DIFFERENTIAL EFFECTS OF THE TRANSIENT OUTWARD K⁺ CURRENT ACTIVATOR NS5806 IN THE CANINE LEFT VENTRICLE

Abstract

Objective:

To examine the electrophysiological and molecular properties of the transient outward current (I_to) in canine left ventricle using a novel I_to activator, NS5806.

Methods and Results:

I_to was measured in isolated epicardial (Epi), midmyocardial (Mid) and endocardial (Endo) cells using whole-cell patch-clamp techniques. NS5806 activation of K_v4.3 current was also studied in CHO-K1 cells and Xenopus laevis oocytes. In CHO-K1 cells co-transfected with K_v4.3 and KChIP2, NS5806 (10 μM) caused a 35 % increase in current amplitude and a marked slowing of current decay with τ increasing from 7.0±0.4 to 10.2±0.3 ms. In the absence of KChIP2, current decay was unaffected by NS5806. In ventricular myocytes, NS5806 increased I_to density by 80%, 82%, and 16% in Epi, Mid, and Endo myocytes, respectively (at +40 mV) and shifted steady-state inactivation to negative potentials. NS5806 also significantly slowed decay of I_to, increasing total charge to 227%, 192% and 83% of control in Epi, Mid and Endo cells, respectively (+40 mV, p<0.05). Quantification of K_v4.3 and KChIP2 mRNA in the 3 ventricular cell types revealed that levels of K_v4.3 message was uniform but those of KChIP2 were significantly greater in Epi and Mid cells. The KChIP2 gradient was confirmed at the protein level by Western blot.

Conclusions:

Our results suggest that NS5806 augments I_to by increasing current density and slowing decay and that both depend on the presence of KChIP2. I_to and its augmentation by NS5806 are greatest in Epi and Mid cells because KChIP2 levels are highest in these cell types.

INTRODUCTION

Repolarization of the cardiac action potential is initiated and controlled by activation of a number of time- and voltage-dependent K⁺ currents. In dog heart at least four K⁺ currents play important roles in regulating the cardiac action potential duration: (i) a Ca²⁺-independent transient outward K⁺ current (I_to); (ii) an inwardly rectifying K⁺ current (I_K1) and (iii) the rapid and slow forms of the delayed rectifier K⁺ current (I_Kr and I_Ks, respectively). An I_to has been identified in the myocardium of most mammalian species (for review see [1]). Ventricular epicardial (Epi) tissue has a more prominent I_to compared to endocardial (Endo) tissue [2-5]. Recently, it has been demonstrated that I_to can be modulated by several proteins such as K⁺-channel interacting protein (KChIP) [6,7], IRX [8], calcineurin/NFAT [9], DPP's [10] and various KCNE subunits [11,12].

Although an I_to gradient between Epi and Endo has been identified, the precise molecular identity of I_to in the canine ventricular myocardium remains unclear. It is generally believed that K_v4.3 channels comprise the majority of transient outward K⁺ channels in canine heart [13]. Previous studies have also identified K_V1.4 and K_V1.5 gene products in ventricular tissue [13,14]. However, the precise role of these alpha subunits and their contribution to canine I_to remains to be determined. Recent evidence also suggests that several β-subunits including KChIP2 can alter peak I_Kv4.3 density, slow decay of the current and accelerate recovery from inactivation[7]. However, the relative abundance of K_v4.3 and KChIP2 in canine ventricle remains controversial. Several studies suggest that K_v4.3 levels are uniform throughout the canine left ventricle and the gradient in I_to expression is due to a gradient in KChIP2 [15,16]. In contrast, another study found that KChIP2 protein was uniform throughout the left ventricle suggesting that I_to gradient in ventricle is not due to a gradient in KChIP2 levels [7]. Finally, Zicha et al. [17] found that both K_v4.3 and KChIP2 exhibit a transmural gradient, with Epi expression being greater than Endo expression in canine ventricle.

