Different subtypes of alpha1-adrenoceptor modulate different K+ currents via different signaling pathways in canine ventricular myocytes.

Multiple subtypes (alpha1A, alpha1B, and alpha1D) of alpha1-adrenoreceptors (alpha1ARs) co-exist in the heart and mediate a variety of cellular functions. We studied alphaAR modulation of inward rectifier (IK1) and transient outward (Ito) K(+) currents in canine ventricular myocytes. Phenylephrine at 10 microM depressed only Ito without affecting IK1 and at 100 microM inhibited both Ito and IK1. The effect of phenylephrine on Ito was abolished by (+)niguldipine (10 nm) to inhibit alpha1AARs but not by chloroethyclonidine (10 microM) to inactivate alpha1BARs nor by BMY-7378 to antagonize alpha1DARs. In contrast, phenylephrine inhibition of IK1 was reversed only by BMY-7378 (1 nm). PDD (100 nm, phorbol ester activator of protein kinase C (PKC)) simulates and bisindolylmaleimide (50 nm, PKC inhibitor) weakens phenylephrine modulation of Ito but not IK1. Ca(2+)/calmodulin-dependent protein kinase II (CaMKII) inhibitor KN-93 and inhibitor peptides abolished the effects of phenylephrine on IK1. Enhancement of PKC or CaMKII activities was seen in alpha1aAR- or alpha1dAR-transfected HEK293 cells and in myocytes pretreated with 10 or 100 microM phenylephrine, respectively. Our data suggest that different subtypes of alpha1ARs selectively modulate different cardiac K(+) currents via different signal transduction mechanisms; alpha1AARs mediate Ito regulation via PKC, and alpha1DARs mediate IK1 regulation via CaMKII.

Over the past decade, evidence from pharmacological studies and molecular cloning has been accumulating indicating that ␣ 1 -adrenoreceptors (␣ 1 ARs) 1 are actually a heterogeneous group of distinct but related protein subsets. Many cellular responses to ␣ 1 ARs are mediated by multiple subtypes (␣ 1A , ␣ 1B , and ␣ 1D ) (1)(2)(3)(4). In the heart, whereas the ␣ 1A and ␣ 1B subtypes have been well characterized, the presence of ␣ 1D AdR was indicated only recently (5)(6)(7). Moreover, although the pathophysiological roles of ␣ 1A and ␣ 1B receptors have been well appreciated, those of ␣ 1D subtype in the heart remain to be determined.
Enhanced ␣ 1 AR activity has been implicated in various types of arrhythmias, particularly those in the pathogenesis of myocardial ischemia, ischemia-reperfusion and preconditioning, cardiac hypertrophy, etc. (1,3). Drug intervention with ␣ 1 ARs has thus become an attractive issue for developing new compounds for potential therapy. A significant mechanism underlying ␣ 1 AR-induced alteration of cardiac electrical activity is attributable to the ability of ␣ 1 ARs to modulate ion channels. To date, no less than seven cardiac ionic currents are on the list of ␣ 1 AR modulation, including inward rectifier K ϩ current (I K1 ), transient outward K ϩ current (I to ), delayed rectifier K ϩ current (I K ), ultrarapid delayed rectifier K ϩ current (I Kur ), acetylcholine-induced K ϩ current (I KACh ), calcium current (I Ca ), and chloride current (1, 3, 8 -11). However, it is not known whether the effects are the results from participation of all three different subtypes of ␣ 1 ARs or of a particular individual subtype, although evidence is accumulating that different subtypes may have different roles in regulating cardiac contraction and electrical activities (12)(13)(14)(15)(16). Moreover, recent studies also demonstrated subtype differences in the signal transduction (17)(18)(19). In light of these studies, we speculated that different subtypes of ␣ 1 ARs may have distinct effects on ion channels. Understanding subtype specificity of ␣ 1 ARs in ion channel regulation is of theoretical and practical importance.
K ϩ currents play critical roles in determining cardiac electrical activities. Besides stabilizing resting potential, I K1 in cardiac cells also plays an important role in modulating cellular excitability and regulating membrane repolarization, therefore an important determinant of action potential initiation. Another important cardiac K ϩ current is transient outward K ϩ current (I to ), which is known to be critical for initiating cardiac repolarization in the early phase of action potentials. Both of these currents have been implicated in the pathology of cardiac electrophysiological disorders and heart failure (20). For these reasons, we explored the potential subtype selectivity and signal transduction mechanisms of ␣ 1 ARs in regulating I to and I K1 in isolated canine ventricular myocytes using whole cell patch clamp techniques.

