Human ether-a-go-go–related gene (hERG) and KCNQ channels are two classes of voltage-gated potassium channels. Specific mutations have been identified that are causal for type II long QT (LQT2) syndrome, neonatal epilepsy, and benign familial neonatal convulsions. Increasing evidence from clinical studies suggests that LQT2 and epilepsy coexist in some patients. Therefore, an integral approach to investigating and treating the two diseases is likely more effective. In the current study, we found that NS1643 [1,3-bis-(2-hydroxy-5-trifluoromethyl-phenyl)-urea], a previously reported hERG activator, is also an activator of KCNQ channels. It potentiates the neuronal KCNQ2, KCNQ4, and KCNQ2/Q3 channels, but not the cardiac KCNQ1. The effects of NS1643 on the KCNQ2 channel include left shifting of voltage for reaching 50% of the maximum conductance and slowing of deactivation. Analysis of the dose-response curve of NS1643 revealed an EC50 value of 2.44 ± 0.25 μM. A hydrophobic phenylalanine (F137) located at the middle region of the voltage-sensing domain was identified as critical for NS1643 activity on KCNQ2. When testing NS1643 effects in rescuing LQT2 hERG mutants and the KCNQ2 BFNC mutants, we found it is particularly efficacious in some cases. Considering the substantial relationship between LQT2 and epilepsy, these findings reveal that NS1643 is a useful compound to elucidate the causal connection of LQT2 and epilepsy. More generally, this may provide a strategy in the development of therapeutics for LQT2 and epilepsy.
Voltage-gated potassium (Kv) channels are a large family of membrane proteins, including approximately 40 members that are critical for a variety of cellular processes in electrically excitable cells, such as control of action potential and excitability in nerve and muscle cells, and regulation of hormone secretion, cell volume, and T-lymph cell activation (Gutman et al., 2005; Judge et al., 2006; Cooper and Yeung, 2007). Kv channels offer tremendous opportunities for the development of new drugs to treat diverse diseases (Overington et al., 2006; Wulff et al., 2009). Kv channels are tetramers, and in general each subunit comprises six transmembrane segments (S1∼S6) with the intracellular N and C termini, of which S1∼S4 contribute to the voltage-sensing domain (VSD) and S5∼S6 are key elements forming the pore region (Long et al., 2007; Jensen et al., 2012). Both the pore region and the VSD are sites targeted by small-molecule drugs (Catterall et al., 2007; Swartz, 2007; Xiong et al., 2008; Wulff et al., 2009; Peretz et al., 2010; Miceli et al., 2011; Li et al., 2013).
The human ether-a-go-go–related gene (hERG or KCNH2) encodes one of the delayed rectifier Kv channels (Gutman et al., 2005). The hERG channel is the major molecular determinant for native cardiac Ikr current (the fast component of potassium rectifying current), and plays a crucial role in repolarization of the action potential of cardiomyocytes in the human heart (Vandenberg et al., 2012). Mutations that reduce hERG expression or conductance may cause the congenital long QT (LQT) syndrome, a life-threatening severe arrhythmia characterized by a prolonged QT interval (Larsen et al., 2000; Millat et al., 2006). So far, at least 13 genes have been associated with congenital LQT syndrome (Millat et al., 2006; Zumhagen et al., 2012). The LQT syndrome associated with the dysfunction of hERG was named type II LQT (LQT2). To date, nearly 300 hERG mutations causing LQT2 have been reported (http://www.fsm.it/cardmoc).
KCNQ (Kv7) channels are low-threshold activated voltage-gated potassium channels. Of the five known KCNQ isoforms, KCNQ2–5 are expressed throughout the nervous system, whereas KCNQ1 is predominantly expressed in cardiac tissue (Brown and Passmore, 2009), but is also found in other tissues. As a primary player that mediates neuronal muscarinic (M) currents, opening of the KCNQ2 channel or of heterogeneous KCNQ2/KCNQ3 complexes inhibits initiation of action potential and thus suppresses neuronal excitability (Brown and Adams, 1980; Stewart et al., 2012). Mutations of human KCNQ2 genes that result in loss or reduction of channel activity are causal for one type of neonatal epilepsy, benign familial neonatal convulsions (BFNC) (Jentsch, 2000; Lerche et al., 2005).