The present study compares the electrophysiological and molecular properties of the Ca²⁺-independent transient outward K⁺ currents in single myocytes isolated from the canine left ventricle. Results of our study indicate that the biophysical and molecular properties of I_to differ significantly in endocardial cells compared to midmyocardial (Mid) or epicardial cells. Analysis of the molecular subunits revealed that KChIP2 mRNA levels are lower in the endocardium, contributing to some of the observed biophysical differences. Furthermore, we found that the levels of putative I_to subunits K_v4.3 and K_v1.4 are equally distributed throughout the left ventricle. Utilizing the I_to activator, NS5806 [18], we confirm by pharmacological methods that KChIP2 is functionally important for I_to activation and that KChIP2 levels are lower in endocardial cells. Application of this I_to activator resulted in a significant increase in I_to, with the greatest effect observed in Epi and Mid cells. The regional variations in subunit contribution are responsible for the observed biophysical differences in magnitude and kinetics of I_to.

METHODS

Expression of K_v4.3 and KChIP2 in CHO-K1

Human (h) K_v4.3 (NM_172198) and hKChIP2.1 (NM_173192) were transiently expressed in CHO-K1 in a 1:3 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 L-Proline at 37°C in 5% CO₂.

Expression of K_v4.3 and KChIP2 in Xenopus laevis oocytes

Female Xenopus laevis frogs were anaesthetized with Tricain (2 g/l, Sigma) and ovarian lobes were removed. Oocytes were defolliculated enzymatically in 1% collagenase (Boehringer Mannheim) and 0.1% trypsin inhibitor (Sigma) in Kulori solution for 1 h followed by wash in Kulori containing 0.1% BSA (Sigma).

cRNA was prepared from hK_v4.3 and hKChIP2.1 using the mMESSAGE mMACHINE T7 kit (Ambion). 50 nl cRNA was injected using a Nanoject microinjector (Drummond Scientific, Broomall, PA) in molar ratios of K_v4.3 and KChIP2. The concentration of K_v4.3 was kept constant at 0.1 ng. Oocytes were kept at 19°C and currents measured 2 days after injection.

Isolation of adult myocytes

Myocytes from Epi, Endo and Mid regions were prepared from canine hearts using techniques previously described [19,20]. Adult mongrel dogs were anesthetized with sodium pentobarbital (35 mg/kg i.v.), their hearts were rapidly removed and placed in nominally Ca²⁺-free Tyrode's solution. A wedge consisting of the left ventricular free wall was cannulated and perfused with nominally Ca²⁺-free Tyrode's solution containing 0.1% BSA for about 5 minutes. The wedge preparations were then subjected to enzyme digestion with the nominally Ca²⁺-free solution supplemented with 0.5 mg/ml collagenase (Type II, Worthington), 0.1 mg/ml protease (Type XIV, Sigma) and 1 mg/ml BSA for 8-12 minutes. After perfusion, thin slices of tissue from the Epi (<2 mm from the epicardial surface), Mid (about 5-7 mm from the epicardial surface), and Endo (<2 mm from the endocardial surface) were shaved from the wedge using a dermatome. The tissue slices were then placed in separate beakers minced and incubated in fresh buffer containing 0.5 mg/ml collagenase, 1 mg/ml BSA and agitated. The supernatant was filtered, centrifuged at 200 rpm for 2 minutes and the pellet containing the myocytes was stored in 0.5 mM Ca²⁺ HEPES buffer at room temperature.

Solutions

K_v4.3 currents in CHO-K1 was measured using an extracellular NaCl Ringer solution (mM): NaCl 140, KCl 4, CaCl₂ 2, MgCl₂ 1, HEPES 10, pH=7.4 adjusted with NaOH and an intracellular solution (mM): KCl 110, KOH/EDTA 31/10, CaCl₂ 5.17, MgCl₂ 1.42, HEPES 10, MgATP 4, pH=7.2 with KOH. Kulori solution contained (mM): NaCl 90, KCl 4, MgCl₂ 1, CaCl₂ 1, HEPES 5, pH=7.4 with NaOH. The nominally Ca²⁺-free dissecting buffer contained (mM): NaCl 129, KCl 5.4, MgSO₄ 2.0, NaH₂PO₄ 0.9, glucose 5.5, NaHCO₃ 20 and was bubbled with 95% O₂/5% CO₂. Ventricular cells were superfused with HEPES buffer (mM): NaCl 126, KCl 5.4, MgCl₂ 1.0, CaCl₂ 2.0, HEPES 10, glucose 11, pH=7.4 with NaOH. The pipette solution consisted of (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.