EXPERIMENTAL PROCEDURES
Myocyte Isolation-Single canine myocytes were isolated from the epicardium of left ventricles with techniques as described previously * This work was supported in part by funds from the Canadian Institute of Health Research, the Heart and Stroke Foundation of Quebec, and the Fonds de la Recherche de l'Institut de Cardiologie de Montreal (to Z. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  Whole Cell Patch Clamp Recording-Patch clamp recording techniques used have been described in detail elsewhere (22)(23)(24). Borosilicate glass electrodes (outer diameter, 1 mm) had tip resistances of 1-3 M⍀ when filled with pipette solution. Junction potentials were zeroed before formation of the membrane-pipette seal in Tyrode's solution. The capacitance and series resistance was electrically compensated to minimize the duration of the capacitive surge on the current recording and the voltage drop across the clamped cell membrane. I to was defined as the peak current amplitude, and I K1 was measured as the amplitude at the end of 400-ms pulses. The experiments were conducted at 36°C.
Solutions and Drugs-The bath solution for whole cell patch clamp recording had the following composition 136 mM NaCl, 5.4 mM KCl, 1 mM MgCl 2 , 0.33 mM NaH 2 PO 4 , 5 mM HEPES, 10 mM glucose, and 1 mM CaCl 2 ; pH was adjusted to 7.4 with NaOH. Unless otherwise specified, the pipette solution contained 0.1 mM GTP, 110 mM potassium aspartate, 20 mM KCl, 1 mM MgCl 2 , 5 mM Mg-ATP, 10 mM HEPES, and 5 mM phosphocreatine, pH 7.3. Sodium current was prevented by holding the cells at Ϫ20 or Ϫ50 mV, and calcium current was blocked by inclusion of Cd 2ϩ (200 M) in the bathing solution. All chemicals were purchased from Sigma.
PKC Activity Assay and Immunoblotting Analysis of CaMKII Activity-The HEK293 cells stably transfected with ␣ 1a ARs and ␣ 1d ARs were a kind gift from Dr. Kenneth P. Minneman (Emory University, Atlanta, GA). The cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin. The cells were passaged regularly and subcultured to ϳ90% confluence before experimental procedures. For drug treatment, the cells were incubated with Phen at 10 or 100 M or vehicle for 20 min before being collected for protein preparation. Similarly, isolated canine ventricular myocytes were also treated with Phen or vehicle in KB solution for 20 min. The proteins were prepared following the same procedures as described previously (25,26). The protein content was determined with Bio-Rad protein assay kit using bovine serum albumin as a standard.
PKC activity was assayed with the PKC enzyme assay system from Amersham Pharmacia Biotech, and the assay procedures were performed according to the system instructions. The Anti-ACTIVE ® CaM KII polyclonal antibody system from Promega was used to assess CaMKII activity, containing the rabbit polyclonal antibody, which recognizes autophosphorylation of Thr 286 of all isoforms of CaMKII. Experiments were performed according to the manufacturers' protocol. Bound antibodies were detected with Western blot Chemiluminescence Reagent Plus (PerkinElmer Life Sciences) and quantified by densitometry, as detailed previously (25)(26). Coomassie staining was performed to verify the amount of protein inputs, as described previously (25)(26).
Statistical Analysis-The group data are expressed as the means Ϯ S.E. Statistical comparisons were made with Student's nonpaired t test. All enzyme activity and Western blot determinations were performed in parallel with cells from each group for each experiment to minimize contamination by inter-day and inter-lot reagent variation, and the results for each experiment are expressed as the values normalized to control group determinations.

RESULTS
␣ 1 AR Modulation of I to and I K1 -Application of depolarizing or hyperpolarizing voltage steps (voltage protocols shown in the non-graphical parts of Fig. 1) elicited outward I to with transient properties (rapid activation and inactivation) or I K1 with strongly inward rectification. Bath application of Phen at 10 M produced marked reduction of I to amplitude without apparent alterations in its activation and inactivation kinetics but did not affect I K1 (Fig. 1). When Phen concentration was elevated to 100 M, I K1 was significantly decreased, I to was further reduced, and the inactivation kinetics was slightly decelerated. Co-application with prazosin (1 M) completely converted the depressed I to and I K1 caused by Phen back to predrug base-line values, indicating the role of ␣ 1 ARs in mediating the effects of Phen.
