INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

NIFEDIPINE HAS BEEN WIDELY USED for almost two decades in the control of cardiac chest pain and hypertension but has been associated, in a dose-dependent manner (10), with unfavorable side effects like negative inotropy and hypotension, proarrhythmia, and (in some studies) increased mortality (10, 31). This may occur especially with administration of short-acting forms of nifedipine in patients who are already hemodynamically compromised (13, 21). Part of the mechanism for proarrhythmia caused by nifedipine may be the block of myocardial K⁺ channels, which has been described both in mammalian cardiac myocytes with equilibrium dissociation constants (K_d) of 0.5-1 µM and in cloned channels (12, 32). The plateau phase of the cardiac action potential is normally terminated by repolarizing outward potassium fluxes. Therefore, nifedipine block can prolong the action potential, cause a dispersion of refractoriness as these channels differ in their regional distribution across the myocardial wall (18), and lead to instability of the resting potential of the muscle.

Despite the potential importance of these K⁺ channel actions of nifedipine, the mechanisms by which it exerts its actions on K⁺ channels are not fully defined. Many compounds block K⁺ channel currents by binding at or near the internal or external mouth of the ion-conducting pore, where they may act as direct open channel blockers by physically occluding the ion-conducting pore to prevent ion passage. Examples of this kind of blocker are quinidine and the quaternary ammonium compounds like tetraethylammonium (TEA), as well as the NH₂-terminal inactivation particle in Shaker channels (8). It is known that nifedipine acts as an open channel blocker of single Shaker mutant channels (1) and of the mammalian homolog Kv1.5 channels with a K_d of 6.3 µM (32). Whether nifedipine acts at the internal or external mouth of the K⁺ channel pore, though, is presently uncertain. Other details of the action of permeant ions on nifedipine block and the ability of the drug to bind to channel-inactivated states are also poorly understood. Kv channels show a rapid N-type inactivation or a slower C-type inactivation, affected by factors such as elevation of the extracellular K⁺ concentration ([K⁺]_o) (2, 20), external application of TEA (6, 11), and mutations of particular amino acids in the channel pore (7, 20, 24). These effects have been incorporated into a physical model where C-type inactivation is caused by the constriction of the outer mouth of the channel pore in a cooperative action of all four subunits (30), which restricts K⁺ flux (23, 24). This rearrangement of the outer mouth of the pore greatly reduces the permeability of K⁺ relative to the permeability of Na⁺, altering the ion selectivity of the channel (27). Recently, it has been further proposed that, during C-type inactivation, the channels dwell in at least three conformational states: an initial open state that is highly selective for K⁺, a state that is less permeable to K⁺ and more permeable to Na⁺, and then a state that is nonconducting (17, 19).

Kv1.5 is a prominent cardiac K⁺ channel -subunit because it is expressed in the human heart, particularly in the human atrium, and possibly in other regions of the human cardiovascular system (9, 28). Nifedipine blocks Kv1.5 as an open channel blocker with a K_d of 6.3 µM (32). Here, we present evidence that the nature of the permeant ion does affect the dose dependence and kinetics of nifedipine block of Kv1.5 channels. This suggests that nifedipine block involves coupling between permeation and the blocking site and locates the binding of nifedipine within the permeation pathway of K⁺ channels. Outer pore mutations of R487 reduce the affinity for Kv1.5 channel block in manner consistent with other channels that have a valine residue at equivalent sites. We also show that nifedipine block, although affected by mutations similar to those known to affect outer pore C-type inactivation, can be shown to be kinetically separate from this slow inactivation process. This is based on nifedipine block of Kv1.5 carried out in different [K⁺]_o and also on the effects of nifedipine on the recovery rates from C-type inactivation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell culture. Human Kv1.5 channels stably expressed in HEK-293 cell lines were used in all experiments. Kv1.5 in the plasmid expression vector pCDNA3 was mutagenized using the Quickchange Kit (Stratagene, La Jolla, CA) to convert arginine residue 487 to valine (R487V) or tyrosine (R487Y). HEK-293 cells were stably transfected with wild-type Kv1.5 or Kv1.5-R487(V/Y) cDNAs using LipofectACE reagent (Canadian Life Technologies, Bramalea, ON, Canada) in a 1-to-10 (wt/vol) ratio. Patch pipettes contained (in mM) 135 KCl, 5 EGTA, and 10 HEPES and were adjusted to pH 7.2 with KOH. When KCl was substituted with CsCl, RbCl, or N-methyl-D-glucamine (NMG⁺), the pH was adjusted with CsOH, RbOH, or HCl, respectively. The base bath solution contained (in mM) 135 NMG⁺, 5 KCl, 10 HEPES, 1 MgCl₂, and 1 CaCl₂ and was adjusted to pH 7.4 with HCl. For recordings in the presence of different external Cs⁺, Rb⁺, or K⁺ concentrations, the NMG⁺ base external solution was used, and KCl was substituted by the appropriate ions. The elevated external K⁺ solutions contained (in mM) 5, 135, or 300 KCl, 10 HEPES, and 1 MgCl₂; pH 7.4. Nifedipine was dissolved in ethanol at a stock concentration of 20 mM and was protected from exposure to light.

