Unexpected and Differential Effects of Cl− Channel Blockers on the Kv4.3 and Kv4.2 K+ Channels
Implications for the Study of the Ito2 Current
Abstract The Kv4.3 K+ channel is thought to underlie the Ca2+-insensitive transient outward current (Ito1) in ventricular myocytes of canine and human heart and to contribute to the Ito1 in rat myocytes. It has been suggested that there is a second component of the transient outward current in some species that is contributed by a Ca2+-activated Cl− current (known as Ito2). The evidence for the existence of the Ito2 current is based, in part, on the pharmacological effects of various Cl− channel blockers. To test for possible interactions between these compounds and Ito1, the effect of several different Cl− channel blockers on the Kv4.3 channel was examined. The fenamates (niflumic and flufenamic acid) were found to have large effects on the position of the steady state inactivation curve of the Kv4.3 channel. The disulfonic stilbenes (DIDS and SITS) had markedly different effects and were found to greatly reduce the rate of recovery from inactivation of the Kv4.3 channel without large changes in the position of the activation and steady state inactivation curves. Both classes of drugs produced an apparent blockade of the Kv4.3 channel under some recording conditions. Surprisingly, the closely related Kv4.2 channel was found to be markedly less sensitive to these drugs. Caffeine was found to block both the Kv4.3 and Kv4.2 channels to a similar extent. These nonspecific drug effects have implications for the study of the two components of the transient outward current and suggest that purely pharmacological criteria cannot be used to define the physiological role of Ito2.
The transient outward current in cardiac myocytes is thought to comprise two components.1 2 3 One component (known as Ito1) is a rapidly activating and inactivating voltage-gated current that is carried by K+ ions and is sensitive to blockade by 4-aminopyridine (4-AP).2 3 The second component (known as Ito2) is a Ca2+-activated4 Cl− current,5 6 7 which is typically much smaller than Ito1. The relative contribution of these two currents to the transient outward current has been a matter of some debate, as has the physiological importance of Ito2.1 2 3 7 8 One strategy that has been used to address the role of Ito2 in the heart is to examine the effect of Cl− channel blockers on the shape of the action potential waveform.9 The basic assumption of this strategy is that the Cl− channel blockers used are specific for Cl− channels and do not affect other channels. This assumption is not always easily tested in cardiac myocytes, where multiple currents overlap in time and voltage activation ranges.
It has recently been suggested that the bulk of the 4-AP–sensitive outward current (Ito1) in canine and human heart is carried by a channel encoded by the Kv4.3 gene.10 In rat heart, both the Kv4.2 and Kv4.3 channels are thought to contribute to Ito1.10 11 12 It is now possible, therefore, to study the pharmacological properties of the channels that underlie the Ito1 in isolation, using a heterologous expression system.
In the present study, we have examined the effects of several common Cl− channel blockers on the Kv4.2 and Kv4.3 channels. Two chemically distinct classes of compounds were tested: fenamates (niflumic and flufenamic acid) and disulfonic stilbenes (DIDS and SITS). It was found that these Cl− channel blockers can produce large changes in the kinetic properties of Kv4 K+ channels, which can mimic channel blockade under some recording conditions. Surprisingly, we also find that the sensitivity of two closely related channels, Kv4.2 and Kv4.3, to these drugs is markedly different. These unexpected drug effects may complicate the interpretation of experiments designed to determine the contribution of Cl− channels to the shaping of the cardiac action potential.
Materials and Methods
Preparation of Kv4.2 and Kv4.3 cRNA
The full-length rat Kv4.310 and Kv4.2 (rShal13 ) cDNA clones were subcloned into the Bluescript vector (Stratagene). The plasmids were linearized, and cRNA transcripts were transcribed using T7 RNA polymerase.
Expression in Xenopus Oocytes
Oocytes were prepared from mature female Xenopus laevis using established procedures.14 Defolliculation was performed by incubation for 2 hours in 2 mg/mL collagenase (type VIII, Sigma Chemical Co) in Ca2+-free OR2 oocyte medium with gentle agitation. Oocytes were stored in OR3 solution (50% L-15 medium [GIBCO], 1 mmol/L glutamine, 15 mmol/L sodium HEPES [pH 7.6], and 0.1 mg/mL gentamicin) at 18°C. Oocytes were injected with 50 nL of Kv4.2 or Kv4.3 cRNA (≈0.3 ng/nL) using a microdispenser (Drummond) and a micropipette with tip diameter of 10 to 20 μm. Injected oocytes were incubated at 18°C for 24 to 48 hours before analysis.
