点击显示 收起
Department of Safety Pharmacology, Pfizer Global Research and Development, Groton, Connecticut (J.Z., J.A.B., X.C., B.J.K., W.A.V., Z.S., J.S.C.)
Central Nervous System Chemistry, Pfizer Global Research and Development, Ann Arbor, Michigan (C.E.A.-S.)
Abstract
A variety of drugs has been reported to cause acquired long QT syndrome through inhibition of the IKr channel. Screening compounds in early discovery and development stages against their ability to inhibit IKr or the hERG channel has therefore become an indispensable procedure in the pharmaceutical industry. In contrast to numerous hERG channel blockers discovered during screening, only (3R,4R)-4-[3-(6-methoxyquinolin-4-yl)-3-oxo-propyl]-1-[3-(2,3,5-trifluoro-phenyl)-prop-2-ynyl]-piperidine-3-carboxylic acid (RPR260243) has been reported so far to enhance the hERG current. In this article, we describe several potent mechanistically distinct hERG channel enhancers. One example is PD-118057 (2-{4-[2-(3,4-dichloro-phenyl)-ethyl]-phenylamino}-benzoic acid) which produced average increases of 5.5 ± 1.1, 44.8 ± 3.1, and 111.1 ± 21.7% in the peak tail hERG current at 1, 3, and 10 e, respectively, in human embryonic kidney 293 cells. PD-118057 did not affect the voltage dependence and kinetics of gating parameters, nor did it require open conformation of the channel. In isolated guinea pig cardiomyocytes, PD-118057 showed no major effect on INa, ICa,L, IK1, and IKs. PD-118057 shortened the action potential duration and QT interval in arterially perfused rabbit ventricular wedge preparation in a concentration-dependent manner. The presence of 3 e PD-118057 prevented action potential duration and QT prolongation caused by dofetilide. "Early after-depolarizations" induced by dofetilide were also completely eliminated by 3 e PD-118057. Although further investigation is warranted to evaluate the therapeutic value and safety profile of these compounds, our data support the notion that hERG activation by pharmaceuticals may offer a new approach in the treatment of delayed repolarization conditions, which may occur in patients with inherited or acquired long QT syndrome, congestive heart failure, and diabetes.
Drug-induced (acquired) long QT syndrome, an effect manifested as prolongation of the QT interval on the surface electrocardiogram (ECG), has drawn increasing attention from regulatory agencies and the pharmaceutical industry in recent years (De Ponti et al., 2000, 2001; Fermini and Fossa, 2003). The presence of delayed repolarization favors the genesis of "early after-depolarization" (EAD), which can initiate an arrhythmia referred to as "triggered activity" (Zabel et al., 1997). In addition, prolongation of the QT interval by drugs is often associated with an increased heterogeneity of cardiac repolarization (Antzelevitch, 2004), a potential substrate for a re-entry mechanism responsible for the maintenance of arrhythmia. One particular type of arrhythmia, torsade de pointes, may cause syncope events and/or degenerate into ventricular fibrillation and death.
Because the majority of these drugs prolong the QT interval through inhibition of the rapidly activating delayed rectifier potassium current, IKr, evaluating the propensity of a compound to inhibit this channel or its molecular counter-part, human ether-e?go-go-related gene (hERG) encoded channel, has become an indispensable screening procedure in drug development (Fermini and Fossa, 2003). It is interesting that the hERG channel seems to be the most promiscuous among all of the known voltage-gated potassium channels, with drugs from a wide range of chemical structures and therapeutic categories capable of inhibiting the channel (De Ponti et al., 2000, 2001). In the pharmaceutical industry, thousands of compounds are screened each day in the hERG assay and an estimated 30% of compounds tested possess hERG-blocking activity to varying degrees. However, only one compound, RPR260243, has recently been reported (Kang et al., 2005) to enhance the hERG channel. This compound increases the hERG current mainly by slowing its deactivation kinetics. In this article, we report several new hERG channel enhancers with dramatic current-enhancing effects and a distinct mechanism. PD-118057, as a representative among these compounds, is able to prevent and reverse QT prolongation and associated arrhythmias (EAD) induced by a selective IKr blocker, dofetilide, in the arterially perfused rabbit ventricular wedge preparation.
Materials and Methods
All of the animal experiments were conducted in accordance with the regulations of the U. S. National Institutes of Health (NIH Publication 8523, revised 1996) and European guidelines and were approved by the Pfizer Institutional Animal Care and Use Committee.