Electrophysiological procedures. Coverslips containing cells were removed from the incubator before experiments and placed in a superfusion chamber (volume 250 µl) containing the control bath solution at 22-23°C. The bath solution was exchanged by switching the perfusates at the inlet of the chamber, with complete bath solution changes taking 5-10 s. Whole cell current recording and data analysis were done using an Axopatch 200A amplifier and pCLAMP version 6.0 software (Axon Instruments, Foster City, CA). Patch electrodes were fabricated using thin-walled borosilicate glass (World Precision Instruments; Sarasota, FL). Capacity compensation and 80-95% series resistance compensation were routinely used. The averaged cell membrane capacitance was 15.1 ± 0.5 pF, n = 128, and measured series resistance was between 0.5-5.5 M for all recordings (averaged series resistance was 2.3 ± 0.1 M, n = 128). When this changed during the course of an experiment, data were discarded. Data were filtered at 5-10 kHz and sampled at 10-20 kHz. Gating currents were recorded as described previously (29). The data for analysis and presentation were off-line leak subtracted if required, and data were discarded if the leakage conductance was >1 nS. Throughout the text, data are shown as means ± SE.

Data analysis. The concentration-response curves for permeating K⁺, Rb⁺, and Cs⁺ ions were computer fitted to the Hill equation

(1)

where f is the fractional current (I_drug/I_control) at drug concentration D, IC₅₀ is the concentration producing half-maximal inhibition, and n_H is the Hill coefficient. The time constant (₂) of the rapid component of current decay in the presence of nifedipine was used as an approximation of the drug channel interaction kinetics, as described previously (4, 32) according to the equation

(2a)

and

(2b)

in which ₂ is the current decay time constant caused by the drug, k₊₁ and k₁ are the apparent rate constants of binding and unbinding for the drug, respectively, and K_d represents the affinity of the drug for its binding site.

(3a)

(3b)

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Nifedipine blocks K⁺, Rb⁺, and Cs⁺ currents through Kv1.5 channels. The currents in Fig. 1 illustrate the actions of nifedipine on K⁺, Rb⁺, and Cs⁺ currents through Kv1.5 channels. Currents were recorded using a physiological ion gradient across the cell membrane [135 mM intracellular K⁺ concentration ([K⁺]_i)/5 mM [K⁺]_o] or a solution with external and internal K⁺ replaced by equimolar Rb⁺ or Cs⁺. The control traces in Fig. 1A were elicited from a holding potential of 80 mV to potentials between 30 and +40 mV. In the control, the main difference between currents carried by the different ions was a decrease in current amplitude that reflected the decreased permeability of Rb⁺ or Cs⁺ compared with K⁺ and a reduction in the amount of slow C-type inactivation in Cs⁺ (8), most visible at the more positive potentials. External application of 10 µM nifedipine markedly inhibited both peak and steady-state K⁺ currents with an apparent acceleration in the decay rates of outward currents positive to +10 mV (Fig. 1B). The characteristics of nifedipine block of K⁺ currents through Kv1.5 channels have been previously studied in detail (32). It is noticeable that the acceleration of current decay only becomes apparent at more positive potentials where the channel activation rate significantly exceeds the nifedipine block rate. At more negative potentials, block reduces current amplitude without significantly altering kinetics. External application of 10 µM nifedipine also blocked Rb⁺ and Cs⁺ currents (Fig. 1B), but the potency of block was significantly less. Application of 50 µM nifedipine (Fig. 1C) produced a more marked block of current carried by all three cations and a significant acceleration of current decay that reflected open channel block caused by the drug. Steady-state current-voltage relations from these data are shown in Fig. 1D for the three ions and show the increased potency of nifedipine action when K⁺ was carrying current through Kv1.5 rather than when Rb⁺ or Cs⁺ were carrying current. At +40 mV, external application of 10 µM nifedipine blocked more than 60% of K⁺ current. However, the same concentration of nifedipine only blocked 35% of Rb⁺ current and 20% of Cs⁺ current. External application of 50 µM nifedipine blocked 86% of K⁺ current, 70% of Rb⁺ current, and 60% of Cs⁺ current. As we (32) have previously noted, nifedipine block was minimal until potentials around 10 mV when channel open probability was significant. At potentials where the channel open probability was high, block was only mildly dependent on pulse potential, with a small reduction of block at more positive potentials.