Oocytes were voltage-clamped using a two-microelectrode voltage clamp (Axoclamp 2A, Axon Instruments). Intracellular electrodes filled with 3 mol/L KCl with resistances of 0.5 to 3 MΩ were used. The standard extracellular recording solution (OR2) contained 80 mmol/L NaCl, 5 mmol/L KCl, 1.8 mmol/L CaCl2, 1 mmol/L MgCl2, and 5 mmol/L sodium HEPES (pH 7.6). Data collection and analysis were performed using Pclamp software (Axon Instruments), and linear leakage currents were digitally subtracted. For the calculation of the K+ reversal potential, an intracellular K+ concentration of 100 mmol/L was assumed.15
Niflumic acid and flufenamic acid were dissolved in ethyl alcohol as 0.1 mol/L stock solutions and were added to the control solution to give final concentrations. DIDS and SITS were prepared in dimethyl sulfoxide as 0.1 mol/L stock solutions. DIDS and SITS were prepared and used in a darkened room. All recording solutions containing drugs were checked and adjusted to pH 7.6 if necessary. Properties of the Kv4.3 or Kv4.2 channels were not affected by control solutions containing up to 1% dimethyl sulfoxide or 1% ethyl alcohol (data not shown). Drugs were obtained from Sigma.
Comparison of activation curves in the presence and absence of a drug was complicated by the fact that the conductance-voltage curves never completely plateaued over the voltage range studied, making it impossible to accurately measure the maximal conductance. For this reason, activation curves were compared by measuring the distance between the curves at several different current levels (typically four) in the steepest part of the curve and then taking the average of these values as a measure of the shift in the activation curve.
Group data are presented as mean±SEM, unless otherwise stated. Statistical tests of drug effects were performed using paired two-tailed Student’s t tests. A t value giving P<.01 was considered to be significant.
Voltage-clamp recordings were initially made from Xenopus oocytes injected with cRNA encoding the Kv4.3 channel. The basic biophysical and pharmacological properties of the Kv4.3 channel have been described previously.10 It is a rapidly activating and inactivating K+ current that is sensitive to blockade by 4-AP and the antiarrhythmic drug flecainide. In an effort to further characterize the pharmacological properties of this channel, we have studied the effects of several common Cl− channel blockers, which have been used previously to examine the properties or physiological function of the cardiac Cl− current, Ito2.
Effect of Niflumic and Flufenamic Acid
An unanticipated initial observation was that application of 100 μmol/L niflumic acid by bath perfusion caused a large reduction in the Kv4.3 current amplitude elicited in response to depolarizing steps from a holding potential of −60 mV (Fig 1A⇓). The related compound, flufenamic acid (100 μmol/L), had very similar effects on the Kv4.3 current (data not shown).
We tested the possibility, although unlikely, that the Kv4.3 current was contaminated by a Cl− conductance by studying its ionic selectivity. In 5 mmol/L external K+, the mean reversal potential of the Kv4.3 current was −74.4±0.7 mV (n=5) (Fig 1B⇑). The reversal potential shifted with changes in [K+]o (Fig 1C⇑), and the mean value of this shift in reversal potential per 10-fold change in [K+]o was 56±1 mV (n=3). These values are in good agreement with predictions from the Nernst equation, suggesting that the Kv4.3 channel is a highly selective K+ channel, as would be expected from the high similarity of the pore sequence to other members of the K+ channel family.10 This result also demonstrates that the current records are not significantly contaminated by the presence of a Cl− conductance endogenous to the oocyte.
Further investigation revealed that the apparent blocking effect of niflumic acid was due to a large shift in the steady state inactivation curve of the Kv4.3 current, resulting in increased steady state inactivation of the current at a holding potential of −60 mV (Fig 2A⇓). Application of 100 μmol/L niflumic acid shifted the midpoint for steady state inactivation to a more negative potential by 9.6±0.4 mV, from −58±0.4 mV to −68± 0.6 mV (P<.01, n=15). The activation curve of the Kv4.3 current was also affected by niflumic acid (Fig 2B⇓). The activation curve was shifted more negative by an average of 9.7±0.9 mV (P<.01, n=9) in the presence of 100 μmol/L niflumic acid (Fig 2B⇓).
