The calmodulin inhibitor N-(6-aminohexyl)-5-chloro-1-naphthalene sulphonamide directly blocks human ether à-go-go-related gene potassium channels stably expressed in human embryonic kidney 293 cells

Abstract

BACKGROUND AND PURPOSE

N-(6-aminohexyl)-5-chloro-1-naphthalene sulphonamide (W-7) is a well-known calmodulin inhibitor used to study calmodulin regulation of intracellular Ca²⁺ signalling-related process. Here, we have determined whether W-7 would inhibit human ether gene (hERG or K_v11.1) potassium channels, hK_v1.5 channels or hK_IR2.1 channels expressed in human embryonic kidney (HEK) 293 cells.

EXPERIMENTAL APPROACH

The hERG channel current, hK_v1.5 channel current or hK_IR2.1 channel current was recorded with a whole-cell patch clamp technique.

KEY RESULTS

It was found that the calmodulin inhibitor W-7 blocked hERG, hK_v1.5 and hK_IR2.1 channels. W-7 decreased the hERG current (I_hERG) in a concentration-dependent manner (IC₅₀: 3.5 µM), and the inhibition was more significant at depolarization potentials between +10 and +60 mV. The hERG mutations in the S6 region Y652A and F656V, and in the pore helix S631A, had the IC₅₀s of 5.5, 9.8 and 25.4 µM respectively. In addition, the compound inhibited hK_v1.5 and hK_IR2.1 channels with IC₅₀s of 6.5 and 13.4 µM respectively.

CONCLUSION AND IMPLICATIONS

These results indicate that the calmodulin inhibitor W-7 exerts a direct channel-blocking effect on hERG, hK_v1.5 and hK_IR2.1 channels stably expressed in HEK 293 cells. Caution should be taken in the interpretation of calmodulin regulation of ion channels with W-7.

Introduction

Calmodulin is a ubiquitous Ca²⁺-binding protein that plays an important role in Ca²⁺-signalling pathways of eukaryotic cells (Means et al. 1982; Rakhilin et al., 2004). The Ca²⁺-activated calmodulin-dependent protein kinase II (CaMKII) is an important regulator of cardiac ionic currents including L-type Ca²⁺ current (I_Ca.L) (Dzhura et al., 2000; Wu et al. 2004) and Na⁺ current (I_Na) (Deschenes et al., 2002; Wagner et al., 2006; Maltsev et al. 2008), and several types of K⁺ currents (Schonherr et al., 2000; Colinas et al., 2006; Li et al. 2006; 2007; Ziechner et al. 2006; Qu et al. 2007). The inhibition of calmodulin and/or CaMKII is believed to be effective in the treatment of ventricular arrhythmias (Wu et al., 1999; Gbadebo et al. 2002; Anderson, 2005). N-(6-Aminohexyl)-5-chloro-1-naphthalene sulphonamide (W-7) is a commonly used calmodulin inhibitor that blocks access to the activation site (Hait and Lazo, 1986; Osawa et al. 1998), and has been reported to effectively suppress ventricular arrhythmias in animal models (Wu et al., 1999; Gbadebo et al. 2002; Anderson, 2005).

Human ether-à-go-go-related gene (hERG or K_v11.1) (Alexander et al., 2009) encodes the α-subunit of the rapidly delayed rectifier K⁺ current (I_Kr) which contributes importantly to repolarization of the cardiac action potential in the human heart. Dysfunction of I_Kr has been implicated in long QT syndrome that can predispose individuals to lethal arrhythmias. Either inherited mutations of hERG or hERG channel block by a variety of drugs can cause the long QT syndrome (Tseng, 2001; Sanguinetti and Mitcheson, 2005). Therefore, hERG channels have been widely used to examine pro-arrhythmic potential of therapeutic agents during drug development. As human ether-à-go-go gene (hEAG) K⁺ channels are effectively regulated by calmodulin (Schonherr et al. 2000; Ziechner et al., 2006), the present study was initially designed to investigate whether calmodulin could regulate hERG channels. However, we found that W-7 directly blocked not only hERG channels, but also hK_v1.5 or hK_IR2.1 channels stably expressed in human embryonic kidney (HEK) 293 cells.

Methods

Cell culture and gene transfection

HEK 293 cell lines (Tang et al., 2007; Zhang et al. 2008) stably expressing hERG gene (K_v11.1, provided by Dr G. Robertson, University of Wisconsin, Madison, WI, USA) (Trudeau et al., 1995), hK_v1.5 gene (provided from Dr M Tamkun, Colorado State University, CO, USA) and hK_IR2.1 gene (provided by Dr Carol A. Vandenberg, University of California at Santa Barbara, CA, USA) (Raab-Graham et al., 1994) were cultured in Dulbecco's modified Eagle's medium (Invitrogen, Hong Kong) containing 400 µg·mL⁻¹ G418 (Sigma-Aldrich, St Louis, MO, USA) and 10% fetal bovine serum. Cells used for electrophysiology were seeded on glass coverslips.

