Functional Cross-Talk between the α1- and β1-Adrenergic Receptors Modulates the Rapidly Activating Delayed Rectifier Potassium Current in Guinea Pig Ventricular Myocytes

The rapidly activating delayed rectifier potassium current (IKr) plays a critical role in cardiac repolarization. Although IKr is known to be regulated by both α1- and β1-adrenergic receptors (ARs), the cross-talk and feedback mechanisms that dictate its response to α1- and β1-AR activation are not known. In the present study, IKr was recorded using the whole-cell patch-clamp technique. IKr amplitude was measured before and after the sequential application of selective adrenergic agonists targeting α1- and β1-ARs. Stimulation of either receptor alone (α1-ARs using 1 μM phenylephrine (PE) or β1-ARs using 10 μM xamoterol (Xamo)) reduced IKr by 0.22 ± 0.03 and 0.28 ± 0.01, respectively. The voltage-dependent activation curve of IKr shifted in the negative direction. The half-maximal activation voltage (V0.5) was altered by −6.35 ± 1.53 and −1.95 ± 2.22 mV, respectively, with no major change in the slope factor (k). When myocytes were pretreated with Xamo, PE-induced reduction in IKr was markedly blunted and the corresponding change in V0.5 was significantly altered. Similarly, when cells were pretreated with PE, Xamo-induced reduction of IKr was significantly attenuated. The present results demonstrate that functional cross-talk between α1- and β1-AR signaling regulates IKr. Such non-linear regulation may form a protective mechanism under excessive adrenergic stimulation.


Introduction
The human ether-a-go-go related gene (hERG) encodes the pore-forming subunit of the voltage-dependent potassium channel that conducts the rapidly activating delayed rectifier potassium current (I Kr ) [1,2]. I Kr exhibits slow activation and deactivation kinetics, coupled with rapid voltage-dependent inactivation and recovery from inactivation. These unique features of I Kr make it a critical repolarizing current in ventricular myocytes. Indeed, disruptions in I Kr have been shown to underlie abnormal action potential repolarization and to promote arrhythmogenic early afterdepolarization triggers, leading to sudden death in congenital and acquired cardiovascular disorders, including the long QT syndrome [3].
The complexity of the adrenergic regulation of hERG is underscored by the fact that during emotional or physical stress, catecholamines bind to multiple adrenoreceptors rather than act selectively on specific ones. Rorabaugh et al. demonstrated that stimulation of α 1 -ARs could down-regulate β 1 -AR-mediated inotropy in the mouse heart [11]. Moreover, cross-activation of PKA, PIP2 and PKC can regulate I Na [12], the slowly activating delayed rectifer K + current (I Ks ) [13,14] and I Kr [4,9,15,16]. Indeed, adrenergic signal transduction pathways may exert complex inhibitory effects on cardiac hERG/I kr currents via multiple mechanisms that potentially involve the intracellular second messenger cAMP, protein kinases A and C, and possibly other regulatory components of the hERG macromolecular complex, including minK, MiRP1, and 14-3-3. Signaling "cross talk" between α 1 -and β 1 -adrenergic cascades might also be involved, although direct experimental evidence is lacking.
We hypothesized that due to complex cross-talk between α 1 -and β 1 -adrenergic cascades, I Kr would exhibit non-linear responses to combined adrenergic stimulation. In other words, the combined activation of both pathways in terms of I Kr inhibition would not reflect the sum of the individual effects of stimulating either axis alone. To address this hypothesis, we designed two experimental groups, which are referred to as P pre + X and X pre + P. In the P pre + X group, 1 μM Phenylephrine (PE) was applied for 9 min resulting in a steady state I Kr inhibition, followed by co-application of 10 μM xamoterol (Xamo) and 10 μM PE for another 9 min. Conversely, in the X pre + P group, 10 μM Xamo was pre-applied for 9 min, followed by simultaneous application of both 1 μM PE and 10 μM Xamo for another 9 min. Hence, we were able to compare the reductions and shifts in activation curves of I Kr caused by exposure to PE alone, Xamo alone, and PE plus Xamo. More importantly, we compared PE-induced I Kr reduction in the presence of Xamo, and Xamo-induced I Kr reduction in the presence of PE, as well as shifts in the activation curve of I Kr under varying conditions.

