Differential effects of K_v11.1 activators on K_v11.1a, K_v11.1b and K_v11.1a/K_v11.1b channels

Abstract

BACKGROUND AND PURPOSE

K_v11.1 channels are involved in regulating cellular excitability in various tissues including brain, heart and smooth muscle. In these tissues, at least two isoforms, K_v11.1a and K_v11.1b, with different kinetics, are expressed. K_v11.1 activators are potential therapeutic agents, but their effects have only been tested on the K_v11.1a isoform. In this study, the effects of two different K_v11.1 activators, NS1643 and RPR260243, were characterized on K_v11.1a and K_v11.1b channels.

EXPERIMENTAL APPROACH

K_v11.1a and K_v11.1b channels were expressed in Xenopus laevis oocytes, and currents were measured using two-electrode voltage clamp. I/V curves and channel kinetics were measured before and after application of 30 µM NS1643 or 10 µM RPR260243.

KEY RESULTS

NS1643 increased steady-state currents through Kv11.1b several fold more than through K_v11.1a channels, without affecting EC₅₀ values. NS1643 increased activation rates and decreased rates of inactivation, recovery from inactivation and deactivation for both channels. Except for activation, where effect of NS1643 was comparable, relative changes were greater for Kv11.1b than for K_v11.1a. RPR260243 increased steady-state currents only through Kv11.1a channels, but slowed the process of deactivation for both channels primarily by decreasing time constant of slow deactivation. This effect was greater on K_v11.1b than on K_v11.1a. Effects of both compounds on heteromeric K_v11.1a/K_v11.1b channels were similar to those on K_v11.1a.

CONCLUSIONS AND IMPLICATIONS

Both NS1643 and RPR260243 displayed differential effects on K_v11.1a and K_v11.1b channels, the effects being relatively more pronounced on K_v11.1b channels. This affirms the importance of testing the effect of K_v11.1 activators on different channel isoforms.

Introduction

Potassium currents carried by the K_v11.1 (ether-a-go-go-related gene or ERG1) ion channel affect excitability in a range of tissues including brain (Chiesa et al., 1997; Huffaker et al., 2009), heart (Sanguinetti et al., 1995) and smooth muscle cells (Ohya et al., 2002; Farrelly et al., 2003). In these tissues, at least two different isoforms of K_v11.1, namely K_v11.1a and K_v11.1b, have been found to be co-expressed (Ohya et al., 2002; Jones et al., 2004; Guasti et al., 2005). K_v11.1a is the full-length isoform originally characterized (Sanguinetti et al., 1995; Trudeau et al., 1995), and K_v11.1b is an N-terminal splice variant (Lees-Miller et al., 1997; London et al., 1997). Specifically, compared to K_v11.1a, the N-terminal of K_v11.1b is 340 amino acids shorter. However, the initial 36 amino acids of K_v11.1b are unique to this isoform. Furthermore, the two isoforms differ in their kinetic gating profiles. K_v11.1b displays faster kinetics of activation, recovery from inactivation and most prominently deactivation as compared to K_v11.1a (Larsen et al., 2008).

Recently, K_v11.1 activators have gained interest as potential therapeutic agents mainly as a potential treatment of certain types of cardiac arrhythmias (Kang et al., 2005; Zhou et al., 2005; Hansen et al., 2006a,b; 2007; Diness et al., 2008). Two of these compounds have markedly different modes of action. NS1643 has been shown to increase K_v11.1 currents primarily by right-ward shifting the inactivation curve and by slowing the fast inactivation process (Casis et al., 2006; Hansen et al., 2006a), although effects on other kinetic parameters have also been described (Xu et al., 2008). In contrast, RPR260243 almost exclusively acts by slowing the deactivation process of the channels (Kang et al., 2005). However, these compounds have only been characterized on the K_v11.1a isoform. In the heart, both K_v11.1a and K_v11.1b have been shown to contribute to the generation of I_Kr, the rapid delayed rectifier current (Jones et al., 2004). The functional properties of I_Kr have recently been suggested to depend on the relative abundance of the K_v11.1 isoforms (Larsen and Olesen, 2010). Furthermore, alterations in the relative expression of K_v11.1 isoforms in neurones have been linked to schizophrenia (Huffaker et al., 2009). Thus, it is clear that K_v11.1 isoforms other than K_v11.1a contribute to the physiological function of K_v11.1-mediated currents. Given the inherent kinetic differences between the K_v11.1a and b isoforms (Larsen et al., 2008), it is likely that activators that affect specific kinetic properties of the channels will display differential effects on the K_v11.1 isoforms. Such differential activators may prove valuable tools to address the physiological role of K_v11.1 isoforms in a given tissue. Additionally, given the therapeutic potential of K_v11.1 activators, it is important to investigate possible differential effects of the compounds on the K_v11.1 isoforms.