The mutant hERG genes Y652A, F656V and S631A (Gang and Zhang, 2006; Guo et al., 2006) were transiently transfected into HEK 293 cells in a 60 mm Petri dish using 10 µL of Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) with 4 µg of hERG mutant cDNA in pCDNA3 vector. After 24–48 h, 20–30% of cells expressed hERG channels were used for pharmacological study.

Solutions

Tyrode solution contained (in mM): NaCl 140, KCl 5.0, MgCl₂ 1.0, CaCl₂ 1.8, NaH₂PO₄ 0.33, HEPES 10.0, glucose 10, and pH adjusted to 7.3 with NaOH. The pipette solution contained (in mM): KCl 20, K-aspartate 110, MgCl₂ 1.0, HEPES 10, EGTA 5.0 or BAPTA 5.0 (where specified), and GTP 0.1, Na₂-phosphocreatine 5.0, Mg-ATP 5.0, with pH adjusted to 7.2 with KOH.

Data acquisition and analysis

Cells on coverslips were transferred to an open cell chamber (0.5 mL) mounted on the stage of an inverted microscope (Diaphot, Nikon, Japan) and superfused with Tyrode solution at ∼2 mL·min⁻¹. Experiments were performed at room temperature (22–23°C). The whole-cell patch clamp technique was used as described previously (Tian et al., 2006; Tang et al., 2007; Zhang et al. 2008). Briefly, borosilicate glass electrodes (1.2 mm OD) were pulled with a Brown–Flaming puller (model P-97, Sutter Instrument Co., Novato, CA, USA) and had tip resistances of 2–3 MΩ when filled with the pipette solution. A 3 M KCl–agar bridge was used as reference electrode. Tip potentials were zeroed before the pipette touched the cell. After a gigaohm seal was obtained, the cell membrane was ruptured by gentle suction to establish whole-cell configuration to record hERG, hKv1.5 or hKir 2.1 channel currents. Liquid junction potentials after membrane rupture between the external and pipette solutions (∼10.5 mV) were not corrected for all current recording. The membrane currents were recorded with an EPC-10 amplifier and Pulse software (HEKA, Lambrecht, Germany). Command pulses were generated by a 12-bit digital-to-analog converter controlled by Pulse software. Current signals were low-pass filtered at 5 kHz and sampled at 10 kHz. The data were stored on the hard disk of an IBM computer.

Data are presented as means ± SEM. Non-linear curve fitting was performed using Pulsefit (HEKA) and/or Sigmaplot (SPSS Science, Chicago, IL, USA). Paired and/or unpaired Student's t-tests were used to evaluate the statistical significance of differences between two group means. One-way anova followed by Tukey's test was used for multiple groups. Values of P < 0.05 were considered to be statistically significant.

Materials

W-7 hydrochloride, from Merck Chemicals (Gibbstown, NJ, USA) was prepared as 30 mM stock solutions in dimethyl sulphoxide, and was diluted in experimental solutions to the final concentrations. Calmodulin was purchased from Boppard (Boppard, RP, Germany). All other chemicals were purchased from Sigma-Aldrich.

Results

Inhibition of hERG channel current by W-7

Figure 1A displays the time-course of hERG tail current (I_hERG.tail) recorded in a representative HEK 293 cell stably expressing hERG channels with the voltage protocol shown in the inset, using a pipette solution containing 5 mM EGTA. The current was gradually inhibited by application of 3 µM W-7 in bath solution, and the effect was partially reversed by washout for 25 min. Original current traces at corresponding time-points are shown on the right side of the panel.

Figure 1.

Inhibition of hERG channel current by W-7. (A) Time-course of hERG tail current recorded in a HEK 293 cell stably expressing hERG channels. The current was elicited by a 2 s voltage step to +30 mV from a holding potential of −80 mV, then back to −50 mV (left inset) every 15 s. Original current traces at corresponding time-points are shown in the right of the panel. (B) Voltage-dependent hERG channel currents were recorded in a typical experiment with 5 s voltage steps to between −60 and +60 mV from a holding potential of −80 mV, then back to −50 mV (inset) at 0.05 Hz in the absence and presence of 3 µM W-7. (C) Current–voltage (I–V) relationships of I_hERG.tail in the absence and presence of 1, 3 and 10 µM W-7. I_hERG.tail was measured at the peak of tail current. (D) Voltage dependence of W-7 for inhibiting I_hERG.tail (P < 0.05 or P < 0.01 at potentials of +10 to +50 mV vs. −10 or 0 mV). (E) Voltage dependence of the hERG channel conductance (g/g_max) in the absence and presence of 3 µM W-7. The g/g_max was determined by normalizing I_hERG.tail, and fitted to the Boltzmann distribution: y = 1/{1 + exp[(V_m−V_0.5)/S]}, where V_m is the membrane potential, V_0.5 is the midpoint and S is the slope.