View larger version (35K): [in this window] [in a new window]   Fig. 1.   Block of Kv1.5 K+, Rb+, and Cs+ currents in HEK cells by nifedipine. Whole cell currents were elicited from a holding potential of 80 mV to voltages between 30 and +40 mV in increments of 10 mV. As indicated, K+ (left), Rb+ (middle), and Cs+ (right) currents were recorded in the control (A) and presence of 10 (B) and 50 µM nifedipine (C), respectively. The scale bars in A also apply to B and C, and data for each ion in A-C are from the same cells. D: steady-state current-voltage relations for K+, Rb+, and Cs+ current block by nifedipine from current data above. In the absence (open symbols) and presence of 10 and 50 µM nifedipine (filled symbols), the steady-state K+ (left), Rb+ (middle), and Cs+ (right) currents measured at the end of current traces were plotted against test potential.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Concentration-dependence of nifedipine block of Kv1.5 K+, Rb+, and Cs+ currents. A: concentration-response curves for nifedipine block of steady-state K+, Rb+, and Cs+ currents [current with nifedipine (Inif)/current in control (Icontrol)]. Residual current in nifedipine was measured at the end of 400-ms depolarizations and normalized to current level before nifedipine. Solid lines were fit to the data using a Hill equation (see MATERIALS AND METHODS). The IC50 concentrations for K+, Rb+, and Cs+ currents blocked by nifedipine were 7.3 ± 0.3, 16.0 ± 0.8, and 26.9 ± 1.3 µM, and the Hill coefficients were 0.9 ± 0.1, 1.1 ± 0.1, and 1.2 ± 0.1, respectively. Data are means ± SE (n = 6-14 cells). B: decay time constants of Kv1.5 K+, Rb+, and Cs+ currents in the presence of nifedipine. The reciprocal of the nifedipine-induced fast time constant of block (1/2) at +40 mV for the different ions is plotted against the concentration of nifedipine. The solid line is the best fit to the data using the equation 1/2 = k+1 × [D] + k1 (see MATERIALS AND METHODS). The association rate constants (k+1) were 3.26 × 106, 2.05 × 106, and 1.33 × 106 M1 · s1 for K+, Rb+, and Cs+ currents, respectively, and the apparent dissociation rate constants (k1) were 25.34, 33.88, and 36.16 s1 for K+, Rb+ and Cs+ currents, respectively. The equilibrium dissociation constants (Kd) (k1/k+1) for K+, Rb+, and Cs+ currents were 7.8, 16.5, and 27.2 µM, respectively. Data points are means ± SE (n = 3-14 cells).