Both LQT2 and BFNC are channelopathies (Splawski et al., 2000; Robbins, 2001; Maljevic et al., 2010). Interestingly, seizure is one of the typical symptoms of LQT2 syndrome. In fact, there is a common misdiagnosis of LQT2 versus epilepsy (Johnson et al., 2009; MacCormick et al., 2009). In some cases, seizures of LQT2 are likely due to cardiac syncope (Keller et al., 2009; Jorge et al., 2011; Zamorano-Leon et al., 2012). In addition, increasing evidences suggest LQT2 and epilepsy coexist (Johnson et al., 2009; MacCormick et al., 2009; Jorge et al., 2011). The LQT2 mutation was thought to confer the susceptibility for recurrent seizure activity, which is similar to noncardiac complications observed in other LQT-susceptible genes, such as KCNQ1/deafness and SCN5A/gastrointestinal symptoms (Jespersen et al., 2005; Locke et al., 2006; Attanasio et al., 2007; Johnson et al., 2009).
Small molecular activators of hERG and KCNQ channels are effective tools to develop therapeutics for LQT2 and epilepsy, respectively (Wulff et al., 2009; Perry et al., 2010). Several novel hERG activators and KCNQ activators have been reported (Xiong et al., 2008; Gao et al., 2010; Perry et al., 2010; Zhang et al., 2011; Boehlen et al., 2013). In 2011, retigabine, a KCNQ activator, was approved as an antiepileptic drug by the US Food and Drug Administration (Deeks, 2011; Stafstrom et al., 2011). NS1643 [1,3-bis-(2-hydroxy-5-trifuoromethyl-phenyl)-urea], one hERG activator that potentiates both the recombinant and native hERG channels, was first reported in 2006 (Fig. 1) (Casis et al., 2006; Hansen et al., 2006). The efficacious pharmacologic effects of NS1643 in decreasing the action potential duration and prolonging the post repolarization refractory time make it an interesting new antiarrhythmic (Hansen et al., 2006; Peitersen et al., 2008). However, its role in epileptogenic KCNQ channels has not been reported. Given the important connection between LQT and epilepsy discussed earlier and the key players of hERG and KCNQ channels, we investigated the effects of NS1643 in modulating KCNQ2 channels, and ask if NS1643 could potentially rescue both the LQT2 and BFNC mutants.
Materials and Methods
Cell Culture and Transfection.
Chinese hamster ovary (CHO) K1 cells were cultured in 50/50 Dulbecco’s modified Eagle’s medium/F-12 (Gibco, Grand Island, NY) with 10% fetal bovine serum and 2 mM l-glutamine (Invitrogen, Carlsbad, CA). At 24 hours before transfection, cells were split and plated in 60-mm dishes. The plasmid with cDNA constructs was transfected with Lipofectamine 2000 reagent (Invitrogen) to express the channels, referring to the manufacturer’s instructions. A green fluorescent protein cDNA (Amaxa, Gaithersburg, MD) was cotransfected to identify the transfected cells by fluorescence microscopy.
Plasmid Constructions and Mutagenesis.
The KCNQ1-KCNQ4 cDNAs were gifts from Drs. T. Jentsch (Zentrum für Molekulare Neurobiologie, Hamburg, Germany), D. MacKinnon (State University of New York, Stony Brook, NY), M. Sanguinetti (University of Utah, Salt Lake City, UT), and M. Shapiro (University of Texas Health Science Center, San Antonio, TX). KCNH2 cDNA was a gift from M. Sanguinetti (University of Utah). Point mutations were introduced using the QuikChange II site-directed mutagenesis kit (Stratagene, La Jolla, CA) and verified by DNA sequencing.
Electrophysiology Recording in CHO-K1 Cells.
