropyridine agonist BAY K 8644 (10 µM), which increases I_Ca to the same extent as Iso, ET-1 only slightly inhibited I_Ca (Fig. 3A), whereas it suppressed I_Ca after Iso application to current pretreated with BAY K 8644 (not shown). A similar inhibitory effect of ET-1 was observed when the current was stimulated with the phosphodiesterase inhibitor IBMX (10 µM, n = 6, Fig. 3C). The effects of ET-1 on I_Ca prestimulated by Iso were suppressed by pretreating myocytes with 0.1 µg/l PTX (8 ± 2%, n = 7, P < 0.05, Fig. 3D). In this experiment, the muscarinic agonist acetylcholine was used to check that G_i proteins were effectively inhibited by the toxin. Moreover, in the absence of Iso, a stimulatory effect of ET-1 on I_Ca was observed in 7 of 10 PTX-treated myocytes studied (3.6 ± 1.0 pA/pF in PTX-treated cells). Taken together, these results indicated that the inhibitory effect of ET-1 occurred when I_Ca was stimulated by the ₁-adrenergic signaling pathway, an effect mediated at least in part by coupling of the ET receptor to adenylyl cyclase via a G_i protein.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   ET-1 inhibits ICa increased by isoproterenol (Iso). A: left, superimposition of current traces elicited by test pulses from 60 to 0 mV in a myocyte exposed to Iso (1 µM) and then Iso plus ET-1 (membrane capacitance 27 pF); right, myocyte was exposed to 10 µM BAY K 8644 alone and then to ET-1 plus BAY K 8644 (membrane capacitance 155 pF). B: time course of changes in ICa elicited by a 350-ms test pulse from 60 to 0 mV in control conditions and on exposure to Iso (1 µM) and 10 nM ET-1 plus Iso (membrane capacitance 27 pF). C: time course of changes in ICa (test pulse from 60 to 0 mV) on exposure of the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (IBMX, 10 µM) and 10 nM ET-1 plus IBMX (membrane capacitance 71 pF). D: same protocol as in B with myocytes pretreated with 0.1 µg/l pertussis toxin (PTX); 10 µM acetylcholine (ACh) was applied to check that PTX-sensitive G protein was inhibited (membrane capacitance 49 pF).

ANP inhibits I_Ca by decreasing intracellular cAMP. In the following experiment, we examined whether ANP could be used to antagonize the ₁-adrenergic effect on I_Ca to study the effects of ET-1 on a reduced current. In the absence of GTP in the internal solution, ANP first transiently increased the current and then suppressed it, indicating the GTP dependency of its effect. Thus to optimize the effect of ANP, myocytes were dialyzed with an internal solution containing 0.42 mM GTP (15). As shown in Fig. 4, A and B, in GTP-loaded myocytes, ANP decreased the current by 41 ± 3% (2.2 ± 0.1 pA/pF vs. 4.5 ± 0.3 pA/pF, n = 48, P < 0.0001). Because one likely mechanism of the GTP-dependent inhibitory effect of ANP on I_Ca involves stimulation of PDE (6), these enzymes were inhibited by using IBMX. Figure 4C shows that IBMX stimulated I_Ca (195 ± 30%, n = 6, P < 0.01) and prevented the inhibition of I_Ca by ANP in GTP-loaded myocytes (n = 6). Moreover, ANP totally reversed the increase in I_Ca caused by Iso (n = 5, Fig. 4D). At the steady state of the ANP effect on I_Ca in GTP-loaded myocytes and after the peptide washout, Iso stimulated I_Ca with a magnitude significantly higher than that obtained in control myocytes (263 ± 49 vs. 116 ± 24%, n = 6, P < 0.05), but I_Ca density after Iso application was not different between control and ANP-treated myocytes (10.8 ± 1.5 vs. 8.2 ± 1.5 pA/pF, n = 6, not significant) (data not shown). These results indicated that ANP and Iso had opposite effects on cAMP production in human atrial myocytes. Additional evidence that the inhibitory effect of ANP on I_Ca resulted in large part from the degradation of cAMP was obtained by measuring the concentration of the nucleotide in control myocytes, after 20 min of incubation with ANP and Iso. Figure 5 shows that in control conditions myocytes contained a significant amount of cAMP that varied from 0.06 ± 0.01 to 0.24 ± 0.01 fmol · mg¹ · l¹ depending on the atrial specimen studied. Treating myocytes with ANP caused a marked reduction in cAMP content [0.06 ± 0.01 fmol · mg¹ · l¹ (n = 11) vs. 0.16 ± 0.01 fmol · mg¹ · l¹ (n = 11), P < 0.001], whereas treating myocytes with Iso caused a major accumulation of the nucleotide [0.37 ± 0.06 fmol · mg¹ · l¹ (n = 4) vs. 0.16 ± 0.01 fmol · mg¹ · l¹ (n = 11), P < 0.05].