Isolation of Myocytes from Guinea Pig Hearts. Ventricular myocytes were isolated from male Hartley guinea pigs using an enzyme-digestion method as described previously (Cordes et al., 2005). In brief, a Langendorff heart was established and first perfused with oxygenated Ca2+-free isolation solution (37°C) for 5 min. This was followed by perfusion with an enzyme-containing solution [collagenase type 2 (Worthington Biochemicals, Freehold, NJ) and protease, type XII (Sigma-Aldrich, St. Louis, MO) for 8 to 10 min. The ventricles were then chopped into pieces in storage solution and filtered through a mesh, and the cell suspension was stored at room temperature for at least 1 h before use.
Patch-Clamp Recording. Stable human embryonic kidney (HEK)-293 cells expressing the hERG channel (Zhou et al., 1998) were licensed from Wisconsin Alumni Research Foundation or generated in-house. Methods for cell culture and whole cell patch-clamp studies on the hERG channel were reported previously (Volberg et al., 2002; Sun et al., 2004; Cordes et al., 2005). EPC-9 (HEKA, Lambrecht/Pfalz, Germany) or MultiClamp 700A (Axon Instruments Inc., Union City, CA) amplifiers were employed to record the current controlled by Pulse + PulseFit 8.40 (HEKA) or pClamp 8.2 (Axon Instruments) software through an interface. The hERG, IK1, and IKs currents were measured at 35°C (maintained by a TC-344B temperature controller; Warner Instruments, Hamden, CT). INa, ICa, and action potentials were recorded at room temperature. Voltage protocols used for eliciting each current are described in the text. Action potentials recorded from isolated myocytes were elicited using the whole-cell configuration or perforated-patch technique when 120 e蘥/ml amphotericin B was included in the pipette solution.
Wedge Preparation Study. Under anesthesia induced by 30 to 35 mg/kg ketamine HCl (i.v.) after 5 mg/kg xylazine (i.m.), the heart of a female New Zealand White rabbit (2.5-5.5 kg) was removed and placed in ice-cold (4-10°C) 9% O2- and 5% CO2-saturated cardioplegic solution (129 mM NaCl, 24 mM KCl, 0.9 mM NaH2PO4, 20 mM NaHCO3, 1.8 mM CaCl2, 0.5 mM MgSO4, and 5.5 mM glucose). The left main coronary artery or its major branch (normally circumflex branch) was cannulated and perfused with the cardioplegic solution to wash out the intravascular blood. A transmural left ventricular wedge from the anterior wall was dissected, and the major leaking vessels were ligated. The tissue was then placed in a tissue bath and perfused with 36 ± 0.5°C Tyrode's solution (129 mM NaCl, 4 mM KCl, 0.9 mM NaH2PO4, 20 mM NaHCO3, 1.8 mM CaCl2, 0.5 mM MgSO4, and 5.5 mM glucose; pH 7.35 when buffered with 95% O2 and 5% CO2). The perfusion pressure was maintained at 40 mm Hg and monitored through a pressure transducer connected with the PowerLab/8SP Data Acquisition System (ADInstruments Pty Ltd., Castle Hill, Australia). The tissue was paced with 150% suprathreshold stimuli at 1 Hz by a DS8000 Digital Stimulator (World Precision Instruments, Inc., Sarasota, FL) through platinum bipolar electrodes on the endocardial surface. Floating glass electrodes, with a resistance of approximately 10 to 20 M when filled with 2.7 M KCl, were placed in the epicardial or endocardial myocardium, respectively. Action potentials from both sites were amplified through an IX2-700 Dual Intracellular Preamplifier (Dagan, Minneapolis, MN). The transmural ECG was recorded by using two Ag/AgCl electrodes placed 1 cm away from epicardial and endocardial surfaces and fed into an EX1 Differential Amplifier (Dagan). All of the signals were monitored and recorded using Chart 5 software (ADInstruments) through the PowerLab/8SP system (ADInstruments). An equilibrium period of at least 1 h was allowed in each experiment before any data collection. Action potential duration (APD) parameters were analyzed using peak parameter extension within the Chart 5 program.
