Introduction

Top Summary Introduction Materials & Methods Results Discussion References

Mutations in the KvLQT1 gene have been recognized as the most frequent cause for the autosomal dominant (Romano-Ward) form of the long QT syndrome, a life-threatening familial disorder characterized by prolonged cardiac repolarization (1). KvLQT1 was also demonstrated to be responsible for the recessively inherited Jervell and Lange-Nielsen cardioauditory syndrome characterized by a bilateral deafness associated with prolonged cardiac repolarization (2). Finally, the KvLQT1 gene, which comprises 14 exons and spans a large 300-kilobase pair genomic region at chromosomal region 11p15.5, encompasses chromosomal rearrangements associated with the Beckwith-Wiedemann syndrome, which causes prenatal overgrowth and cancer but no cardiac repolarization anomalies (3). Within the past year, these successive discoveries enlightened the medical importance of KvLQT1, which may be implicated in at least three distinct genetic disease entities.

The KvLQT1 gene product was very recently demonstrated to be a voltage-dependent K⁺ channel with electrophysiological characteristics that are consistently modified by the presence of a membranous regulator protein termed IsK (4, 5). The IsK (also termed minK) gene product, which contains a single transmembrane domain, often was suspected to be itself a K⁺ channel (6). This assumption was derived from the observation that injection of Xenopus laevis oocytes with IsK mRNA alone produced a voltage-dependent K⁺ current (6-8). However, in 1993, Attali et al. (9) proposed that IsK acts as a channel regulator but not as a channel protein. In agreement with this latter interpretation, transfection of mammalian cells with IsK cDNA produced no K⁺ current (10). Discrepant results obtained in different expression systems were later explained by Sanguinetti et al. (5), who demonstrated that X. laevis oocytes constitutively express a KvLQT1 transcript that generates a K⁺ current in the presence of IsK. In mammalian cells, expression of KvLQT1 alone but not of IsK alone produces a voltage-dependent K⁺ current of small amplitude and fast activation kinetics. Coexpression of KvLQT1 plus IsK is associated with a K⁺ current with physiological characteristics reminiscent of that of the endogenous i_Ks current, a major component of the delayed-rectifier sustained K⁺ current in cardiac myocytes (11). Expression of KvLQT1 is not limited to the heart muscle and is also found in human pancreas, lung, kidney, and placenta (4, 5).

The medical importance of KvLQT1 prompted us to identify a pharmacological blocker that can be used as a tool to elucidate the physiological role of KvLQT1 K⁺ channels in various organs. The chromanol compound 293B was a good candidate because it was previously found to block the K⁺ current induced by expression of IsK in X. laevis oocytes (12). Here, we demonstrate that 293B is a potent blocker of human KvLQT1 K⁺ channels and that it binds to the channel protein itself but not to the IsK regulator.

	Materials and Methods

Top Summary Introduction Materials & Methods Results Discussion References

Intranuclear injection of plasmids. The African green monkey kidney-derived cell line COS-7 was obtained from the American Type Culture Collection (Rockville, MD) and cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum and antibiotics (100 IU/ml penicillin and 100 µg/ml streptomycin; all from GIBCO, Paisley, Scotland) at 37° in a humidified incubator. They were subcultured regularly by enzymatic treatment. Cells growth onto plastic Petri dishes with a glass coverslip bottom (Nunclon; InterMed Nunc, Roskilde, Denmark) were microinjected with plasmids at day 1 after plating. Our protocol to microinject cultured cells using the Eppendorf ECET microinjector 5246 system has been reported in detail previously (13). Plasmids were diluted at a final concentration of 5-50 µg/ml in a buffer of 50 mM HEPES, 50 mM NaOH, and 40 mM NaCl, pH 7.4, supplemented with 0.5% fluorescein isothiocyanate-dextran. The human KvLQT1 and IsK cDNAs (a kind gift from J. Barhanin, Sophia-Antipolis, France) were subcloned into the mammalian expression vector, pCI and pCR, respectively (Promega, Madison, WI) under the control of a cytomegalovirus enhancer/promoter (4).