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Atrial natriuretic peptide (ANP) inhibits ICa. A: superimposition of current traces elicited by test pulses from 60 to +10 mV in control condition and at the steady-state effect of ANP exposure (membrane capacitance 62 pF). B: time course of ICa (test pulses from 60 to +10 mV) during the application of 10 nM ANP (membrane capacitance 62 pF). C: time course of ICa (test pulses from 60 to +10 mV) on exposure to 10 µM IBMX and then to ANP in the presence of IBMX. D: time course of ICa (test pulses from 60 to +10 mV) during Iso application and ANP in the continuous presence of the -adrenergic agonist (membrane capacitance 70 pF).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   ANP decreases intracellular cAMP content. Intracellular cAMP content was measured in control (n = 15 ), ANP-treated (n = 11), and Iso-treated (n = 4) myocytes. *P < 0.05; ***P < 0.001.

ET-1 stimulates I_Ca inhibited by ANP. Having established that the main effect of ANP in human atrial myocytes is to reduce intracellular cAMP concentration via the stimulation of cGMP-dependent PDE, the effects of ET-1 were tested on I_Ca recorded in myocytes pretreated with ANP. ET-1 sometimes decreased (27 ± 3%, n = 13) but most often increased (52 ± 7%, n = 28) I_Ca in these conditions (Fig. 6). A typical example of a stimulatory effect of ET-1 on I_Ca pretreated by ANP is shown in Fig. 6B. Figure 7 summarizes the distribution of the dual effects of ET-1 on I_Ca in the different conditions tested. The stimulatory effect of ET-1 predominated in ANP-treated myocytes relative to control myocytes and Iso-treated myocytes. However, when the distribution of the two effects of ET-1 in ANP-treated myocytes was compared with that in control myocytes characterized by a low I_Ca density (<5 pA/pF) there was no longer a significant difference between the two experimental conditions.

View larger version (10K): [in this window] [in a new window]   Fig. 6.   ANP promotes the stimulatory effect of ET-1. A: percent changes in ICa plotted against current density before ET-1 application in myocytes pretreated with ANP. B: example of the time course of ICa elicited by 350-ms test pulses from 60 to 0 mV on application of ANP and ET-1 in the continuous presence of ANP (membrane capacitance 123 pF).

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Relationships between ET-1 effects on ICa and ANP and Iso pretreatment. Distribution of the stimulatory and inhibitory ET-1 effects on ICa among the various experimental conditions tested (control, ANP, and Iso exposure). Results are expressed as number of events/number of myocytes studied. **P < 0.01; ***P < 0.001.