Chemicals and Solutions. For isolation of the myocytes, the calcium-free isolation solution was composed of 137 mM NaCl, 5.4 mM KCl, 10 mM HEPES, 1 mM MgCl2·6H2O, 0.33 mM NH2PO4·H2O, and 10 mM d-glucose, pH adjusted to 7.4 with NaOH. The storage solution contained 50 mM glutamic acid, 0.5 mM EGTA, 10 mM glucose, 10 mM HEPES, 40 mM KCl, 20 mM KH2PO4, 70 mM KOH, 3 mM MgCl2·6H2O, and 20 mM taurine, pH adjusted to 7.4 with NaOH. For the hERG current recording from the hERG-HEK-293 cells, the bath (Tyrode's) solution was composed of 137 mM NaCl, 4 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 10 mM glucose, and 10 mM HEPES, pH adjusted to 7.4 with NaOH. The pipette solution contained 130 mM KCl, 5 mM MgATP, 1 mM MgCl2, 10 mM HEPES, and 5 mM EGTA, pH adjusted to 7.2 with KOH. For the recording of the calcium currents, the bath solution was composed of 137 mM Tris, 1.8 mM CaCl2, 1 mM MgCl2·6H2O, 5 mM glucose, and 20 mM CsCl, pH adjusted to 7.4 with CsOH. The internal pipette solution was composed of 125 mM CsCl, 5 mM MgATP, 10 mM HEPES, 15 mM EGTA, and 20 mM TEA-Cl, pH adjusted to 7.2 with CsOH. For IK1 and IKs recordings, bath solution contained 137 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 10 mM HEPES, 0.33 mM NaH2PO4, and 10 mM glucose, pH adjusted to 7.4 with NaOH. CdCl2 (250 e) and 1 e dofetilide were added to block ICa,L and IKr, respectively. The pipette solution contained 125 mM potassium aspartate, 20 mM KCl, 1 mM MgCl2, 5 mM HEPES, 10 mM EGTA, and 5 mM MgATP, pH adjusted to 7.2 with KOH. For INa recording, the bath solution contained 15 mM NaCl, 0.5 mM CaCl2, 4 mM CsCl, 1 mM MgCl2, 125 mM TEA-Cl, 0.25 mM CdCl2, 10 mM HEPES, and 10 mM glucose, pH adjusted to 7.4 with CsOH. Pipette solution was composed of 5 mM NaCl, 125 mM CsOH, 125 mM aspartic acid, 20 mM TEA-Cl, 10 mM EGTA, 10 mM HEPES, and 5 mM MgATP, pH 7.2 with CsOH. The bath solution used to record the action potentials was the Tyrode's solution, and the pipette solution contained 130 mM KCl, 5 mM MgATP, 1 mM MgCl2, 10 mM HEPES, 5 mM EGTA, and 120 e蘥/ml amphotericin B, pH 7.2 with KOH.
All of the test compounds used in this study were synthesized at Pfizer Global Research and Development. Other chemicals were purchased from Sigma-Aldrich. Compounds were dissolved in dimethyl sulfoxide (DMSO) first as a stock solution and then added into the bath solution to a desired test concentration. DMSO concentration in the drug-containing solutions was limited to 0.3% (at this concentration, DMSO does not have any effect on the ionic currents of interest).
Data Analysis. Data were expressed as the mean ± S.E. Paired Student's t test was used to evaluate the significance of the difference between the means before and after application of drugs. Analysis of variance was applied to evaluate the multiple group data. A value of p < 0.05 was accepted as a statistically significant level. Curve fittings and graphing were performed using Clampfit in pClamp 8.2 bundle and Origin version 7.0 software (OriginLab Corp., Northampton, MA).
Results
Potentiation of the hERG Current by PD-118057 and Its Analogs. PD-118057 was first examined for its effect on the hERG potassium channel in hERG-HEK-293 cells. Currents were elicited by a voltage protocol described previously (Volberg et al., 2002; Cordes et al., 2005), which held the cell at -80 mV, and stimulated at 0.25 Hz with a step pulse to +20 mV for 1 s followed by a 0.5 V/s ramp to -80 mV (Fig. 1A, inset). A slow but significant increase of the current amplitude was observed at 1, 3, and 10 e, respectively (Fig. 1, A-C, n = 4-8, p < 0.01). The holding current and series resistance were monitored to rule out the possibility of a recording artifact. The current traces obtained in the presence of the drug, both at depolarization and repolarization phases, almost paralleled that of the control, indicating no apparent kinetic change and involvement of the endogenous currents in HEK-293 cells. Indeed, PD-118057 was also tested in three wild-type HEK-293 cells and showed either a negligible effect or a slight inhibition of the endogenous current (Fig. 1D). Moreover, the increased current, when measured at the repolarization tail, can be completely blocked by a high concentration of 10 e dofetilide (Fig. 1B), which was previously shown to have little effect on the endogenous current. The hERG current increase by PD-118057 at the concentrations tested required at least 5 min to reach a steady state and often led to a loss of recordings after the current had been substantially elevated (average 111% increase in the tail current at 10 e, n = 8), therefore, higher concentrations were not investigated. It is possible that the small HEK-293 cells were especially sensitive to the activation of the hERG potassium channel, because no apparent cytotoxicity was observed after exposure to the compound at higher concentrations, which were achieved in other in vitro and in vivo studies (data not shown).