Here, we tested the effects of the K_v11.1 channel activators NS1643 and RPR260243 on the human variants of the K_v11.1a and K_v11.1b isoforms. Also, we tested their effects on both homomeric and heteromeric channels. Our results demonstrate that both activators have differential effects on K_v11.1 channels depending on the subunit composition, and suggest that it may be possible to differentially target K_v11.1 isoforms. Thus, our results affirm the importance of testing the effect of K_v11.1 activators on different channel isoforms.

Methods

All animal procedures conformed to the guidelines of the Danish National Committee for Animal Studies. Unless otherwise mentioned, all chemicals were obtained from Sigma-Aldrich (St Louis, MO, USA). The nomenclature of ion channels used in this paper conforms to the standards outlined in Alexander et al. (2008).

DNA constructs

cDNAs encoding human K_v11.1a (acc. number NM_000238) and K_v11.1b (acc. number NM_172057) were inserted into pXOOM for expression in Xenopus laevis oocytes (Jespersen et al., 2002). The K_v11.1b clone was obtained from A. Arcangeli (Università degli Studi di Firenze, Florence, Italy).

Expression in X. laevis oocytes

Preparation and injection of cRNA in XO were performed as previously described (Larsen et al., 2008). Oocytes were injected with 1.25–5 ng of cRNA. For co-expression of K_v11.1a and K_v11.1b, cRNAs were mixed in a molar ratio of 4:1 (K_v11.1a : K_v11.1b) before injection. Experiments were performed 3 days after injection.

Electrophysiological recordings

Measurements on X. laevis oocytes were performed with the two-electrode voltage clamp technique using a Dagan CA-1B amplifier (Minneapolis, MN, USA). All recordings were performed at room temperature (21–22°C) under continuous superfusion with Kulori solution (in mM: 87 NaCl, 4 KCl, 1 MgCl₂, 1 CaCl₂, 5 HEPES, pH 7.4). Oocytes were placed in a perfusion chamber connected to a multi-barrel flow system that allowed for rapid change between solutions. Glass pipettes for recording electrodes were pulled from borosilicate glass capillaries (Module Ohm, Herlev, Denmark) on DMZ Universal Puller (Zeitz Instruments, Munich, Germany), and had a tip resistance between 0.5 and 2.5 MΩ when filled with 2 M KCl. The series resistance was compensated using the internal compensatory circuit on the amplifier. Data acquisition was performed with the Pulse software (HEKA Elektronik, Lambrecht/Pfalz, Germany). For all recordings, the holding potential was −80 mV. The time interval between recordings was adjusted depending on the specific protocol and the drug applied in order to allow channels to fully deactivate before a new protocol was initiated.

Data analysis was performed with Igor Pro (Wavemetrics, Lake Oswego, OR, USA) and GraphPad Prism (GraphPad Software Inc, San Diego, CA, USA). I/V curves were constructed by measuring the current at the end of a 2 s voltage step to potentials ranging from −80 to +40 mV. The data were plotted against the corresponding membrane potentials. Similarly, the peak tail currents, measured at −60 mV following the depolarizing step, were plotted against the membrane potential of the depolarizing step to construct the activation curves. A Boltzmann function (I/I_max = I_min+ (I_max−I_min)/{1 + exp[(V₅₀−V)/k)]} was fitted to the activation curves to obtain the potential of half-maximal activation (V₅₀) and the slope factor (k). The time constant of activation (τ_act) at 0 mV was determined using the standard envelope of tails protocol, measuring the peak tail current at either −60 or −100 mV. A mono-exponential function was fitted to the normalized peak tail currents to obtain τ_act. An estimate of the values of the time constants of channel deactivation (τ_slow and τ_fast) was obtained by fitting a double-exponential function to tail current traces measured at −60 mV following a voltage step to +40 mV for 1 s to achieve maximal activation. Similarly, the time constant of recovery from inactivation (τ_rec) at −60 mV was determined by subtracting the extrapolated fit of the deactivation process from the initial rising phase of the tail currents. The difference between the extrapolated values and the recorded current values represents the time-course of recovery from inactivation. The time constant of this process was estimated by fitting a mono-exponential function to the resulting curve. Inactivation was addressed by first clamping the oocytes at +40 mV for 1 s, then briefly at −120 mV to allow channels to recover from the inactivated state, and finally at 0 mV to investigate the time-course of inactivation at this potential. To avoid significant closure of the channels during the clamp at −120 mV, this step lasted 10 ms for K_v11.1a and 5 ms for K_v11.1b, due to the faster deactivation kinetics of K_v11.1b.

The time constant of inactivation was estimated by fitting a mono-exponential function to the initial steep decrease in current amplitude at 0 mV.