Patch-clamp recordings. Whole-cell currents were recorded as described previously (13). A Petri dish containing cells was placed on the stage of an inverted microscope and superfused continuously with the standard extracellular solution. Patch pipettes with a tip resistance of 2.5-5 M were electrically connected to a patch-clamp amplifier (Axopatch 200A; Axon Instruments, Foster City, CA). Stimulation, data recording, and analysis were performed with a software designed by Gérard Sadoc (DIPSI Industrie, Asnière, France) through an A/D converter (Tecmar TM100 Labmaster; Scientific Solution, Solon, OH). A microperfusion system allowed local application and rapid change of the different experimental solutions warmed at 35°. Patch-clamp measurements are presented as the mean ± standard error of the mean. Statistical significance of the observed effects was assessed by means of the Student's t test.

Solutions and drugs. The standard extracellular medium contained 145 mM NaCl, 4 mM KCl, 1 mM MgCl₂, 1 mM CaCl₂, 5 mM HEPES, and 5 mM glucose, pH adjusted to 7.4 with NaOH. For whole-cell patch-clamp recordings, the intracellular medium contained 145 mM K-gluconate, 5 mM HEPES, 2 mM EGTA, 2 mM Mg_1/2-gluconate (0.1 mM free Mg²⁺), and 2 mM K₂ATP, pH 7.2 with KOH, and the extracellular medium contained 145 mM Na-gluconate, 4 mM K-gluconate, 7 mM Ca_1/2-gluconate (1 mM free Ca²⁺), 4 mM Mg_1/2-gluconate (1 mM free Mg²⁺), 5 mM HEPES, and 5 mM glucose, pH 7.4 with NaOH. For cell-attached recordings, the bath solution contained 145 mM K-gluconate, 4 mM Mg_1/2-gluconate 4 (1 mM free Mg²⁺), 7 mM Ca_1/2-gluconate (1 mM free Ca²⁺), and 5 mM HEPES, pH 7.2 with KOH, and the pipette solution contained 145 mM Na-gluconate, 4 mM K-gluconate, 7 mM Ca_1/2-gluconate (1 mM free Ca²⁺), 4 mM Mg_1/2-gluconate (1 mM free Mg²⁺), and 5 mM HEPES, pH 7.4 with NaOH. Compound 293B [trans-6-cyano-4-(N-ethylsulfonyl-N-methylamino)-3-hydroxy-2,2-dimethyl-chromane] was dissolved in dimethylsulfoxide so the final concentration of the solvent was <1. Cromakalim (Sigma Chemical) was dissolved in dimethylsulfoxide.

	Results

Top Summary Introduction Materials & Methods Results Discussion References

KvLQT1 expression in COS-7 cells. Noninjected COS-7 cells or cells injected with the pCI plasmid alone exhibited no time-dependent current. Fig. 1 illustrates typical whole-cell recordings in COS-7 cells injected with KvLQT1 cDNA in the absence (Fig. 1A) and presence (Fig. 1B) of IsK. Cells injected with KvLQT1 plus IsK cDNAs exhibited a time-dependent outward current with a comparable amplitude but with slower activation kinetics than cells expressing KvLQT1 alone. Deactivating tail currents at 40 mV were fitted by a monoexponential decay that was extrapolated to time 0 to reliably measure tail current amplitude in the absence of contamination by the capacitative current. Under our experimental conditions, the current tail density so measured in cells injected with KvLQT1 cDNA alone (5.91 ± 1.08 pA/pF; prepulse potential to +40 mV; 15 cells) was not significantly different from that recorded in the presence of the IsK regulator (4.19 ± 0.64 pA/pF; 16 cells; p > 0.5). As observed previously by others (4), the activation threshold for the time-dependent current related to KvLQT1 expression was comparable in the presence and absence of IsK. In contrast, in cells coinjected with KvLQT1 plus IsK cDNAs, the potential for half-maximal activation (V_0.5 = 14.1 mV), was shifted in comparison with cells injected with KvLQT1 cDNA alone (V_0.5 = 23.2 mV). Fig. 1C illustrates activation and deactivation kinetics for KvLQT1. In cells expressing KvLQT1 alone, activation was adequately fitted by the sum of two exponential functions. Both _act-fast (which represented 83% of the total current) and _act-slow (which represented 17% of the total current) decreased at more depolarized voltages. In cells coexpressing KvLQT1 and IsK, activation kinetics were convincingly fitted by a single exponential with a time constant (_act) 10-fold greater than the activation time constant (_act-fast) of the major component in cells expressing KvLQT1 alone. In contrast, as reported previously (13), deactivation time constants (_deact) were of the same order of magnitude in the presence and absence of IsK (Fig. 1C).