Similar hERG-enhancing effects were observed in several structurally related analogs of PD-118057 (Table 1). Comparable or slightly stronger effects were noted with PD-198986, PD-307243, and PD-322388, whereas a less potent effect was obtained with PD-202091. Virtually no effect was induced by PD-117780 and PD-201583 at 10 e. The current traces after administration of these compounds demonstrated unchanged kinetics (data not shown) as did the experiments of PD-118057 (Fig. 1A), indicating a similar mechanism of action. Because of cell survival and compound solubility issues at high concentrations, a full concentration-response relationship could not be established. No signs of saturation of the hERG-enhancing effect were indicated for PD-118057 at 10 e, although >100% increase of the hERG current was induced.
Percentage increase of the peak tail current was calculated. Currents were elicited using a voltage protocol shown in Figure 1. Mean ± S.E., n = 4-8.
Additional mechanistic studies were completed using PD-118057 as a representative of the structural series. First, the voltage-dependent activation was investigated by using a series of 1-s depolarization pulses ranging from -70 to +40 mV from a holding potential of -80 mV (Fig. 2A, inset). Tail currents were elicited by a repolarization at -70 mV. The current amplitudes measured at the end of depolarization and the peak tail currents were averaged (n = 6) and plotted against membrane potentials, as shown in Fig. 2, B and C. Depolarization-activated currents increased with voltage initially and then decreased at voltages over -10 mV (Fig. 2B), a characteristic of the hERG channel resulting from voltage-dependent C-type inactivation. PD-118057 at 3 e increased the current at all of the voltages positive to -50 mV (p < 0.05). The voltage dependence of steady-state activation of the channel was described by fitting the tail currents using a Boltzmann function (Fig. 2C): I/Imax = 1/{1 + exp[(V1/2 - Vm)/k]}, where I represents the tail current, Vm is the test membrane potential, V1/2 is the half-maximal activation voltage, and k is the slope factor representing the steepness of the voltage dependence. No statistically significant difference was found in V1/2 between PD-118057 (-31.4 ± 2.3 mV) and the control (-29.6 ± 2.2 mV); however, a slightly steeper activation-voltage relationship was obtained in the presence of PD-118057, as indicated by increased slope factor k (8.1 ± 1.1 versus 6.1 ± 0.6 mV control, n = 6, p < 0.05).
The activation kinetics of the hERG channel in the presence and absence of 3 e PD-118057 was estimated by fitting the rising phase of the currents with a single exponential function. Acknowledging the methodological limitation of this approach at high-membrane potentials because of an overlapping inactivation process (Zhou et al., 1998), we only applied curve fitting to the current traces obtained at approximately -40 to 10 mV when the fast inactivation posed minimal interference to the analysis of much slower activation kinetics. The resulting time constants demonstrated little difference between the drug and control groups (Fig. 3B), indicating that PD-118057 did not affect activation kinetics of the hERG channel. Indeed, when the current sizes before and after drug administration were normalized, the two traces were superimposable (Fig. 3A), including the deactivation (tail) current at -70 mV. Further investigation of the deactivation kinetics were performed in a separate experiment in which a series of 4-s repolarization steps was given after a 1-s depolarization pulse at +60 mV from holding potential of -80 mV. Deactivation kinetics at each voltage was then described using single or double exponentials, and the resulting fast and slow exponentials were averaged and plotted in Fig. 3C. No major difference was found between the PD-118057 and control groups (n = 6, p > 0.05). Using the same protocol, the recovery kinetics from inactivation was obtained by fitting the rising phase of the current upon repolarization with a single exponential function. The resulting time constants in relation to the membrane voltages were presented in Fig. 3D, and the recovery phase of the current traces before and after application of PD-118057 were superimposable after the current amplitude was normalized (Fig. 3D, insets). These results indicated that PD-118057 did not affect the recovery kinetics from steady-state inactivation. However, slightly slower inactivation kinetics was observed after treatment of 3 e PD-118057, represented by statistically significant increases in time constants at all of the testing voltages (Fig. 4A). Here, the currents were elicited by membrane potentials ranging from -40 to +50 mV after a 200-ms prepulse at +60 mV and a 2-ms hyperpolarization at -100 mV. The current decay was then fitted with a single exponential. It should be noted that the instantaneous currents in these experiments were extremely large (often >10 nA postdrug). The details of the fast inactivation phase could easily be confounded by technical limitations (e.g., overlapping capacitance transient, voltage, and dynamic voltage errors), making it difficult to accurately assess the kinetics. Therefore, we followed up the study at room temperature (22°C) when fewer channels were activated and the gating kinetics was slowed down. As shown in Fig. 4B, no difference was seen in the time constants at any test potentials (n = 5, p > 0.05) under this recording condition. Another piece of supporting evidence was that, in a separate set of experiments, 3 e PD-198986, an equipotent close analog of PD-118057 with 3-trifluomethyl group replacing the 3-chlorine, did not exert any effect on the inactivation kinetics (n = 4, p > 0.05) (Fig. 4C) and other biophysical parameters (data not shown).