Drugs

NS1643 [1,3-bis-(2-hydroxy-5-trifluoromethyl-phenyl)-urea; Hansen et al., 2006a] and RPR260243 [(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; Kang et al., 2005] were synthesized at NeuroSearch A/S, Ballerup, Denmark. The synthesis of RPR260243 was complicated involving more than 10 steps and a very low yield in the final purification step. Therefore, we were only able to obtain limited amounts of RPR260243, and consequently we were not able to do a full set of experiments, including a full concentration–response curve, with this drug. NS1643 was dissolved as a 30 mM stock solution in DMSO. RPR260243 was dissolved as a 10 mM stock solution in DMSO. Both compounds were diluted in the Kulori solution to achieve the desired working concentrations.

Data analysis

Data are represented as mean ± SEM, unless otherwise indicated. Paired t-tests or anova followed by Tukey's method of multiple comparisons were used as appropriate to compare data before and after drug application, as well as the relative effects of the compounds on K_v11.1a and K_v11.1b. P values below 0.05 were considered statistically significant.

Results

K_v11.1a and K_v11.1b channels were expressed as homomeric or heteromeric K_v11.1a/K_v11.1b (4:1) channels in X. laevis oocytes. A molar ratio of 4:1 reflects the apparent ratio of the isoforms as determined by real-time PCR of amplified mRNA in adult human heart (Larsen et al., 2008). To verify that co-expression of K_v11.1a and K_v11.1b led to the formation of heteromeric channels, deactivation kinetics were compared to those of homomeric K_v11.1a and K_v11.1b. Using the average time constants for the deactivation process under control conditions (Figures 3 and 7), the observed current decay for K_v11.1a/K_v11.1b (4:1) was compared to the expected theoretical current decay, assuming the formation of only homomeric channels. The theoretical current decay was calculated as a weighted algebraic summation (4:1) of the deactivation kinetics of K_v11.1a and K_v11.1b homomeric channels. Supporting Information Figure S1 shows the normalized decay of both the observed and theoretical situations. In the observed situation, the current decay was faster than in the theoretical situation. Thus, the observed currents cannot be explained by the presence of homomeric channels only. This argument is valid under the assumption that the expression level is comparable for homomeric K_v11.1a and K_v11.1b channels. The majority of the experiments in this study were not designed to compare current density per se; however, statistical comparison of a separate set of experiments, where oocytes were injected with the same amount of either K_v11.1a or K_v11.1b cRNA, showed that the steady-state current measured at the end of a 2 s depolarizing pulse to 0 mV was significantly lower for K_v11.1b than for K_v11.1a homomeric channels (K_v11.1a, I = 1.5 ± 0.3 µA, n = 6; K_v11.1b, I = 0.6 ± 0.05 µA, n = 9; P < 0.05). If these data were incorporated in the calculation of theoretical current decay, the properties of K_v11.1a homomeric channels would be even more dominant. Thus, the comparison of theoretical and observed current decay clearly demonstrates the existence of functional heteromeric channels. This is in line with previous observations on the formation of heteromeric K_v11.1a and K_v11.1b channels in oocytes (London et al., 1997).

Figure 3.

Effect of NS1643 on deactivation kinetics of K_v11.1 channels. (A) Representative current traces from oocytes expressing K_v11.1a (left panel), K_v11.1b (middle panel) or K_v11.a/K_v11.1b (4:1) (right panel) in the absence and presence of NS1643, and an enlargement of the tail currents (normalized to maximum current amplitude). Currents were elicited by the voltage clamp protocol shown in the insert in (A). The time-course of deactivation could be described by a double-exponential function. Time constants of the slow component (B), fast component (C) and the relative contribution of the fast and slow component (D) before and after application of 30 µM NS1643 are shown. In addition, the relative effect of NS1643 on (B, C and D) are shown. (K_v11.1a; K_v11.1b; K_v11.1a/K_v11.1b: n = 9; 12; 7).

Figure 7.

Effect of RPR260243 on deactivation kinetics of K_v11.1 channels. Representative current traces from oocytes expressing K_v11.1a (left panel), K_v11.1b (middle panel) or K_v11.1a/K_v11.1b (4:1) (right panel) in the absence and presence of RPR260243 are shown in the top panel, and an enlargement of the tail currents (normalized to maximum current amplitude) is shown in the bottom panel. Time constants of the slow component (B), fast component (C) and the relative contribution of the fast and slow component (D) before and after RPR260243 (10 µM) are shown. In addition, the relative effect of RPR260243 on (B, C and D) are shown (for further explanation, see Figure 3). (K_v11.1a; K_v11.1b; K_v11.1a/K_v11.1b: n = 5; 6; 8).

