Effect of the I_to activator NS5806 on cloned K_v4 channels depends on the accessory protein KChIP2

Abstract

BACKGROUND AND PURPOSE

The compound NS5806 increases the transient outward current (I_to) in canine ventricular cardiomyocytes and slows current decay. In human and canine ventricle, I_to is thought to be mediated by K_V4.3 and various ancillary proteins, yet, the exact subunit composition of I_to channels is still debated. Here we characterize the effect of NS5806 on heterologously expressed putative I_to channel subunits and other potassium channels.

EXPERIMENTAL APPROACH

Cloned K_V4 channels were co-expressed with KChIP2, DPP6, DPP10, KCNE2, KCNE3 and K_V1.4 in Xenopus laevis oocytes or CHO-K1 cells.

KEY RESULTS

NS5806 increased K_V4.3/KChIP2 peak current amplitudes with an EC₅₀ of 5.3 ± 1.5µM and significantly slowed current decay. KCNE2, KCNE3, DPP6 and DPP10 modulated K_V4.3 currents and the response to NS5806, but current decay was slowed only in complexes containing KChIP2. The effect of NS5806 on K_V4.2 was similar to that on K_V4.3, and current decay was only slowed in presence of KChIP2. However, for K_V4.1, the slowing of current decay by NS5806 was independent of KChIP2. K_V1.4 was strongly inhibited by 10 µM NS5806 and K_V1.5 was inhibited to a smaller extent. Effects of NS5806 on kinetics of currents generated by K_V4.3/KChIP2/DPP6 with K_V1.4 in oocytes could reproduce those on cardiac I_to in canine ventricular myocytes. K_V7.1, K_V11.1 and K_ir2 currents were unaffected by NS5806.

CONCLUSION AND IMPLICATIONS

NS5806 modulated K_V4 channel gating depending on the presence of KChIP2, suggesting that NS5806 can potentially be used to address the molecular composition as well as the physiological role of cardiac I_to.

Introduction

Epi- and midmyocardial cells from larger mammals have a prominent phase 1 repolarization of the action potential due to the presence of a Ca²⁺-independent transient outward potassium current (I_to). In human and canine hearts, I_to is principally mediated by the α-subunit K_V4.3, but other channels such as K_V1.4 may be involved (Dixon et al., 1996; Akar et al., 2004; channel nomenclature follows Alexander et al., 2009). However, heterologously expressed K_V4.3 channels do not reproduce the kinetics of native cardiac I_to and K_V4.3 has been demonstrated to assemble with various ancillary β-subunits in cardiac tissue; the most prominent of which is K Channel Interaction Protein 2 (KChIP2). KChIP2 belongs to a large family of cytosolic Ca²⁺– sensing proteins that contain up to four putative Ca²⁺ binding EF-hands. KChIP2 increases K_V4.3 current density by facilitating trafficking, slowing inactivation and by accelerating recovery kinetics (An et al., 2000). Interestingly, the expression of KChIP2 mRNA is more abundant in epi- and midmyocardium than in endocardium and this differential expression of KChIP2 has been suggested to underlie the transmural I_to gradient (Rosati et al., 2001; 2003; Calloe et al., 2009b). Besides KChIP2, several other proteins expressed in ventricular tissue have been shown to interact with K_V4.3. Dipeptidyl-peptidase (DDP) 6 can facilitate K_V4.3 trafficking, accelerate K_V4.3 inactivation (Nadal et al., 2003) and increase single channel conductance of K_V4.2 (Kaulin et al., 2009). Co-expression of K_V4.3 with KChIP2 and DPP6 in heterologous systems results in currents with kinetics similar to those of I_to in human ventricular myocytes (Radicke et al., 2005).

Members of the KCNE β-subunit family have been shown to modulate K_V4.3 currents (Zhang et al., 2001; Lundby and Olesen, 2006; Radicke et al., 2006). KCNE2 may be a promising candidate for the human I_to channel complex as co-expression of K_V4 channels with KCNE2 induced an overshoot of peak current during recovery from inactivation (Zhang et al., 2001; Radicke et al., 2006), comparable to that described for I_to in human epicardial myocytes (Wettwer et al., 1994). However, other studies have shown that KCNE2 is expressed in low quantities in ventricular tissue compared with Purkinje tissue (Pourrier et al., 2003) suggesting KCNE2 is important mainly in the conduction system of the heart (Sanguinetti and Tristani-Firouzi, 2006). KCNE3 has been shown to inhibit K_V4.3 currents (Lundby and Olesen, 2006) and mutations in KCNE3 resulting in less inhibition of K_V4.3 current have been linked to the Brugada syndrome (Delpón et al., 2008) and atrial fibrillation (Lundby et al., 2008). Furthermore, heterologously expressed K_V4.3 channels are modulated by K_Vβ, KChAP and Na_Vβ accessory subunits (Deschenes and Tomaselli, 2002) but their physiological roles have yet to be determined.

K_V4 currents are selectively inhibited by several spider toxins that modify gating kinetics, including the Heteropoda venatoria toxins (Sanguinetti et al., 1997) HpTX2 (Zarayskiy et al., 2005) and HpTX3 (Brahmajothi et al., 1999), the Phrixotrichus auratus toxins PaTx1 and PaTx2 (Diochot et al., 1999) and the Theraphosa leblondi toxins TLx1-3 (Ebbinghaus et al., 2004). In common with other A-type K_V channels, K_V4.3 is blocked by 4-aminopyridine (4-AP) in millimolar range concentrations (Wang et al., 1995). I_to is blocked by several sodium channel blockers, including flecainide (Radicke et al., 2008) and quinidine (Wang et al., 1995), and several calcium channel blockers, including nifedipine (Hatano et al., 2003; Bett et al., 2006). We have recently added an I_to activator, NS5806, to this list of compounds affecting I_to and described the effect of NS5806 on canine ventricular wedge preparations (Calloe et al., 2009a) as well as on native I_to in isolated cells from canine left ventricular epi-, mid- and endocardium (Calloe et al., 2009b). We found that 10 µM NS5806 increased the magnitude of I_to, slowed current decay, induced a negative shift in steady-state inactivation and accelerated recovery from inactivation for native I_to.

In the present study, we characterized the effects of NS5806 on heterologously expressed putative I_to channel subunits. NS5806 enhanced peak currents for all K_V4 channels and affected channel gating. In the presence of KChIP2, NS5806 slowed the decay of K_V4.2 and K_V4.3 currents significantly, whereas it had little effect in the absence of KChIP2. Co-expression of K_V4.3 with and without KChIP2 with DPP6, DPP10, KCNE2 or KCNE3 β-subunits corroborated that NS5806 only slowed current decay of channel complexes containing KChIP2. Besides the effects on K_V4 channels, NS5806 inhibited K_V1.4 and K_V1.5 mediated currents independently of the presence of KChIP2. Effects of NS5806 on currents generated by K_V4.3/KChIP2/DPP6 with K_V1.4 in oocytes could reproduce those on cardiac I_to in canine ventricular myocytes

Methods

NS5806

NS5806 (1-[2,4-dibromo-6-(1H-tetrazol-5-yl)-phenyl]-3-(3,5-bis-trifluoromethyl-phenyl)-urea) was synthesized at NeuroSearch A/S (Ballerup, Denmark) by reaction of 6-cyano-2,4-dibromoaniline with sodium azide to form the respective 2,4-dibromo-6-tetrazolylaniline, which was condensed with 3,5-bis-trifluoromethyl-phenylisocyanate to provide the final product, NS5806. NS5806 was dissolved in DMSO to give a concentrated stock solution of 30 mM. The final DMSO concentration never exceeded 0.1%, and at this concentration DMSO did not influence the electrical properties of the cells.