In Fig. 5, A and B, 3 e PD-118057 was allowed to super-fuse two cells separately while being either continuously stimulated (Fig. 5A) or held at -80 mV for 10 min before stimulation resumed (Fig. 5B). A similar extent of potentiation of the hERG current was observed after the pause of stimulation (Fig. 5B) compared with the current with continuous stimulation (Fig. 5A), indicating that the hERG-enhancing effect of PD-118057 does not require open conformation of the channel.
PD-118057 Does Not Affect INa, ICa,L, IK1, and IKs in Guinea Pig Ventricular Myocytes. To examine whether the effect of PD-118057 was selective to the hERG channel, we isolated ventricular myocytes from guinea pig hearts and used whole-cell patch-clamp technique to record INa, ICa,L, IK1, and IKs currents. These are the major currents in shaping the action potential besides IKr. ICa was elicited by a 250-ms depolarization to +10 mV after a 200-ms prepulse at -40 mV to inactivate the T-type calcium current, if present. Cells were held at -70 mV. As shown in Fig. 6, A and B, an 5-min application of 10 e PD-118057 did not significantly affect the L-type calcium current. To record IKs, we held the myocytes at -40 mV and stimulated them with a series of 2-s depolarization pulses at -20 to approximately +60 mV with a 20-mV increment (Fig. 6C). The tail IKs current was elicited at -40 mV. The IK1 current illustrated in Fig. 6D used a ramp voltage protocol (from -100 to +50 mV, 75 mV/s). To record INa, the extracellular Na+ concentration was reduced to 15 mM and the holding potential was -120 mV. A series of depolarization pulses were applied, ranging from -100 to 0 mV. Similar to the results on the ICa,L, 10 e PD-118057 did not produce any significant effect on IKs, IK1, and INa as shown in Fig. 6, C-E.
Effect of PD-118057 on Action Potential Duration. The specific hERG-enhancing effect by PD-118057 indicates its ability to shorten the action potential duration in native cardiac myocytes. To test this hypothesis, we first recorded action potentials in guinea pig ventricular myocytes by using the perforate-patch technique under current-clamping mode. Because of inherent technical limitations in achieving stable recordings of action potentials from single myocytes, we were unable to measure the effect. However, a trend of significant shortening was observed by PD-118057 at both 3 and 10 e, as shown in Fig. 6F for the current traces chosen from medians of the steady-state range of action potentials in an experiment. Similar results were observed in several other cells. To quantitatively determine the effect, we used floating electrodes to simultaneously record the action potentials from epicardial and endocardial sites and transmural ECG in arterially perfused rabbit ventricular wedge preparations at a frequency of 0.5 Hz. As shown in Fig. 7A, perfusion with solutions containing 3 and 10 e PD-118057 caused shortening of both epicardial and endocardial APDs and the QT interval on the transmural ECG. This effect normally developed slowly at 3 e and could continue on for approximately 2 to 3 h, and it was difficult to elute. At 10 e, the shortening often progressed until no action potentials could be induced if no other intervention was allowed. Data shown in Fig. 7A were chosen at 2 h after administration of 3 e PD-118057 and 30 min after 10 e. Shortenings of epicardial APD90, endocardial APD90, and QT interval by 10.8 ± 1.1, 17.7 ± 1.6, and 10.4 ± 2.6%, respectively, were observed at 3 e, and the decreases by 26.0 ± 5.9, 35.6 ± 6.0, and 25.6 ± 1.6%, respectively, were observed at 10 e (n = 4). It seems that the effect of PD-118057 was more dramatic in the endocardial region than in the epicardial region, resulting in a decreased transmural dispersion of repolarization [APD was 14 ± 7 and 7 ± 11 ms, respectively, at 3 and 10 e versus 39 ± 4 in control; endocardial (Endo) to epicardial (Epi)]. At 10 e, PD-118057 could invert the normal heterogeneity to a shorter epicardial APD than endocardial APD, which was manifested as a missing or inverted T wave on the transmural ECG (Fig. 7A).