Effects of NS1643 on Kv11.1 channels

The effects of 30 µM NS1643 on steady-state and tail currents were investigated. In Figure 1A, representative current recordings from oocytes expressing K_v11.1a (left side), K_v11.1b (middle) and K_v11.1a/K_v11.1b (right side) channels clamped at 0 mV for 2 s, and then at −60 mV for 2 s under control conditions and after application of NS1643 are shown. The data demonstrate that while NS1643 increased the steady-state current for both K_v11.1 homomeric and heteromeric channels, the increase was more noticeable for K_v11.1b at 0 mV. In Figure 1B, the I/V relationships are shown for K_v11.1a (left), K_v11.1b (middle) and K_v11.1a/K_v11.1b (right side). The protocol used to determine the I/V relationship is shown in the inset in Figure 1B. The current amplitude at the end of the 2 s voltage step was plotted against the command potential. For comparison, the data were normalized to the maximum current amplitude under control conditions. NS1643 consistently increased currents through both channels for voltages above –50 mV. The data show that the effect of NS1643 on steady-state current amplitude was greater on K_v11.1b channels over this voltage range. To determine the half-maximal voltage of activation (V₅₀), Boltzmann functions were fitted to the normalized tail current amplitudes measured at −60 mV (Figure 1C). For K_v11.1a (left side), NS1643 shifted the V₅₀ value from −21.9 ± 0.8 to −30.6 ± 0.9 mV (P < 0.05). For K_v11.1b (middle), NS1643 shifted the V₅₀ value from −12.5 ± 1.2 to −27.4 ± 1.6 mV (P < 0.05). For heteromeric channels, application of NS1643 shifted the V₅₀ value from −25.5 ± 0.9 to −32.3 ± 1.9 mV (P < 0.05).

Figure 1.

Effect of NS1643 on current–voltage relationships of K_v11.1 channels. (A) Representative current traces recorded at a test potential of 0 mV from Xenopus laevis oocytes expressing either K_v11.1a (left panel), K_v11.1b (middle panel) or K_v11.1a/K_v11.1b (4:1) (right panel) in the absence and presence of 30 µM NS1643. To construct current–voltage (I/V) relationships, currents were elicited using the voltage clamp protocol shown in the insert (A). Steady-state currents measured at the end of each step were normalized to the maximum current in control condition, and plotted as a function of the test potential (B). Voltage dependence of activation was investigated by plotting the normalized peak tail current measured immediately after stepping back to −60 mV against the preceding step potential. V₅₀ values were calculated from individual Boltzmann fits to the normalized tail current–voltage relationships (C) (K_v11.1a; K_v11.1b; K_v11.1/K_v11.1b: n = 10; 13; 7).

To determine whether differences in the potency of NS1643 could account for the more noticeable effect on K_v11.1b, concentration–response curves were constructed for both isoforms (Supporting Information Figure S2). There was no significant difference between the EC₅₀ values for NS1643 on K_v11.1a (9.4 ± 1.1 µM) and K_v11.1b (8.6 ± 0.7 µM).

The effect of NS1643 on activation kinetics at 0 mV was evaluated using an envelope of tails protocol (Figure 2). The protocol was carried out by activating channels at 0 mV for various durations of time (from 10 to 1000 ms), and then measuring the tail current at either −60 or −100 mV. The amplitude of the tail current represents the relative amount of activated channels released from inactivation at a given time-point. The protocols (inset) along with representative current recordings from oocytes expressing K_v11.1a (left side), K_v11.1b (middle) and K_v11.1a/K_v11.1b (right side) are shown in Figure 2A. For simplicity, only currents under control conditions are shown. In Figure 2B, the peak tail currents normalized to maximum amplitude are shown as a function of the duration of the activating step. The solid lines correspond to mono-exponential fits to the data. The data show that NS1643 accelerates activation (decreases τ) of K_v11.1a, K_v11.1b and K_v11.1a/K_v11.1b (4:1) channels. The time constant of activation (τ_act) was obtained from the fitted functions, and quantified in Figure 2C. The τ_act for the K_v11.1a, K_v11.1b and K_v11.1a/K_v11.1b channels was significantly decreased by NS1643. The relative change in τ_act was comparable for Kv11.1a, Kv11.1b and K_v11.1a/K_v11.1b as shown in Figure 2C, demonstrating no differential effect of NS1643 on this parameter.

Figure 2.

Effect of NS1643 on activation kinetics of K_v11.1 channels. To avoid contribution from fast inactivation, an envelope of tails voltage clamp protocol (insert) was used to study the activation kinetics of K_v11.1a (left panel), K_v11.1b (middle panel) or K_v11.1a/K_v11.1b (4:1) (right panel). Representative current traces are depicted in (A). Peak currents measured when stepping back to −100 mV (K_v11.1a and K_v11.1a/K_v11.1b) or −60 mV (K_v11.1b) were normalized and plotted as a function of activation time. Mean time constants from single exponential fits to the current–time relationships before and after application of 30 µM NS1643, and the relative effect of NS1643 on the time constants for K_v11.1a, K_v11.1b and K_v11.1a/K_v11.1b are shown in (C) and (D). (K_v11.1a; K_v11.1b; K_v11.1a/K_v11.1b: n = 10; 9; 7).