Because both the Kv4.3 and the Kv4.2 channels contribute to the Ito1 in rat,10 we also tested the effects of niflumic acid on the Kv4.2 channel, using identical procedures to those used for the Kv4.3 channel. Surprisingly, the effect of niflumic acid was significantly smaller on the Kv4.2 channel than on the Kv4.3 channel (Fig 2C⇑ and 2D⇑). The midpoint for steady state inactivation of the Kv4.2 current was shifted only 3.4 ±0.2 mV (P<.01, n=7) after application of 100 μmol/L niflumic acid, from −68±0.5 to −71±0.4 mV. The effect of niflumic acid on the activation curve of the Kv4.2 current (Fig 2D⇑) was also relatively small, with an average shift of 3.3±0.8 mV (P<.02, n=5).
The dose-response relationship for the shift in steady state inactivation curves of the Kv4.2 and Kv4.3 currents by niflumic acid is shown in Fig 3⇓. The maximum shift for the Kv4.3 current was 24.4 mV, with a Kd of 160 μmol/L. The equivalent values for the Kv4.2 current were 9.1 mV and 90 μmol/L, respectively. Shifts were observed at drug concentrations as low as 30 μmol/L.
Niflumic acid also affected the rate of recovery from inactivation of the Kv4.3 channel (Fig 4⇓). The time constant for recovery from inactivation of the Kv4.3 current was increased 1.7-fold in the presence of 100 μmol/L niflumic acid, from 145±3 to 248±8 ms (P<.01, n=6). In marked contrast, the time course of recovery of the Kv4.2 current was unaffected by niflumic acid (Fig 4⇓).
Effect of DIDS and SITS
Two other commonly used Cl− channel blockers, DIDS and SITS, also modified the kinetic properties of the Kv4.2 and Kv4.3 channels. Application of 100 μmol/L DIDS had a dramatic effect on the peak Kv4.3 current during repeated voltage steps to 0 mV (Fig 5A⇓). In the presence of the drug, the peak Kv4.3 current declined during repeated pulse protocols at stimulation rates as slow as 0.2 Hz (Fig 5B⇓). The related compound, SITS, produced similar effects on the Kv4.3 channel at equivalent drug concentrations (data not shown).
The time constant for recovery from inactivation of the Kv4.3 channels increased 3.3-fold in the presence of 100 μmol/L DIDS (Fig 6A⇓), from 137±5 to 458±27 ms (P<.01, n=7). The effect on the Kv4.2 channels was smaller (Fig 6B⇓), with the time constant for recovery changing from 150±13 to 228±15 ms (P<.01, n=5). A dose-response analysis of the effect of DIDS on the recovery rate of the Kv4.3 channel yielded a Kd of 25 μmol/L with a 4.0-fold maximum increase (Fig 6D⇓). Significant changes in recovery times were seen at drug concentrations as low as 10 μmol/L.
DIDS affected other kinetic properties of both channels, although the effects were generally less dramatic. Application of 1 mmol/L DIDS produced shifts in the steady state inactivation curves of the Kv4.2 and Kv4.3 currents in the positive direction by 5.4±1.3 mV (n=3) and 3.3±0.5 mV (P<.01, n=4), respectively (Fig 7A⇓). This shift was in the direction opposite that observed after the application of niflumic acid. DIDS also increased the peak conductance of the Kv4.3 and Kv4.2 currents without affecting the threshold for activation (Fig 7B⇓). The increase in peak conductance produced by 1 mmol/L DIDS may be secondary to a decline in the inactivation rate (Fig 7C⇓ and 7D⇓), because there was no apparent affect on the time course of activation. At −25 mV, the half-time of decay was increased from 101±2 to 141±7 ms for the Kv4.3 channels (P<.02, n=4) and from 39±1 to 55±4 ms for the Kv4.2 channels (n=3).
It has previously been reported that 2 mmol/L SITS produces a significant decrease in the size of the transient outward current in rat ventricular myocytes, a result that was interpreted to demonstrate that a large fraction of the transient outward current in rat myocytes was a Cl− current.16 This previous observation appears to be at variance with our results showing an increase in peak conductance of the Kv4.2 and Kv4.3 channels in the presence of DIDS (Fig 7⇑). An explanation for this apparent discrepancy can be found, however, by examining the recovery rate of the Kv4.3 channel in the presence of 2 mmol/L SITS (Fig 8A⇓). This concentration of SITS produced a dramatic decrease in the recovery rate. When stimulation frequencies (0.1 Hz) and voltage-clamp protocols identical to those used in the rat myocyte study were used,16 there was a significant reduction in the peak current in the presence of 2 mmol/L SITS (Fig 8B⇓). This reduction was similar in magnitude to that produced by SITS on the transient outward current in rat myocytes.16 The fact that SITS does not act by a channel-blocking mechanism can be shown, however, by applying the same voltage-clamp protocol at very slow stimulation rates (0.03 Hz), a frequency at which there is a slight increase in peak current (Fig 8B⇓). These results obviously call into question the conclusion of this previous study that the transient outward current in rat myocytes contains a significant Cl− current component.