Membrane sideness of cation modulation of nifedipine action. One of the primary aims of the present study was to determine the site of action of nifedipine on the Kv1.5 channels. Because the potency of block varied with different cation species permeating the channel, the following experiments were designed to test whether or not Kv1.5 block by nifedipine was mediated within the pore region itself. The rationale was based on the preceding experiments, where it appeared that the presence of larger cations within the pore, like Cs⁺ and Rb⁺, reduced the potency of nifedipine binding compared with K⁺. The intracellular Cs⁺ concentration was kept constant at 130 mM, and the external ion concentration and species were changed to alter the direction of ion flow across the pore. A two-pulse protocol was used, with an initial 200-ms prepulse to index block of Kv1.5 outward Cs⁺ current at +80 mV by 20 µM nifedipine and a 200-ms test pulse immediately afterward to +30 mV to measure nifedipine block of current under different extracellular ion conditions (Fig. 3, A-D). Mean currents measured at the end of the prepulse and the test pulse are shown in Table 2. At the end of the 200-ms prepulse with an outward Cs⁺ current in all cases, there was no significant difference in the nifedipine block between extracellular Cs⁺ or K⁺ (Table 2). During the test pulse at +30 mV, as shown in Fig. 3, A, C, and D, an outward Cs⁺ current is present, and there is little difference in the action of nifedipine.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Nifedipine block is modulated by the ion species occupying the pore. The same twin-pulse voltage protocol was used in A-D in the presence of 130 mM intracellular Cs+ (Cs) concentration (A-D) and 70 mM extracellular Cs+ (Cs) concentration (A), 70 mM extracellular K+ (K) concentration (B), 5 mM Cs concentration (C), and 5 mM K concentration (D). Outward Cs+ current was elicited at +80 mV from a 80 mV holding potential during the first 200-ms pulse. During the second pulse, cells were depolarized to +30 mV for 200 ms. Current tracings are shown in the absence and presence of 20 µM nifedipine. Dashed lines in each panel denote the zero current level.

Nifedipine action on Kv1.5 mutant channels. The above experiments strongly support a pore-blocking action of nifedipine in Kv1.5. We wished to extend these experiments in a significant way by utilizing deep pore and outer pore mutants of Kv1.5. The first of these was a nonconducting mutant of Kv1.5, W472F. This single point mutant prevents ion permeation through the pore but otherwise gates normally, as demonstrated in Kv1.5 (5) and in Drosophila Shaker channels (25). This mutant allows ions to be changed independently on each side of the membrane with the knowledge that they cannot cross the pore to influence nifedipine block on the other side. Nonconducting mutant function is measured by recording gating currents as an index of channel gating and opening, as we (15, 29) have shown before.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Ion modulation of nifedipine (nif) block of gating current in Kv1.5-W427F. A-F: off-gating currents in control and in the presence of 100 µM nifedipine under different ionic conditions, as indicated above each panel. In each case, on-gating currents were unchanged in the presence of nifedipine (data not shown). NMG and NMG, extracellular and intracellular N-methyl-D-glucamine concentration, respectively; K and K, intracellular K+ concentration; Rb and Rb, intracellular and extracellular Rb+ concentration, respectively; Cs, intracellular Cs+ concentration.

View larger version (31K): [in this window] [in a new window]   Fig. 5.   An external pore R487V mutant reduces nifedipine block. In all cases, pulses were from 80 mV to +40 mV for 400 ms. A: K+ currents; B: Rb+ currents; C: Cs+ currents. As indicated adjacent to the tracings in A-C, currents show block by different concentrations of nifedipine between 2 and 500 µM. In A-C, data are from the same cell. D: concentration-response curve for steady-state block of K+, Rb+, and Cs+ currents. Lines were fit to the data using the Hill equation. The IC50 concentrations for K+, Rb+, and Cs+ currents blocked by nifedipine were 20.2 ± 0.5, 30.4 ± 0.6, and 50.7 ± 2.0 µM, and the Hill coefficients were 1.6 ± 0.1, 1.6 ± 0.04, and 1.6 ± 0.1, respectively. Data are means ± SE (n = 4-14 cells). The dotted lines represent the concentration-response relations for wild-type channels redrawn from Fig. 2.

Less potent action of nifedipine on Kv4.2 The observation that substitution of valine or tyrosine for arginine in the outer pore mouth of Kv1.5 reduced the IC₅₀ for nifedipine action significantly to ~20 µM, which prompted us to examine other Kv channels with a valine at this position. In a Shaker B T449V mutant channel, an IC₅₀ of ~30 µM was noted (1). The Shal channel Kv4.2 also has a valine at this equivalent position, and so we tested the ability of nifedipine to block this channel (Fig. 6). The rat Kv4.2 gene was stably expressed in HEK cells, and currents were recorded in the whole cell configuration as for Kv1.5. The Kv4.2 gene encodes an A-type rapidly inactivating outward K⁺ current, and block by nifedipine can be measured as peak current reduction (as nifedipine block is of rapid onset) or as an overall reduction in charge during the pulse. When currents were integrated over time, we obtained charge records, as shown in Fig. 6B obtained from the current recordings in Fig. 6A. The concentration-response relations for peak current block and charge reduction by nifedipine are shown in Fig. 6C. Relations have similar IC₅₀ values of 32 and 29 µM for peak current and charge reduction, respectively, with Hill coefficients of 0.9. We applied the same single binding site model to the Kv4.2 data as was applied to results from Kv1.5 (Fig. 2). For Kv4.2 K⁺ currents, biexponential functions were required to fit the current decay both in the control and presence of nifedipine. As shown in Table 3, the slow component of Kv4.2 inactivation was not changed with the addition of nifedipine, suggesting that the drug did not change the inactivation rate. However, the time constant of the initial fast component was decreased depending on drug concentration, suggesting an overlap between drug block and the initial fast inactivation component. If nifedipine block and inactivation are independent, the rate constants of the initial fast-decaying phase of Kv4.2 currents (1/_decay) in the presence of nifedipine should be a sum of the rate constants of channel inactivation (1/_inactivation) and of channel block (1/_block) (26) as follows