Both ET_A and ET_B receptors mediate the effects of ET-1 on I_Ca. BQ-123, a specific ET_A receptor antagonist, was used to study the specificity of the effects of ET-1 on I_Ca and the type of receptor involved. We first examined the inhibitory effect of ET-1 on Iso-prestimulated I_Ca in presence of BQ-123. As illustrated Fig. 8A, on the top of the stimulation of I_Ca by 1 µM Iso, application of BQ-123 (1 µM) prevented its inhibition by ET-1 (n = 8). To test the involvement of ET_A receptors in the stimulatory effect of ET-1, we studied the effect of BQ-123 on current inhibited by ANP to enhance the incidence of the stimulatory effect. In these conditions, BQ-123 did not block the stimulatory effect of ET-1 (Fig. 8B), which, in contrast, was suppressed by application of the ET_B receptor antagonist BQ-788 (1 µM, n = 6, Fig. 8C). For control experiments, we checked that I_Ca recorded in myocytes isolated from the same sample responded positively to ET-1 after prolonged ANP exposure (n = 6). Taken together, these results indicated that ET_A receptors were responsible for the inhibitory effect of ET-1, whereas the stimulatory effect was mediated by the ET_B receptors.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Two types of ET receptors mediate the effects of ET-1 on ICa. A: time course of ICa after initial stimulation by Iso (1 µM) and exposure to BQ-123 (1 µM) and ET-1 in the continuous presence of Iso and BQ-123 (membrane capacitance 40 pF). B: time course of changes in ICa on exposure to ANP plus BQ-123 and then to ET-1 (membrane capacitance 150 pF). C: time course of changes in ICa on exposure to ANP plus the ETB receptor antagonist BQ-788 (1 µM) and then to ET-1 (membrane capacitance 150 pF).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The complex effects of ET-1 on the I_Ca of cardiac myocytes of various species is well established and has been attributed to the coupling of ET-1 receptors to different signaling pathways. The new contributions of our study to this field are as follows: 1) this is the first observation of complex and opposite effects of ET-1 on I_Ca in human atrial myocytes; 2) both ET-1 effects appear to depend on regulation of basal I_Ca by -adrenergic pathways; and 3) whereas the ET-A receptor mediates the inhibitory effect of ET-1 on I_Ca via negative coupling with cAMP-dependent signaling pathway, ET_B mediates the stimulatory effect of the peptide.

In various species and cell types, including human atrial myocytes, application of nanomolar concentration of ET-1 decreases the intracellular cAMP concentration via G_i protein-mediated inhibition of adenylyl cyclase (8, 22). A similar regulatory mechanism probably accounts for the inhibition of I_Ca by ET-1 in human atrial myocytes, because this effect was consistently observed when adenylyl cyclase was stimulated by the ₁-agonist Iso and was blunted in PTX-treated myocytes. Because the incidence of the inhibitory effect of ET-1 on I_Ca was not enhanced by pretreating the current with the dihydropyridine agonist BAY K 8644, a direct blocking effect of the peptide on the channel is very unlikely. An antiadrenergic effect of ET-1 on I_Ca has already been reported in rabbit (39) and in dog (38) ventricular myocytes. However, because ET-1 inhibits the calcium current stimulated by IBMX, it is possible that its antiadrenergic effect also involves mechanisms distinct from the modulation of cAMP generation, such as protein kinase C (PKC) activation (43), interaction at the level of phosphatase, and/or direct modulation of the channels (39).