Confirmation of the Rapidly Activating Delayed Rectifier Potassium Current (I Kr ) Identity and Stability
When 1 μM dofetilide, a specific I Kr blocker, was applied to the extracellular solution, the I Kr tail current was almost completely eliminated ( Figure 1A). The I Kr tail current at the return pulse of −40 mV after depolarizing to +40 mV, exhibited a marked time-dependent decrease to negligible levels within 10 min ( Figure 1B). We plotted the tail current-voltage relationship at 0 and 20 min during application of dofelitide ( Figure 1C) and found that the I Kr tail current was completely abolished by dofetilide. These data indicated that no other contaminating currents contributed to the I Kr tail current under our experimental conditions. The I Kr tail current at the return pulse of −40 mV from a test potential of +40 mV, was stable for at least 20 min of perfusion with control extracellular solution ( Figure 1D,E). The tail current-voltage curves at 0 and 20 min were almost identical ( Figure 1F) indicating presence of minimal tail current rundown under our experimental conditions.

Cell Capacitance and Basic Gating Data in Different Groups
The two groups (P pre + X and X pre + P) that were defined above were studied. Myocyte size, measured as cell capacitance, was comparable (p = 0.82, Figure 2A) between groups (161.80 ± 9.23 pF for P pre + X and 165.70 ± 13.93 pF for X pre + P, n = 7 each). Two key characteristics of the activating curve (V 0.5 and k) were measured in both groups during exposure of myocytes to the control bath solution. As shown in Figure 2B, V 0.5 was not significantly different between the P pre + X (−1.93 ± 2.58 mV) and the X pre + PE (−5.22 ± 1.73 mV) groups (p = 0.31, n = 7). Similarly, K was also comparable (p = 0.82, Figure 2C) between groups (11.92 ± 1.33 for P pre + X and 12.35 ± 1.54 for X pre + P).

Figure 2.
Cell capacitance and basic gating data in different groups. (A) Cell capacitance in the P pre + X and X pre + P groups, showing no significant difference (n = 7); (B,C) The half-maximal activation voltage (V 0.5 ) and the slope factor (k) of myocytes at basal conditions in the two groups also exhibit no significant difference (n = 7). These data indicate that myocytes used in either group exhibit comparable properties at baseline. "ns" indicates "not significant".

Cellular Electrophysiology Data of the P pre + X Group
When myocytes were exposed to 1 μM PE alone, the I Kr tail current decreased rapidly, reaching a steady-state nadir within 4-6 min ( Figure 3A,B). After 9 min of PE exposure, I Kr tail current density measured at +40 mV decreased from 0.80 ± 0.04 to 0.62 ± 0.01 pA/pF, which means I Kr tail current decreased to (78.33 ± 3.19)% compaired to the basal I Kr tail current (n = 7; p < 0.001; Figure 3C). Voltage-dependent activation of I Kr exhibited a trend towards a shift in the negative direction ( Figure 3D), with the half-maximal activation voltage (V 0.5 ) changing from −1.93 ± 2.58 to −8.24 ± 1.88 mV (n = 7; p = 0.06; Figure 3E). Similarly, the change in slope factor (k) from 11.92 ± 1.33 to 10.63 ± 1.10 (n = 7; p = 0.42; Figure 3F) did not reach statistical significance. Co-application of Xamo decreased I Kr tail current density even further to 0.54 ± 0.01 pA/pF. This additional reduction in current density was significant when compared to baseline control levels (n = 7; p < 0.001; Figure 3C) as well as to those achieved with PE alone (n = 7; p < 0.05; Figure 3C). As such, during α 1 -AR activation, the Xamo-mediated decrease was only 0.13 ± 0.01 (n = 7; Figure 7). The voltage-dependent activation curve of I Kr remained unchanged compared with PE alone (n = 7, Figure 3D), with V 0.5 changing to −11.15 ± 2.15 mV, (n = 7; p < 0.05, compared to the basal V 0.5 ; Figure 3E), and k changing to 10.82 ± 0.80 (n = 7; p = 0.49, compared to the basal k; Figure 3F). Shown in Figure 4 are I Kr tail current density measurements obtained over a wide range of test voltages before and after treatment of myocytes with PE alone or combined PE + Xamo. Neither PE alone nor PE + Xamo were sufficient to alter I Kr tail current density compared to pre-treatment levels at −40, −20, and 0 mV. Interestingly, at +20 mV, the combined treatment (PE + Xamo) but not PE alone was associated with a significant reduction in tail current density (n = 7; p < 0.05; Figure 4E). Finally, at +40 mV, both PE alone and PE + Xamo were sufficient to elicit significant reductions in tail current density as compared to basal pre-treatment levels (n = 7; each p value was less than 0.001; Figure 4F). . Cellular electrophysiology data of the P pre + X group; (A) shows the representative traces of time-dependence of the relative current reduction by phenylephrine (PE, 1 μM) and combined PE plus Xamo (Xamo, 10 μM); (B) Typical original I Kr tail currents recorded at return pulse after depolarizing to +40 mV at basal conditions and after application of PE and PE + Xamo; (C) Comparison of the relative current at baseline, following 9-min exposure to PE, and following 9-min exposure to PE + Xamo (n = 7; *** p < 0.001, vs. basal; # p < 0.05, vs. PE); (D) The plots of I tail /I tail.max vs. membrane voltage at three different conditions, fit with the single-power Boltzmann equation: reflecting the activation kinetics. Here, I Kr tail currents were induced by protocol II: holding potential −40 mV, test pulses from −40 to +40 mV in 10 mV increments (duration 225 ms), return pulse to −40 mV (duration 775 ms); (E,F) V 0.5 and k of the myocytes measured at three different conditions (n = 7). "ns" indicates "not significant"; * indicates p < 0.05). . I Kr tail currents at different depolarization levels in group P pre + X group. (A) I Kr tail current densities (pA/pF) measured at different membrane voltages before and after treatment with PE alone or PE + Xamo; (B-F) Comparison of I Kr tail current densities at baseline and following PE alone and PE + Xamo treatment. I Kr tail currents were seperately measured when the test pulse was at −40, −20, 0, +20, and +40 mV (n = 7). "ns" indicates "not significant"; * p < 0.05; *** p < 0.001.
Furthermore, the I Kr tail currents at different depolarization levels were tested and compared before and after exposure of myocytes to PE alone or PE + Xamo ( Figure 6A). As shown in Figure 6B,C, the I Kr tail currents measured at −40 and −20 mV did not exhibit statistical differences across groups (n = 7). At 0 mV, exposure of myocytes to combined PE + Xamo but not PE alone resulted in a significant reduction in current density compared to basal pretreatment levels (n = 7; p < 0.05; Figure 6D). In contrast, at +20 and the I Kr tail current was statistically different to basal one at +20 and +40 mV, both PE alone as well as PE + Xamo were sufficient to elicit significant decreases in current density as compared to basal pretreatment levels (n = 7; all p value was less than 0.05; Figures 6E,F).