The deactivation kinetics were determined at −60 mV. In Figure 3A, representative current recordings are shown for K_v11.1a (left side), K_v11.1b (middle) and K_v11.1a/K_v11.1b (right side) before and after application of NS1643. Also shown are enlargements of the tail currents normalized to the peak tail current to allow comparison of the time-course of deactivation. NS1643 slowed the deactivation time-course of both channel isoforms. However, due to the intrinsic fast deactivation of K_v11.1b, the effect is relatively more pronounced on this channel isoform. By fitting a double-exponential function to the current traces, the two time constants of deactivation and their relative contribution to the process could be determined. On K_v11.1a, NS1643 did not significantly change the slow time constant of deactivation (τ_slow), whereas on K_v11.1b and K_v11.1a/K_v11.1b, τ_slow was increased significantly (P < 0.001) as shown in Figure 3B (left side). As shown in Figure 3C (left side), the fast time constant of deactivation (τ_fast) was significantly increased by NS1643 for K_v11.1a (P < 0.05), K_v11.1b (P < 0.001) and K_v11.1a/K_v11.1b (P < 0.01). However, comparison of the relative change in τ_fast revealed that the effect was greater on K_v11.1b as compared to K_v11.1a and K_v11.1a/K_v11.1b (P < 0.001, Figure 3C, right side). The relative contribution of the fast component of deactivation was not changed by NS1643 for K_v11.1a. For K_v11.1b and K_v11.1a/K_v11.1b, the relative contribution of the fast component was significantly decreased by NS1643 (P < 0.001 and P < 0.01 respectively). Quantification of the relative changes in all three parameters clearly demonstrates that the effects of NS1643 on deactivation kinetics are isoform dependent (K_v11.1a vs. K_v11.1b: P < 0.01 for τ_slow, P < 0.001 for τ_fast, P < 0.05 for the relative contribution; K_v11.1b vs. K_v11.1a/K_v11.1b: P < 0.001 for τ_fast; Figure 3B–D, right side). Thus, NS1643 slows deactivation of K_v11.1b relatively more than deactivation of K_v11.1a and K_v11.1a/K_v11.1b.

Inactivation properties were addressed using a three-pulse protocol designed to isolate the process of inactivation (see Methods and inset in Figure 4A for details). In Figure 4A, representative current recordings of K_v11.1a (left side), K_v11.1b (middle) and K_v11.1a/K_v11.1b (right side) channels using such a protocol are shown. The currents have been enlarged and normalized to maximum amplitude for comparison of the time-course of inactivation. For both homomeric and heteromeric channels, NS1643 slows the process of inactivation as compared to control. Quantification of the time constant of inactivation (τ_inact) shows that NS1643 increases τ_inact significantly (P < 0.001 for both K_v11.1a, K_v11.1b and K_v11.1a/K_v11.1b). However, comparison of the relative change in τ_inact shows that the effect is greater on K_v11.1b channels as compared to K_v11.1a or K_v11.1a/K_v11.1b (P < 0.001 and P < 0.001, respectively, Figure 4C).

Figure 4.

Effect of NS1643 on inactivation kinetics of K_v11.1 channels. Oocytes expressing K_v11.1a (left panel), K_v11.1b (middle panel) or K_v11.1a/K_v11.1b (4:1) (right panel) were voltage clamped at +40 mV for 1 s and stepped to −120 mV for 10 ms (K_v11.1a and K_v11.1a/K_v11.1b) or 5 ms (K_v11.1b) to allow the channels to recover from inactivation, and subsequently clamped at 0 mV to investigate the time-course of inactivation. Representative traces of K_v11.1a are shown in (A, left). Also shown in (A) are enlargements of the time-course of inactivation normalized to maximum current amplitude for K_v11.1a, K_v11.1b and K_v11.1a/K_v11.1b (4:1) respectively. (B) The time-course of inactivation was well-fit by a single exponential function. Time constants of inactivation are shown in (B), and the relative effect of NS1643 in (C). (K_v11.1a; K_v11.1b; K_v11.1a/K_v11.1b: n = 10; 11; 6).