The effect of W-7 on voltage-dependent I_hERG was determined in a typical experiment with the voltage protocol shown in the inset (Figure 1B). The current was markedly inhibited by 3 µM W-7, and the inhibitory effect recovered partially upon drug washout (Figure 1B). Figure 1C illustrates the current–voltage (I–V) relationships of I_hERG.tail during control and after application of 1, 3 and 10 µM W-7. The current was blocked by W-7 in a concentration-dependent manner. Figure 1D displays the percentage of hERG channel inhibition by W-7 relative to control, showing a voltage-dependent inhibition of hERG channels. The inhibition of I_hERG.tail by 1, 3 and 10 µM W-7 was stronger at positive potentials between +10 and +60 mV than that at potentials between −20 and 0 mV (n = 6, P < 0.01 or P < 0.05).

Voltage dependence of hERG channels activation (g/g_max) was determined by normalizing I_hERG.tail in the absence and presence of 3 µM W-7 (Figure 1E). The g/g_max curves were fitted to a Boltzmann distribution to obtain the midpoint (V_0.5) of activation potential and slope factor. The V_0.5 of hERG channel activation conductance was negatively shifted by 9.3 mV (from −5.1 ± 1.3 mV of control to −14.4 ± 1.9 mV, n = 11, P < 0.01) by 3 µM W-7, and no change was observed for the slope factor (8.4 ± 0.3 mV for control, 8.5 ± 0.4 mV for W-7, P = NS).

To determine whether the pipette inclusion of W-7 would show more profound effect on hERG current, 10 µM W-7 was applied in pipette solution. Figure 2A shows that the pipette inclusion of W-7 induced a slow reduction of I_hERG.tail, and the inhibitory effect was weaker than bath application of W-7. I_hERG.tail was decreased by 23 ± 6% (n = 5, P < 0.05 vs. initial membrane rupture) with 10 µM W-7 dialysis for 16 min, while by 71 ± 3% (Figure 1D, n = 7) with 10 µM W-7 application in bath solution for 10 min (P < 0.01 vs. pipette solution application). The pipette inclusion of 10 µM W-7 exhibited 49% less inhibitory effect on hERG channels compared to bath application. To exclude the possibility that the weaker inhibition resulted from the incomplete drug dialysis during the recording period, the hERG channel blocker E4031 (1 µM) was included in pipette solution (Figure 2B). I_hERG was gradually inhibited, and steady-state inhibition was observed with 10–12 min dialysis. E4031 inclusion inhibited I_hERG.tail by 91.3% (n = 3), indicating that the drug dialysis is complete in HEK 293 cells during the recording period. These results suggest that W-7-induced inhibition of hERG is likely to be related to direct binding to both extracellular and intracellular sites of the channels, because intracellular calmodulin activity is inhibited by the EGTA-buffered pipette solution, in which free Ca²⁺ concentration is close to 0 nM. The experiment with a 5 mM BAPTA-buffered pipette solution showed that W-7 exhibited a similar inhibitory effect on hERG current to that with 5 mM EGTA pipette solution (Supporting Information Figure S1).

Figure 2.

Pipette inclusion of W-7 or calmodulin on hERG channel. (A) Time-course of I_hERG.tail recorded with voltage step as shown in the inset in a typical experiment with pipette inclusion of 10 µM W-7 (Rs = 3.4 MΩ), original current traces at corresponding time-points are shown in right of the panel. (B) Time-course of I_hERG.tail recorded in a representative cell with pipette inclusion of 1 µM E4031 (Rs = 3.8 MΩ), original current traces at corresponding time-points are shown in right of the panel. (C) Time-course of I_hERG.tail recorded with voltage step as shown in the inset in a representative cell with pipette inclusion of 100 nM free Ca²⁺ and 500 nM calmodulin (Rs = 3.6 MΩ), original current traces at corresponding time-points are shown in right of the panel.

To examine whether calmodulin regulated hERG channels, we included 500 nM calmodulin in a 5 mM EGTA pipette solution containing 100 nM free Ca²⁺ (calculated using the Cabuf software created by Dr G. Droogmans in the Department of Physiology, KU Leuven, Leuven, Belgium (http://ftp://ftp.cc.kuleuven.ac.be/pub/droogmans), which is close to the physiological intracellular free Ca²⁺ level in HEK 293 cells (Tong et al., 1999). Based on Ca²⁺–calmodulin binding data (Teruel et al., 2000), theoretically about 100 nM calmodulin should be activated. Figure 2C displays the time course of I_hERG.tail during a 20 min dialysis of calmodulin after the cell membrane rupture. No significant change in the current was observed in dialysed cells with calmodulin. Similar results were obtained in a total of five cells. In addition, calmodulin inclusion in the pipette solution did not prevent the channel inhibition by W-7 (data not shown). This result indicates that calmodulin may not have the expected effect on hERG channels, consistent with our previous observation that activity of hERG channels is not related to intracellular Ca²⁺ activity (Tang et al., 2007). Therefore, the suppression of the current by W-7 is likely to be due to a direct blocking effect.