View larger version (28K): [in this window] [in a new window]   Fig. 1.   KvLQT1-related K+ currents in COS-7 cells in the absence and presence of the IsK regulator. A and B, Top, superimposed current tracings recorded in a noninjected control COS-7 cell (A) and a cell injected with 5 µg/ml pCR-IsK alone (B). Middle, current tracings in a cell injected with 5 µg/ml pCI-KvLQT1 without (A) and with (B) 5 µg/ml pCR-IsK. The voltage potential was stepped, in 10-mV increments, to various voltages (pulse potential) between 100 and +60 mV (for clarity, current traces are shown recorded in response to voltage steps of 20-mV increments) and then stepped back to 40 mV, where tail currents are visible. Holding potential, 80 mV. Vertical bars, 2.5 pA/pF. Horizontal bars, 500 msec. Bottom, current-voltage relationships for the time-dependent current activated by depolarization. A, Averaged data from noninjected cells () and cells injected with pCI-KvLQT1 alone () (five cells). B, Averaged data from cells injected with pCR-IsK alone () and cells injected with pCR-IsK plus pCI-KvLQT1 () (six cells). C, Activation and deactivation time constants for the KvLQT1 current in relation to pulse potential. Values are mean ± standard error. Top, Activation time constants act-fast () and act-slow () for the rapid and slow components of the KvLQT1 current recorded in the absence of IsK (n = 9). , Activation time constant act for the KvLQT1 current recorded in the presence of IsK (four cells). Same voltage protocol as in A and B. Bottom, deactivation time constants deact for the KvLQT1 current in relation to pulse potential in the absence (; four cells) and presence (; four cells) of IsK. The voltage potential was stepped to +60 mV and then stepped back to various voltages between 100 and +60 mV, where tail-current deact were measured.

293B dose-dependently blocks KvLQT1-related K⁺ current. We then tested the effects of 293B on the K⁺ current related to KvLQT1 expression. In a concentration range of 0.1-100 µM, 293B dose-dependently blocked KvLQT1 tail currents. The dose-effect relations shown in Fig. 2A were obtained by measuring the tail current density at 40 mV in the presence of various concentrations of the drug. Exponential fits forced to time 0 were used to achieve this measurement. Fig. 2A shows that 293B blocked KvLQT1 currents regardless of the presence of IsK. The half-maximum concentration (IC₅₀) values for 293B to block tail currents related to KvLQT1 expression were 9.8 and 9.9 µM in the presence and absence of IsK, respectively. The blocking effects of 293B were fully reversible when the drug was washed off. The current blocked by 10 µM 293B obtained by digital subtraction and its current-voltage relationship are depicted in Fig. 2, B and C. 

View larger version (24K): [in this window] [in a new window]   Fig. 2.   The chromanol compound 293B blocks KvLQT1 currents. A, Dose-effect relationships for 293B to block KvLQT1 current in the absence () and presence () of IsK. The protocol consisted of depolarizing voltage steps applied from 80 to +40 mV and then back to 40 mV. Data points represent the normalized current tail amplitudes measured in the presence of various concentrations of 293B. Data points, mean ± standard error of four to seven experiments. Semilogarithmic plot in which data were fitted to a Hill equation, with a half-activation point at 9.9 µM (IsK) and 9.8 µM (+IsK) and a steepness of 1.1 (IsK) and 0.9 (+IsK). Inset, superimposed current traces recorded in control (C) and in the presence of 10 µM 293B in the presence (left; +IsK) and absence (right; IsK) of IsK. Vertical bars, 5 pA/pF. Horizontal bars, 200 msec. B, 293B-sensitive K+ current obtained by digital subtraction in the absence (left; IsK) and presence (right; +IsK) of IsK. Vertical bars, 4 pA/pF; horizontal bars, 200 (IsK) and 500 (+IsK) msec. C, Current-voltage relationship for the 293B-sensitive current recorded in the cells shown in B. Same protocol as in legend to Fig. 1A.