Effect of Caffeine
In previous studies, it has been shown that 10 mmol/L caffeine can block Ito2 in cardiac myocytes.5 9 For this reason, the effect of this concentration of caffeine on the Kv4.2 and Kv4.3 channels was tested (Fig 9A⇓). After application of 10 mmol/L caffeine, there was a 46±2% decrease in the average peak Kv4.2 current (P<.01, n=4) and a 40±2% decrease in the Kv4.3 current (P<.001, n=6). No significant effects of caffeine on the positions of the activation curve (Fig 9B⇓) or the steady state inactivation curve or on the recovery from inactivation were observed (data not shown). It has previously been reported that caffeine can directly block Ito1 under recording conditions in which Ito2 was not present.8 17 The blockade of the Kv4.2 and Kv4.3 channels by caffeine lends support to this conclusion.
It is conceivable that caffeine reduces the peak Kv4.2 and Kv4.3 currents via a mechanism other than channel block. One possibility would be inhibition of phosphodiesterases and the activation of A kinase, resulting in the phosphorylation of the channel protein. We tested this possibility by incubating the oocytes with a membrane-permeant analogue of cAMP to directly activate A kinase. No effect on peak current amplitude was found after incubation for up to 30 minutes with 1 mmol/L dibutyryl cAMP (data not shown). Application of caffeine can also increase the levels of intracellular Ca2+ in many cell types. There was, however, no evidence that significant Ca2+ release occurred in the oocytes after application of caffeine, because the drug did not activate the endogenous Ca2+-activated Cl− channel, which is a sensitive indicator of changes in internal Ca2+ levels in these cells.
Current flow through Kv4.3 K+ channels appears to underlie a significant fraction of Ito1 in human and canine ventricular myocytes and contributes to Ito1 in rat ventricular myocytes.10 In the present study, we report that several Cl− channel blockers have unexpectedly large effects on the kinetic properties of the Kv4.3 channel. This is an interesting finding because there has been a long running debate regarding the functional roles and/or existence of the two components of the transient outward current: the 4-AP–sensitive K+ current (Ito1) and the Ca2+-activated Cl− current (Ito2).1 2 7 8 Pharmacological agents that can be used to reliably block either one of these two components are of some interest in addressing questions regarding the relative physiological importance of these two currents in controlling phase-1 action potential repolarization. The results reported in the present study suggest that both the fenamates, niflumic and flufenamic acid, and the disulfonic stilbenes, DIDS and SITS, present some significant problems in studies designed to compare the relative roles of Ito1 and Ito2 because these drugs significantly modify the kinetic properties of the Kv4.3 channel. These unintended nonspecific effects could greatly complicate the interpretation of results obtained in experiments in which these drugs were assumed to act solely by blockade of Cl− channels.
Because of the very steep steady state inactivation curve of the Kv4.3/Kv4.2 channels, even small changes in the position of the steady state inactivation curve can have quite large effects on the number of channels available for activation in voltage ranges that influence the shape of the cardiac action potential. Similarly, a decrease in the rate of recovery from inactivation can significantly reduce the number of channels available for activation, depending upon stimulation frequencies. A decrease in the recovery rate of Ito1 channels could account for a previous report of changes in action potential waveform that occurred after the application of DIDS to block Ito2 in rabbit atrial myocytes.9 The results reported here suggest that the use of Cl− channel blockers to study the contribution of Cl− channels to cardiac action potential repolarization should be carefully monitored to assess potential nonspecific effects of these agents on voltage-gated K+ channels.