Electrophysiology

I_to recordings from myocytes were performed as previously described [21]. All myocyte and CHO-K1 experiments were performed at 36°C. Voltage-clamp and conventional recordings were made using a MultiClamp 700A amplifier and MultiClamp Commander (Axon Instruments). Patch pipettes were fabricated from borosilicate glass capillaries (1.5 mm O.D., Fisher Scientific, Pittsburg, PA). Pipettes were pulled using a gravity puller (Narishige Corp) and the resistance ranged from 0.9-3 MΩ when filled with the internal solution. Cell capacitance was measured by applying −5 mV voltage steps. Electronic compensation of series resistance to 60-70% was applied. All analog signals were acquired at 10-25 kHz, filtered at 4-6 kHz, digitized with a Digidata 1322 converter (Axon Instruments) and stored using pClamp9 software.

Recordings from oocytes were performed using a two-electrode voltage-clamp amplifier (Dagan CA-1B; Chicago, IL). Borosilicate glass recording electrodes (Module Ohm, Denmark) were made using a DMZ-Universal Puller (Zeitz Instruments, Germany) and had a resistance of 0.5 to 1 MΩ when filled with 2 M KCl. Oocytes were superfused with Kulori solution and experiments performed at room temperature.

Analysis of mRNA Levels in the Left Ventricle

RNA preparation and cDNA synthesis

RNA was prepared from canine left ventricular tissue. Tissue samples were shaved using a dermatome and stored in an RNA stabilizing solution (RNAlater®-ICE, Ambion). Total RNA was purified from the homogenized tissue specimens (homogenizer Kinemtica, Buch & Holm, Switzerland) with Tri Reagent® (Sigma-Aldrich) according to the manufacturer's instructions. Total cDNA was synthesized from 2 μg total RNA using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystem) with random hexamer primers following the manufacturer's instructions.

Real-time PCR

Real-time reverse transcriptase-polymerase chain reaction (RT-PCR) was performed on the 7300 RT-PCR System (Applied Biosystem) and data were collected by SDS1.2 software. The selected genes were investigated using TaqMan assays (TaqMan® MGB probe and primers). The pre-designed gene expression assays from Applied Biosystems were as follows: cf02698011 (KCND3), cf02624497 (KCNIP2), cf02640342 (DPP10), cf02690512 (KCNE1), cf02625138 (KCNE2), cf02646775 (KCNE3), cf02650669 (KCNE4), cf02698674 (IRX5), cf02657295 (SMYD1), cf0263023048 (KCNA5), cf02659079 (β2-microglobulin), and cf02629556 (Cyclophilin B). The primers and probes targeting KCND2, KCNA4, KCNE5, DPP6, and HPRT were designed and synthesized by Applied Biosystems, following submission of intron spanning sequences using Primer Express 3.0 software. Thermal profile for all the real-time PCR reactions was as followed: 50°C (2 min), 95°C (10 min) and then 40 cycles with 95°C (15 sec) and 60°C (1 min).

Prior to the experiments, PCR efficiency for each assay was calculated using a standard curve constructed by plotting range of log cDNA input against Ct (threshold cycle) value. The slope of the plot was used to calculate the percentage amplification efficiency (PE). PE values ranged between 90% and 110% and an amplification efficiency of 2 (100%) per cycle has therefore been used in all the calculations. For each tissue specimen, each gene was quantified in triplicates. Three reference genes were tested: Cyclophilin B, HPRT, and β2-microglobulin. Quantification of these three genes gave almost similar result and Cyclophilin B was chosen for normalization. Normalized gene expression levels were calculated by the 2^ΔΔCt method [22]. Delta cycle threshold (ΔC_t) values were calculated by subtracting the Ct value of a target gene from the Ct value of Cyclophilin B for each sample. Values for the relative expression were obtained by using the following formula: 1/(2^−ΔCt)*100.