Standard whole-cell recording was used to record current of the transiently expressed KCNQ channels in CHO cells. Pipettes were pulled from borosilicate glass capillaries (Sutter Instrument, Novato, CA). When filled with the intracellular solution, the pipettes have 3- to 5-megaohm resistances. During recording, extracellular solution was constantly perfused by a bath perfusion system (ALA Scientific Instruments, Farmingdale, NY). Intracellular solution contained (in mM) 145 KCl, 1 MgCl2, 5 EGTA, 10 HEPES, and 5 MgATP (pH = 7.3 adjusted with KOH); extracellular solution contained (in mM) 140 NaCl, 3 KCl, 2 CaCl2, 1.5 MgCl2, 10 HEPES, and 10 glucose (pH = 7.4 adjusted with NaOH). Whole-cell currents were recorded using an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA), filtered at 2 kHz, and digitized using a DigiData 1440A with pClamp 10.2 software (Molecular Devices). Series resistance compensation was used and set to 60–80%.
Based on the closed-state conformation of the Kv1.2/Kv2.1 “paddle chimera” channel reported by Jensen et al. (2012), we previously constructed a structural model of the closed-state KCNQ2 channel (Zhang et al., 2013). The structural model of the closed-state mutant F137W used for the docking was built from this closed-state KCNQ2 model by mutating F137 into tryptophan. Docking was performed by using Glide (Schrödinger, LLC, New York, NY). The receptor was prepared by using the Protein Preparation and Grid Preparation tools in the Schrödinger Maestro interface. Hydrogen orientations were optimized in a protein preparation module. The docking box was set to cover the assumed binding pocket and centered at F137W, which was supposed to be the middle of the potential pocket according to the primary mutagenesis and electrophysiologic results. The default settings were adopted for the cutoff, neutralization, scaling, and dimension of the binding pocket. Homology modeling and mutation could not predict proper conformations for the residues interacting with a ligand. Therefore, we applied the Induced Fit module encoded in the Schrödinger modeling package. F137W and residues around it in the central part of the VSD were treated as flexible residues.
Patch-clamp data were processed using Clampfit 10.2 (Molecular Devices) and then analyzed in GraphPad Prism 5 (GraphPad Software, San Diego, CA). Voltage-dependent activation curves were fitted with the Boltzmann equation, where Gmax is the maximum conductance, Gmin is the minimum conductance, V1/2 is the voltage for reaching 50% of the maximum conductance, and S is the slope factor.
Dose-response curves were fitted with the Hill equation, where EC50 is the drug concentration producing half of the maximum response, and P is the Hill coefficient. Data are presented as means ± S.E.M. Significance was estimated using unpaired two-tailed Student’s t tests. Statistical significance: *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.
NS1643 was purchased from Tocris Bioscience (Bristol, UK) and prepared as a 20 mM stock solution in water. Final drug concentrations were prepared daily by dilution of stock solutions kept at 25°C.
Effects of NS1643 on hERG Channels
Noticeable differences in the effects of NS1643 on hERG channels were observed depending on expression systems (Casis et al., 2006; Hansen et al., 2006; Schuster et al., 2011; Durdagi et al., 2012). Thus, potentiation effects of NS1643 on hERG channels expressed in CHO-K1 cells, the transient expression system we used in this study, were characterized first. At the 30-μM concentration, effects of NS1643 on the outward current and kinetics of hERG channels were tested by a voltage protocol as indicated in Fig. 2A. In brief, the outward current of hERG channels was elicited by a 2-second depolarization step to +20 mV from a holding potential of −80 mV followed by a 3-second repolarization step to −60 mV to measure the tail current using a standard protocol (Li et al., 2012; Garg et al., 2013). The potentiation on the steady-state current at +20 mV was much more potent than that on the peak tail current at −60 mV (Fig. 2B). The activation and deactivation phases were then fit by a single exponential equation, which revealed that the deactivation kinetics were dramatically slowed, while the activation was not significantly changed (Fig. 2C). The voltage dependence of the effects of NS1643 was then examined. At a 30-μM concentration, NS1643 increased the outward current under the entire range of test voltages (Fig. 2, D and E) and left-shifted the Itail (tail current)–V curve largely with V1/2 values shifting from −5.46 ± 1.42 to −52.47 ± 1.21 mV (Fig. 2F), which was similar to the observation in other mammalian expression systems, such as human embryonic kidney 293 cells (Casis et al., 2006; Schuster et al., 2011), but different from that in Xenopus laevis oocytes (Hansen et al., 2006).