Discussion
The main findings of the present report are as follows: (1) acute activation of α 1 -ARs produce comparable effects on I Kr tail current density to β 1 -ARs; (2) acute activation of α 1 -AR in the presence of β 1 -AR activation elicits a minor decrease in I Kr tail current, which is statistically different from that achieved by α 1 -AR activation alone; (3) similarly, acute activation of β 1 -AR in the presence of α 1 -AR activation induces a very small decrease in I Kr tail current, which again is statistically different from that achieved by β 1 -AR activation alone. The blunted I Kr response to concomitant adrenergic activation is suggestive of a protective feedback regulatory mechanism that acts to maintain the I Kr tail current density in the wake of excessive catecholamine stress, modeled in vitro in our study as combined PE and Xamo exposure. tail current after application of various AR agonists. Column α 1 represents the percent decrease in I Kr tail current by application of α 1 -AR agonist alone; and column (α 1 )β 1 represents the β 1 -AR mediated percent decrease of I Kr after pre-activation of α 1 -AR; similarly, column β 1 stands for β 1 -AR alone; column (β 1 )α 1 stands for α 1 -AR mediated percent inhibition of I Kr after pre-stimulation of β 1 -AR (n = 7); (B,C) Comparison of activation shifts in corresponding V 0.5 and k under different conditions. The four columns, α 1 , (α 1 )β 1 , β 1 , and (β 1 )α 1 represent conditions of α 1 -AR activation alone, β 1 -AR activation in presence of α 1 -AR activation, β 1 -AR activation alone, and α 1 -AR activation in presence of β 1 -AR, respectively (n = 7). "ns" indicates "not significant"; * p < 0.05; ** p < 0.01; *** p < 0.001.
In the present study, the finding that PE or Xamo is able to decrease the I Kr tail current via stimulation of α 1 -ARs or β 1 -ARs is consistent with previous reports [5,7,9,[17][18][19][20], including our own published work [21] in HEK-293 cells, CHO cells, Xenopus oocytes, and native ventricular cardiomyocytes. The I Kr tail current exhibits a concentration-dependent decrease after acute stimulation of α 1 -ARs or β 1 -ARs within 4-6 min, reaching steady-state levels within 7-9 min. Our choice of drug concentrations was guided by previous work, in which we determined the IC 50 values of PE and Xamo to be approximately 0.9 and 6.4 μM, respectively [22]. The I Kr tail current induced by depolarization to +40 mV was 0.78 ± 0.03 or 0.72 ± 0.01 of the basal I Kr tail current after application of 1 μM PE or 10 μM Xamo, respectively (p < 0.05). The I Kr tail current activation curve exhibited a minor shift in the repolarizing direction in response to α 1 -AR stimulation that was comparable to that achieved by β 1 -AR activation ( Figure 5B,C), suggesting that the gating kinetics of the hERG channel was not markedly affected by acute α 1 -AR or β 1 -AR stimulation separately. Characterized by slow activation and deactivation kinetics, and rapid voltage-dependent inactivation and recovery from inactivation kinetics, I Kr encoded by hERG is indeed a critical component of action potential repolarization in both atrial and ventricular myocytes of most species, including humans [2,3,23,24]. Thus, excessive I Kr inhibition causes marked repolarization delays that result in prolongation of the action potential at the cellular level, and the QT-interval of the electrocardiogram at the body surface level; thereby promoting the incidence of arrhythmogenic early afterdepolarizations, and polymorphic ventricular tachycardia. As such, our in vitro findings have direct relevance to clinical scenarios, in which patients with inherited or acquired long QT syndrome (LQTS) experience stress-related arrhythmias, in many cases leading to sudden cardiac death.