(4)

Thus the value of 1/_block can be estimated by subtracting 1/_inactivation from 1/_decay. Assuming the inactivation process was not changed by nifedipine, we calculated 1/_block and plotted it as a function of nifedipine concentration in Fig. 6D. We fit the data with Eq. 2a and extracted k₊₁ (2.09 × 10⁶ M¹ · s¹), k₁ (57.9 s¹), and the K_d (27.7 µM) from the straight line fits. The K_d value was in good agreement with that obtained from the concentration-response relation (Fig. 6C) and suggested that the assumption of the model was correct. These Kv4.2 results support the idea that a valine (or a tyrosine in Kv1.5) rather than an arginine at the outer pore mouth can reduce the affinity of Kv channels for nifedipine.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Nifedipine block of Kv4.2. A: Kv4.2 K+ currents at +60 mV from the 80 mV holding potential in control and the presence of different nifedipine concentrations are as indicated. B: Kv4.2 charge at each concentration obtained by time integration of current records in A. C: concentration-response curve for peak Kv4.2 current and charge block by nifedipine. Lines were fit to the data using the Hill equation with IC50 and Hill coefficient values, respectively, of 28.7 ± 2.0 µM and 0.9 ± 0.1 (integrated current) and 32.2 ± 3.2 µM and 0.9 ± 0.1 (peak current). Data are means ± SE (n = 2-7 cells). D: time constants of Kv4.2 current block by nifedipine. The rate constant of channel block (1/block) at +60 mV is plotted against concentration. From the fit line, k+1 is 2.09 × 106 M1 · s1, and k1 is 57.9 s1. The Kd is 27.7 µM. Data points are means ± SE (n = 2-10 cells)

Effects of [K⁺]_o on nifedipine block Whole cell K⁺ currents during 5-s depolarizations to +40 mV are shown in Fig. 7A. The current trace in 5 mM [K⁺]_o shows a slow inactivation that is not complete at the end of a 5-s depolarizing step. The inactivation process can be fit using a double-exponential function (Eq. 3b in Data analysis). The resulting fast (₂) and slow (₁) time constants were 250 and 1,500 ms, respectively. Increasing [K⁺]_o to 135 mM slowed the inactivation process in parallel with an obvious reduction in current amplitude due to the decreased K⁺ driving force. After changing to 135 mM [K⁺]_o, the time constants of inactivation were 400 (₂) and 1,490 ms (₁), respectively, by applying the same double-exponential function. The averaged time constants ₂ and ₁ in 5 and 135 mM [K⁺]_o were 265 ± 23 and 364 ± 32 ms (P < 0.05) and 1,750 ± 141 and 1,945 ± 145 ms, respectively. Data are means ± SE from three to four experiments. These data support the conclusion that C-type inactivation was slowed somewhat by elevation of [K⁺]_o and that this effect was caused by an effect on ₂. Current traces from Fig. 7A were normalized to peak current and shown in Fig. 7B. This clearly showed that the C-type inactivation process was slower in 135 than 5 mM [K⁺]_o. In contrast with the effects of external K⁺ on C-type inactivation, activation kinetics were not significantly altered by different external K⁺ concentrations (data not shown).