Western blotting

Epi, Mid and Endo tissue from 5 dogs was snap-frozen in liquid nitrogen and stored in −80 C prior to protein isolation. Total membrane and cytosolic proteins were isolated using a method modified after Han et al. [23]. Briefly, the tissue (20-30 mg) was pulverized in liquid nitrogen and suspended in 500 μL of ice-cold TE buffer (containing Tris 20 mM, EDTA 1 mM supplemented with a cocktail of protease inhibitors: 10 μM 4-(2-aminoethyl)benzenesulphonyl fluoride, 0.2 μM leupeptin, 0.4 μM bestatin, 0.15 μM pepstatin A, 0.14 μM M E-64 and 8 pM aprotinin, all from Sigma). The tissue suspension was treated with 2% of Triton X-100 for 2 hours at 4°C, centrifuged at 15 000 g for 15 minutes at 4°C. The soluble fraction was retained and stored in −80 C prior to Western blotting. Protein concentration was measured (in triplicates for each sample) using Bradford method (DC Protein Assay, Sigma). Samples (50 μg/lane) were separated on precast polyacrylamide 4-15% SDS-PAGE gels (BioRad) at 100 V for 2 hours and transferred to hybond-P PVDF membranes (Amersham Biosciences, 0.45 μm) at 400 mA for 2 hours. Membranes were blocked with 5% non-fat milk in TBST (Tris 10 mM, NaCl 150 mM, 0.1% Tween 20, pH=7.4) for 1-2 hour at RT and then incubated overnight at 4°C in monoclonal antibody against KChIP2b (1:200, clone K60/73, obtained from NeuroMab Facility, supported by NIH grant U24NS050606 and maintained by the Department of Pharmacology, School of Medicine, University of California, Davis, CA 95616). The proteins were detected by HRP-conjugated donkey anti-mouse antibody (1/10000, Jackson Immunosearch Laboratories) and visualized by ECL staining (Supersignal West Pico Chemiluminescent detection system, Pierce). Immunoblots were exposed (for 5-10 minutes) on hyperfilm ECL (Amersham Biosciences). To assess equal loading, the membranes were stripped in Restore™ Western Blot Stripping Buffer (Pierce) for 20 minutes in RT and re-probed with a mouse anti-actin antibody (MAB1501, Chemicon International). Band density was quantified by Quantity-One software as Gaussian trace quantity. KChIP2 signal was detected as two bands with molecular weight between about 25-32 kDa.

Statistics

Pooled data are presented as Mean±SEM. Statistical analysis was performed using an ANOVA test followed by a Student-Newman-Keuls test or Student t-test, as appropriate, using SigmaStat software. Statistical analysis of RNA and protein expression was performed with a repeated measures ANOVA followed by a Tukey's post test. p<0.05 was considered statistically significant.

RESULTS

Our previous studies showed that NS5806 activated native I_to in canine ventricular tissue[18]. To address the mechanism of this activation, we tested the effect of the compound on K_v4.3 channels transiently expressed in CHO-K1 cells. As an initial basis of comparison, the enhancement of I_Kv4.3 by NS5806 (10 μM) was determined in the absence and presence of KChIP2 (Figure 1). Application of NS5806 to CHO-K1 cells expressing K_v4.3 and KChIP2 resulted in a 35 % increase in I_Kv4.3 peak current amplitude and a dramatic slowing in decay (τ), from 7.0±0.4 to 10.2±0.3 ms (Figure 1A-C). In contrast, application of NS5806 to CHO-K1 cells expressing only K_v4.3 did not change peak I_Kv4.3 (Figure 1A-B). The decay of the current was unaffected over a range of potentials (Figure 1C). These results suggest that NS5806 can be used to discriminate between currents mediated by K_V4.3 alone and K_V4.3 together with KChIP2 as the effect of NS5806 is dependent on the presence of KChIP2. We next evaluated the effect on steady state gating parameters using a prepulse-test pulse voltage clamp protocol (Figure 1D). Peak current following a 0.5 sec prepulse was normalized to the maximum current and plotted as a function of the prepulse voltage to obtain the availability of the channels and a Boltzmann function was fitted to the data. NS5806 caused a significant negative shift in the mid-inactivation voltage for K_V4.3 from −51.6±0.8 to −60.7±0.8 and the respective slope factors were k=−6.8±0.7 and k=−6.4±0.7 (not significant). For K_V4.3 in the presence of KChIP2, mid-inactivation significantly shifted from −38.6±0.1 to −43.6±0.3 mV and the respective slope factors were significantly different with k=−5.1±0.1 and k=−6.3±0.2 (Figure 1D). Recovery from inactivation was addressed by a double-pulse protocol (Figure 1E). For K_v4.3 channels the time constant (τ) was significantly slowed from τ =47.7±3.6 ms to 74.0±7.7 ms in the presence of NS5806. Similar results were found for K_v4.3 in the presence of KChIP2, where NS5806 slowed the recovery from τ=6.1±0.5 ms to 22.8±1.5ms.