Effects of NS1643 on KCNQ2 Channels
ztz240 [N-(6-chloro-pyridin-3-yl)-4-fluorobenzamide] is a KCNQ2 activator (Gao et al., 2010). Examination of the chemical structures of ztz240 and NS1643 shows the scaffolds of the two compounds can be described as two aryl groups connected by a functional linker (Fig. 1). Effects of NS1643 on KCNQ2 were then tested. NS1643 indeed potentiated KCNQ2 channels at the 30-μM concentration, identical to that used for hERG testing (Fig. 3A). Under stepping the holding potential from −90 to −10 mV, the outward current of KCNQ2 channels was significantly increased 1.41 ± 0.10–fold (n = 6) (Fig. 3B). Analysis of the dose-response curve on current amplitude revealed an EC50 value of 2.44 ± 0.25 μM (Fig. 3E), which is similar to the previously reported EC50 value for hERG channels (Hansen et al., 2006). Similar to effects on hERG, NS1643 also left-shifted the G-V curve of KCNQ2 and slowed the deactivation rate without affecting the activation rate (Fig. 3, C and D).
To understand the state dependence of NS1643, three different paradigms of stimulation were used (Fig. 4). In the first set of experiments, KCNQ2 currents were elicited by −10 mV depolarization from the holding potential −100 mV every 3.25 seconds; the potentiation developed progressively and reached the steady state in 10 seconds. In the second set of experiments, KCNQ2 currents were elicited and then held at −10 mV to keep the channel in open state, where application of NS1643 resulted in similar potentiation as that of the first set of experiments. These data are consistent with the notion that NS1643 accesses the open-state channel. In the third set of experiments, after the control trace was recorded, the cell was held at −100 mV to keep the channel in closed state for 10 seconds, and steady-state potentiation with almost the same level was immediately achieved, which supports the possibility that NS1643 may interact with the closed-state channel. These results suggest that the potentiation of NS1643 on KCNQ2 channels lacks state dependence.
Subtype Selectivity of NS1643 on KCNQ Channels
Subtype selectivity of NS1643 on KCNQ channels was examined in CHO-K1 cells transiently expressing KCNQ1, KCNQ3, KCNQ2/KCNQ3, and KCNQ4 channels, respectively. As shown in Fig. 5A, NS1643 potentiated all KCNQ2/KCNQ3, KCNQ4, and KCNQ2 channels under stepping to −10 mV from the holding potential −90 mV. In contrast, KCNQ3 channels were not sensitive, and KCNQ1 channels were inhibited. The KCNQ1/KCNE1 complex mediates the native IKs (slow component of potassium rectifying current) (Wang et al., 2011; Van Horn et al., 2011; Wrobel et al., 2012). Consistent with a previous study, 30 μM NS1643 exhibited significant inhibitory effects on the outward currents of the KCNQ1/KCNE1 complex (Hansen et al., 2006). Interestingly, for KCNQ2/KCNQ3 and KCNQ4 channels, the left-shifting of G-V curves was also observed, whereas the slowing of deactivation was no more significant (Fig. 5, B–D). Note that ztz240 also lacks potentiation activity on KCNQ1 and KCNQ3 channels (Gao et al., 2010). The two compounds exhibit similar subtype selectivity.