Voltage- and time-dependence of the block produced by 293B. KvLQT1 tail current-voltage relationships in a cell injected with KvLQT1 cDNA alone and recorded in the absence and presence of 10 µM 293B are superimposed in Fig. 3A. These curves suggested that the block produced by 293B was voltage-dependent in that for activation voltages of <30 mV, the tail current amplitude in the presence of 293B was comparable in size to that recorded in control conditions. Averaged data (Fig. 3B) further substantiated the voltage-dependent effects of 293B. This curve was obtained by plotting the percentage of block produced by 293B at a concentration close to its IC₅₀ (i.e., the current tail amplitude in the presence of 10 µM 293B subtracted from the current amplitude in control and divided by the tail current amplitude in control) for various prepulse potentials. Obviously, the amount of KvLQT1 block produced by 293B increased as the activation prepulse was set more positive. To further investigate the voltage-dependency of the block produced by 293B, additional experiments were performed to determine dose-effect relations at different membrane potentials in cells expressing KvLQT1 alone. The protocol consisted of depolarizing voltage steps, in 10-mV increments, applied to various voltages between 100 and +60 mV and followed by a step to 40 mV, where tail current amplitude was measured in the absence and presence of various concentrations of 293B. From the results of five different experiments, the calculated IC₅₀ values were 15.6 µM at 20 mV, 10.1 µM at 0 mV, 7.2 µM at + 20 mV, and 7.1 µM at +40 mV.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Voltage- and time-dependence of the effects of 293B. A, KvLQT1 tail current-voltage relationship in control (C) and with 10 µM 293B. Data are averaged current tail amplitudes from four cells injected with KvLQT1 in the absence of IsK. The voltage potential was stepped to various voltages (prepulse potential) between 100 and +60 mV and then stepped back to 40 mV (pulse potential) where tail currents were measured. B, Voltage-dependence of 293B block at steady state. Same voltage protocol as in A. For each prepulse potential, the maximal tail current amplitude (measured using monoexponential fits forced to time 0) in the presence of 10 µM 293B was subtracted from that in control and divided by the maximal tail current amplitude in control. This value, which represents the percent of block produced by the drug at steady state, was plotted against the prepulse potential. Data are mean ± standard error from five different cells. C, Time-dependence for 293B block. The voltage protocol consisted of depolarizing steps from 80 to + 40 mV for 1 sec. The KvLQT1 current in the presence of 293B (10 µM) was digitally subtracted from that recorded in control and then divided by the current recorded in control [(icontrol  i293)/icontrol]. This calculation gave the relative amount of block (in percent) produced by 293B as a function of time during a depolarizing step. D, Current deactivation at 80 mV. Bottom, voltage protocol. Cells were depolarized from 80 to +40 mV for 1 sec and then clamped at 80 mV for increasing periods of time, in 5-msec increments, before a second depolarizing pulse was applied. For clarity, only the currents recorded during the second depolarizing pulse are superimposed in control (left) and with 293B (10 µM; right). Initial current amplitudes at the onset of the second depolarizing pulse (open symbols), which reflected deactivation kinetics, were fitted by a monoexponential function, as shown. The time constant was 18 msec in control and 34 msec with 293B. Vertical bars, 3 pA/pF; horizontal bars, 50 msec. Current amplitude at the end of the conditioning prepulse: 32 pA/pF in control and 22 pA/pF with 293B. E, Time dependence for 293B unblock. The value for (icontrol  i293)/icontrol was calculated from the traces shown in D, plotted as percent against time, and fitted to a monoexponential (time constant, 17 msec). This relation reflects unblock kinetics.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Lack of use-dependence of the chromanol 293B effects. A and B, Outward current amplitude at +40 mV in a cell injected with 5 µg/ml KvLQT1 is plotted against time. 293B (10 µM) applied as indicated. A, Cell was sequentially depolarized at 0.2 Hz from 80 to +40 mV. In the first part of the experiment, voltage stimulation (open horizontal bars) was stopped for 30 sec, started again for 40 sec, and then stopped for 30 sec. In the second part of the experiment, the voltage stimulation was not interrupted. B, In the first part of the experiment (crosses), the membrane voltage was maintained at +40 mV. In the second part of the experiment, the cell was stimulated as in A but the cell was a different cell. Open horizontal bars, voltage stimulation.