Molecular biology

α-subunits

cDNAs coding for human (h) Kv4.1 (NM_004979) and hKv4.2 (NM_012281) were a kind gift from D. Isbrandt (U Hamburg, Germany). cDNA coding for hKv4.3 (short isoform, NM_172198) and hKv1.5 (NM_002234) were kindly provided by O. Pongs (U Hamburg, Germany). hKv4.3 cDNA was subcloned into the expression vector pXOOM, Kv1.5 was cloned into pXOON. hKv4.1 and hKv4.2 were PCR-amplified and cloned into the expression vector pGEM-HEJuel. cDNA encoding hKv1.4 (NM_002233) was amplified from EST HU_p940D11203D (ImaGenes, Berlin, Germany) and cloned into pGEM-HEJuel.

Ancillary subunits

hKChIP2.1 (NM_173192) was amplified from IMAGE clone 2430271 and subcloned into pXOOM. cDNA coding for hDPP6 iso 2 (NM_001936) and hDPP10 iso 1 (NM_020868) were kindly provided by E. Wettwer (TU Dresden, Germany), PCR-amplified and cloned into pGEM-HEJuel. KCNE2 (NM_172201) in pSGEM and hKCNE3 (NM_005472) in pXOOM have been described previously (Lundby and Olesen, 2006). All constructs were verified by sequencing.

Heterologous expression in Xenopus laevis oocytes

All animal care and experimental procedures were in accordance with Danish National Committee for Animal Studies guidelines. Female Xenopus laevis frogs were anesthetized (2 g·L⁻¹ Tricaine; Sigma) and ovarian lobes cut off through a small abdominal incision. The oocytes were manually dissected into smaller groups and defolliculated using collagenase (Type 1, Sigma-Aldrich) for 1 h. Oocytes were kept in Kulori solution (mM); NaCl 90, KCl 4, MgCl₂ 1, CaCl₂ 1, HEPES 5, pH 7.4 with NaOH at 19°C for 24 h before injection of cRNA. cRNA was prepared from linearized plasmid DNA using the T7 mMessage mMachine kit (Ambion) according to the manufacturer's instructions. 50 nl containing 0.5 ng K_V4 cRNA and ancillary subunits added in a 1:1 molar ratio was injected using a Nanoject microinjector (Drummond Scientific, Broowell, PA, USA). K_V4.3/KChIP2 and K_V1.4 channels were co-expressed in a 1:1 current ratio corresponding to 0.15 ng K_V4.3 + 0.15 ng KChIP2 + 1.3 ng K_V1.4 pr oocyte. The oocytes were kept in Kulori solution at 19°C, which was changed daily and currents were recorded after 2 to 3 days.

Two-electrode voltage clamp

Recordings were at room temperature in Kulori solution using a two-electrode voltage-clamp amplifier (Dagan CA-1B; Chicago, IL, USA). Borosilicate glass recording electrodes (Module Ohm, Denmark) were fabricated using a DMZ-Universal Puller (Zeitz Instruments, Munich, Germany) and had a resistance of 0.5 to 1 MΩ when filled with 3 M KCl.

CHO-K1 cell culture and transfection

CHO-K1 cells were transiently transfected with hK_V4.3 (0.5 µg to a 25 cm² cell flask), hKChIP2.1, hDPP6 in a 1:3:3 molar ratio using Lipofectamine and Plus Reagent according to manufacturer's instruction (Gibco, Invitrogen). The cells were cultured in Dulbecco's modified Eagle's medium (DMEM, Substrate Department, University of Copenhagen, Denmark) supplemented with 10% fetal calf serum (Gibco, Invitrogen) and 40 mg·L⁻¹ L-proline at 37°C in 5% CO₂.

Electrophysiological recordings of transiently transfected CHO-K1

Whole-cell currents were recorded using an EPC-10 amplifier (HEKA Electronics, Lambrecht, Germany). Data were sampled with Pulse software (HEKA Electronics) and analysed with IGOR software (Wavemetrics, Lake Oswego, OR, USA). The series resistance (R_s) was compensated 80 % and did not exceed 3 MΩ. Electrodes were pulled from borosilicate glass capillaries (Module Ohm, Herlev, Denmark) and had tip resistances between 1.5 and 2.5 MΩ. For the K_V4.3/KChIP2/DPP6 experiments shown in Figure 1, recording conditions were identical to those previously used for measuring native I_to (Calloe et al., 2009a). Cells were superfused with a HEPES buffer of the following composition (mM): NaCl 126, KCl 5.4, MgCl₂ 1.0, CaCl₂ 2.0, HEPES 10 and glucose 11, pH adjusted to 7.4 with NaOH. The patch pipette solution had the following composition (mM): K-aspartate 90, KCl 30, glucose 5.5, MgCl₂ 1.0, EGTA 5, MgATP 5, HEPES 5, NaCl 10, pH = 7.2 with KOH. In myocytes, I_to was recorded in the presence of 300 µM Cd²⁺ which was used to block I_CaL. We therefore included 300 µM Cd²⁺ in the present study to allow a more direct comparison.

Figure 1.

Effect of NS5806 on K_V4.3/KChIP2/DPP6. K_V4.3/KChIP2/DPP6 were transiently expressed in CHO-K1 cells and currents were measured in the absence and in the presence of 10 µM NS5806. (A) Representative recordings of K_V4.3/KChIP2/DPP6 currents elicited by the protocol shown in panel B. (B) Relation between peak current density and voltage, n= 5. (C) Mono-exponential functions were fitted to the current decays, and the time constants (τ) are shown as a function of voltage, n= 7). (D) Steady-state inactivation of K_V4.3/KChIP2/DPP6 currents. Normalized tail current amplitudes recorded at +10 mV are plotted as a function of the prepulse potential and Boltzmann equations are fitted to the data, n= 5. (E) Time-dependent release from inactivation. K_V4.3/KChIP2/DPP6 currents were activated by the depicted two-pulse protocol. Current amplitudes at the second test pulse were normalized to that of the first test pulse and plotted as a function of the inter-pulse interval, and a single exponential equation was fitted to the data, n= 6). *P < 0.05, **P < 0.01, ***P < 0.001; significantly different in the absence and presence of NS5806. CHO-K1, Chinese hamster ovary cells; DPP, dipeptidyl-peptidase; KChIP2, K channel interacting protein 2.

For CHO-K1 cells expressing K_V4.3/KChIP2/DPP6, the membrane capacitance was 14.0 ± 2.2 pF and the mean peak current 7.5 ± 2.0 nA. For the concentration-response experiments on K_V4.3/KChIP2 (Figure 3), a standard extracellular NaCl solution consisted of (mM): NaCl 140, KCl 4, CaCl₂ 2, MgCl₂ 1, HEPES 10, pH = 7.4 adjusted with NaOH was used and the intracellular solution contained (mM): KCl 110, KOH/EDTA 31/10, CaCl₂ 5.17, MgCl₂ 1.42, HEPES 10, MgATP 4, pH = 7.2 with KOH. The membrane capacitance was 14.9 ± 1.4 pF and the mean peak current in control 7.1 ± 1.4 nA at +20 mV. All experiments were performed at 37 ± 1°C.

Figure 3.