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Nifedipine block and C-type inactivation of the Kv1.5 channel. A: K+ currents were elicited from 80 mV to +40 mV for 5 s. Current traces were recorded with normal external solution (5 mM K) and in the presence of 20 µM nifedipine (5 K + Nif) or with 135 mM K and in the presence of 20 µM nifedipine (135 K + Nif). Inset: traces on an enlarged time scale. B: current traces from A scaled to the peak current level to illustrate inactivation in 5 and 135 mM K solutions or block in the presence of 20 µM nifedipine under both K conditions. C: nifedipine-sensitive currents in 5 and 135 mM K solutions. Current traces were obtained from A by subtracting control currents in each K and corresponding currents in the presence of 20 µM nifedipine. D: scaled nifedipine-sensitive traces from C indicating the similar blocking rate in 5 and 135 mM K solutions.

View larger version (12K): [in this window] [in a new window]   Fig. 8.   Summary of external K+ dependence of nifedipine block of Kv1.5. A: currents during 400-ms voltage steps to +40 mV from 80 mV. Currents were recorded in 5, 135, and 300 mM K during exposure to 10 µM nifedipine, and the tracings were normalized to peak current and superimposed. B: at different concentrations of K, bars show relative block at the end of the voltage pulse. This was calculated from steady nifedipine block {1  [steady-state current in nifedipine (ss Nif)/steady-state current in control (ss Ctrl)]}. Bars are means ± SE (n = 6-14 cells), and differences were not significant at the 5% level (one-way ANOVA).

Effects of nifedipine on recovery from C-type inactivation. To study recovery from C-type inactivation in the absence and presence of nifedipine, a double-pulse voltage protocol was used. The purpose of the first pulse (prepulse) was to predominantly activate (during short prepulses) or to activate and then inactivate the channel (during long prepulses). After a variable interpulse duration, the second pulse (test pulse) was applied to test how many channels had recovered from the inactivated state induced by the prepulse. The fractional recovery was therefore defined as follows: fractional recovery = I_peak2/I_peak1, where I_peak1 and I_peak2 represent the peak currents elicited by the pre- and test pulse depolarizations, respectively. The current traces shown in Fig. 9 were elicited by double-pulse protocols with long and short pulse durations over a range of interpulse intervals. All pulses were to +40 mV. The holding potential and the potential during the interpulse intervals was 80 mV. The interval between each trace was 30 s to ensure that no inactivation accumulated between each cycle of the protocol. In Fig. 9, A and B, the currents are in response to a long 5-s prepulse with interpulse intervals of variable duration from 180 ms to 3 s in increments of 400 ms. In the absence of nifedipine (Fig. 9A), more than 50% of the current inactivated during the 5-s prepulse at +40 mV. After a brief interval, peak current amplitudes recorded during test pulses (I_peak2) were smaller than those of the first prepulse (I_peak1). I_peak2 recovered slowly with increasing interpulse intervals and reached only 85% of I_peak1 after an interpulse interval of 3 s. When 20 µM nifedipine was applied externally (Fig. 9B), both peak and steady-state currents at +40 mV were rapidly inhibited during the prepulse.

View larger version (30K): [in this window] [in a new window]   Fig. 9.   The effects of nifedipine on recovery from C-type inactivation. A and B: superimposed whole cell current traces were elicited by a double pulse to +40 mV separated by a variable interpulse interval from 180 ms to 3 s with increments of 400 ms. The duration of the prepulse was 5 s, and the test pulse was 60 ms. The interval between pairs of pulses was 30 s at 80 mV. Current traces shown were recorded in the absence (A) and presence (B) of 20 µM nifedipine. C and D: currents elicited by pairs of 60-ms steps to +40 mV separated by a variable interpulse interval from 120 to 1,920 ms in increments of 200 ms. The interval between pairs of pulses was 30 s. Current traces were recorded in control (C) and 20 µM nifedipine (D). Results in A-D were from the same cell, and protocols are illustrated below each pair. E and F: fractional recovery (Ipeak2/Ipeak1) from inactivation plotted as a function of interpulse interval. Data points in E and F were averaged from data using the same voltage protocols as in A-B and C-D, respectively. Data points represent means ± SE from 3-19 experiments. Solid lines are best fits to data points using Eq. 3a in E (control) and F (nifedipine) and Eq. 3b in E (nifedipine).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Influence of permeating ions on Kv1.5 channel block by nifedipine. It has been reported before that permeating ions can affect the efficacy of channel block by charged drugs like TEA (3, 16). In cloned Kv2.1 channels, external application of 30 mM TEA can block K⁺ currents by 87%, whereas the same external concentration of TEA has no effects on Na⁺ currents through the same channel (16). Here, we have shown that, like TEA, nifedipine also has a different potency for block of K⁺, Rb⁺, and Cs⁺ currents through another Kv channel, Kv1.5 (Figs. 1 and 2). Because TEA block is coordinated by all four K⁺ channel subunits, although Na⁺ ions are smaller than K⁺, it seems most likely that binding of K⁺ rather than Na⁺ within the permeation pathway allows the formation of the TEA or nifedipine binding sites. Alternatively, it is possible that nifedipine binds with the same affinity to channels that pass both K⁺ and Cs⁺ and that the larger cation can more easily escape the blocker and pass through the channels. The pore configuration that would allow this possibility is uncertain, and the data in Fig. 3B showed increasing block of the inward current during the switch from outward Cs⁺ to inward K⁺ current. This suggests that the nifedipine binding affinity was different in the presence of different ions.