Of note, co-application of both β 1 -and α 1 -AR agonists resulted in a relatively small additional inhibitory effect on I Kr tail current compared to the selective activation of either receptor alone. Importantly, however, the decrease in I Kr tail current produced by PE in the presence of Xamo was significantly different from that achieved by α 1 -AR activation alone (0.14 ± 0.03 vs. 0.22 ± 0.03, p < 0.05). Similarly, the decrease in I Kr tail current by β 1 -AR stimulation in the wake of α 1 -AR activation was significantly different from that produced by β 1 -AR activation alone (0.13 ± 0.01 vs. 0.28 ± 0.01, p < 0.001). This suggests that pre-activation of β 1 -ARs markedly suppresses the inhibitory effect of α 1 -AR activation on I Kr tail current, and that pre-activation of α 1 -ARs produces an even stronger modulatory effect on the inhibition of I Kr by β 1 -AR. Indeed, pre-activation of one adrenoreceptor subclass dramatically restricts the inhibitory effect of the other subclass on I Kr tail current. As such, there appears to be significant cross-talk in the regulation of I Kr by acute adrenergic stimulation of AR receptors. We propose that this tight regulatory mechanism acts to protect against excessive I Kr inhibition, and therefore action potential prolongation, under conditions of extreme stress and associated catecholamine release.
Acute nonselective α 1 -AR activation suppresses the positive inotropic effect of β-AR activation and associated cAMP generation [25]. Under certain conditions, α 1 -and β-AR signaling pathways exhibit synergistic effects [26]. Moreover, α 1 -and β-AR interactions modulate the L-type calcium current [27], and sustained activation of PKC-epsilon leads to a blunted response of the current to α 1 -and β-AR signaling [28]. In addition, PKC activation cross-activates PKA to modulate the cardiac Na + current [12]. Moreover, both PKA and PKC regulate the cardiac I Ks in a mutually exclusive manner [13], and the channel phosphorylation by PKA cross-activates PLC-dependent regulation, which activates downstream PKC [14]. Therefore, the linear signaling paradigm has given way to a complex multidimensional "signalome" in which an individual adrenoceptor can dynamically couple to multiple signaling proteins in a temporally and spatially regulated manner resulting in pharmacologically and functionally distinct receptor populations [29][30][31][32]. Remarkably, mechanistic studies focusing on the regulation of ionic currents by AR signaling have been rapidly translated to the bedside [33]. We have previously demonstrated that I Kr tail current is inhibited by acute stimulation of α 1 -ARs using PE [21]. Furthermore, PE-mediated inhibition of the current was significantly attenuated by the PKC inhibitor chelerythrine and the PKA inhibitor KT5720, suggesting that activation of PKA and/or PKC may play an important role in mediating the effects of α 1 -ARs on native I Kr current in guinea-pig ventricular myocytes. Coincidentally, Thomas and colleagues [9] reported similar phenomena in Xenopus laevis oocytes heterologously coexpressing hERG channels and human α 1A -ARs. In contrast, Bian et al. [17] found that pretreatment with chelerythrine augmented the inhibitory effects of α 1 -ARs on I Kr current. In β 1 -AR regulation of I Kr , Karle et al. found that inhibiting PKA may attenuate the reduction effect of xamotorol on I Kr [5], which is consistant with our previous study. Moreover, we also found that PKC inhibition effectively attenuated the effect of Xamo on I Kr [34]. This suggests that, in addition to PKA, PKC is also activated as part of the β 1 -AR signaling pathway in the regulation of I Kr tail current. Despite that, both PKA and PKC are among the most important signal transduction pathways that modulate the I Kr /hERG current, and they may act in a cross-activation manner, an important issue which will require further investigation.