Identification of a Residue Critical for NS1643 Activity
ztz240 binds in the VSD of KCNQ2 channels (Li et al., 2013). The hydrophobic phenylalanine (F137) has been demonstrated critical for ztz240 activity. Considering the similar phenotypes and subtype selectivity of ztz240 and NS1643, we tested the importance of F137 for NS1643 potentiating KCNQ2. NS1643 activity was not affected when F137 was mutated to hydrophobic amino acids with smaller side chains of volume, such as glycine, alanine, and leucine. Mutation of F137 to the tyrosine, a hydrophobic amino acid with comparable side chain of volume, also did not affect NS1643 activity (Fig. 6, C–E). However, when F137 was mutated to tryptophan, the hydrophobic amino acid with bigger side chain of volume than F, NS1643 largely lost potentiation activity on KCNQ2 channels (Fig. 6, A and C). Note that the left-shifting of the G-V curve observed on wild-type channels was also completely abolished by F137W (Fig. 6, B and E). Experimental data suggest that NS1643 may interact with the closed state of the KCNQ2 channel (Fig. 4). To understand the potential mechanism underlying the inhibitory effects of NS1643 on F137W, we proposed that NS1643 may also bind to the closed state of the mutant F137W. Based on the model of the closed-state KCNQ2 channel we previously used (Zhang et al., 2013), we built the structure of the closed state of the mutant F137W and docked NS1643 to the pocket near the residue F137W. Interestingly, NS1643 shows strong binding potency to the VSD region (Fig. 6F). In contrast, ztz240, the activator which lacks sensitivity to F137W (Li et al., 2013), has weak binding affinity. Although further direct evidence of the interaction is necessary, these modeling results are consistent with the notion that NS1643 may bind with the closed state of F137W and result in the inhibitory effects.
NS1643 Rescues Both LQT2 and BFNC Mutants
NS1643 Rescues LQT2 Mutants.
More than 300 LQT2 mutants have been reported. Most of them are trafficking-deficient, giving rise to reduced currents in the exogenous recombinant system, and thus whether NS1643 can rescue the homomultimer LQT2 mutants has not been elucidated. Among eight tested LQT2 mutants (A561V, G572R, I593R, G614V, G628S, R752Q, S818L, and V822M) in our study, only G572R and R752Q were functional. G572 and R752 locate at the pore region and the C terminus, respectively (Splawski et al., 2000). We first examined NS1643 effects on the two mutants with the single testing stimulation as indicated in Fig. 7A. At a 30-μM concentration, NS1643 showed potentiation on the outward currents comparable with that on wild-type hERG channels (Fig. 7, A and B). Similar to wild-type channels, the deactivation phase, although not the activation phase, was affected by NS1643 (Fig. 7C). The current of R752Q was bigger than G572R, and thus allowed us to further investigate the effects of NS1643 on the Itail-V relationship of the mutant. NS1643 at a 30-μM concentration increased the outward current of R752Q under the entire range of test voltages (Fig. 7, D and E) and left-shifted the Itail-V curve significantly with V1/2 shifting to −34.00 ± 6.35 mV from −20.50 ± 1.70 mV (Fig. 6F).
NS1643 Rescues BFNC Mutants.
Encouraged by the effects on LQT2 mutants, we then proceeded with testing whether NS1643 can rescue BFNC mutants. Different from LQT2 mutants, most of the BFNC mutants can form functional homomultimers in CHO-K1 cells. At 30 μM concentration, NS1643 exhibited comparable or even much stronger potentiation effects on the four tested BFNC mutants in all three observed parameters, i.e., I/I0 (the ratio of ion channel current after and before drug perfusion), V1/2, and the time constant of deactivation. As concluded in Fig. 8, the outward currents of R207W, M208V, R214W, and Y284C were able to be increased significantly under the perfusion with NS1643, and the I/I0 ratios were 8.93 ± 1.84 (n = 5), 2.29 ± 0.17 (n = 8), 1.85 ± 0.24 (n = 7), and 1.29 ± 0.05 (n = 8), respectively. The measured outward currents were elicited by stepping to −10 mV from the holding potential −90 mV. In addition, NS1643 made it much easier for the channel opening process; therefore, the G-V curve of the mutants moves notably toward hyperpolarization direction under the compound perfusion (Fig. 8, B and D). For example, for the mutant Y284C, the left-shift of V1/2 was around −50.19 ± 5.32 mV (n = 5) versus −30.43 ± 1.02 mV (n = 5, P < 0.01) for wild-type channels. Furthermore, the deactivation time of these mutants was prolonged by the chemical to a different extent, which was just like the wild-type channel response. It is notable that NS1643 dramatically slowed down the deactivation kinetics of R207W from 32.52 ± 5.24 to 256.04 ± 29.95 milliseconds (Fig. 8E).