Concentration dependent effect of NS5806 on K_V4.3, KChIP2 and DPP6 in Xenopus laevis oocytes. K_V4.3, KChIP2 and/or DPP6 were expressed in Xenopus laevis oocytes. Currents were repeatedly activated from −80 mV by steps to +40 mV. (A) Representative recordings of K_V4.3, K_V4.3/KChIP2 and K_V4.3/KChIP2/DPP6 currents in presence of 0 (black), 1 (blue), 3 (red) 10 (green) or 30 (orange) µM NS5806. (B) Single exponential functions were fitted to the first 150 ms of K_V4.3, K_V4.3/KChIP2 and K_V4.3/KChIP2/DDP6 current decays and the time constants (τ) are plotted as a function of NS5806 concentration (n= 6–8). *P < 0.05, significantly different in the absence and presence of NS5806. (C) K_V4.3/KChIP2 time-course experiment. The black trace shows the effect of increasing concentrations of NS5806 on K_V4.3/KChIP2 peak current decay (τ; lefthand Y axis) and the grey trace shows the effect on the peak current amplitude (I, righthand Y axis), representative of n= 6. DPP, dipeptidyl-peptidase; KChIP2, K channel interacting protein 2.

Statistics

Mean ± SEM are shown. Statistical significance was evaluated by Student's t-test and one way anova with Dunnett's post test as appropriate using GraphPad Prism (GraphPad Software, San Diego, CA, USA).

Results

Effect of NS5806 on K_V4.3/KChIP2/DPP6 channels expressed in CHO-K1 cells

K_V4.3 together with KChIP2 and DPP6 produces a current resembling native I_to (Radicke et al., 2005). As an initial basis for comparison, we co-expressed K_V4.3/KChIP2/DPP6 in CHO-K1 cells and recorded currents in solutions identical to those previously used for native I_to measurements (Calloe et al., 2009a). As we found with native I_to, 10 µM NS5806 induced a 65% increase of K_V4.3/KChIP2/DPP6 peak current amplitudes (Figure 1A and B). The time constant of decay (τ) of I_to was significantly slowed as reflected in 80% increase in τ-values over a range of voltages (Figure 1C). Steady-state gating parameters were evaluated using a pre-pulse-test pulse voltage clamp protocol. The peak current following a 2 s prepulse was normalized to the maximum current and plotted as a function of prepulse voltage to obtain the availability of channels. A Boltzmann curve was fitted to the data and the mid-inactivation voltage (V_1/2) was determined. NS5806 induced a significant left-shift in mid-inactivation from −34.5 ± 0.6 mV to −40 ± 0.4 mV (Figure 1D), indicating a tendency for a larger fraction of channels being in the inactivated state in presence of NS5806, as observed for native I_to (Calloe et al., 2009b).

Time-dependent recovery from inactivation of K_V4.3/KChIP2/DPP6 was evaluated by a two-pulse protocol with increasing interpulse intervals. The fraction of recovered current was plotted as a function of interpulse interval and an exponential equation fitted to the data. NS5806 significantly slowed recovery of K_V4.3/KChIP2/DPP6 current, with time constants of 9.3 ± 0.6 ms prior to NS5806 application versus 16.4 ± 1.2 ms after application of 10 µM NS5806. These results were opposite to the effect of NS5806 on I_to recorded in isolated canine ventricular cardiomyocytes where recovery of I_to was faster in the presence of NS5806.

Concentration-dependence of the effect of NS5806 on K_V4.3/KChIP2/DDP6 channels

To characterize NS5806, we initially tested the concentration-dependent effect of NS5806 on peak current amplitude and time course of current inactivation of Kv4.3/KChIP2 expressed in CHO-K1 cells. NS5806 increased peak-current amplitudes concentration-dependently with an EC₅₀ value of 5.3 ± 1.5 µM and the time course of inactivation (τ) was slowed with an EC₅₀ value of 25.4 ± 1.1 µM (Figure 2).

Figure 2.

Concentration dependent effect of NS5806 on K_V4.3/KChIP2 in CHO-K1 cells. K_V4.3 and KChIP2 were transiently expressed in CHO-K1 cells and measured in presence of 0 to 100 µM NS5806 as indicated in the figure. (A) Representative current recordings at different NS5806 concentrations. (B) Peak current amplitudes normalized to current amplitudes in absence of drug and plotted as function of NS5806 concentration. A sigmoidal dose-response curve was fitted to the data in the range from 0–30 µM and an EC₅₀ of 5.3 ± 1.5 µM with and a Hill slope of 1.1 ± 0.05 and minimum and maximum values of 1.04 ± 0.02 and 1.6 ± 0.06 respectively. (C) Effect of different NS5806 concentrations on Kv4.3/KChIP2 decay. The time-constants, τ, were plotted as function of NS5806 concentration and revealed an EC₅₀ of 25.4 ± 1.1 µM. The Hill slope was 1.08 ± 0.06 with minimum and maximum values of 7.0 ± 0.1 ms and 20.0 ± 0.5 ms respectively (n= 7–12). *P < 0.05, **P < 0.01, ***P < 0.001; significantly different in the absence and presence of NS5806. CHO-K1, Chinese hamster ovary cells; KChIP2, K channel interacting protein 2.

To further test the effect of NS5806 on different ion channels and multiple combinations of ion channel subunits we used Xenopus laevis oocytes as cRNA encoding ion channel subunits can be injected directly into the oocytes, ensuring better control over subunit ratios than the transfection procedure of mammalian cell lines. In this series of experiments, K_V4.3 was expressed in Xenopus laevis oocytes in absence or presence of KChIP2 and DPP6 and currents measured in presence of 0 to 30 µM NS5806 (Figure 3A). Interestingly, the K_V4.3 peak current amplitude was reduced by NS5806, whereas NS5806 caused a minor increase in K_V4.3/KChIP2 and K_V4.3/KChIP2/DPP6 peak current amplitude as well as a pronounced slowing of current decay. In contrast to K_V4.3/KChIP2/DPP6 expressed in CHO-K1 cells and native I_to (Calloe et al., 2009a,b;) the effect of NS5806 on K_V4.3/KChIP2 peak current amplitude in Xenopus laevis oocytes was minor. The slowing of current decay was concentration-dependent and qualitatively similar with and without DPP6. For K_V4.3 expressed alone, there was no effect on the decay even at higher concentrations of NS5806 (Figure 3B). A sigmoidal concentration-response curve was fitted to K_V4.3/KChIP2 current decay and an EC₅₀ 18.8 ± 2.0 µM was calculated. For all subunit combinations, the effect of NS5806 was rapid and fully reversible, as illustrated by the representative K_V4.3/KChIP2 time-course experiment (Figure 3C).

Effect of NS5806 on K_V4 family channels in presence or absence of KChIP2

To investigate if NS5806 affected other members of the K_V4 family and whether this effect also depended on KChIP2, all K_V4 family α-subunits were expressed with and without KChIP2 in oocytes. Currents were activated by depolarizing voltage steps from a holding potential of −100 mV. For K_V4.2 and K_V4.3, NS5806 had no effect on current amplitude, in contrast to the results presented for K_V4.3 in Figure 3, where a holding potential of −80 mV was used. For all K_V4 channels, co-expression with KChIP2 resulted in higher current amplitudes and a significantly slower current decay (Figure 4 and Table 1). NS5806 had no effect on current decay of K_V4.2 and K_V4.3, whereas it markedly slowed the decay in presence of KChIP2 (Figure 4B and C). For K_V4.1, NS5806 slowed current decay both in presence and absence of KChIP2 (Figure 4A).

Figure 4.