Guinea Pig Ventricular Myocyte Isolation and Electrophysiological Recordings
All experiments were performed in accordance with Animal Care Protocols approved by the Nanjing Medical University Institutional Animal Care and Use Committee (SCXK2002-0031, Experimental animal production license of Jiangsu Province). Single left ventricular myocytes were isolated from the hearts of healthy male adult guinea pigs (300 ± 50 g) as described previously [21]. Cells were transferred to a temperature controlled recording chamber in which they were continuously perfused with the bath solution. Temperature was maintained at 37 ± 0.5 °C by a temperature control system (TC-324B, Warner, Hamden, CT, USA). Whole-cell patch-clamp recordings were performed using an EPC-9 amplifier (HEKA Electronics, Lambrecht/Pfalz, Germany). Pipettes, filled with the pipette solution, had resistances of 1-3 MΩ. The flow rate of the bath solution through the chamber was maintained at 1-2 mL/min.

Cellular Electrophysiology Protocols
I Kr /hERG currents were measured using a two-step protocol, referred to as Protocol I. From a holding potential of −40 mV, currents were activated by a variable test pulse from −40 to +40 mV (in 20 mV steps, 225 ms duration), followed by a return pulse to −40 mV (duration 775 ms) to evoke outward tail currents. Signals were analog-filtered at 2380 Hz, and the sampled interval was 10,000 Hz. The effects of α 1 -and β 1 -AR agonists were investigated on peak tail currents after the return pulse to +40 mV.
An additional two-step protocol (Protocol II) was also used. Specifically, from a holding potential of −40 mV, currents were activated by a variable test pulse from −40 to +40 mV (in 10 mV steps, 225 ms duration), followed by a return pulse to −40 mV (duration 775 ms) to evoke outward tail currents. This protocol was applied to measure each peak tail current at repolarization after different test voltages. Activation curves were fit to a single-power Boltzmann equation: I tail = I tail.max /[1 + exp(V 0.5 − V)/k], where I tail indicates the tail current, V represents the test pulse potential, V 0.5 refers to the half-maximal activation voltage, and k is the slope factor.
For stock solutions, nifedipine and chromanol 293B were dissolved in dimethyl sulfoxide (DMSO) (10 mM concentration); Xamo in distilled water (10 mM concentration); PE and dofetilide in distilled water (1 mM). All stock solutions were stored at −20 °C except nifedipine which was stored at 4 °C. Before experiments, aliquots of the stock solutions were diluted with the extracellular solution to the desired test concentrations of 1 μM (PE and dofetilide) and 10 μM (Xamo, nifedipine and chromanol 293B). The final concentration of DMSO was less than 0.5% in extracellular solution, which was determined to have no confounding effects on the currents of interest.

Statistical Analysis
Currents were acquired with Pulse + Pulsefit V8.53 (HEKA Electronics, Lambrecht/Pfalz, Germany) and analyzed by SPSS 18.0 software (SPSS Inc., Chicago, IL, USA). Data were expressed as mean ± S.E.M. Statistical significance was evaluated using the unpaired Student's t test. Multiple comparisons were analyzed using one-way analysis of variance (ANOVA), with a post-hoc comparison using a Newman-Keuls test. A p value less than 0.05 was considered statistically significant.

Conclusions
hERG channel activation is strongly modulated by acute stimulation of α 1 -ARs and β 1 -ARs. Remarkably, stimulation of either receptor class markedly attenuates the inhibitory effects of the other class on I Kr Further investigations are required to elucidate the molecular mechanisms that underlie the functional cross-talk or to identify putative intermediate proteins in the signal transduction pathways involved in modulation of I Kr current via adrenergic activation.