W-7 effects on kinetics of inactivation and reactivation of hERG current

The inactivation of hERG channels is believed to play an important role in high-affinity drug binding to hERG channels (Zhang et al., 1999). To examine whether W-7 would affect inactivation time-course of hERG channels, the current was fully activated and inactivated by a depolarization to +60 mV followed by −100 mV for 10 ms to allow the channels to recover from inactivation, but not enough for channel deactivation (Smith et al., 1996; Spector et al., 1996; Guo et al. 2006), and test steps were then applied to different voltages to observe inactivation time-courses before and after 3 µM W-7 (left panel of Figure 3A). The inactivation time constant was obtained by fitting the current decay to a single exponential function. The mean values of voltage-dependent inactivation time constant are illustrated in the right panel of Figure 3A before and after application of 3 µM W-7. W-7 significantly reduced inactivation time constant at voltages from −30 to +40 mV (n = 9, P < 0.05 or P < 0.01 vs. control), which suggests that inactivation of hERG channels is accelerated.

Figure 3.

Effects of W-7 on hERG channel kinetics. (A) Left panel, current traces recorded with the protocol were used to evaluate inactivation time-constant of hERG channels. Right panel, mean values of voltage-dependent inactivation time constant (tau) in control and after application of 3 µM W-7 (n = 9, *P < 0.05, **P < 0.01 vs. control). (B) Left panel, current traces obtained with the protocol were used to assess steady-state inactivation. After a 1 s inactivation step to +30 mV, the rapid inactivation of hERG channels was relieved by application of 20 ms test pulses to potentials ranging from −120 to +40 mV. Right panel, steady-state inactivation curves were fitted to a Boltzmann distribution. (C) Left panel, current traces recorded with the protocol were used to evaluate recovery time-constant of hERG channels. The recovery phase of the current was fitted to a mono-exponential function. Right panel, time-constant (tau) of recovery from inactivation was reduced by 3 µM W-7 at voltages from −40 to −20 mV (n = 9, **P < 0.01 vs. control).

The voltage protocol and recorded currents before and after 3 µM W-7 (left panel of Figure 3B) were used to determine steady-state inactivation (availability) of hERG channels as previously described (Smith et al., 1996; Tang et al. 2007). The current curves (right panel of Figure 3B) in the absence (control) and presence of 3 µM W-7 were corrected by extrapolating the exponential decay phase back to the start of the negative voltage step and applying the same relative correction to the initial outward tail current as described previously (Smith et al., 1996; Tang et al. 2007). The normalized inactivation curves of corrected I_hERG were fitted to a Boltzmann distribution. The half potential (V_0.5) of hERG channel inactivation was negatively shifted (from −57.5 ± 2.4 mV of control to −67.4 ± 2.7 mV, n = 7, P < 0.01), and the slope factor was slightly, but not significantly, changed from −20.2 ± 0.4 for control to −18.2 ± 0.4 for W-7 (P > 0.05).

The recovery of hERG channels from inactivation was examined before and after 3 µM W-7 application using the standard dual-pulse protocol (left panel of Figure 3C) as previously described (Spector et al., 1996). The rising phase of the current was fitted to a mono-exponential function, and the recovery time constant was plotted against the repolarization potentials (right panel of Figure 3C). The recovery time-constant was significantly reduced by 3 µM W-7 at potentials between −40 and −20 mV (n = 9, P < 0.01 vs. control).

Time-dependent block of hERG channels by W-7

The time dependence of development of I_hERG block by W-7 was evaluated by holding the potential at −80 mV (upper panel of Figure 4A) to ensure that all hERG channels were in a closed state, followed by a long duration (10 s) voltage step to 0 mV to open the channels. The protocol was applied in control, and discontinued during superfusion of 3 µM W-7 (holding potential −80 mV), then re-applied after 10 min drug exposure. W-7 inhibited I_hERG from the beginning of the current activation to end of depolarization step, suggesting that this compound blocked both closed and open channels.

Figure 4.

Development of hERG channel block by W-7. (A) Voltage clamp pulse protocol and representative recordings of hERG current before and after exposure of the cell to 3 µM W-7. (B) Drug-sensitive current expressed as a proportion of the current in the absence and presence of 3 µM W-7. Raw data (points) were fitted to a single exponential function with a time-constant of 830 ± 182 ms (n = 6). (C) Envelope tail protocol and representative hERG current before (control) and after application of 3 µM W-7. Cells were held at a holding potential of −80 mV and pulsed to depolarizing voltage (+30 mV) for variable durations from 50 to 4950 ms in 250 ms increments. I_hERG.tail was recorded upon repolarization to −50 mV. (D) A plot of relative tail current with 3 µM W-7 versus the depolarizing duration. The time-dependent decay in relative tail current was fitted to a single exponential function.