PD-118057 Reverses Dofetilide-Induced APD/QT Prolongation and EADs. The shortening of APD and QT by PD-118057 indicated a potential of this compound in treating drug-induced QT prolongation and arrhythmias. As shown in Fig. 7, B and C, PD-118057 (B, right, and C, ) prevented the significant prolongation of APDs and QT interval caused by 3, 10, and 30 nM dofetilide, which was evident in the time-matched vehicle control (B, left, and C, ). Similar results were observed in three more experiments (stimulation frequency: 1 Hz). In this study, 3 e PD-118057 or vehicle was present from 30 min before 3 nM dofetilide to the end of 30 nM dofetilide treatment. In the vehicle-treated preparation, dofetilide at 3, 10, and 30 nM produced epicardial APD90 prolongations of 13, 32, and 56%, respectively, and endocardial APD90 prolongations of 15, 35, and 59%, respectively. In contrast, in the preparation pretreated with 3 e PD-118057, dofetilide at 3, 10, and 30 nM only produced average changes of 5.6, 2, and -20% in the epicardial APD90, respectively, and -1.5, -5.1, and -18.4% in endocardial APD90, respectively. Compared with the effect of PD-118057 alone shown in Fig. 7A, it seemed that PD-118057 was more effective in shortening the action potentials that had been prolonged. The lack of increase or even a decrease in APD by higher concentrations of dofetilide most probably reflected a continuously increased effect of PD-118057 during the course of the experiment. It is worthy to note that dofetilide and other selective IKr blockers, such as dl-sotalol (Shimizu et al., 1999), increase the transmural dispersion of repolarization (TDR) by more prominently increasing the APD in endocardium (or M cells in large animals) than in epicardium (Fig. 7B, vehicle group). PD-118057 (3 e) was shown to be effective in reversing this action, leading to a decreased TDR. Overcorrected TDR could result in a shorter endocardial than epicardial APDs and missing or inverted T waves (Fig. 7B).
In another set of experiments, the wedge was exposed to 10 nM dofetilide first while being stimulated at 0.5 Hz, a condition that typically facilitates the genesis of EAD (Xu et al., 2003; Joshi et al., 2004). As shown in Fig. 8, a phase 2 EAD was clearly present in both epicardial and endocardial action potential waveforms after application of 10 nM dofetilide. The APD prolongation and EADs induced by dofetilide were shown in previous studies to persist during the course of an experiment if no intervention was given. We then applied 3 e PD-118057 to the perfusate in the presence of dofetilide. After approximately 1 to 2 h of perfusion, the action potentials were almost restored to its control level and the EAD was completely eliminated. Similar results were obtained in two other experiments.
Discussion
PD-118057 and several of its analogs were shown in our study to significantly enhance the hERG current in HEK-293 cells within the test concentrations of 1 to 10 e. No major changes were found in the gating and kinetic properties of the hERG channel by 3 e PD-118057 with the exception of a small but statistically significant slowing of the inactivation time course. This small change in the inactivation kinetics is most probably attributed to technical limitation in accurately measuring the superfast inactivation process at 35°C, supported by the negative results from the room temperature experiment and from its close analog PD-198986 in a separate set of experiments.
The seven analogs reported in this article possess differing potencies with respect to increasing the hERG channel current. Although it is difficult to draw a conclusive structure-activity relationship based on the limited number of compounds, it seems that both the electrostatic properties of molecules and the distance between the two phenyl rings may play a role in the effect. For example, both PD-307243 and PD-322388 have a 3,4-disubstitution pattern on the halogenated phenyl ring, similar to the substitution pattern of PD-118057 and PD-198986. Removal of the halogen groups from the phenyl ring resulted in a near complete loss of the hERG-enhancing activity, as in the case with PD-117780 and PD-201583. The major difference here is that the substituents on the "Western" ring in these two compounds are electron-donating in character compared with the other active analogs that have halogen substituents known to be electronegative. On the other hand, a less effect was also induced by PD-202091, although it did have a halogenated phenyl ring. It seems that the lengthened linker (three atoms) between the halogenated and the attached phenyl rings, compared with a two-atom linker in PD-118057 and PD-198986, resulted in a less optimal structure in exerting the hERG-enhancing effect.