293B blocks KvLQT1 current in cell-attached patches. In six additional experiments, KvLQT1 K⁺ currents were recorded in the maxipatch cell-attached configuration (Fig. 5). When applied in the extracellular medium, 293B also blocked the time- and voltage-dependent K⁺ current related to KvLQT1 cDNA expression, suggesting that 293B penetrates the cell to bind at a cytoplasmic site of the channel protein. In the cell-attached configuration, onset and offset kinetics of current block by 293B were in the same order as in whole-cell recordings.

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Cell-attached maxipatch recordings in a COS-7 cell injected with 50 µg/ml pCI-KvLQT1 plus 50 µg/ml pCRIsK. Superimposed current traces recorded in control (C) and in the presence of 10 µM and 100 µM 293B applied in the extracellular medium. The potential was stepped from 80 to +40 mV and then back to 40 mV. Potentials are expressed relative to the bath. The cell membrane potential considered as 0 mV when bathed in the high K+ extracellular medium. Vertical bar, 100 pA; horizontal bar, 200 msec.

Cromakalim blocks KvLQT1 K⁺ current. Finally, we explored the effects on KvLQT1 currents of cromakalim [trans-6-cyano-3-hydroxy-2,2-dimethyl-4-(2-oxo-1-pyrrolidinyl)-chromane], another chromanol derivative. The main property of cromakalim is to act as a K⁺ channel opener targeted to ATP-sensitive K⁺ channels (15). However, in a previous study conducted in guinea-pig ventricular myocytes (16), we observed that potassium channel openers not only activate a time-independent current corresponding to the ATP-sensitive K⁺ current but also suppress outward tail currents triggered by cell repolarization. Our findings with 293B prompted us to search for an effect of cromakalim on recombinant KvLQT1 current. As illustrated in Fig. 6, we found that cromakalim, like 293B, also acted as a KvLQT1 current blocker. On average, cromakalim, when used at 10 µM (the concentration used to activate ATP-sensitive K⁺ channels in cardiac myocytes), blocked 30 ± 6% of the tail current amplitude related to KvLQT1 expression in COS-7 cells.

View larger version (10K): [in this window] [in a new window]   Fig. 6.   The effects of cromakalim on the KvLQT1-related current. Superimposed current traces recorded in control (C) and in the presence of cromakalim (10 µM) in a COS-7 cell injected with 5 µg/ml pCI-KvLQT1 without IsK. The protocol consisted of depolarizing voltage steps applied from 80 to +40 mV and then back to 40 mV. Vertical bar, 2 pA/pF; horizontal bar, 200 msec.

	Discussion

Top Summary Introduction Materials & Methods Results Discussion References

Our results show that (i) the chromanol 293B binds directly to KvLQT1 channel protein but not to the IsK channel regulator, (ii) expression of IsK does not modify 293B affinity for KvLQT1 channels, and (iii) 293B exerts strong voltage- and time-dependent-blocking effects on KvLQT1 channels. This pharmacological profile is consistent with an open-channel block. Another possible explanation is that 293B binds at some site on the channel protein with accessibility that is state-dependent. The kinetics for 293B to associate with its binding site during depolarization is not instantaneous but takes a few tenths of a msec in the absence of IsK and a few hundredths of a msec in the presence of IsK. In contrast, unbinding of the drug to its receptor as induced by repolarization is fast and ~5-10-fold faster than deactivation kinetics of the current. The lack of use-dependence of the effects of 293B is not surprising in regard to the very rapid unblock kinetics of the drug. Finally, we show that another chromanol compound, cromakalim, which is well known for its K⁺ channel-activating properties targeted to ATP-sensitive K⁺ channels, also exerts measurable KvLQT1 blocking effects at concentrations compatible with its K⁺ channel-activating properties in cardiac muscle.