To analyse the contribution of open channel block, onset of channel block was analysed by plotting the time-course of the inhibition by W-7. The onset of open channel block was analysed using the drug-sensitive current formula (Gao et al., 2004): [(I_C−I_B)/I_C], where I_C and I_B are the currents in the absence and presence of W-7 (Figure 4B). The time-constant of the rate of block development was 830.2 ± 182.2 ms (n = 6) with 3 µM W-7 (n = 5), suggesting W-7 also exerted an open channel blocking effect on hERG channels.

The time-course for the development of W-7 block of hERG channels was further assessed using an envelope of tail test (Tang et al., 2007). Cells were held at a holding potential of −80 mV and pulsed to depolarizing voltage (+30 mV) for variable durations from 50 to 4950 ms in 250 ms increments. I_hERG.tail was recorded upon repolarization to −50 mV (upper panel of Figure 4C). The envelope tail test was performed in the same cell before and after addition of 3 µM W-7 (10 min).

The envelope tail current with 3 µM W-7 was expressed relative to control (Figure 4D), and the relative tail current decayed in a pulse duration-dependent manner. The time-course of this decay was fitted to a single exponential function. The onset of hERG channel block by 3 µM W-7 developed rapidly with a time-constant of 328.5 ± 9.9 ms (n = 5) at +30 mV.

Tonic block of hERG channel current by W-7 was estimated by the initial value of the relative tail current activated by the envelope protocol. The initial relative tail current at 50 ms with 3 µM W-7 was 67 ± 4% of control, and the steady-state tail current at 4.5 s was 45 ± 4%. Therefore, the fraction of tonic block of hERG channels was about 33% with 3 µM W-7 compared with 23% in open channel block. This result suggests that W-7 blocks hERG channels by binding to both the resting and open states.

Molecular determinants of hERG channel block by W-7

The molecular determinants of W-7 block of hERG channels were investigated using three hERG mutants, S631A, F656V and Y652A. S631A is a well-documented pore helix, inactivation-deficient hERG mutant that has been used for investigating the role of inactivation in hERG block (Zhang et al., 1999; Guo et al. 2006), and the Y652 and F656 are co-located on the S6 transmembrane domain of the hERG channels, both mutations have been shown to attenuate I_hERG block by a number of drugs (Zhang et al., 1999; Su et al. 2004). To test the molecular determinant of hERG channel block by W-7, the inhibition of S631A, F656V or Y652A channels by this compound was studied in HEK 293 cells transiently expressing these mutants.

Figure 5A illustrates the hERG current traces recorded with a single-voltage protocol as shown in the inset in cells expressing WT hERG, Y652A, F656V and S631A channels, respectively, in the absence and presence of a high concentration of W-7 (30 µM). WT hERG channels were completely blocked by 30 µM W-7, whereas Y652A and F656V channels showed a slight attenuation of I_hERG block by W-7. S631A was less sensitive to block by 30 µM W-7.

Figure 5.

Effects of hERG channel mutations on W-7-induced block. (A) Current traces recorded in the absence (control) and presence of 30 µM W-7 in HEK 293 cells expressing wild-type (WT) or mutant hERG channels: Y652A, F656V and S631A respectively. (B) Concentration–response relationships of W-7 on tail current of WT and various hERG mutant channels were fitted to a Hill equation (n = 4–15 cells for each concentration). The IC₅₀ values of W-7 for inhibiting hERG mutants were increased by 1.6- to 7.3-fold, relative to that for WT hERG channels.

The IC₅₀ of W-7 for inhibiting hERG channels was assessed by fitting the concentration–response curves (Figure 5B) to the Hill equation: E =E_max/[1 + (IC₅₀/C)^b], where E is the inhibition of current in percentage at concentration C, E_max is the maximum inhibition, IC₅₀ is the concentration for a half-maximum action and b is the Hill coefficient. The IC₅₀ values of W-7 were 3.5, 5.5, 9.8 and 25.4 µM, respectively, for inhibiting the tail current (+30 mV) of WT hERG, Y652A, F656V, S631A mutant channels. These results further demonstrate that W-7 directly binds to hERG channels and blocks the channel.