Figure 1.

A: Representative traces of K_v4.3 currents recorded in the absence (left, n=6) and presence of KChIP2 (right, n=5-10) under control conditions and in presence of 10 μM NS5806. From a holding potential of −80 mV, cells were stepped to +40 mV in 10 mV increments. Panel B: Mean I-V relations for peak current density Panel C: Mean τ's in absence and presence of NS5806. Panel D: Voltage dependence of inactivation and Boltzmann curves showing mid-inactivation Panel E: Time-dependent recovery from inactivation was evaluated using a standard double pulse protocol from a holding of −80 mV.

The enhancement of I_Kv4.3 as well as the slowing of I_Kv4.3. decay appeared to be dependent on KChIP2 and previous studies have suggested a differential distribution of KChIP2 levels across the canine left ventricle [15,16] In the next series of experiments, the effect of NS5806 on Xenopus laevis oocytes injected with different ratios of K_v4.3 and KChIP2 was evaluated (Figure 2). Application of N S5806 resulted in a significant increase in peak current by 12.1, 5.9 and 11.8% in channels with 4:2, 4:4, and 4:12 ratios respectively (Figure 2A) and resulted in a significant slowing of current decay only in the presence of KChIP2 (Figure 2B). This resulted in an increase in charge movement (Figure 2C). The effect of KChIP2 appeared to saturate at a 4:4 cRNA ratio.

Figure 2.

The effect of NS5806 on different ratios of K_v4.3 and KChIP2 expressed in Xenopus laevis oocytes. Mean data showing peak current (Panel A), decay (τ) of the current (Panel B) and total charge (Panel C). n=7-8 oocytes

The data obtained on K_V 4.3 expressed in CHO-K and Xenopus laevis oocytes could suggests that NS5806 can be used to discriminate whether KChIP2 is present in native I_to channels as K_V4.3 current decay was slowed by NS5806 only when KChIP2 was co-expressed. Thus if K_V4.3 and KChIP2 interact in ventricular cardiomyocytes we would expect an increase in current and a slowing of the decay. We next determined the effect of the I_to activator NS5806 in Epi Mid and Endo cardiomyocytes. The density of I_to was examined in the absence and presence of 10 μM NS5806. Cd²⁺ (300 μM) was added to the extracellular solution to block the calcium current (I_CaL). Following a brief step to −50 mV to discharge sodium channels, voltage steps from −40 to +50 mV applied to all three cell types elicited fast activating and rapidly inactivating I_to currents. Representative currents measured in Epi and Endo are shown in Figure 3A-B. Application of NS5806 resulted in an increase in the magnitude of I_to as well as slowing the decay of the current. Analysis of the current-voltage (I-V) relation of peak I_to showed that the current density was significantly greater in Epi and Mid cells compared to Endo, as previously reported [3,17]. Application of 10 μM NS5806 significantly increased the peak current amplitude of I_to in Epi and Mid cells but had no effect in Endo cells (Figures 3C-E). Evaluation of I_to in ventricular myocytes showed that application of NS5806 increased the magnitude of current by 80%, 82%, and 16% in Epi, Mid, and Endo myocytes, respectively (at +40 mV).

Figure 3.

Representative I_to traces recorded from an Epi (Panel A) and an Endo cell (Panel B) under control conditions and in the presence of NS5806 (10 μM). The voltage clamp protocol is shown at the top of the figure and Cd²⁺ was present to block I_CaL. Mean I-V relation for peak I_to from Epi (Panel C), Mid (Panel D) and Endo cells (Panel E) in absence and presence of NS5806 (10μM). The density of I_to was greater in Epi and Mid cells and application of NS5806 caused a substantial increase in the magnitude of I_to.

The time constant of decay (τ) of I_to following application of NS5806 was significantly slower in all 3 cell layers (Figure 4A-C). This slowing of I_to coupled with the increase in current magnitude resulted in a marked increase in total charge (assessed as area under the current trace, Figure 4D-F). While the area was significantly increased in all 3 cell layers, Epi and Mid cells showed the greatest effect. NS5806 increased total charge by 227%, 192% and 83% compared to control in Epi, Mid, and Endo cells respectively (at +40 mV, p<0.05).