FIG. 1. ␣ 1 AR modulation of transient outward K ؉ current (I to ) and inward rectifier K ؉ current (I K1 ) in canine ventricular myocytes. A and B, raw traces recorded from representative cells showing the inhibition of I to and I K1 produced by Phen at concentrations of 10 and 100 M, respectively. Voltage protocols for current recordings are shown in the insets. C and D, current-voltage (I-V) relationships of I to (n ϭ 19 cells) and I K1 (n ϭ 22 cells). When Phen concentration was 10 M, statistically significant inhibition (p Ͻ 0.05) of I to was seen at potentials positive to 0 mV, and the effect on I K1 was not statistically significant (p Ͼ 0.05). When Phen concentration was raised to 100 M, I to blockade was statistically significant (p Ͻ 0.05) at potentials positive to Ϫ20 mV, and I K1 blockade was statistically significant (p Ͻ 0.05) at potentials more negative than Ϫ80 mV.
Subtype Specificity of ␣ 1 AR Modulation of I to and I K1 -To determine which receptor subtypes, ␣ 1A , ␣ 1B , or ␣ 1D , mediate the effect of ␣ 1 AR modulation of I to and I K1 , we performed experiments using subtype-selective antagonists. Current recordings made under control conditions were repeated 10 min after exposure of the cells to Phen (10 M). Subsequently, Nig (10 nM) was concurrently applied with Phen to the bath solution. As illustrated in Fig. 2, Nig reversed Phen-induced I to depression. Nig also reduced the further reduction of I to caused by raising Phen concentration to 100 M. On the other hand, Nig failed to change the depressed I K1 caused by Phen (100 M).
We then turned to study the effects of CEC (10 M), an alkylating agent selective toward ␣ 1B ARs over other subtypes. Following a 30-min incubation with CEC to inactivate the ␣ 1B ARs, the cells were superfused with drug-free Tyrode's solution for 20 min, and then measurements of I to and I K1 were made as base-line values. Then Phen was added to the superfusate. The same degrees of I to and I K1 diminishment as in cells without pretreatment with CEC were consistently seen in a total of six cells (Fig. 2).
Signal Transduction Mechanisms of ␣ 1 AR Modulation of I to and I K1 -␣ 1 ARs are known to stimulate activation of PKC (17,19,28), which in turn can phosphorylate the channel proteins. To investigate whether PKC can account for the ␣ 1 AR modulation of I to and I K1 , the effects of PKC inhibitor Bis and activator PDD were assessed (10 -11). 10 min after superfusion with Phen (10 M) to verify the effect on I to , Bis (50 nM) was co-applied. As illustrated in Fig. 3 suppression of I to amplitude. Moreover, no significant effect of 4␣-PDD (100 nM, the inactive stereoisomer of PDD) on I to was seen, and when in the presence of Bis, PDD had little effect on I to . Slight reduction of canine I K1 (about 7%) was observed when treated with PDD (100 nM).
CaMKII has recently been reported to be involved in regulating ion currents (29 -31). Accordingly, we also studied the potential participation of CaMKII in ␣ 1 ARs or Phen modulation of I to and I K1 . As illustrated in Fig. 4, Phen at 10 M produced the similar effects on I to with and without KN-93 (3 M, a potent CaMKII inhibitor) present in the bath solution. However, with 100 M Phen, I to reduction was slightly smaller in the presence of KN-93 than in the absence of the compound (Fig. 4B). The CaMKII inhibitor peptides P281-309 or ACP dialyzed into the cells did not affect I to neither the ability of Phen to inhibit I to (Fig. 4, A and C). By comparison, the reduction of I K1 caused by Phen (100 M) alone was completely prevented by KN-93 (Fig. 5). We then examined whether the effects of KN-93 on I K1 were related to the inhibition of CaMKII activity or to a direct effect on K ϩ channels. KN-92 (10 M), the inactive analog of KN-93, failed to affect I K1 reduction caused by Phen (Fig. 5B). Further evidence that I K1 is regulated by CaMKII was obtained by dialyzing cells with the inhibitor peptides P281-309 or ACP. The cells were bathed in the solution containing Phen (100 M) for at least 10 min before the formation of whole cell configuration. The currents recorded immediately after membrane rupture with minimal dialysis were considered the base-line data, and the data acquired 15 min after membrane rupture with complete dialysis were taken as the effects of CaMKII inhibition by ACP or P281-309 (Fig.  5). In addition, addition of EGTA (10 mM) to the pipette solution also substantially weakened the ability of Phen to suppress I K1 (data not shown). The effects of KN-93 on Phen-induced de-crease in I K1 were also assessed in myocytes pretreated with the calmodulin inhibitor calmidazolium. Similar to KN-93, calmidazolium converted the depressed I K1 to the base-line amplitude (Fig. 5B). External application of KN-93 (3 M) when the steady-state effect of calmidazolium had been achieved failed to cause further changes on I K1 (data not shown).