A second confounding factor in studies of Ito1 and Ito2 is the complicated mechanism of action of 4-AP on the Ito1 channel. This drug is included in the recording solutions used in many studies of Ito2. Unfortunately, considerable unblocking of the Ito1 channel by 4-AP can occur at positive potentials when medium to fast stimulation frequencies are used; a process known as reverse use dependence.18 19 This occurs because the drug has a low binding affinity for open or inactivated channels and only binds effectively to the closed states of the channel.19 A similar reverse use dependence of 4-AP block is seen for the Kv4.3 and Kv4.2 channels (Reference 2020 and H.S. Wang and D. McKinnon, unpublished data, 1997). This phenomenon results in the appearance of an apparently 4-AP–insensitive current in cells that do not possess Ito2.8 Based on the results presented in the present study, it is possible that Cl− channel blockers can produce an apparent blockade of this unblocked fraction of Ito1 channels by modifying the kinetic properties of the Kv4.3 channel. It is also likely that caffeine can block this apparently 4-AP–insensitive component, given that caffeine blocks both the native Ito1 channel8 and the Kv4.3 channel. It is possible, therefore, that some previous reports on the presence and properties of Ito2 may be compromised by the less than ideal nature of the actions of 4-AP, Cl− channel blockers, and caffeine on the Kv4 channels that underlie Ito1, particularly those studies that used low concentrations of 4-AP (<3 mmol/L). This is not to imply that Ito2 is not present in some cells; the evidence for this current is quite strong in rabbit atrial myocytes.9 The existence of Ito2 in different types of myocytes is, however, very variable. Many types of myocytes may not possess this current at all, and even identical classes of cardiac myocytes, such as atrial myocytes, from different species may differ significantly in their expression of Ito2.8 It would seem that considerable caution is required in defining Ito2 in any given system using pharmacological criteria because of the inadequacy of the pharmacological tools currently available.
The mode of action of either the fenamates or the disulfonic stilbenes on the Kv4.3 K+ channel is uncertain. These compounds appear to function as allosteric modifiers of channel gating, presumably by binding directly to the channel. Since these drugs are amphipathic molecules, it is likely that they interact with hydrophobic residues of the channel. There is a high degree of structural similarity between the Kv4.3 and Kv4.2 channels; the Kv4.3 channel is 75% identical to the Kv4.2 channel at the amino acid level.10 Given this structural conservation, it was surprising to find that the disulfonic stilbenes had significantly greater effects on the Kv4.3 channels than on the Kv4.2 channels. The hydrophobic domains of these channels are generally the most highly conserved regions, making the isoform specificity even less expected.
Although niflumic and flufenamic acid have been shown to be effective blockers of several kinds of Cl− channels,21 22 23 their usefulness as pharmacological tools may be greatly limited by their nonspecific effects on several different classes of ion channels in addition to Cl− channels. Fenamates have been shown to modify the kinetics of the hyperpolarization-activated current found in the sinoatrial node.24 These drugs have been shown to activate a K+ conductance in smooth muscle cells of the gut25 and a large-conductance Ca2+-activated K+ channel.26 27 Inhibition of nonselective cation channels has also been reported.28 29 These results, together with our observations on the Kv4 K+ channels, suggest that the fenamates, as a class of agents, have a broad spectrum of nonspecific actions on ion channel proteins.
DIDS and SITS had a different set of effects on the Kv4.3 channel than did the fenamates. There was a much larger increase in the time constant for recovery from inactivation, and the shift in the steady state inactivation curve was relatively small and in the opposite direction. The peak current and time constant of inactivation were also increased. Similar to the fen-amates, the effects were isoform selective. The increase in the time course of recovery in the presence of DIDS was much greater for the Kv4.3 than the Kv4.2 channel. The effects of DIDS on the other kinetic properties of the two channel types were relatively similar. Both DIDS and SITS have been previously reported to modify the kinetics of voltage-gated K+ currents in squid axon.30
One commonly used preparation for the expression and study of voltage-gated K+ channel cDNA clones is the Xenopus oocyte system.15 These cells can express an endogenous Ca2+-activated Cl− channel,31 although the level of expression is quite variable. In our cell preparations, this current was not significantly activated in experiments using voltage-step protocols to activate the expressed channels, but other laboratories have reported significant Cl− currents that have had to be eliminated with Cl− channel blockers.21 Interactions between commonly used Cl− channel blockers and heterologously expressed voltage-gated K+ channels have not been reported previously, although this has not been tested systematically to our knowledge.
It is very unlikely that the effects of Cl− channel blockers that we observe on the Kv4 channels are due instead to an effect on the endogenous Cl− channel. There are several observations that argue against this possibility. First, the voltage-clamp protocols used in these experiments did not activate a significant Cl− conductance in uninjected oocytes. Second, the current induced by injection of Kv4.2/Kv4.3 cRNA was highly selective for K+ ions. Third, different classes of Cl− channel blockers had distinct effects: niflumic and flufenamic acid shifted the steady state inactivation curve to the left, whereas SITS and DIDS produced a small shift to the right.