NS1643 was among the first reported activators for hERG channels (Casis et al., 2006; Hansen et al., 2006). It exhibits potentiation effects on both recombinant and native hERG channels, which makes it a promising candidate for treating LQT2 (Hansen et al., 2006; Diness et al., 2008). In this study, we found that NS1643 is also an activator of KCNQ2 channels, a proven antiepilepsy drug target.
The potentiation effects of reported small molecular activators of KCNQ2 channels may include all or part of the following three biophysical properties: increasing of the overall conductance, left-shifting of the G-V curve, and slowing of the deactivation phase. Many activators, including ztz240, exhibit activity on all three of these aspects (Gao et al., 2010). However, retigabine, the approved antiepilepsy drug, does not increase the overall conductance. The major potentiation effects of retigabine are left-shifting of the G-V curves and slowing of the deactivation. The potentiation effects of NS1643 on KCNQ2 channels seem to mimic those of retigabine. As shown in Fig. 3, at a 30-μM concentration, the most dramatic effects of NS1643 were left-shifting of G-V curves and slowing of the deactivation. Similar to retigabine, the overall conductance of the KCNQ2 channel at +50 mV testing potential was not significantly increased after application of NS1643.
Although the potentiation phenotype of NS1643 on KCNQ2 channels is similar to retigabine, the subtype selectivity of NS1643 is similar to ztz240. Both NS1643 and ztz240 lack potentiation activity on KCNQ1 and KCNQ3. In contrast, retigabine potentiates all KCNQ isoforms except KCNQ1. Identification of F137, critical for NS1643 potentiation activity on the KCNQ2 channel, further raises the speculation that NS1643 may act on the voltage-sensing domain as ztz240 does (Li et al., 2013). In the previous study by Durdagi et al. (2012), some critical residues for NS1643 potentiating hERG channels have been identified. One putative binding site is N629 at the outer mouth of the channel. Mutation of N629H abrogates the NS1643-induced left-shifting of the G-V curves (Durdagi et al., 2012). Another predicted critical residue, D540, locates at the intracellular surface of the S4–S5 linker. The inward conductance of D540K was completely inhibited by NS1643 treatment. None of these reported residues locates at the voltage-sensing domain of hERG. For KCNQ2 channels, we have demonstrated that ztz240 binds in the VSD (Li et al., 2013). The hydrophobic phenylalanine F137, locating at the middle of S2, is one of the key binding sites. Mutation of F137A dramatically reduced ztz240 potentiation activity, which includes increasing of the outward current amplitude, left-shifting of the G-V curve, and slowing of deactivation. Interestingly, F137 is also critical for NS1643 activity on KCNQ2 channels. Mutation of F137W completely abolished the left-shift of G-V curves by 30 μM NS1643. Note that mutation of F137A did not affect NS1643 activity, which suggests NS1643 may act on the residue with a different interaction manner from ztz240. The direct interaction between NS1643 and the residues lining the VSD, including F137, will be a worthwhile topic for future investigation. Our finding revealed for the first time the importance of the VSD for NS1643 activity, and thus is helpful in elucidating the binding mode of NS1643 on KCNQ or hERG channels in future studies.