PKC Activity Assay and Immunoblotting Analysis of CaMKII Activity-Based on the results from the above functional studies, we believed that suppression of I to by Phen at 10 M is primarily mediated by PKC activation as a result of ␣ 1A AR activation, whereas an increase in CaMKII activity caused by ␣ 1D ARs stimulation by 100 M Phen leads to I K1 inhibition. To test this point, we performed analyses for PKC and CaMKII activities. Fig. 6 shows the results from PKC activity assay. A pronounced increase in PKC activity was seen in HEK293 cells stably transfected with ␣ 1a ARs and pretreated with Phen, relative to nontransfected and untreated cells. Only a minor increase in PKC activation was observed in ␣ 1d AR-transfected cells, even with elevated Phen concentration to 100 M. Coincidentally, canine ventricular myocytes pretreated with 10 M (or higher) Phen also resulted in a significant increase in PKC activity relative to untreated cells.
The antibody to the autophosphorylation site Thr 286 of CaMKII recognized a band of 51 kDa in both HEK and a band of 48 kDa in canine ventricular cells, which is in agreement with the size of CaMKII reported by other laboratories (29 -31). As shown in Fig. 7, a more prominent band was seen only in

DISCUSSION
We have shown in this study that ␣ 1 AR stimulation by phenylephrine produces important suppression of I to and I K1 in dog ventricular myocytes. The effects on I to and I K1 can be separated by the strength of ␣ 1 AR stimulation (Phen concentrations), and the use of subtype-selective antagonists suggests a possibility of differential modulation of I to and I K1 by activation of different subtypes of ␣ 1 ARs: ␣ 1A AR for I to and ␣ 1D AR for I K1 . Moreover, our data also suggest that the differential modulation of I to and I K1 by ␣ 1 ARs is mediated by different signaling systems: I to primarily by PKC and I K1 mainly by CaMKII. Our data therefore provide evidence that different subtypes of ␣ 1 ARs modulate different cardiac K ϩ currents via different signal transduction mechanisms. Our results also represent, to our knowledge, the first to define the functional role of ␣ 1D ARs in modulating cardiac ion channels and the role of CaMKII in modulating inward rectifier K ϩ current.
With the development of pharmacological tools and molecular cloning, it is now possible to study ␣ 1 AR subtypes separately. The ␣ 1A ARs show a high affinity to (ϩ)niguldipine that has been demonstrated to be 100ϳ4000-fold more selective toward ␣ 1A AR over the other two subtypes (4,32). The ␣ 1B ARs are readily and irreversibly inactivated by the alkylating agent CEC (4,33). BMY-7378 distinguishes ␣ 1D AR from the other two subtypes (␣ 1A and ␣ 1B ) with a binding affinity at least 2 orders of magnitude higher for the former over the latter (4,7,27). We show here that neither Nig nor CEC produced any appreciable influences on the inhibitory effects of Phen on I K1 . In sharp contrast, BMY-7378, at a concentration of as low as 1 nM, readily reversed Phen actions on I K1 . These data strongly suggest that ␣ 1D ARs mediate ␣ 1 AR/Phen modulation of I K1 in canine ventricular myocytes. ␣ 1 AR modulation of I K1 in rabbit (34) and human (9) atria has also been previously studied, and consistent with the present study ␣ 1 AR stimulation was found to decrease atrial I K1 . However, the involvement of particular subtypes of ␣ 1 ARs was not investigated in these studies. Although our study probably is the first to establish the functions of ␣ 1D ARs in the heart, its roles may not be restricted to the modulation of I K1 .
The present study appears to favor the notion that modulation of I to in canine ventricular cells by ␣ 1 AR stimulation is primarily mediated by a subclass of ␣ 1 ARs, that is, ␣ 1A ARs. A similar decrease in I to upon ␣ 1 AR stimulation was also previously documented in other species such as rabbits (36) and rats (37,38), but no information regarding involvement of particular subtypes of ␣ 1 ARs was provided. In another study reported by Wang et al. (39) in rat ventricular cells, phenylephrine at 30 M was shown to reduce peak I to , and both of the ␣ 1A ARselective antagonists 5-methylurapidil and (ϩ)niguldipine (0.1 M each) and the irreversible ␣ 1B AR-subtype antagonist CEC (100 M) blocked the phenylephrine effect on I to . They concluded that stimulation of both ␣ 1 AR subtypes contributes to the phenylephrine-induced reduction in I to of rat myocytes. However, it should be noted that CEC concentration used in this study was 100 M, high enough to inactivate subtypes (e.g. ␣ 1A and ␣ 1D ) other than ␣ 1B ARs.