The stimulation of I_Ca by ET-1 appears to depend on the species (34), the cell type (21, 41), or the presence of intracellular factors easily dialyzed by the patch pipette (11). Although no clear mechanism has been identified to explain the stimulation of I_Ca by ET-1, a number of studies point to a role of phospholipase-dependent signaling pathway activation (31). Phospholipases are coupled to ET-1 receptors via PTX-insensitive G proteins and, in response to ET-1 application, activate a number of major second messengers, including diacylglycerol, D-myo-inositol 1,4,5-trisphosphate [Ins(1,4,5)P₃], arachidonic acid (5), and PKC, which can directly or indirectly regulate L-type Ca²⁺ channels. For instance, it has been proposed that ET-1 stimulates the Na⁺/H⁺ exchanger via a phospholipase C-PKC signaling pathway and, in turn, alkalinizes the intracellular medium (41). In human atrial myocytes too, stimulation of I_Ca by ET-1 appears to result from activation of a signaling pathway distinct from that involved in current suppression, mainly because the effect is PTX insensitive. In addition, there is biochemical evidence that, in human right atrial myocardium, ET-1 is coupled to at least two distinct signal transduction pathways leading to adenylyl cyclase inhibition and Ins(1,4,5)P₃ formation via the stimulation of phospholipase C (36). The lack of convenient pharmacological tools hinders studies of the contribution of these phospholipase-dependent pathways to the effect of ET-1 on I_Ca in human atrial myocytes. The main information that our study adds to the characterization of the stimulatory effect of ET-1 on I_Ca is that this effect depends indirectly on cAMP-dependent regulation of the Ca²⁺ current. This was first suggested by our observation that ET-1 stimulated currents of low density, an effect associated with a leftward shift of the current-voltage relationship similar to that induced by Iso. It was further supported by the finding that an experimental reduction in the intracellular cAMP concentration by myocyte treatment with ANP markedly increased the frequency of the stimulatory effect relative to the inhibitory effect. Our results indicate that, in human atrial myocytes, ANP decreases intracellular cAMP by activating phosphodiesterase stimulated by cGMP (13). However, it is possible that part of the inhibitory effect of ANP on I_Ca also results from coupling of the ANP receptor to a G_i protein or from activation of cGMP-dependent protein kinases, as previously suggested (15). ANP modulation of the effects of ET-1 did not seem to result from a direct synergistic or cooperative action on a signal transduction pathway but rather from the ability of ANP to decrease intracellular cAMP and, in turn, to maintain the current in a status appropriate for its stimulation by ET-1. This is also consistent with the lack of significant difference in the distribution of the dual effects of ET-1 between ANP-treated myocytes and myocytes with a low density of baseline I_Ca, i.e., myocytes that were probably already dephosphorylated. When myocytes were phosphorylated by Iso, ANP always decreased I_Ca but failed to restore the sensitivity of Ca²⁺ channels to the stimulatory effects of ET-1, suggesting that, when adenylyl cyclase is stimulated, ET-1 preferentially inhibits I_Ca. This may indicate preferential coupling of ET-1 receptors to PTX-sensitive G protein-dependent regulatory pathways. Our assumption that the variation among myocytes in the density of baseline I_Ca could be due to changes in the current phosphorylation status is also supported by previous reports of a basal production of cAMP in human atrial myocytes, unstimulated by peptides and hormones (13, 26). However, in addition to cAMP-dependent regulation of I_Ca, other mechanisms may contribute to the heterogeneous effects of ET-1 including PKC- (3) or tyrosine kinase-dependent (37, 2) regulatory cascades or alterations in the composition of channel subunits (17, 24), which are also good substrates for intracellular second messengers (25).

In human atrial myocytes, as in most other species (22, 12), the inhibitory effect of ET-1 on I_Ca is mainly mediated by type A receptors. In contrast, the increase in I_Ca produced by ET-1 appears to be mediated by the ET_B receptor subtype in human atrial myocytes. Previous studies have shown that part of the effects of endothelin on the heart are due to activation of the ET_B receptor subtype. In rabbit ventricular myocytes, ET_B receptors are involved in the anti-adrenergic effect of ET-1 on I_Ca (39), whereas they mediate the stimulatory effect of ET-3 on I_Ca (12). There is another recent report that, in contrast to rabbit ventricular myocytes (39), ET_B receptor mediated in part the positive inotropic effect of ET-1 in human atria trabeculae (28). Finally, the fact that ET_A receptor mediates the inhibitory effects of ET-1 and that, in human atrial myocytes, this receptor subtype constitutes the majority of endothelin receptors (18) could explain the predominantly inhibitory effect of ET-1 on I_Ca observed here.

Conflicting results have been reported on the regulation of cardiac contractility by ET-1, with both positive and negative inotropic effects. These discrepancies are attributed to differences in the species, tissue, or development stage, and in experimental procedures. Our results provide an additional explanation, namely, that the type of Ca²⁺ current response to ET-1 depends on the phosphorylation status of channels and/or the various proteins that regulate I_Ca in human atrial myocytes. Opposite inotropic effects of ET-1 have also been observed among rat ventricular myocytes, a finding interpreted as indicating variations in receptors and regulatory pathway activity. In situ, the phosphorylation status of the atrial myocardium is continuously regulated by a subtle balance among neuromediators, peptides and hormones, which may largely determine the effect of ET-1. For instance, in pathological setting characterized by a degree of Ca²⁺ channel dephosphorylation (15, 35), ET-1 may predominantly stimulate I_Ca. ANP and ET-1 are two major regulators of cardiovascular function and often counteract each other at various levels. Thus it is possible that by increasing I_Ca ET-1 prevents excessive suppression of the current caused by prolonged ANP exposure. Further studies of animal models and tissue culture systems should help to determine the significance of these dual effects of ET-1 on atrial myocardial function, including the regulation of ANP secretion by ET-1.