Figure 4.

I_to decay as a function of voltage for Epi, Mid and Endo cells under control conditions and in the presence of NS5806 (10 μM) (Panels A-C). Panels D-F: Total I_to charge (area under the curve) was increased by NS5806 in all 3 cell layers.

We next determined if the difference in current density between the three cell types was due to changes in steady state gating parameters. Steady state inactivation of I_to was evaluated using a prepulse-test pulse voltage clamp protocol (top of Figure 5) in presence of Cd²⁺. The peak current following a 2 sec prepulse was normalized to the maximum current and plotted as a function of the prepulse voltage to obtain the availability of the channels and a Boltzmann function was fitted to the data. In the absence of drug, the mid-inactivation voltage and slope factors were −46.0±1.0 mV, k=4.75±0.44 for Epi (Figure 5C), −43.7±0.8 mV, k=4.17±0.41 for Mid (Figure 5D) and −52.9±1.1, k=8.45±1.34 mV for Endo cells (Figure 5E). Application of NS5806 (10 μM) caused a significant shift in the mid-inactivation potential in all cell layers with mid-inactivation voltage and slope factors of −52.0±0.9 mV, k=4.18±0.37 for Epi (Figure 5C), −48.7±0.7 mV, k=4.04±0.59 for Mid (Figure 5D) and −56.3±0.8 mV, k=7.73±1.07 for Endo cells (Figure 5E). The mid-inactivation potential of I_to is significantly shifted to more negative potentials and this shift could not account for the increase in current magnitude observed in the presence of NS5806.

Figure 5.

Representative traces recorded under control conditions (Panel A) and after application of 10 μM NS5806 (Panel B) showing voltage dependence of inactivation of I_to. Boltzmann curves showing mid-inactivation voltages for Epi (Panel C), Mid (Panel D) and Endo (Panel E) cells in the absence and presence of drug.

We next tested if I_to recovery from inactivation was different in the three ventricular cell layers. Representative traces of the frequency-dependent changes in I_to in the absence and presence of NS5806 are shown in Figure 6A-B. Reactivation of I_to at −80 mV for all three cell types showed a fast and a slow phase of recovery as follows: i) τ1=43.8±6.7 ms and τ2=256.6±18.9 ms for Epi cells, ii) τ1=39.7±8.3 ms and τ2=278.4±19.2 ms for Mid, and iii) τ1=75.2±8.6 ms and τ2=803.9±79.8 ms for Endo (Figure 6C-E). In addition, under control conditions (absence of NS5806) the recovery of I_to was much slower in Endo cells compared to Epi and Mid cells. In the presence of NS5806, the reactivation of I_to in Epi and Mid cells could be fit with a single exponential and was markedly faster with τ=55.6±1.7 ms for Epi and τ=71.3±2.8 ms for Mid (Figure 6C-D). However, in the presence of NS5806 reactivation time course of I_to in Endo cells was unchanged with τ1=87.2±8.2 ms and τ2=705.6±81.8 ms (Figure 6E).

Figure 6.

Representative traces recorded under control conditions (Panel A) and after 10 μM NS5806 (Panel B) showing recovery of I_to. Cd²⁺ was present to block I_CaL. The recovery time-course of I_to recorded from Epi (Panel C), Mid (Panel D) and Endo (Panel E) cells in the absence and presence of drug.

The differential effects of the I_to activator in the 3 ventricular cell layers could suggest that KChIP2 levels may be high in Epi and Mid cells and low in Endo tissue corresponding to the effects of NS5806 on K_v4.3 expressed with different ratios of KChIP2 (Figure 2). However, as KChIP2 slows K_v4.3 current decay and the decay of Endo I_to is slower than that of Epi and Mid (Figure 4), it seems unlikely that transmural differences in I_to kinetics, and response to NS5806 is solely due to differential KChIP2 expression. Other K_v channel subunits may contribute to I_to, in particular in Endo tissue. To test this hypothesis, the expression level of the genes encoding proteins suspected to mediate I_to as well as I_Kur was investigated (Figure 7). The relative mRNA expression levels of KCND2 and KCND3, encoding K_v4.2 and K_v4.3, respectively, showed an almost 10,000-fold higher expression of KCND3 as compared to KCND2, demonstrating that K_v4.3 is the predominant α-subunit. We also tested the expression of several ancillary subunits and found a high expression of KChIP2, DPP6 and KCNE1. DPP6 was found to be expressed in an approximately 100-fold higher level than DPP10. For the KCNE β-subunits, KCNE1 mRNA was expressed at a high level, KCNE2-4 in intermediate levels, and KCNE5 at a low level. The expression of 2 transcription factors, SMYD1 and IRX5, believed to be important for the I_to gradient in other species [24,25] was also investigated. SMYD1 (Bop) transcripts were found in a much higher abundance than IRX5 transcripts. KCNA4 (K_v1.4) and KCNA5 (K_v1.5) transcripts were found to be expressed in 10% and 30-50%, respectively, of the level of KCND3 (K_v4.3).