For the signal transduction mechanisms underlying ␣ 1 AR modulation of I to , studies on cloned channels that generate I to -like K ϩ currents in heterologous expression systems demonstrated that activation of PKC reduces Kv4.2 and Kv4.3 (40).  This result is in good agreement with ours, which points to an important role of PKC activation mediated by ␣ 1 AR stimulation in regulating I to , particularly when considering that Kv4.3 is the major molecular component of native I to in dogs (41). The results from previous studies on PKC modulation of inward rectifier K ϩ channels (Kir) have been controversial. For the cloned Kirs, the study from Henry et al. (42) convincingly demonstrated that PKC activation by phorbol 12-myristate 13-acetate or phorbol 12,13-dibytyrate significantly inhibited Kir2.3 but did not alter Kir1.1 and Kir2.1. In light of our previous finding that Kir2.1 is the most abundantly expressed Kir subunit in human hearts (22), our present data are in line with the results from the study by Henry et al. Yet Fakler et al. (43) reported that stimulation of PKC by 12-O-tetradecanoylphorbol 13-acetate suppressed Kir2.1. It is unclear whether this is due to the use of different PKC activators by the two laboratories. Similarly, in native cells no consistent data have been reported. In the study performed by Braun et al. (34) it was shown that ␣ 1 AR inhibition of I K1 in rabbit atrial cells did not depend on PKC activation, and direct PKC activation or inhibition did not affect I K1 either. In contrast, one study conducted in human atrial myocytes suggested that ␣ 1 AR inhibition of I K1 is mediated by PKC activation (9). One possible explanation for the discrepancy is that different species might have distinct molecular compositions of I K1 because to date no less than 10 different Kir cDNAs have been cloned (44).
Several lines of evidence from the present study suggest that suppression of I K1 by ␣ 1 AR stimulation in canine ventricular myocytes is mediated by CaMKII activation. To date, no other studies have published regarding CaMKII modulation of I K1 . However, CaMKII modulation of I to or the cloned channels expressing I to -like currents has been documented in several studies. Consistent in all these studies is the reduction of current amplitude. In the present study, we show that Phen at a concentration of 10 M mainly activates PKC pathway, as indicated by inhibitor experiments, PKC assay, and CaMKII immunoblotting analysis. We therefore speculate that I to modulation by 10 M Phen is mainly mediated by PKC phosphorylation, whereas I to reduction by 100 M Phen might be the consequence of combined PKC and CaMKII activation. This is supported by our data showing that inhibition of CaMKII partially reversed I to reduction caused by Phen at 100 M but not at 10 M. In addition, 100 M Phen slightly slowed the inactivation kinetics of canine I to (Fig. 1A), which was not seen with 10 M Phen.
The ␣ 1A AR is generally far more efficient in stimulating PKC activation than the ␣ 1D AR (19). For example, Taguchi et al. (17) showed that Phen significantly stimulated PKC in rat-1 fibroblasts stably expressing ␣ 1A ARs and ␣ 1B ARs but not ␣ 1D ARs. Our data are consistent with this notion. Intriguingly, a study performed in a vascular smooth muscle cell line (AC01) (18) demonstrated that the ␣ 1D ARs, although representing the minor population compared with the ␣ 1B ARs in this cell, are the main mediators of phosphoinositide/Ca 2ϩ signaling. Moreover, based on their experimental data from isolated hepatocytes, Butta et al. (45) concluded that there are at least two major ␣ 1 AR signaling pathways; one is PKC-dependent and independent of variations in free cytosolic Ca 2ϩ , and the other one is dependent on variations in free cytosolic Ca 2ϩ but is PKCindependent. Actually, the ability of ␣ 1 AR to activate CaMK has been previously realized. The study reported by Guo et al. (46) found that CaMK contributed to the ␣ 1 AR-mediated decrease in Kv1.5 K ϩ channel expression in cultured newborn rat ventricular cells. However, it was not characterized which subtype of ␣ 1 ARs is responsible for the effect in these studies. Our data suggest that ␣ 1A ARs are mainly associated with PKC activation, whereas ␣ 1D ARs are primarily coupled to CaMKII activation. Yet it should be noted that our data do not allow us to reach a conclusion on how ␣ 1A ARs are coupled to PKC and how ␣ 1D ARs are coupled to CaMKII. More detailed studies are necessary for verifying this notion and for delineating the subtype-specific signaling coupling mechanisms.