Effect of NS5806 on K_V4 channels in presence or absence of KChIP2. K_V4 channels were expressed with and without KChIP2 in Xenopus laevis oocytes and currents were elicited from a holding potential of −100 mV by a step to +40 mV in absence and presence of 10 µM NS5806. The first 150 ms of current decay was fitted to a mono-exponential function, and the time constants (τ) are shown as a function of voltage. (A) Representative K_V4.1 (n= 5) and K_V4.1/KChIP2 currents (n= 6) and τ-values as a function of voltage before and after application of drug. (B) Representative K_V4.2 (n= 6) and K_V4.2/KChIP2 (n= 10) currents and τ-values as a function of voltage. (C) Representative K_V4.3 (n= 8) and K_V4.3/KChIP2 (n= 8) currents and τ-values as a function of voltage. *P < 0.05, **P < 0.01, ***P < 0.001; significantly different in the absence and presence of NS5806. KChIP2, K channel interacting protein 2.

Table 1.

Effect of 10 µM NS5806 on K_v4 channels in absence or presence of KChIP2

	Kv4.1	Kv4.1/KChIP2	Kv4.2	Kv4.2/KChIP2	Kv4.3	Kv4.3/KChIP2
Peak current (µA) control	12.2 ± 1.1	23.6 ± 1.3	1.9 ± 0.3	6.6 ± 1.4	13.8 ± 1.2	23.2 ± 1.2
10 µM NS5806	12.6 ± 1.0 (5)	27.0 ± 1.3 (6)***	2.0 ± 0.3 (6)	7.5 ± 1.3 (10)*	15.3 ± 1.4 (11)	26.9 ± 1.3 (9)***
Decay τ (ms) control	45.4 ± 1.2	63.6 ± 5.5	24.2 ± 1.1 (6)	56.8 ± 2.7	44.2 ± 1.9	67.2 ± 3.4
10 µM NS5806	51.7 ± 1.8 (5)*	101 ± 6.2 (8)***	22.6 ± 0.7 (6)*	68.1 ± 3.8 (10)*	43.5 ± 1.5 (11)	83.2 ± 4.7 (9)***
Area (µA·ms–1) control	814 ± 70	2260 ± 181	66 ± 9.8	378 ± 74	762 ± 63.9	1450 ± 103
10 µM NS5806	1250 ± 109 (5)**	3150 ± 166 (8)***	67 ± 10 (6)	533 ± 101 (10)**	825 ± 84 (11)	1890 ± 142 (8)**
V1/2 (mV) control	−60.8 ± 0.85	−44.7 ± 0.7	−65.2 ± 1.1	−57.3 ± 0.9	−55.4 ± 0.5	−48.6 ± 0.4
10 µM NS5806	−75 ± 1.8 (5)***	−59.5 ± 0.5 (8)***	−75.7 ± 1.9 (6)***	−67.1 ± 1.3 (10)***	−66.0 ± 1.0 (10)***	−54.0 ± 1.1 (8)**
Recovery τ (ms) control	80.8 ± 3.5	8.0 ± 2.8	110 ± 29	12.6 ± 3.0	95.8 ± 2	15.9 ± 0.5
10 µM NS5806	205 ± 31 (5)***	37.1 ± 10 (8)***	202 ± 54 (6)**	26.4 ± 1.8 (10)***	221 ± 8 (8)***	23.7 ± 0.48 (8)***

K_V4 channels expressed with and without KChIP2 (1:1 molar ratio) in Xenopus laevis oocytes. Currents were activated from a holding potential of −100 mV by a +40 mV step before and during application of 10 µM NS5806 and peak current amplitudes, time constants (τ) of current decay and total charge movement for the first 150 ms after complete activation was measured. Steady-state mid-inactivation (V_1/2) was evaluated from a holding potential of −100 mV by a series of 2 s prepulses from −100 to +40 mV followed by a +40 mV step. Boltzmann equations were fitted to normalized tail current amplitudes plotted as a function of the prepulse potential. Time constants (τ) for recovery from inactivation was determined by a two-pulse protocol; from a holding potential of −80 mV currents were activated by two test potentials to +40 mV with increasing interpulse time. The peak current at the second pulse was normalized to the first pulse and single exponential functions were fitted to the data points. The different K_V4 α-subunits were recorded on different days; however, the experiments in absence and presence of KChIP2 were performed in parallel and thus current amplitude and biophysical characteristics can be directly compared.

^*P < 0.05,

^**P < 0.01

^***P < 0.001; significantly different in the absence and presence of NS5806.

KChIP2, K channel interacting protein 2.

Slowing of K_V4/KChIP2 current decay resulted in an increase in total charge movement in presence of the drug, as assessed by the area under the current traces (Table 1 and Supplementary Figures S1 and 2). To evaluate steady-state mid-inactivation (V_1/2) of K_V4 and K_V4/KChIP2 channels, peak tail-currents following a 2 s prepulse were normalized to the maximal current and a Boltzmann equation was fitted to the data. For all channel complexes, 10 µM NS5806 caused a prominent left-shift in V_1/2 (Tables 1 and 2 and Supplementary Figures S1 and S2), and as illustrated for K_V4.3 and K_V4.3/KChIP2 in Figure 5A and B. This negative shift in V_1/2 explains the disparate effects of NS5806 on K_V4.3 peak currents (Figures 3B and 4C, Tables 1 and 2) when using −80 mV or −100 mV as holding potential. As −80 mV is close to physiological resting membrane potentials, we used −80 mV as holding in the following studies. Recovery from inactivation for K_V4 and K_V4/KChIP2 channels expressed in Xenopus laevis oocytes was addressed by a standard two-pulse protocol and was found to be slowed by 10 µM NS5806 for all K_V4 and K_V4/KChIP2 channels (Table 1, Figure 5C and D and Supplementary Figures S1 and 2).

Table 2.

Effect of 10 µM NS5806 on K_v4.3 co-expressed with different subunits

Kv4.3+	−	DPP6	KCNE2	KCNE3	DPP6/KCNE3	KChIP2/DPP10
Peak current (µA) control	14.4 ± 2.0	20.4 ± 1.4	12.1 ± 1.9	1.8 ± 0.18	37.8 ± 6.8	14 ± 1.4
10 µM NS5806	6.6 ± 1.3 (8)**	11.0 ± 1.7 (9)***	7.5 ± 1.5 (6)**	1.3 ± 0.28 (8)*	23.6 ± 5.8 (8)**	10.4 ± 1.1 (5)**
Decay τ (ms) control	42.4 ± 2.1	37.7 ± 2.7	66.4 ± 13.8	55.1 ± 4.7	31.0 ± 1.0	17.6 ± 0.8
10 µM NS5806	45.6 ± 3.6 (8)	28.8 ± 0.73 (9)*	52.4 ± 7–3 (6)	48.1 ± 1.9 (4)	28.2 ± 7.7 (8)*	11.9 ± 0.5 (5)***
Area (µA·ms–1) control	783 ± 102	939 ± 120	959 ± 251	176 ± 45	1415 ± 311	409 ± 23
10 µM NS5806	367 ± 59 (8)***	501 ± 88 (9)***	408 ± 88 (6)**	169 ± 49 (4)	877 ± 20 (9)**	188 ± 194 (5)***
V1/2 (mV) control	−56.8 ± 1.3	−75.2 ± 0.14	−59.6 ± 0.9	−60.1 ± 0.57	−78.6 + 0.3	−72.8 ± 0.5
10 µM NS5806	−80.7 ± 2.7 (8)***	−82.2 ± 0.14 (8)***	−79.9 ± 1.1 (6)***	−78.4 ± 1.1 (8)***	−81.7 ± 0.2 (9)***	−80.6 ± 0.4 (5)***
Recovery τ (ms) control	480 ± 71	159 ± 3.8	235 ± 13.4	543 ± 24	176 ± 1.9	93.9 ± 2.1
10 µM NS5806	1459 ± 10 (8)***	239 ± 2.2 (5)***	492 ± 94.6 (6)***	1237 ± 123 (8)***	252 ± 2.4 (10)***	86.6 ± 6.3 (5)