Earlier reports concerning the chromanol compound 293B have demonstrated the ability of this drug to block a cAMP-activated K⁺ conductance in colonic cells (18, 19). Subsequently, it was shown that 293B blocked a time-dependent K⁺ conductance in X. laevis oocytes injected with IsK cRNA (12) with an IC₅₀ value of 6.7 µM (i.e., close to the value for 293B to block KvLQT1 current in the current report). In contrast, the drug was ineffective on the K⁺ current related to Kir2.1 or Kv1.1 expressions (12). Most recently, however, it was shown that IsK expression in X. laevis oocytes does not create recombinant K⁺ channel proteins per se but instead reveals an endogenous K⁺ channel with strong homologies with the KvLQT1 gene product identified in the human heart (5). On the basis of experiments performed previously in X. laevis oocytes, it thus remained obscure whether 293B binds to the IsK regulator or to the KvLQT1 channel protein itself. Our work addresses this important question because it demonstrates that the affinity for 293B to block KvLQT1 current is independent of the presence of IsK. The KvLQT1 gene is expressed not only in the heart but also in numerous other organs, including the kidney, pancreas, adrenal and salivary glands, and stomach (for a more complete list, see Ref. 17). Depending on the tissue, the IsK regulator should or should not be coexpressed together with KvLQT1. Our results demonstrate that 293B is a valuable tool with which to explore the role of KvLQT1 channels in a tissue regardless of the presence of IsK.

This assumption is caricatural as far as the heart muscle is concerned. Class III antiarrhythmic agents of the Vaughan-Williams classification, which is based on the effects of drugs on the morphology of the cardiac action potential (20), prolong the plateau duration and thus prolong refractoriness. The simplest way to achieve this goal is to block myocardial K⁺ channels because less net outward current will be available to repolarize the cells. Virtually all members of pure class III antiarrhythmic drugs retain voltage-dependent K⁺ channel block as their mechanism of action (21); more precisely, these drugs usually block the rapid component of the delayed rectifier termed i_Kr (11), as is the case with D-sotalol, dofetilide, sematilide, risotilide, E4031, and almokalant. At the molecular level, the i_Kr current is conducted through the HERG channel protein (22), which is encoded by a gene located on chromosome 7q35-36 and is responsible for the congenital long QT 2 phenotype (23). In opposition to class I antiarrhythmic drugs, class III agents targeted to HERG have an unfavorable frequency-dependent profile (so-called reverse frequency dependence) because they are more active at slow than at fast rates (24). Theoretically, this may limit their activity during tachycardia and produce excessive prolongation when the cardiac rate is slow, leading to torsades de pointe arrhythmias. Thus, novel class III antiarrhythmic drugs lacking an unfavorable frequency-dependent profile and therefore not targeted to HERG are needed. During an action potential, the i_Ks current slowly activates and participates to cell repolarization in concert with i_Kr and other K⁺ currents. Repolarization in turn switches off the i_Ks current through deactivation mechanisms. Because deactivation is slow during diastole, at high frequencies deactivation is incomplete before the next action potential arises, leading i_Ks channels to remain permanently in the open state (25). Thus, KvLQT1 channels, which are expressed at a high level in the heart together with the IsK regulator (4, 5, 17), may represent a molecular target for novel class III antiarrhythmic agents with a more favorable rate-dependent profile. The pharmacological profile for 293B as reported here makes this drug a valuable tool for investigation of the antiarrhythmic properties of an i_Ks blocker. It can be anticipated that at a slow heart rate, the effect of the drug on the cardiac action potential should be modest because (i) i_Ks activation kinetics are slow in relation with the duration of the action potential, (ii) 293B binding kinetics on opened i_Ks channels also are slow, and (iii) 293B unbinding kinetics are fast and much faster than deactivation kinetics leading i_Ks tail current available for terminal phase 3 repolarization. At high rates of stimulation, the effect of the drug on action potential duration should be more pronounced because a significant proportion of i_Ks channels are permanently opened. As yet, little information is available as to whether 293B is specific for KvLQT1 K⁺ channels in cardiac cells. Busch et al. (26) suggested recently that in the guinea-pig heart, 293B exerts no significant effects on the rapid component of the delayed rectifier i_Kr, on the inwardly rectifying instantaneous K⁺ current or the L-type Ca²⁺ current (26). Regardless of the specificity of 293B for KvLQT1 in the heart muscle, KvLQT1 channels are widely distributed in the organism and KvLQT1 blocking drugs that should be developed in the context of arrhythmias are suspected to have side effects that would result from their action on KvLQT1 channels in other tissues.