In conclusion, at least two chemically distinct classes of Cl− channel blockers can produce significant changes in the kinetic properties of the Kv4.3 channel and, to a lesser extent, the Kv4.2 channel, giving the appearance of channel blockade under some recording conditions. These effects have implications regarding the identity of the channels that underlie the two components of the transient outward current that have been described in cardiac myocytes.
This study was supported by grants NS-29755 and NS-01718 from the National Institutes of Health and a grant from the American Heart Association, New York State Affiliate.
- Received February 14, 1997.
- Accepted August 1, 1997.
- © 1997 American Heart Association, Inc.
Kenyon JL, Gibbons WR. 4-Aminopyridine and the early outward current of sheep cardiac Purkinje fibers. J Gen Physiol. 1979;73:139-157.
Coraboeuf E, Carmeliet E. Existence of two transient outward currents in sheep cardiac Purkinje fibers. Pflugers Arch. 1982;392:353-359.
Tseng G-N, Hoffman BF. Two components of transient outward current in canine ventricular myocytes. Circ Res. 1989;64:633-647.
Zygmunt AC, Gibbons WR. Calcium-activated chloride current in rabbit ventricular myocytes. Circ Res. 1991;68:424-437.
Zygmunt AC. Intracellular calcium activates a chloride current in canine ventricular myocytes. Am J Physiol. 1994;267:H1984-H1995.
Li G-R, Feng J, Wang Z, Fermini B, Nattel S. Comparative mechanisms of 4-aminopyridine-resistant Ito in human and rabbit atrial myocytes. Am J Physiol. 1995;269:H463-H472.
Wang Z, Fermini B, Feng J, Nattel S. Role of chloride current in repolarizing rabbit atrial myocytes. Am J Physiol. 1995;268:H1992-H2002.
Dixon JE, Shi W, Wang H-S, McDonald C, Yu H, Wymore RS, Cohen IS, McKinnon D. Role of the Kv4.3 K+ channel in ventricular muscle: a molecular correlate for the transient outward current. Circ Res. 1996;79:659-668.
Dixon JE, McKinnon D. Quantitative analysis of potassium channel mRNA expression in atrial and ventricular muscle of rats. Circ Res. 1994;75:252-260.
Barry DM, Trimmer JS, Merlie JP, Nerbonne JM. Differential expression of voltage-gated K+ channel subunits in adult rat heart: relationship to functional K+ channels? Circ Res. 1995;77:361-369.
Colman A. Translation of eukaryotic messenger RNA in Xenopus oocytes. In: Hames BD, Higgins SJ, eds. Transcription and Translation. Oxford, UK: IRL Press; 1984:271-302.
Campbell DL, Qu Y, Rasmusson RL, Strauss HC. The calcium-independent transient outward current in isolated ferret right ventricular myocytes. J Gen Physiol. 1993;101:603-626.
Tseng GN, Jiang M, Yao JA. Reverse use dependence of Kv4.2 blockade by 4-aminopyridine. J Pharmacol Exp Ther. 1996;279:865-876.
White MM, Aylwin M. Niflumic and flufenamic acids are potent reversible blockers of Ca2+-activated channels in Xenopus oocytes. Mol Pharmacol. 1990;37:720-724.
McCarty NA, McDonough S, Cohen BN, Riordan JR, Davidson N, Lester HA. Voltage-dependent block of the cystic fibrosis transmembrane conductance regulator Cl− channel by two closely related arylaminobenzoates. J Gen Physiol. 1993;102:1-23.
Collier ML, Levesque PC, Kenyon JL, Hume JR. Unitary Cl− channels activated by cytoplasmic Ca2+ in canine ventricular myocytes. Circ Res. 1996;78:936-944.
Gribkoff VK, Lum-Ragan JT, Boissard CG, Post-Munson DJ, Meanwell NA, Starrett JE, Kozlowski ES, Romine JL, Trojnacki JT, McKay MC, Zhong J, Dworetzky SI. Effects of channel modulators on cloned large-conductance calcium-activated potassium channels. Mol Pharmacol. 1996;50:206-217.
Inoue I. Modification of K conductance of the squid axon by SITS. J Gen Physiol. 1986;88:507-520.