The mutations of hERG and KCNQ2 channels are causal for LQT2 and BFNC, respectively. Because seizure is one of the typical symptoms of LQT2, LQT2 is easily misdiagnosed as epilepsy. Many patients with recurrent convulsions who were treated for epilepsy for years were finally diagnosed with LQT (Gatto et al., 1993; Scheepers et al., 1998; Medina-Villanueva et al., 2002; Hunt and Tang, 2005; Johnson et al., 2009; MacCormick et al., 2009; Jorge et al., 2011). Although the responsible genes of the two channelopathies are different, besides misdiagnosis, increasing evidence argues for a close pathogenic relationship between LQT2 and epilepsy (Johnson et al., 2009; Keller et al., 2009; MacCormick et al., 2009; Omichi et al., 2010; Zamorano-Leon et al., 2012). For example, in a chart review study (Johnson et al., 2009), a seizure phenotype was observed in 98 patients among 343 LQT patients (98/343, 29%). However, the seizure phenotype is obviously more common in LQT2 (36/77, 47%) than average and all other subtypes of LQT syndrome. Inhibition of the hERG channel by drugs causes acquired LQT2 (Roden, 2004; Perrin et al., 2008; Berger et al., 2010; Zhou et al., 2011).
Interestingly, inhibition of the hERG channel by blockers, such as cisapride, may also make the patients more vulnerable to epilepsy (Attanasio et al., 2007). The possibility of coexistence of LQT2 and epilepsy is further strengthened by the identification of expression of the hERG channel in the brain. Recently, hERG1b, a splicing isoform of hERG1a, was found to be expressed in a certain cerebral cortex area in humans (Jonsson et al., 2012). hERG1b affects both the firing rate and firing pattern of neurons in the brain (Huffaker et al., 2009). Although there is no direct evidence for the relationship between hERG1b and epilepsy, deletion of KCNH3, a homolog channel to hERG, has been demonstrated to cause hippocampal hyperexcitability and epilepsy (Zhang et al., 2010). Being an activator of both hERG and KCNQ2 channels, in combination with its capability in rescuing LQT2 and BFNC mutants, NS1643 might be a beneficial lead compound in the development of therapeutics for LQT2 patients with seizure phenotype. Note that NS1643 was also found to potentiate large-conductance Ca2+-activated K+ channels (BK) channels obvious in pituitary tumor (GH3) cells (Wu et al., 2008). Whether and to what extent the less optimal selectivity of the compounds activating both hERG and KCNQ2, including NS1643, prevents its pharmacologic activities should be evaluated in future studies.
In summary, this study has provided evidence that NS1643 is an activator of both hERG and KCNQ2 channels. In combination with the capability of NS1643, restoring the LQT2 mutants and the BFNC mutants argues for the notion that, at the molecular level, rescue of mutants causing LQT2 and those causing seizures may be dealt with by a single pharmacologic agent. Hence this study provides in vitro pharmacologic validation for an integral approach to the two seemingly unrelated disease phenotypes. Our finding may make NS1643 a useful compound to investigate the relevance between LQT2 and epilepsy.
The authors thank Dr. Min Li for helpful discussions of the project.
Participated in research design: Gao, Jiang.
Conducted experiments: Li, Chen, Zhang, Zheng.
Contributed new reagents or analytic tools: Yang.
Performed data analysis: Li, Zhang, Zheng, Yang, Gao.
Wrote or contributed to the writing of the manuscript: Gao, Yang.
- Received June 18, 2014.
- Accepted September 16, 2014.
This work received supported by the State Key Program of Basic Research of China [Grant 2013CB910604]; the National Science and Technology Major Project on “Key New Drug Creation and Manufacturing Program” [Grant 2013ZX09103001-016]; the National Natural Science Foundation of China Grant for Excellent Key Laboratory [Grant 81123004]; the National Natural Science Foundation of China [Grants 61327014, 61175103, and 81173027]; Shanghai Municipal Science and Technology Commission [Grant 13JC1406700]; and the External Cooperation Program of BIC, Chinese Academy of Sciences [Grant 1536631KYSB20130003].
- benign familial neonatal convulsions
- Chinese hamster ovary
- human ether-a-go-go–related gene
- the ratio of ion channel current after and before drug perfusion
- tail current
- voltage-gated potassium
- long QT
- type II long QT
- voltage for reaching 50% of the maximum conductance
- voltage-sensing domain
- Copyright © 2014 by The American Society for Pharmacology and Experimental Therapeutics