The process of recovery from inactivation can be addressed by looking at the initial rising phase of the tail currents after a depolarizing pulse. The data obtained using the deactivation protocol (see above) were used to address the time-course of recovery from inactivation at −60 mV. In Figure 5A, enlargements of the initial rising phase of representative currents recorded under control conditions and in the presence of NS1643 are shown for K_v11.1a (left side), K_v11.1b (middle) and K_v11.1a/K_v11.1b (right side). The dotted lines correspond to the fit of the deactivation process extrapolated back to the point where the command voltage was changed from +40 to −60 mV. To avoid any potential influence of fast deactivation, especially for K_v11.1b, on the recovery kinetics, the recorded current was subtracted from the extrapolated values to obtain the true time-course of recovery. In Figure 5B, the resulting (difference) currents are shown. The currents were normalized to maximum amplitude for comparison of the time-course. NS1643 changed the time-course of recovery only a little for K_v11.1a and K_v11.1a/K_v11.1b channels, while a clear slowing of the process was observed for K_v11.1b. Data for the time constant of recovery (τ_recovery) are shown in Figure 5C. The data show that for K_v11.1a and K_v11.1a/K_v11.1b, NS1643 induced a small, but significant increase in τ_recovery (P < 0.001 and P < 0.001, respectively), while for K_v11.1b the increase (P < 0.001) was more noticeable. Comparing the relative change in τ_recovery, the effect of NS1643 is clearly more noticeable for K_v11.1b than for K_v11.1a or K_v11.1a/K_v11.1b (P < 0.001 and P < 0.001 respectively).

Figure 5.

Effect of NS1643 on kinetics of recovery from inactivation of K_v11.1 channels. (A) Representative current trace of the recovery from inactivation from Xenopus laevis oocytes expressing K_v11.1a (left panel), K_v11.1b (middle panel) or K_v11.1a/K_v11.1b (4:1) (right panel) in the absence and presence of 30 µM NS1643. The fit of the deactivation (dotted line) was extrapolated to the time-point when the membrane potential was changed from +40 to −60 mV. By subtracting the initial rising phase of the current from the extrapolated fit of deactivation, the time-course of the recovery from inactivation can be isolated (B). A single-exponential function was fitted to the isolated time-course of recovery from inactivation. The resulting time constants of recovery from inactivation are shown in (C), and the relative effect of NS1643 in (D). (K_v11.1a; K_v11.1b; K_v11.1a/K_v11.1b: n = 10; 13; 7).

Effects of RPR260243 on K_v11.1 channels

Having investigated a compound with a relative broad mechanism of action, we turned to the K_v11.1 activator RPR260243, a compound which has been reported to primarily affect deactivation kinetics (Kang et al., 2005). The effects of 10 µM RPR260243 on the K_v11.1 isoforms are illustrated in Figure 6. From the I/V relationships in the absence and presence of RPR260243, it can be observed that RPR260243 increased the steady-state current of K_v11.1a homomers and K_v11.1a/ K_v11.1b (4:1) heteromers, whereas it had no effect on Kv11.1b steady-state current (Figure 6A,B). This was also observed for higher concentrations of RPR260243 (Supporting Information Figure S3). For both homomeric and heteromeric channels, RPR260243 had no significant effect on the voltage dependence of activation (Figure 6C). Likewise, RPR260243 had no effect on the kinetics of inactivation or recovery from inactivation (data not shown). However, it did cause a slight but significant slowing of activation for K_v11.1a [τ_act: 189 ± 14 ms (control) vs. 209 ± 14 ms, P < 0.01], but not for K_v11.1b [τ_act: 143 ± 13 ms (control) vs. 148 ± 17 ms, P = ns] or K_v11.1a/K_v11.1b [τ_act: 166 ± 15 ms (control) vs. 166 ± 10 ms, P = ns].

Figure 6.

Effect of RPR260243 on current–voltage relationships of K_v11.1 channels. (A) Representative current traces recorded at a test potential of 0 mV from Xenopus laevis oocytes expressing K_v11.1a (left panel), K_v11.1b (middle panel) or K_v11.1a/K_v11.1b (4:1) (right panel) in the absence and presence of RPR260243 (10 µM). Normalized current–voltage relationships (B) and voltage dependence of activation (C) before and after application of RPR260243 (for further details, see Figure 1). (K_v11.1a; K_v11.1b; K_v11.1a/K_v11.1b: n = 7; 7; 9).

From the representative current traces in Figure 6A, it can be observed that RPR260243 has noticeable effects on deactivation kinetics. On both channel isoforms, RPR260243 caused a concentration-dependent slowing of deactivation (Supporting Information Figure S3). To further investigate the effect of RPR260243 on deactivation kinetics, we fitted the time-course of deactivation measured at −60 mV to a double-exponential function. The time constant of the slow component of deactivation (τ_slow) was significantly increased for both homomeric and heteromeric channels after RPR260243. The relative effect of RPR260243 on τ_slow was significantly greater on K_v11.1b compared to K_v11.1a and K_v11.1a/K_v11.1b (P < 0.001 and P < 0.001, respectively) (Figure 7B, right panel). The marked effect of RPR260243 on the slow time constant for Kv11.1 is also noticeable from the enlargement of the time-course of deactivation in Figure 7A (middle panel). Similarly, the time constant of the fast component of deactivation (τ_fast) was significantly increased for K_v11.1b and K_v11.1a/K_v11.1b after application of RPR260243. In addition, the relative contribution of the fast component of deactivation significantly decreased after application of RPR260243 for K_v11.1b (Figure 7D).