Kv4.3+	KChIP2	KChIP2/DPP6	KChIP2/KCNE2	KChIP2/KCNE3	KChIP2/DPP6/KCNE3	KChIP2/DPP6/KCNE2
Peak current (µA) control	33.9 ± 2.9	33.1 ± 8.5	13.9 ± 1.47	1.8 ± 0.36 (5)	24.7 ± 3.9	33.9 ± 1.3
10 µM NS5806	35.9 ± 3.1 (7)**	33.0 ± 9.5 (9)	10.7 ± 1.09 (9)***	1.5 ± 0.36 (5)*	25.3 ± 4.0 (9)	25.9 ± 1.4 (5)**
Decay τ (ms) control	65.5 ± 6.3	33.6 ± 1.6	72.0 ± 6.5	187 ± 38	37.0 ± 2.1	44.3 ± 3.0
10 µM NS5806	90.0 ± 14.9 (7)***	45.3 ± 1.0 (10)***	78.0 ± 6.2 (9)*	199 ± 100(5)	42.1 ± 1.3 (9)*	44.9 ± 1.9 (5)
Area (µA·ms–1) control	2011 ± 219	1111 ± 304	1140 ± 98	195 ± 86	1190 ± 224	1804 ± 310
10 µM NS5806	2702 ± 316 (7)***	1752 ± 498 (10)*	822 ± 101 (9)*	153 ± 84 (5)	1622 ± 325 (9)**	1616 ± 98 (5)*
V1/2 (mV) control	−50.2 ± 0.7	−63 ± 0.4	−52.44 ± 1.3	−60.4 ± 1.3 (5)	−71.4 ± 0.6	−70.2 ± 0.4
10 µM NS5806	−58.9 ± 0.7 (7)***	−69 ± 0.3 (4)***	−71.9 ± 2.2 (9)***	−79.9 ± 1.5 (5)***	−77.8 ± 0.43 (9)***	−78.2 ± 0.2 (5)***
Recovery τ (ms) control	76.6 ± 2	42.6 ± 1.8	189.1 ± 5.5	522 ± 56 (5)	62.6 ± 1.6	68.6 ± 1.1
10 µM NS5806	274 ± 8.0 (7)***	66.8 ± 3.1 (8)***	547 ± 151 (9)***	1304 ± 134 (5)***	99.1 ± 1.6 (5)***	106 ± 4.5 (5)***

K_V4.3 channels expressed with different accessory subunits in Xenopus laevis oocytes. Currents were activated from a holding potential of −80 mV by a +40 mV step before and during application of 10 µM NS5806 and peak current amplitudes, time constants (τ) of current decay and total charge movement for the first 150 ms after complete activation was measured. Steady-state mid-inactivation (V_1/2) was evaluated from a holding potential of −100 mV by a series of 2 s prepulses from −100 to +40 mV followed by a +40 mV step. Boltzmann equations were fitted to normalized tail current amplitudes plotted as a function of the prepulse potential. Time constants (τ) for recovery from inactivation was determined by a two-pulse protocol; from a holding potential of −80 mV currents were activated by two test potentials to +40 mV with increasing interpulse time. The peak current at the second pulse was normalized to the first pulse and single exponential functions were fit to the data points. The different K_V4.3 β-subunit combinations were recorded on different days; however, the experiments in absence and presence of KChIP2 were in all cases performed in parallel and thus current amplitude and biophysical characteristics can be directly compared.

^*P < 0.05,

^**P < 0.01,

^***P < 0.001; significantly different in the absence and presence of NS5806.

DPP, dipeptidyl-peptidase.

Figure 5.

Effect of NS5806 on steady-state inactivation and time dependent recovery of K_V4.3 and K_V4.3/KChIP2 channels. K_V4.3 and K_V4.3/KChIP2 were expressed in Xenopus laevis oocytes. Steady-state inactivation was evaluated before (squares) and after application of 10 µM NS5806 (triangles). Normalized tail current amplitudes plotted as a function of the prepulse potential for currents elicited by the depicted voltage-clamp protocol and Boltzmann equations were fitted to the data. (A) Steady-state inactivation of K_V4.3, n= 10. (B) Steady-state inactivation of K_V4.3/KChIP2, n= 9. Time-dependent recovery from inactivation at −100 mV was addressed by a two-pulse protocol with an increasing inter-pulse interval as shown. The peak current at the second pulse was normalized to the current at the first pulse and a single exponential equation was fitted to the data. (C) Time-dependent recovery of K_V4.3, n= 8. (D) Time-dependent recovery of K_V4.3/KChIP2, n= 8. KChIP2, K channel interacting protein 2.

The negative shift in mid-inactivation V_1/2 suggests that NS5806 increases closed-state inactivation. Closed-state inactivation of K_V4.3 and for K_V4.3/KChIP2 expressed in Xenopus laevis oocytes was addressed by a modified double-pulse protocol (Bahring et al., 2001). After the first test-pulse, inactivation was completely released by a hyperpolarizing step to −100 mV followed by a step to a voltage below the activation threshold (−50 mV) allowing for onset of closed-state inactivation for increasing time intervals before application of a second test pulse. The current amplitude at the second pulse was plotted as a function of interpulse interval and a mono-exponential decay function was fitted to the data (Figure 6). Both in the presence and absence of KChIP2, NS5806 significantly accelerated the onset of closed-state inactivation. For K_V4.3, the time constant (τ) of closed-state inactivation was 701 ± 203 ms before and 164 ± 29 ms after application of 10 µM NS5806 and for K_V4.3/KChIP2, τ= 838 ± 338 ms before, and 310 ± 80 ms after application of NS5806.

Figure 6.

Effect of NS5806 on closed-state inactivation of K_V4.3 and K_V4.3/KChIP2 channels. K_V4.3 and K_V4.3/KChIP2 were expressed in Xenopus laevis oocytes. Closed-state inactivation was evaluated before and after application of 10 µM NS5806. Currents were activated from a holding of −100 mV by a brief +40 mV step. Inactivation was completely released by a −100 mV step and followed by a step at a voltage below the activation threshold (−50 mV) at increasing time-intervals before currents were activated by a second +40 mV step. Current amplitude at the second pulse is plotted as a function of pulse interval time for K_V4.3, n= 9 (A) and for K_V4.3/KChIP2, n= 5 (B). Representative K_V4.3/KChIP2 currents are shown in panel B. KChIP2, K channel interacting protein 2.

Effect of NS5806 on K_V4.3 expressed with various ancillary subunits

For K_V4.3/KChIP2 currents in CHO-K1 cells and in Xenopus laevis oocytes, the time dependent recovery from inactivation was slowed by NS5806 (Figure 1E and Figure 5C and D); however, for native I_to the recovery was accelerated by 10 µM NS5806 (Calloe et al., 2009a). As ancillary subunits can dramatically change the effect of exogenous compounds (Bett and Rasmusson, 2008), we wondered if the difference between I_to and K_V4.3/KChIP2 currents could be due to presence of additional ancillary subunits in the native I_to channel complex. In the next series of experiments we evaluated the effect of 10 µM NS5806 on K_V4.3 expressed with and without KChIP2 and the putative I_to ancillary subunits DPP6, DPP10, KCNE2 and KCNE3 and the results are summarized in Table 2.