The hERG potassium channel has been reported to be regulated by a number of intracellular molecules and pathways, including cAMP and protein kinase A (Thomas et al., 1999; Cui et al., 2000), protein kinase C (Barros et al., 1998; Thomas et al., 2004), phospholipase C (Bian et al., 2001, 2004; Gomez-Varela et al., 2003), and Src tyrosine kinase (Cayabyab and Schlichter, 2002). Increase of the hERG current can be induced by the substrate of phospholipase C, phosphatidylinositol 4,5-bisphosphate (Bian et al., 2001, 2004), activation of Src tyrosine kinase (Cayabyab and Schlichter, 2002), and perhaps by modulations of other pathways. However, these regulatory mechanisms often lack specificity and, therefore, affect functions of other ion channels. In addition, regulations through these cellular processes are often associated with gating and/or kinetic changes and the magnitude of the effect is often limited in cell lines (e.g., HEK-293) with only endogenous expression of these proteins, if present. Although we cannot rule out the possibility that PD-118057 and its analogs act through indirect mechanisms to increase the hERG current, the unchanged biophysical property of the channel and such a dramatic enhancement of the current amplitude seem to support that PD-118057 is able to bind to the channel directly and increase its open probability. Further work at the single-channel level should help to confirm this hypothesis.
The magnitude of the hERG current potentiation by these compounds and the mechanism, if proved, seemed to contrast drastically with RPR260243, another hERG enhancer reported very recently (Kang et al., 2005). At 10 e, PD-118057 produced an average of 111% increase of the hERG current compared with only 15% by RPR260243 at the same concentration. RPR260243 dramatically slowed the deactivation kinetics, but PD-118054 had basically no major effect on gating or kinetic properties of the hERG channel. PD-118057 significantly shortened the action potential duration at the tested concentrations (3-10 e) in guinea pig ventricular myocytes, but RPR260243 had almost no effect until 30 e. Similar to RPR260243 in the guinea pig myocytes but more effectively, PD-118057 at 3 e was able to prevent and treat 10 nM dofetilide-induced APD or QT prolongation, increasing heterogeneity of repolarization and phase 2 EAD in arterially perfused rabbit left ventricular wedge. Phase 2 EADs are believed to be the major mechanism of the triggered activity responsible for initiation of the arrhythmia, and the increased TDR is considered to be a substrate for re-entry responsible for the maintenance of the arrhythmia (Antzelevitch, 2004; Shryock et al., 2004). Both parameters can be easily monitored in the ventricular wedge preparation (Shimizu et al., 1999; Medina-Ravell et al., 2003; Antzelevitch, 2004; Joshi et al., 2004) and shown to be highly predictive to the clinical outcome of drug-induced arrhythmia. Therefore, our study results have undoubtedly provided important preclinical evidence that potentiation of the hERG potassium channel by a pharmaceutical might be beneficial to individuals with delayed repolarization. Potential therapeutic indications may include prevention and treatment of arrhythmias in patients with long QT syndrome (congenital or acquired), congestive heart failure, and diabetes. Combinations of these compounds with QT-prolonging agents may potentially mitigate their adverse cardiac effects while retaining their benefits. On the other hand, it should also be noted that PD-118057 could cause QT interval shortening and decrease the normal heterogeneity of repolarization in rabbit ventricular wedge. It is unclear at this stage whether this effect observed in vitro would translate to a proarrhythmic risk in man. More investigative work is necessary to further assess the potential therapeutic value and safety profile of PD-118057 and its analogs.
Acknowledgements
We thank Dr. Mei-Hua Tu for helpful discussions. We also wish to acknowledge Annette Sakkab-Tan, Chung Choi, and Yingjie Lai for the synthesis of test compounds. These compounds were made through a collaboration with Yamanouchi Pharmaceutical Company, Ltd.
References
Antzelevitch C (2004) Cellular basis and mechanism underlying normal and abnormal myocardial repolarization and arrhythmogenesis. Ann Med 36 (Suppl 1): 5-14.
Barros F, Gomez-Varela D, Viloria CG, Palomero T, Giraldez T, and de la Pena P (1998) Modulation of human erg K+ channel gating by activation of a G protein-coupled receptor and protein kinase C. J Physiol (Lond) 511: 333-346.
Bian J, Cui J, and McDonald TV (2001) HERG K(+) channel activity is regulated by changes in phosphatidylinositol 4,5-bisphosphate. Circ Res 89: 1168-1176.
Bian JS, Kagan A, and McDonald TV (2004) Molecular analysis of PIP2 regulation of HERG and IKr. Am J Physiol 287: H2154-H2163.
Cayabyab FS and Schlichter LC (2002) Regulation of an ERG K+ current by Src tyrosine kinase. J Biol Chem 277: 13673-13681.
Cordes JS, Sun Z, Lloyd DB, Bradley JA, Opsahl AC, Tengowski MW, Chen X, and Zhou J (2005) Pentamidine reduces hERG expression to prolong the QT interval. Br J Pharmacol 145: 15-23.