Discussion and conclusions

In this study, we have investigated two K_v11.1 activators NS1643 and RPR260243, which represent two compounds with marked differences in channel activation mechanisms. Previously, the effects of these compounds have only been characterized on the K_v11.1a isoform (Kang et al., 2005; Casis et al., 2006; Hansen et al., 2006a; Perry et al., 2007; Xu et al., 2008). However, at least two isoforms of K_v11.1 have been found to be co-expressed in native tissues (Ohya et al., 2002; Jones et al., 2004; Guasti et al., 2005). To determine whether NS1643 and RPR260243 have differential effects on K_v11.1 isoforms, we compared the effects of these compounds on K_v11.1a and K_v11.1b channels. Except for a shorter but unique N-terminus, K_v11.1a and K_v11.1b are identical (Lees-Miller et al., 1997; London et al., 1997). In general, homomeric K_v11.1b channels display faster kinetics than K_v11.1a channels (Larsen et al., 2008).

In agreement with earlier studies (Kang et al., 2005; Casis et al., 2006; Hansen et al., 2006a; Perry et al., 2007; Xu et al., 2008), we found that NS1643 enhanced K_v11.1a currents primarily by slowing the rates of deactivation and inactivation, while RPR260243 acted mainly by slowing the rate of deactivation. In general, both NS1643 and RPR260243 affected the same biophysical properties of K_v11.1b as they did of K_v11.1a. However, for both activators, the magnitudes of the effects were found to be isoform dependent. NS1643 was found to have higher efficacy on K_v11.1b compared to K_v11.1a steady-state currents. The greater efficacy of NS1643 on K_v11.1b channels was partly explained by approximately twofold greater increases in the time constants of K_v11.1b deactivation, inactivation and recovery from inactivation processes as compared to K_v11.1a. Similarly, the relative effect of RPR260243 on deactivation kinetics was more marked on the K_v11.1b isoform, showing a ∼10-fold greater decrease of the slow time constant for K_v11.1b compared to the K_v11.1b isoform.

Both NS1643 and RPR260243 have been suggested to bind in or near the channel pore (Perry et al., 2007; Xu et al., 2008). Importantly, by applying NS1643 from either the intra- or extracellular side of the membrane of X. laevis oocytes expressing K_v11.1a channels, Xu et al. (2008) showed that NS1643 interacts with the extracellular region of the pore. In contrast, RPR260243 interacts with the pore from the intracellular side as demonstrated by mutational analysis and computational docking studies (Perry et al., 2007). This difference in binding sites may underlie the selective effect of RPR260243 on deactivation kinetics. In addition, a third K_v11.1 activator PD307243 has also been found to interact with the channel pore (Xu et al., 2008). The region surrounding the channel pore of K_v11.1 therefore appears to be the preferred binding site for K_v11.1 activators. As this region is identical between all described K_v11.1 isoforms (Warmke and Ganetzky, 1994; Lees-Miller et al., 1997; London et al., 1997; Kupershmidt et al., 1998; Guasti et al., 2008; Huffaker et al., 2009), it may, from a structural point of view, seem difficult to achieve isoform selectivity. Nevertheless, our data suggest that at least some degree of selectivity can be achieved by differentially targeting specific kinetic processes inherently different between the isoforms.

The structural basis for the differential effects of NS1643 and RPR260243 on K_v11.1a and K_v11.1b is unclear. Perry et al. (2009) recently reported that the effect of PD118057, another K_v11.1 activator that also binds in the pore region, was greatly increased by mutation of residues C643 and M645 in the S6 helix of K_v11.1a. Results from a molecular docking model suggested that both mutations reduced steric hindrance for drug binding. By analogy, it can be hypothesized that interactions between the long N-terminal of K_v11.1a and the pore region to some degree alter the structure of the binding sites, and thereby restrict NS1643 and RPR260243 from accessing their optimal binding configuration. Consequently, in K_v11.1b channels, the long N-terminal is absent and the compounds have easier access to their binding sites. This mechanism would, however, be expected to also affect the EC₅₀ value unless association and dissociation rates are changed to a similar extent. We were not able to obtain EC₅₀ values for RPR260243 (see Methods), but for NS1643 the EC₅₀ was not different on K_v11.1a and K_v11.1b channels. An alternative hypothesis could be that while the N-terminal of K_v11.1a does not restrict the compounds from binding to the channel per se, it may restrict the movements of the channel complex by stabilizing channel states. Such an effect would also explain the inherent differences in kinetics between K_v11.1a and K_v11.1b channels (Larsen et al., 2008). Following this hypothesis, NS1643 and RPR260243 would be able to induce greater conformational changes in K_v11.1b channels than in K_v11.1a channels where the long N-terminal restricts the movements. This hypothesis needs to be addressed in future studies.