Comparing the control recordings revealed that DPP6 and DPP10 increased K_V4.3 and K_V4.3/KChIP2 currents and accelerated current decay. KCNE2 inhibited K_V4.3 and K_V4.3/KChIP2 current amplitude as also reported by Radicke et al. (2006). In agreement with our previous findings (Lundby and Olesen, 2006; Delpón et al., 2008), KCNE3 reduced both K_V4.3 and K_V4.3/KChIP2 currents. Interestingly, it appeared that this inhibition of current could be abolished by co-expressing K_V4.3/KCNE3 with DPP6 (Table 2). Addressing the effect of NS5806 on the different subunit combinations, we found NS5806 exclusively caused a slowing of the current decay (τ) in channel complexes encompassing KChIP2. Thus, KChIP2 expression appears to be central for the effect of NS5806 on current decay and suggests that KChIP2 is part of the native I_to channel. Co-expression of K_V4.3/KChIP2 or K_V4.3/KChIP2/DPP6 with KCNE2 resulted in a marked slowing in recovery from inactivation following application of NS5806. These results are opposite to the effects seen in canine I_to. NS5806 accelerated K_V4.3/KChIP2/DPP10 current decay suggesting that DPP10 is an unlikely candidate for the native I_to channel. For all subunit combinations tested, application of NS5806 slowed recovery from inactivation, in sharp contrast to the acceleration of recovery observed in native I_to channels.

Effect of NS5806 on other cardiac potassium channels

The effect of 10 µM NS5806 was further tested on α-subunits mediating other important repolarizing cardiac potassium currents. K_V1.4 and K_V1.5 were expressed in Xenopus laevis oocytes in the presence or absence of KChIP2. For K_V1.4, 10 µM NS5806 inhibited more than 80% of the current (Figure 7A). The effect on the related K_V1.5 channel was smaller and mainly on the sustained current (Figure 7C). For both K_V1.4 and K_V1.5, co-expression with KChIP2 did not affect basal currents or the response to NS5806 (Figure 7B and D). There was no effect of NS5806 on currents generated by K_V11.1, K_V7.1 and K_ir2.1-3 channels (Supplementary Figures S3–5).

Figure 7.

Effect of NS5806 on K_V1.4 and K_V1.5 channels in presence or absence of KChIP2. K_V1.4 and K_V1.5 were expressed with and without KChIP2 in Xenopus laevis oocytes. Currents were activated from a holding potential of −80 mV to voltage steps from −100 mV to +40 mV before (squares) and after application of 10 µM NS5806 (triangles). Peak current amplitudes as function of voltage and representative current traces are shown for K_V1.4, n= 5 (A), K_V1.4/KChIP2, n= 5 (B), Kv1.5, n= 9 (C) and K_V1.5/KChIP2, n= 10 (D). *P < 0.05, **P < 0.01, ***P < 0.001; significantly different in the absence and presence of NS5806. KChIP2, K channel interacting protein 2.

Are other α-subunits contributing to canine ventricular I_to?

I_to measured in isolated cells from canine left ventricular epi- and midmyocardium recovers from inactivation with a bi-exponential time-course. Besides increasing currents and accelerating the recovery, the time-course of the recovery is changed from bi-exponential to mono-exponential in the presence of NS5806 (Calloe et al., 2009b).This might be explained by NS5806 enhancing a fast recovering current and inhibiting a slower recovering current. Thus, to test how NS5806 would affect a current generated by a mixture of the slowly recovering K_V1.4 (Supplementary Figure S6A) and the fast recovering K_V4.3/KChIP2/DPP6 channels, we co-expressed the channel constructs in Xenopus laevis oocytes in a 1:1 current ratio. Peak-current amplitude was unaffected by 10 µM NS5806 (Figure 8A), current decay was slowed (Figure 8B) and there was a significant left shift in steady-state mid-nactivation V_1/2, from −53.0 ± 0.8 mV to −67.7 ± 1.3 mV (Figure 8C). Recovery from inactivation was addressed as previously described and the fraction of recovered current was plotted as a function of interpulse time and a bi-exponential function was fitted to the data (Figure 8D and E). For control, the recovery showed a slow and a fast phase, τ₁= 20.1 ± 2.6 ms (relative weight of the pre-exponential factor, A₁= 0.30 ± 0.04) and τ₂= 1155 ± 448 ms (relative weight of the pre-exponential factor, A₂= 0.70 ± 0.03) respectively. In the presence of NS5806 the reactivation time course was markedly faster with τ₁= 26.0 ± 1.9 ms (A₁= 0.95 ± 0.04) and τ₂= 411 ± 242 ms (A₂= 0.05 ± 0.03). In the presence of NS5806, the pre-exponential factor of the slow component was small and the reactivation could be fitted with a mono-exponential equations with τ= 30.8 ± 1.2 ms. Thus, the recovery of the K_V4.3/KChIP2/DPP6/K_V1.4 current from inactivation was qualitatively similar to that of native I_to in the absence and presence of NS5806. Similar results were found for K_V4.3/KChIP2/K_V1.4 (Supplementary Figure S6B).

Figure 8.

Effect of NS5806 on K_V4.3/KChIP2/DPP6 co-expressed with K_V1.4 channels. K_V4.3/KChIP2/DPP6 and K_V1.4 were co-expressed in Xenopus laevis oocytes in a 1:1 current ratio. Currents were elicited using the depicted voltage protocol and measured in the absence and in the presence of 10 µM NS5806. (A) Relation between peak current density and voltage, n= 8. (B) Single exponential equations were fitted to the current decays, and the time constants (τ) are shown as a function of voltage, n= 8. (C) Steady-state inactivation of K_V4.3/KChIP2/DPP6/K_V1.4 currents. Normalized tail current amplitudes are plotted as a function of the prepulse potential and Boltzmann equations are fitted to the data, n= 8. (D) Time-dependent recovery from inactivation. K_V4.3/KChIP2/DPP6/K_V1.4 currents were activated by a two-pulse protocol. Current amplitudes at the second test pulse were normalized to that of the first test pulse and plotted as a function of the inter-pulse interval and double-exponential equations were fitted to the data, n= 7. (E) Bar graph showing the time-constants of K_V4.3/KChIP2/DPP6/K_V1.4 recovery from inactivation, τ₁ and τ₂ as well as pre-exponential factors. *P < 0.05, **P < 0.01, ***P < 0.001; significantly different in the absence and presence of NS5806. DPP, dipeptidyl-peptidase; KChIP2, K channel interacting protein 2.

Discussion and conclusion

Effects of NS5806 on K_V4.3 channel gating

NS5806 increased peak current amplitude of Kv4.3/KChIP2 expressed in CHO-K cells with an EC₅₀ of 5.3 ± 1.5 µM and slowed the time course of current inactivation (τ) with an EC₅₀ of 25.4 ± 1.1 µM. For Kv4.3/KChIP2 expressed in Xenopus laevis oocytes, the effect of the compound was mainly on current decay. Ancillary subunits can modify channel gating and may induce conformational changes that can affect access or affinity for exogenous compounds (Bett and Rasmusson, 2008) and the effect of NS5806 on current decay appeared to be linked to KChIP2. KChIP2 is a cytosolic protein that interacts with the intracellular N-termini of K_V4 subunits (Callsen et al., 2005); however, NS5806 also affected K_V4 channels in the absence of KChIP2, as evident from the shift in mid-inactivation, suggesting the binding site is on the α-subunit.