Site of nifedipine action on Kv1.5. Nifedipine action on Kv1.5 appears to be closely tied to the ion conduction pathway. The drug does not block closed channels but rapidly blocks open channels causing a rapid current decay superimposed on slower inactivation when activation is sufficiently rapid to allow this to be seen (Fig. 1). The influence of different permeating cations on the rate and potency of the block of ionic and gating currents is further evidence for effects within the open pore. In conducting channels, when K⁺ was the permeating cation (as in Fig. 3B at +30 mV), block was more extensive than when Cs⁺ was the permeating cation. Overall, the data showed that the permeating cation was the most important determinant of block (Table 1). When ion permeation was prevented in the W472F nonconducting mutant and gating currents were measured (Fig. 4), it was shown that extracellular K⁺ allows greater block of Kv1.5 by nifedipine than other extracellular cations. Interestingly, though, in these experiments, addition of only intracellular K⁺ or Rb⁺ was able to coordinate a relatively potent block of off-gating current by nifedipine. This suggested that even intracellular ions were able to have long range effects on overall pore conformation and affect the formation of the nifedipine binding site in the outer channel mouth without being present there.

Pore mutants and their effects. Two outer pore mutants of Kv1.5, R487V and R487Y, affected the concentration dependence of nifedipine block of the channel. Low concentrations of nifedipine were much less effective on these mutant channels regardless of the ionic species, but there was a steepening of the concentration-response relationships so that at higher concentrations nifedipine apparently blocked the channel to the same degree as the wild-type channels, and the overall efficacy was unchanged (Fig. 5). The ion-dependent shift in the IC₅₀ of the concentration-response relations was maintained but reduced in relative terms, so that whereas Cs⁺ was 4 times less potent in the wild-type channels, it was 2.5 times less potent in the R487V mutant. This result suggested that the ion effects on nifedipine potency could be separated to a large degree from effects of outer pore mutations on nifedipine binding to Kv1.5 and that ion species effects were not caused by interaction directly at the nifedipine binding site. This argues against a common site for nifedipine block and ion modulation of block in Kv1.5.

Nifedipine block and C-type inactivation. The principal action of nifedipine on Kv1.5 was to accelerate the rate of current decay once the channel-opening rate significantly exceeded blocking rate. The possibility existed, because permeant cations modulated the degree of nifedipine block, that the action of nifedipine was in some way related to an acceleration of C-type inactivation. This is based on the knowledge that permeant cations, and especially extracellular cation concentrations, are effective modulators of the rate of C-type inactivation (20) and the recovery from inactivation. In addition, the residue R487 in Kv1.5 is analogous to T449 in Shaker channels in that both are important determinants of the inactivation rate. In Kv1.5, though, the role for this residue is less critical (8), and, as discussed below, quite large changes in [K⁺]_o have less pronounced effects on the inactivation rate than in Shaker channels. Modulation of C-type inactivation by channel blockers has been extensively studied by internal or external applications of TEA, where external TEA competes for C-type inactivation (6, 11). Interactions between C-type inactivation and channel blockers are manifested in many ways. In Kv1.3 channels in T lymphocytes, slowed C-type inactivation via exposure to 160 mM [K⁺]_o greatly diminished the potency of CP-339,818, a channel blocker of Kv1.3 (22), suggesting that CP-339,818 preferentially bound to the inactivated state of Kv1.3. In mutant ShB 6-46 expressed in Xenopus oocytes, nifedipine block was found to be insensitive to [K⁺]_o (1), suggesting that it could be separated from C-type inactivation.