Figure 7.

qPCR of left ventricular canine tissue. Epi, Mid, and Endo tissue was isolated from 5 canine left ventricles. Taqman based assays (Applied Biosystems) were used to quantify mRNA of interest which was then normalized to cyclophilin B expression (Panel A). Dot plots of KCND3, KCNIP2, DPP6, IRX5, and SMYD1 expression (Panel B). The expression level values were obtained as described in materials and methods.

Transmural differences in the expression level were found for several genes (Figure 8). As reported previously, there was no gradient for KCND3 and a large gradient for KCNIP2 with higher expression in the EPI compared to the ENDO [15,16]. For DPP6 a difference in expression level was observed between Epi and Mid, with more transcript in the Mid. For both the transcription factors IRX5 and SMYD1 a higher expression level was found in Endo compared to the Epi consistent with suggested function as negative regulators of expression of I_to related genes.

Figure 8.

A: Transmural gradient of KChIP2 protein in canine left ventricle. Representative blot of KChIP2 expression in Endo-, Mid- and Epi. KChIP2 expressing and non-transfected HEK-293 cell lysates were included as controls. For Epi, Mid and Endo, two bands sized 25-32 kDa were consistently detected for all protein isolation methods tested. The same blot was re-probed with anti-actin antibody. Panel B: Both KChIP2 bands were quantified and KChIP2 expression normalized to actin expression. Tissue from 5 dogs was analysed.

The transmural gradient of KChIP2 transcript was paralleled at protein level as assessed by Western blotting (Figure 8). We repeatedly obtained two bands, using three other different methods of protein purification (data not shown). Antigenic peptide could not be obtained to assess the specificity of the KChIP2 signal but no unspecific bands were found in HEK-293 cell lysate. As the KChIP2 antibody was raised against a highly conserved region, the two bands could represent different KChIP2 splice variants as previously reported in cardiac tissue [26] and both bands were quantified (Figure 8).

DISCUSSION

Summary of Main Findings

In CHO-K1 cells and in Xenopus laevis oocytes, the effect on I_Kv4.3 decay by NS5806 was dependent on the presence of KChIP2. In the presence of KChIP2, application of NS5806 increased peak current and slowed current decay. This was reflected as a significant increase in total charge movement in the presence of KChIP2. In canine ventricular cells, Endo I_to was smaller, the decay significantly slower, mid-inactivation voltage was more negative and the recovery from inactivation substantially slower when compared to Mid and Epi I_to. Application of NS5806 caused an increase in the magnitude of Mid and Epi I_to but had no effect on Endo I_to magnitude. In all three layers, the decay of the current was slowed. In the presence od NS5806, recovery from inactivation was faster in Mid and Epi cells but reactivation of Endo I_to was unaffected. These observations could suggest that different subunits contribute to Endo I_to.

Analysis of mRNA expression levels of putative I_to subunits across the canine left ventricular wall revealed a differential distribution of KChIP2 levels with Endo tissue expressing the lowest levels. K_v4.3 levels appeared uniform and no other putative subunits exhibited a transmural gradient. These results suggest that a KChIP2 gradient is responsible for the I_to gradient observed in the canine left ventricle.