The effect of K_v11.1 activators will probably be dependent on the distribution of K_v11.1 isoforms in a given tissue and whether the isoforms form heteromeric channel complexes in that tissue. Endogenous co-expression of K_v11.1a and K_v11.1b has been described in heart, brain and smooth muscle (Ohya et al., 2002; Jones et al., 2004; Guasti et al., 2005). From mRNA levels, it has been found that the human heart left ventricle contains on average 10–20% K_v11.1b (Larsen et al., 2008). Consequently, differential expression of K_v11.1 isoforms should be taken into consideration when developing and investigating novel K_v11.1 activators. To address this issue, the effects of NS1643 and RPR260243 were investigated on heteromeric K_v11.1a/K_v11.1b channels. Work from our group has shown that the kinetic properties of heteromeric K_v11.1a/K_v11.1b macroscopic currents are dependent on the relative abundance of the isoforms (Larsen et al., 2008; Larsen and Olesen, 2010). Accordingly, we observed that heteromeric expression resulted in currents intermediate of those observed for homomeric K_v11.1a and K_v11.1b channels. Comparison of the deactivation kinetics of homomeric channels and the resulting kinetics of the co-expressed channels strongly suggested that heteromeric channel formation did occur. When co-expressed in a molar ratio of 4:1 (K_v11.1a/K_v11.1b), the effects of both NS1643 and RPR260243 were similar to the effects seen for homomeric K_v11.1a channels. If the hypothesis that interaction between the K_v11.1a-type N-terminal with the pore domain restricts the conformational changes that can be induced by binding of an activator is true, the presence of K_v11.1a N-terminals in a heteromeric channel is likely to determine the efficacy of the activators. The observed effects of NS1643 and RPR260243 on co-expressed channels are in agreement with such a hypothesis.

Recently, Huffaker et al. (2009) described a link between schizophrenia and up-regulation of a brain-specific fast-deactivating K_v11.1 isoform (KCNH2-3.1). The up-regulation of the fast-deactivating isoform leads to increased excitability in primary cortical neurones. In theory, a targeted gating modifier may convert a fast-deactivating K_v11.1 channel isoform to a more K_v11.1a-like phenotype, thereby reducing excitability to non-pathological levels. The differential effects of two K_v11.1 channel activators, especially RPR260243, on deactivation properties that we observed in this study suggest that this would indeed be a possibility. Although the effect of K_v11.1 activators has not been tested on the KCNH2-3.1 isoform, our data suggest that in principle a fast-deactivating isoform (K_v11.1b) can be converted to a more K_v11.1a-like phenotype. The use of differentially acting K_v11.1 activators may represent a novel treatment of diseases caused by changes in the abundance of K_v11.1 isoforms.

A limitation to our study is that we did not perform computational modelling to verify that the effects of NS1643 and RPR260243 on Kv11.1 currents can be fully represented by the observed changes in kinetic rates. Therefore, we cannot completely rule out the possibility that changes in surface density of the channels during drug application may also play a role.

In conclusion, we demonstrated that two previously described activators on K_v11.1a channels have qualitatively similar effects on K_v11.1b channels. However, the efficacy of the activators is isoform dependent. Our results suggest that development of isoform-selective activators might be possible, and highlight the importance of the N-terminus in gating of K_v11.1 channels. Furthermore, the demonstration of isoform-dependent effects of K_v11.1 activators stresses the importance of establishing the relative abundance of isoforms in the targeted tissue and investigating the effect of activators on the relevant combination of K_v11.1 isoforms.

Glossary

Abbreviations

I_Kr

the rapid component of the delayed-rectifier potassium channel

I/V

current/voltage

K_V

voltage-gated potassium channel

Conflict of interest

M.G. is an employee at NeuroSearch A/S.

Supporting Information

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

Figure S1 Comparison of deactivation kinetics of experimental and theoretical currents. The theoretical current decay was calculated under the assumption that only homomeric K_v11.1a and K_v11.1b channels were formed in a 4:1 ratio. The experimental current decay was calculated from the observed macroscopic deactivation properties of oocytes expression K_v11.1a/K_v11.1b in a 4:1 ratio (Figures 3 and 7).

Figure S2 Concentration–response relationship of NS1643 for K_v11.1a (left) and K_v11.1b (right). (A) Current traces in the absence or presence of 10 and 100 μM NS1643. (B) Concentration–response curve for K_v11.1a and K_v11.1b. EC₅₀ values were calculated from individual sigmoidal fits to the normalized steady-state current measured at 0 mV as a function of increasing concentrations of NS1643. (K_v11.1a; K_v11.1b: n = 5; 5).

Figure S3 Concentration–response relationship of RPR260243 for K_v11.1a (left) and K_v11.1b (right). (A) Representative current traces recorded at a test potential of 0 mV in the absence and presence of 10 and 100 mM RPR260243.

Supporting Information: Teaching Materials; Figs 1–7 as PowerPoint slide.

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.