Inhibition of hK_v1.5 channels by W-7

A previous study demonstrated that hK_v1.5 was inhibited by the CaMKII inhibitor KN-93 (2-[N-(2-hydroxyethyl)]-N-(4-methoxybenzenesulphonyl)]amino-N-(4-chlorocinnamyl)-N-methylbenzylamine) in a CaMKII-independent manner (Rezazadeh et al., 2006). Here, we determined whether the calmodulin inhibitor W-7 could regulate hK_v1.5 channels, stably expressed in HEK 293 cells. Figure 6A shows the time-course of hK_v1.5 current recorded in a representative cell in the absence and presence of 10 µM W-7. The compound gradually decreased hK_v1.5 current, and the effect was partially reversed by drug washout for 30 min. Original current traces at corresponding time-points are shown in right of the panel. Figure 6B displays the effects of W-7 on voltage-dependent hK_v1.5 current recorded in a representative cell. W-7 at 10 µM substantially suppressed hK_v1.5 current, and the effect was partially reversed by washout (30 min). Figure 6C illustrates the I–V relationships of hK_v1.5 channels in the absence and presence of 3 and 10 µM W-7. W-7 significantly suppressed activity of hK_v1.5 channels at test potentials of 0 to +50 mV (P < 0.05 or P < 0.01 vs. control). The inhibitory effect of W-7 on hK_v1.5 channels, as in hERG channels (Figure 1), exhibited a significant voltage dependence (Figure 6D). Channel block increased with depolarization to more positive potentials. The concentration–response relationship of W-7 is illustrated in Figure 6E, and fitted to the Hill equation. The IC₅₀ of W-7 for inhibiting hK_v1.5 channels (at +50 mV) was 6.5 µM, and Hill coefficient was 1.1. Moreover, dialysis of cells with 500 nM calmodulin had no effect on hK_v1.5 channels (n = 5, data not shown).

Figure 6.

Inhibition of hK_v1.5 channels by W-7. (A) Time-course of hK_v1.5 current recorded in a HEK 293 cell stably expressing hK_v1.5 channels. Membrane current was elicited by a 300 ms voltage step to +50 mV from a holding potential of −80 mV, then back to −50 mV (left insert) every 10 s. Original current traces at corresponding time-points are shown in the right of the panel. (B) Voltage-dependent hK_v1.5 current recorded in a representative HEK 293 cell with 300 ms voltage steps to between −50 and +50 mV from a holding potential of −80 mV (inset) at 0.1 Hz in the absence and presence of 10 µM W-7. (C) I–V relationships of hK_v1.5 (measured at end of the depolarization steps) in the absence and presence of 3 and 10 µM W-7. (D) Voltage dependence of W-7 for inhibiting hK_v1.5 channels (P < 0.05 or P < 0.01 at potentials of +10 to +60 mV vs. −10 or 0 mV). (E) Concentration–response relationship of W-7 for inhibiting hK_v1.5 current was fitted to a Hill equation (n = 7–12 cells for each concentration).

Effect of W-7 on hK_IR2.1 channels

The inward rectifier K⁺ current (I_K1 or K_IR2.1) was reported to be up-regulated by the inhibition of CaMKII in mice with chronic myocardial CaMKII inhibition resulting from transgenic expression of a CaMKII inhibitory peptide (AC3-I) (Li et al., 2006). Here, we determined whether acute suppression of calmodulin could up-regulate hK_IR2.1 channels stably expressed in HEK 293 cells. Figure 7A displays the effect of W-7 on hK_IR2.1 current recorded with the voltage protocol shown in the inset in a representative cell in the absence and presence of 10 µM W-7. W-7 did not increase, instead, suppressed hK_IR2.1 current, and the inhibitory effect was partially reversed by drug washout (30 min). Figure 7B illustrates the I–V relationships of hK_IR2.1 current before and after application of 3, 10 and 30 µM W-7. The hK_IR2.1 current was significantly inhibited by W-7 at test potentials of −120 to −80 mV and −60 to −10 mV (n = 6–15, P < 0.05 or P < 0.01 vs. control) in a concentration-dependent manner. The concentration–response curve (Figure 7C) of W-7 was fitted to a Hill equation. The IC₅₀ of W-7 for inhibiting hK_IR2.1 current (at −120 mV) was 13.4 µM, and Hill coefficient was 1.4. These results indicate that in addition to K_v channel block, W-7 may directly inhibit hK_IR channels.

Figure 7.

Inhibition of hK_IR2.1 channels by W-7. (A) Voltage-dependent hK_IR2.1 current recorded in a representative HEK 293 cell with 200 ms voltage steps to between −120 and 0 mV from a holding potential of −40 mV (inset) at 0.1 Hz in the absence and presence of 10 µM W-7. (B) I–V relationships of hK_IR2.1 current (measured at end of the depolarization steps) in the absence and presence of 3, 10 and 30 µM W-7. (C) The concentration–response relationship of W-7 for inhibiting hK_IR2.1 current fitted to a Hill equation (n = 5–8 cells for each concentration).