Cui J, Melman Y, Palma E, Fishman GI, and McDonald TV (2000) Cyclic AMP regulates the HERG K(+) channel by dual pathways. Curr Biol 10: 671-674.
De Ponti F, Poluzzi E, and Montanaro N (2000) QT-interval prolongation by non-cardiac drugs: lessons to be learned from recent experience. Eur J Clin Pharmacol 56: 1-18.
De Ponti F, Poluzzi E, and Montanaro N (2001) Organising evidence on QT prolongation and occurrence of torsades de pointes with non-antiarrhythmic drugs: a call for consensus. Eur J Clin Pharmacol 57: 185-209.
Fermini B and Fossa AA (2003) The impact of drug-induced QT interval prolongation on drug discovery and development. Nat Rev Drug Discov 2: 439-447.
Gomez-Varela D, Barros F, Viloria CG, Giraldez T, Manso DG, Dupuy SG, Miranda P, and de la Pena P (2003) Relevance of the proximal domain in the amino-terminus of HERG channels for regulation by a phospholipase C-coupled hormone receptor. FEBS Lett 535: 125-130.
Joshi A, Dimino T, Vohra Y, Cui C, and Yan GX (2004) Preclinical strategies to assess QT liability and torsadogenic potential of new drugs: the role of experimental models. J Electrocardiol 37 (Suppl): 7-14.
Kang J, Chen XL, Wang H, Ji J, Cheng H, Incardona J, Reynolds W, Viviani F, Tabart M, and Rampe D (2005) Discovery of a small molecule activator of the human ether-e?go-go-related gene (hERG) cardiac K+ channel. Mol Pharmacol 67: 827-836.
Medina-Ravell VA, Lankipalli RS, Yan GX, Antzelevitch C, Medina-Malpica NA, Medina-Malpica OA, Droogan C, and Kowey PR (2003) Effect of epicardial or biventricular pacing to prolong QT interval and increase transmural dispersion of repolarization: does resynchronization therapy pose a risk for patients predisposed to long QT or torsade de pointes Circulation 107: 740-746.
Shimizu W, McMahon B, and Antzelevitch C (1999) Sodium pentobarbital reduces transmural dispersion of repolarization and prevents torsades de pointes in models of acquired and congenital long QT syndrome. J Cardiovasc Electrophysiol 10: 154-164.
Shryock JC, Song Y, Wu L, Fraser H, and Belardinelli L (2004) A mechanistic approach to assess the proarrhythmic risk of QT-prolonging drugs in preclinical pharmacologic studies. J Electrocardiol 37 (Suppl): 34-39.
Sun Z, Milos PM, Thompson JF, Lloyd DB, Mank-Seymour A, Richmond J, Cordes JS, and Zhou J (2004) Role of a KCNH2 polymorphism (R1047 L) in dofetilide-induced torsades de pointes. J Mol Cell Cardiol 37: 1031-1039.
Thomas D, Wu K, Wimmer AB, Zitron E, Hammerling BC, Kathofer S, Lueck S, Bloehs R, Kreye VA, Kiehn J, Katus HA, Schoels W, and Karle CA (2004) Activation of cardiac human ether-e?go-go-related gene potassium currents is regulated by alpha(1A)-adrenoceptors. J Mol Med 82: 826-837.
Thomas D, Zhang W, Karle CA, Kathofer S, Schols W, Kubler W, and Kiehn J (1999) Deletion of protein kinase A phosphorylation sites in the HERG potassium channel inhibits activation shift by protein kinase A. J Biol Chem 274: 27457-27462.
Volberg WA, Koci BJ, Su W, Lin J, and Zhou J (2002) Blockade of human cardiac potassium channel human ether-e?go-go-related gene (hERG) by macrolide antibiotics. J Pharmacol Exp Ther 302: 320-327.
Xu X, Yan GX, Wu Y, Liu T, and Kowey PR (2003) Electrophysiologic effects of SB-237376: a new antiarrhythmic compound with dual potassium and calcium channel blocking action. J Cardiovasc Pharmacol 41: 414-421.
Zabel M, Hohnloser SH, Behrens S, Li YG, Woosley RL, and Franz MR (1997) Electrophysiologic features of torsades de pointes: insights from a new isolated rabbit heart model. J Cardiovasc Electrophysiol 8: 1148-1158.
Zhou Z, Gong Q, Ye B, Fan Z, Makielski JC, Robertson GA, and January CT (1998) Properties of HERG channels stably expressed in HEK 293 cells studied at physiological temperature. Biophys J 74: 230-241.