The negative shift in the mid-inactivation observed for all subunit combinations tested indicates that NS5806 increased closed-state inactivation, which was confirmed in a set of experiments addressing the onset of closed-state inactivation for K_V4.3 in the presence and absence of KChIP2. For K_V4.3 and K_V4.3/KChIP2, NS5806 increased the fraction of channels that entered closed-state inactivation.

Recovery from closed-state inactivation has been suggested to be the limiting factor for K_V4 channel recovery (Bahring et al., 2001; Amadi et al., 2007). In agreement, we show that the time-dependent recovery of K_V4 channels was slowed in the presence of NS5806. The faster onset of closed state inactivation and slower recovery from inactivation results in a larger fraction of K_V4 channels being in an inactivated state, mainly in the absence of KChIP2. Thus, NS5806 modified Kv4.3/KChIP2 gating in several ways that inhibit current. This dual mode of activation and inhibition may explain why 100 µM NS5806 increased Kv4.3/KChIP2 peak-currents less than 30 µM did in CHO-K1 cells. However, it should be emphasized that in canine ventricular cardiomyocytes and in CHO-K1 cells, the prominent effect of 10 µM NS5806 was to increase the size of current. The mechanisms behind the observed increase in peak current are unknown, for K_V4.3/KChIP2 channels it could be due to slowed inactivation; however, this is an unlikely mechanism for the increase in Kv4.3 currents in the absence of KChIP2 as current decay is unaffected in the absence of KChIP2 and the exact mechanism of drug binding and how drug binding is transmitted to a response is not known.

Differential effect of NS5806 on K_V4.3 and K_V4.3/KChIP2 channels – can NS5806 be used to discriminate the molecular composition of native I_to?

We have previously characterized the effect of NS5806 on native canine I_to in isolated cardiomyocytes. For mid- and epicardial cells, NS5806 increased I_to peak current, slowed current decay, caused a negative shift in the steady-state mid-inactivation and accelerated the time dependent recovery from inactivation. As an initial basis for comparison, K_V4.3 was co-expressed with KChIP2 in CHO-K1 cells and Xenopus laevis oocytes (Calloe et al., 2009b). In neither expression system did the measured current recapitulate native I_to fully, in terms of the response to NS5806. Interestingly K_V4.3 current decay was slowed when KChIP2 was co-expressed, but unaffected in the absence of KChIP2.

As K_V4.3 co-expressed with KChIP2 and DPP6 in heterologous systems has been demonstrated to result in currents with kinetics similar to those of I_to in human ventricular cardiomyocytes (Radicke et al., 2005), we first co-expressed K_V4.3 with KChIP2 and DPP6 in CHO-K1 cells in the present study. Currents were measured using similar solutions as for native I_to. Similar to native I_to, peak current density was increased by NS5806, the current decay slowed and mid-inactivation V_1/2 shifted to more negative voltages. However, in contrast to the effect on native I_to, the time dependent recovery from inactivation was markedly slowed for K_V4.3/KChIP2/DPP6 currents in the presence of NS5806.

We therefore tested the effect of NS5806 on K_V4.3 expressed with KChIP2 and other ancillary subunits in Xenopus laevis oocytes. In agreement with our previous observations, current decay was slowed by NS5806 only for K_V4.3 co-expressed with KChIP2, independently on the presence of other ancillary subunits. This implies that NS5806 potentially can be used to investigate the molecular composition of native I_to channels. Furthermore, for K_V4.3 channels expressed in the absence of KChIP2, NS5806 may inhibit currents at physiological resting membrane potentials due to the negative shift in the steady-state mid-inactivation (Figure 5 and Table 2).

NS5806 induced a marked increase in peak I_to and slowed I_to decay in canine mid- and epicardial I_to (Calloe et al., 2009a,b;) suggesting that KChIP2 is an integral part of the I_to channel in canine ventricle. However, none of the tested ancillary subunit combinations completely reconstituted native canine mid- and epicardial I_to with regard to the effect of NS5806. For native I_to, the recovery from inactivation was accelerated by NS5806, but for all subunit combinations tested, recovery was slowed by NS5806.

Besides the various ancillary subunits that have been implied to contribute to native I_to, other channel-forming subunits have also been suggested; for other species, including humans, K_V1.4 has been reported to contribute to I_to (Patel and Campbell, 2005). Previous studies have found K_V1.4 mRNA (Dixon et al., 1996; Rosati et al., 2001) and K_V1.4 protein (Akar et al., 2004) in the canine left ventricle. However, the functional role of K_V1.4 and I_{to, slow} in canine tissue is uncertain. Akar and colleagues found no appreciable contribution of K_V1.4 current to canine ventricular I_to (Akar et al., 2004). Interestingly, for canine mid- and epicardial I_to, application of NS5806 accelerated the recovery but also changed the kinetics from bi-exponential to mono-exponential (Calloe et al., 2009b). It was tempting to speculate whether this change in recovery kinetics could be due to enhancement of a fast recovering current component in combination with inhibition of a slow recovering component of I_to. To test this hypothesis, the fast recovering K_V4.3/KChIP2/DPP6 channels were co-expressed with slowly recovering K_V1.4 channels in Xenopus laevis oocytes. The current generated by K_V4.3/KChIP2/DPP6/K_V1.4 recovered from inactivation with a bi-exponential time-course (Figure 8), similar to native I_to. In addition, NS5806 accelerated the recovery and changed the time-course from bi-exponential to mono-exponential. This suggests that K_V1.4 or some other slowly recovering current could contribute to native canine I_to. Whether this is the case in vivo remains to be determined. NS5806 slowed K_V4.3/KChIP2/DPP6/K_V1.4 current decay yet did not affect peak current amplitude. We consistently observed that NS5806 had little effect on amplitude of currents recorded in Xenopus laevis oocytes whereas current amplitudes were increased when similar constructs were expressed in CHO-K1 or HEK-293 cells (data not shown). Increasing the temperature to 37°C for the oocytes experiments and replacing the extracellular solution with that used for native I_to recordings did not alter the effect of NS5806 on K_V4.3/KChIP2 peak current amplitudes in Xenopus laevis oocytes (data not shown) suggesting that the observed difference was due to intrinsic differences between expression systems.

Comparison with other I_to affecting compounds

The effects of several compounds on K_V4.3 have been reported to be modulated by co-expression of ancillary subunits. The I_Kr blocker tedisamil inhibits K_V4.3/KChIP2 with different IC₅₀ values depending on co-expression with KCNE1, KCNE2 or DPP6 due to differential modulation of current kinetics (Radicke et al., 2009). Similarly for the class 1C anti-arrhythmic agent flecainide, the IC₅₀ values for the peak current amplitudes of K_V4.3/KChIP2 are dependent on co-expression with KCNE1, KCNE2 or DPP6 (Radicke et al., 2008). The inhibition of K_V4.3 current by the local anesthetic bupivacaine is reduced by co-expression of KChIP2 (Solth et al., 2005). However, for these compounds the differences in drug action due to ancillary subunits are so subtle that they limit their usefulness to investigate the molecular constituents of native I_to or I_A channel complexes. The calcium channel blocker nifedipine is an open-pore blocker of K_V4.3 and K_V4.3/KChIP2 channels when used in high concentrations (150 µM). Interestingly, the time-dependent recovery of K_V4.3 is markedly slowed by nifedipine whereas in the presence of KChIP2 the recovery is unaffected by the compound (Bett et al., 2006). This suggests that nifedipine can be used to identify the presence of KChIP2 in native channels; however, the effect on I_CaL may be an issue in some preparations. For NS5806, a small inhibiting effect on I_CaL and I_Na in isolated canine ventricular cells was observed (Calloe et al., 2009a), and in the present work we found that NS5806 inhibits K_V1.4 and K_V1.5 currents. Nevertheless, even with these caveats in mind, NS5806 provides a tool to address the physiological role and the molecular composition of I_to and I_A.