Inhibition of hERG, hK_v1.5 and hK_IR2.1 currents by W-13

N-(4-aminobutyl)-5-chloro-2-naphthalene sulphonamide hydrochloride (W-13) is another membrane-permeable calmodulin inhibitor, with a different chemical structure from that of W-7, and its effects on hERG, hK_v1.5 and hK_IR2.1 currents were determined. Figure 8 shows that W-13 also suppressed these K⁺ currents; however, the inhibitory effect was weaker than W-7. The IC₅₀ of W-13 for inhibiting I_hERG.tail was 19.8 µM (Hill coefficient, 1.2); 75.5 µM for inhibiting hK_v1.5 (Hill coefficient, 0.8), 43.7 µM for inhibiting hK_IR2.1 currents (Hill coefficient, 2.1).

Figure 8.

Inhibition of hERG, hK_v1.5 and hK_IR2.1 currents by W-13. (A) hERG current recorded in a representative HEK 293 cell with a 5 s voltage step to +30 mV from a holding potential of −80 mV, then back to −50 mV at 0.02 Hz in the absence and presence of 10 and 30 µM W-13 for 8 min incubation. (B) The concentration–response relationship of W-13 for inhibiting hERG current fitted to a Hill equation (n = 5–9 cells for each concentration). (C) hK_v1.5 current recorded in a representative HEK 293 cell with a 300 s voltage step to +50 mV from a holding potential of −80 mV, then back to −40 mV at 0.02 Hz in the absence and presence of 30 and 100 µM W-13. (D) The concentration–response relationship of W-13 for inhibiting hK_v1.5 current fitted to a Hill equation (n = 5–7 cells for each concentration). (E) hK_IR2.1 current recorded in a representative HEK 293 cell with a 200 ms voltage step to −120 mV from a holding potential of −40 mV at 0.02 Hz in the absence and presence of 30 and 100 µM W-13. (F) The concentration–response relationship of W-13 for inhibiting hK_IR2.1 current fitted to a Hill equation (n = 5–6 cells for each concentration).

Discussion

It is believed that increased activity of calmodulin and CaMKII is involved in pathophysiology of different types of cardiac disorders, including ventricular arrhythmias (Anderson, 2005; Lu et al., 2006). Inhibition of calmodulin kinase activity was found to be anti-arrhythmic with the calmodulin inhibitor W-7 (Wu et al. 1999; Gbadebo et al., 2002; Anderson, 2005; Lu et al., 2006). It has been recognized that W-7 inhibits calmodulin kinases and regulates additional enzymes (Rakhilin et al., 2004) and/or ion channels including Na⁺ channels (Herzog et al. 2003; Lee-Kwon et al., 2007), K⁺ channels (e.g. K_v4.3, K_v1.3, KCNQ1/KCNE1) (Chang et al. 2001; Shamgar et al., 2006; Qu et al., 2007), pacemaker channels (Rigg et al., 2003; Chatelier et al. 2005) and Ca²⁺ channels (Caulfield et al. 1991; Anderson et al., 1994; Dzhura et al., 2000; Wu et al. 2004) in different types of cells. The effects of W-7 on ion channels were found to be calmodulin dependent for K_v1.3 channels in activated human T lymphocytes (Chang et al., 2001), Na_v1.3 channels expressed in rat descending vasa recta (Lee-Kwon et al., 2007) and KCNQ1/KCNE1 channels expressed in Chinese hamster ovarian cells or Xenopus oocytes (Shamgar et al., 2006), and guinea pig sinus-atrial I_f (Rigg et al., 2003) and ventricular I_Ca.L (Anderson et al., 1994; Dzhura et al., 2000; Wu et al. 2004). On the other hand, the effects of W-7 were calmodulin independent for rabbit sinus-atrial I_f (Chatelier et al., 2005) and neuronal K⁺ currents and Ca²⁺ current (Caulfield et al., 1991). Both calmodulin-dependent and independent effects were also reported for K_v4.3 channels expressed in Xenopus oocytes (Qu et al., 2007).

The IC₅₀ values of W-7 for inhibiting calmodulin-regulated enzymes are variable and tissue dependent in the range of 11–39 µM (Nakajima and Katoh, 1987; Osawa et al. 1998). However, no evidence was detected in the present study for calmodulin-dependent processes in tissue-cultured cell lines. We demonstrated that the calmodulin inhibitor W-7 blocked hERG channels stably expressed in HEK 293 cells (Figures 1–5) in a voltage-dependent and concentration-dependent manner (Figure 1). The IC₅₀ of W-7 for inhibiting hERG channels was 3.5 µM (Figure 5), 3–11 times less than the IC₅₀ values for inhibiting calmodulin-regulated enzymes (Nakajima and Katoh, 1987; Osawa et al. 1998). Moreover, another calmodulin inhibitor, W-13, also inhibited hERG channels. These results suggest that hERG channels, unlike hEAG channels (Schonherr et al. 2000; Ziechner et al., 2006), might not be regulated by calmodulin in HEK 293 cell line. This notion is supported by the following observations: (i) inclusion of W-7 in pipette solution induced a lesser degree of current inhibition than that with bath solution application (Figure 2B); (ii) calmodulin (500 nM) pipette inclusion produced no effect on hERG current (Figure 2B); and (iii) like other hERG channels blockers (Milnes et al., 2003; Guo et al., 2006; Ridley et al., 2006; Su et al. 2006; Tang et al. 2008), mutations of hERG channels reduced the blocking effect of W-7 (Figure 5).