Molecular Identification of I_to in the Left Ventricle

The precise molecular identity of I_to in the canine ventricular myocardium remains unclear. It is generally believed that K_v4.3 channels comprise the majority of transient outward K⁺ channels in canine heart with lower levels of K_v1.4 also being identified [13,27]. Recent evidence suggests that several β-subunits including KChIP2 can alter peak I_Kv4.3 density, slow decay of the current and accelerate recovery from inactivation [7]. However, the relative abundance of K_v4.3 and KChIP2 in canine ventricle remains controversial. Our qPCR results confirm that the I_to gradient across the left ventricle is likely due to a KChIP2 gradient which was also confirmed at protein level. We also investigated if other putative I_to subunits exhibited a transmural gradient but found no significant differences. For KCNE1 and KCNE3 there was a tendency toward higher expression in Endo compared to Mid and Epi. In CHO-K1 cells, KCNE1 did not significantly affect K_v4.3 + KChIP2 current amplitude [12] whereas in HEK cells, KCNE1 has been demonstrated to increase K_v4.3 currents [28] and thus is an unlikely candidate responsible for the I_to gradient. KCNE3 has been demonstrated to inhibit K_v4.3±KChIP2 current [29,30] which is consistent with the smaller I_to in Endo.

We also observed the presence of K_v1.4 mRNA, in agreement with previous experiments [13]. Interestingly, results of our molecular analysis revealed the presence of K_v1.5 mRNA in higher amounts than that of K_v1.4 as previously reported [13]. While it has been shown that K_v1.5 is abundantly expressed in the atria and is thought to mediate I_Kur [31], recent studies have suggested that I_Kur is also present in ventricle [14]. If Kv1.4 and Kv1.5 contribute to I_to, then the relative contribution of K_v1.4 and K_v1.5 may be of greater importance in Endo where KChIP2 mRNA levels are low and the contribution of current generated by K_v4.3 smaller. In our study, the slower τ values as well as the slow recovery from inactivation observed in Endo support this.

Biophysical Analysis of I_to in the Left Ventricle

The electrophysiological properties of K_v4.3 channels are modulated by several β-subunits including KChIP2, which increases peak K_v4.3 current density, accelerates recovery from inactivation, and slows the decay (τ) of the current [16]. In the present study we demonstrated that the effect of NS5806 on heterologously expressed K_v4.3 channels is dependent on KChIP2. Epi and Mid cells had the highest KChIP2 levels and exhibited the largest I_to (Figure 2). Interestingly, the t of decay was faster in Epi and Mid than in Endo as also previously reported [5,32,33]. As KChIP2 slows K_v4.3 current decay, these findings may suggest that Endo I_to is not merely K_v4.3 in absence of KChIP2 and suggests that other subunits may contribute to Endo I_to. In further support of this notion is the observation that the decay of I_to in Epi, Mid and Endo was slowed by NS5806 suggesting that KChIP2 is a functional component of the I_to channels all three cell layers. The currents generated by K_v4.3 and KChIP2 channels did not recapitulate all the features of the native I_to. This is particularly evident when comparing the effect of NS5806 on recovery from inactivation. As also previous studies have demonstrated [5,32,33], the recovery from inactivation is much slower of Endo I_to than of Epi or Mid. The recovery of Epi and Mid I_to was accelerated by NS5806 whereas Endo I_to was unaffected. In contrast, NS5806 slowed recovery for heterologously expressed Kv4.3 independently of co-expression with KChIP2.

In agreement with previous publications, we found that KChIP2 increased Kv4.3 currents in both CHO-K1 cells and Xenopus laevis oocytes. The qPCR demonstrated a low abundance of KChIP2 in the endocardium and this would suggest less Kv4.3 mediated current. As the distribution of K_V1.4 and K_V1.5 message is uniform across the ventricular wall this would likely result in a greater relative contribution of current mediated by K_V1.4 and K_V1.5 channels to endocardial I_to. It is tempting to speculate that the relatively larger contribution from K_V1.4 or K_V1.5 channels to Endo I_to compared to Epi and Mid I_to accounts for the differences in current kinetics as well as response to NS5806.

In summary, we have identified a novel I_to activator that appears to have differential effect on I_Kv4.3 in the presence or absence of KChIP2. Application of this compound to canine ventricular cells resulted in a differential augmentation on I_to. Our results suggest that the KChIP2 gradient is mainly responsible for the I_to gradient observed in the canine left ventricle. Since Epi and Mid cells have the highest KChIP2 levels, application of NS5806 had the largest effect on the decay of I_to and produced the largest increase in I_to in those 2 cell layers.