Limitations of this study

In the present study we compared the effect of NS5806 on heterologously expressed putative human I_to channel subunits. The results were compared with results obtained in a previous paper where we characterized the effect of NS5806 on canine ventricular I_to (Calloe et al., 2009b). Canine I_to has a faster current decay and recovers slower from inactivation than human I_to. Additionally, human I_to recovery follows a single exponential time course, whereas canine I_to follows a biexponential time course (Akar et al., 2004). We cannot exclude the possibility that the observed discrepancies between the effect of NS5806 on the heterologously expressed channels in this study and native canine I_to are due to species differences. However, more importantly, the expression systems may lack important regulatory factors, additional ancillary subunits, etc. found in native cells and the molecular composition of the native I_to channel remains speculative. Further Xenopus laevis oocyte experiments were performed at room temperature which slows the kinetic parameters compared with currents measured at 37°C. However, for screening as well as for expression of multiple constructs, the oocytes are preferable to mammalian cell lines.

In conclusion, NS5806 activates native I_to and K_V4 channels heterologously expressed with KChIP2 in both Xenopus laevis oocytes and CHO-K1 cells. The differential effect of NS5806 in the absence and presence of KChIP2 suggests that NS5806 besides providing an experimental model of the Brugada syndrome can be used as a tool to address the physiological role of I_to and I_A and potentially to identify the molecular components of the channels mediating the currents in vivo; however, this should be tested by complementary approaches.

Glossary

Abbreviations

CHO-K1

Chinese hamster ovary cells

DPP

dipeptidyl-peptidase

I_A

A-type potassium current

I_to

transient outward potassium current

KChIP2

K channel interacting protein 2

Supporting information

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

Figure S1 K_V4.1 with and without KChIP2 were expressed in Xenopus laevis oocytes. Currents were elicited by protocols shown in Figure 5 in the absence (black) and presence of 10 μM NS5806 (red). (A) Peak current amplitude as a function of voltage, n = 5. (B) Total charge movement evaluated as area under current trace during the first 150 ms after complete activation, n = 5. (C) Steady-state mid-inactivation. Current amplitudes at this step was normalized, plotted as a function of the pre-pulse potential and fitted to Boltzmann equation. For K_V4.1, V_1/2 was −61 ± 0.4 mV in the absence and −75 ± 0.8 mV in the presence of NS5806, n = 5. For K_V4.1/KChIP2, V_1/2 was −45 ± 0.2 mV in the absence and −59 ± 0.2 mV in the presence of NS5806, n = 6. (D) The time constant of recovery was determined by a two-pulse protocol. The peak current at the second pulse was normalized to the first pulse and the data points obtained were fit to a single exponential function. For K_V4.1, τ = 109 ± 3 ms in the absence and τ = 492 ± 24 ms in the presence of NS5806, n = 6. For K_V4.1/KChIP2, τ = 16.8 ± 1.3 ms in the absence and τ = 52.8 ± 3.3 ms in the presence of NS5806, n = 5.

Figure S2 K_V4.2 with and without KChIP2 were expressed in Xenopus laevis oocytes. Currents were elicited by protocols shown in Figure 5 in the absence (black) and presence of 10 μM NS5806 (red). (A) Peak current amplitude as a function of voltage, n = 6. (B) Total charge movement evaluated as area under current trace during the first 150 ms after complete activation, n = 6. (C) Steady-state mid-inactivation. Current amplitudes at this step was normalized, plotted as a function of the pre-pulse potential and fitted to Boltzmann equation, For K_V4.2, V_1/2 was −65 ± 0.5 mV in the absence and −76 ± 0.8 mV in the presence of NS5806, n = 6. For K_V4.2/KChIP2, V_1/2 was −57 ± 0.3 mV in the absence and −67 ± 0.4 mV in the presence of NS5806, n = 10. (D) The time constant of recovery was determined by a two-pulse protocol. The peak current at the second pulse was normalized to the first pulse and the data points obtained were fitted to a single exponential function. For K_V4.2, τ = 176 ± 12 ms in the absence and τ = 328 ± 16 ms in the presence of NS5806, n = 5. For K_V4.2/KChIP2, τ = 27.6 ± 1.1 ms in the absence and τ = 44.2 ± 1.9 ms in the presence of NS5806, n = 10.

Figure S3 K_V11.1 was expressed in Xenopus laevis oocytes. Currents were elicited from a holding potential of −80 mV by a series of steps from −80 to +40 mV in 10 mV increments. Tails current were measured at −60 mV. Recordings were made in the absence (black) and presence of 10 μM NS5806 (red). (A) Representative currents of n = 5 (B) Steady-state current amplitude plotted as a function of voltage, n = 5. (C) Peak tail-current amplitude as a function of voltage, n = 5.

Figure S4 K_V7.1 was expressed in Xenopus laevis oocytes. Currents were elicited from a holding potential of −80 mV by a series of steps from −80 to +40 mV in 10 mV increments. Tails current were measured at −60 mV. Recordings were made in the absence (black) and presence of 10 μM NS5806 (red). (A) Representative currents of n = 5 (B) Peak-current amplitude as a function of voltage, n = 5.

Figure S5 K_ir2.1, 2.2 or 2.3 were expressed in Xenopus laevis oocytes. From a holding potential of 0 mV, currents were activated a ramp protocol from −120 mV to +50 mV. Recordings were made in the absence (black) and presence of 10 μM NS5806 (red). (A) Representative recordings of K_ir2.1, n = 5. (B) Representative recordings of K_ir2.2, n = 5. (C) Representative recordings of K_ir2.3, n = 5.

Figure S6 Time-dependent recovery from inactivation of K_V1.4 and K_V4.3/KChIP2 co-expressed with K_V1.4 in Xenopus laevis oocytes. The time constant of recovery was determined by a two-pulse protocol. The peak current at the second pulse was normalized to the first pulse and the data points obtained were fitted to a single exponential function before (black) and after application of 10 μM NS5806 (red). (A) Reactivation of K_V1.4 in the absence of NS5806 showed a bi-exponential time course with τ₁ = 286 ± 125 ms and τ₂ = 2055 ± 798 ms pre-exponential factors of 0.65 ± 0.12 and 0.35 ± 0.002 for controls. In the presence of NS5806 a mono-exponential function could be fitted to the data with τ = 11 297 ± 3945 ms, n = 5. (B) K_V4.3/KChIP2 co-expressed with K_V1.4 followed a bi-exponential time course for reactivation with τ1 = 40 ± 11ms and τ2 = 1384 ± 260 ms with pre-exponential factors of 0.8 ± 0.05 and 0.2 ± 0.01 respectively. In the presence of NS5806, τ1 = 79 ± 10 ms and τ2 = 6906 ± 1450 ms with pre-exponential factors of 0.96 ± 1.4 and 0.04 ± 0.001, respectively, n = 7.

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.