The blocking properties of hERG channels by W-7 are similar, but not identical, to those of the selective 5-HT re-uptake inhibitor, fluvoxamine, and the metabolite of the aldosterone antagonist spironolactone, canrenoic acid (Caballero et al. 2003; Milnes et al., 2003). W-7, like fluvoxamine or canrenoic acid (Caballero et al. 2003; Milnes et al., 2003), blocked hERG channels by binding to both closed and open states of the channel (Figure 4), negatively shifted the activation conductance of the channel and exhibited a voltage-dependent block of the channel (Figure 1C–E). W-7 also negatively shifted the voltage dependence of hERG channel inactivation, and reduced the channel reactivation time-constant at positive potentials (Figure 3), unlike canrenoic acid which had no effect on these parameters (Milnes et al., 2003). On the other hand, similar to fluvoxamine, the blocking effect of W-7 was reduced on the mutants in the S6 Y652A and F656V, and the pore helix mutant S631A (Figure 5). Although the changes in onset and recovery of inactivation were statistically significant, these changes were hardly profound. Thus, W-7 unlikely inhibits hERG channel activity through affecting the inactivation gating of the channel.

It is generally recognized that S631 is located at the outer mouth of the pore helix and S631A is an inactivation-deficient mutant of hERG channels. The dramatically reduced blocking effect (Figure 5) suggests that the structural changes of S631A mutant associated with lack of inactivation may influence W-7 binding to the channel. On the other hand, the mutants Y652A and F656V exhibited only a slight attenuation of channel block by W-7, suggesting some binding to drug receptor sites within the pore-S6 region, as established in other hERG channel blockers (Milnes et al., 2003; Guo et al., 2006; Ridley et al., 2006; Su et al. 2006; Tang et al. 2008). The possible contribution of closed-channel blockade to the overall action of W-7 may also be of relevance here. The mutations F656V, Y652A and S631A are anticipated to influence blockade contingent upon channel gating and binding within the pore.

The calmodulin inhibitor W-7 (also W-13), like the CaMKII inhibitor KN-93 (Rezazadeh et al., 2006), not only directly suppressed hERG channels, but also inhibited hK_v1.5 channels (Figure 6). Moreover, the direct channel block is also applicable to hK_IR2.1 channels stably expressed in HEK 293 cells (Figure 7). W-7 inhibited hK_IR2.1 current in a concentration-dependent manner. The IC₅₀ values of W-7 for blocking hERG (3.5 µM), hK_v1.5 (6.5 µM) and hK_IR2.1 channels (13.4 µM) were lower than that for inhibiting calmodulin (Osawa et al., 1998). However, the possibility that the result from HEK 293 cells may not be identical to that from native cardiac tissue cannot be excluded.

In summary, in addition to the inhibition of hK_v1.5 and hK_IR2.1 channels, the calmodulin inhibitor W-7 directly blocked hERG channels by binding to inactivated and activated states of the channels in a HEK 293 cell line. Therefore, caution should be taken in interpreting observations of calmodulin regulation of ion channels, using W-7.

Glossary

Abbreviations

CaMKII

Ca²⁺-activated calmodulin-dependent protein kinase II

hEAG

human ether-à-go-go gene

hERG

human ether-à-go-go-related gene

I_Kr

rapidly delayed rectifier K⁺ current

W-7

N-(6-aminohexyl)-5-chloro-1-naphthalene sulphonamide

W-13

N-(4-aminobutyl)-5-chloro-2-naphthalene sulphonamide hydrochloride

Conflict of interest

The authors state no conflict of interest.

Supplemental information

Additional Supporting Information may be found in the online version of this article:

Figure S1 Inhibitory effect of W-7 on hERG current recorded with a pipette solution containing 5 mM BAPTA. (A) Voltage-dependent hERG channel currents were recorded in a typical experiment with 5 s voltage steps to between −60 and +60 mV from a holding potential of −80 mV, then back to −50 mV (inset) at 0.05 Hz in the absence and presence of 3 μM W-7 for 8 min incubation. W-7 remarkably inhibited hERG current, and the effect is partially reversed on washout. (B) The concentration–response relationship of W-7 for inhibiting hERG current fitted to a Hill equation (n = 4–6 cells for each concentration). The IC50 (at +30 mV) of W-7 for inhibiting hERG tail current was 3.4 μM with a Hill coefficient of 1.1.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.