Centre for Experimental Surgery and Anaesthesiology (R.M.,
S.V., K.M.) and Laboratory of Experimental Cardiology (K.R.S.),
University of Leuven, Leuven, Belgium
2-Methyl-3-(3,5-diiodo-4-carboxymethoxybenzyl)benzofuran (KB130015
or KB) is a new drug, structurally related to amiodarone and to thyroid
hormones. Its effects on cardiac voltage-dependent Na+
current (INa) were studied in pig single
ventricular myocytes at 22°C using the whole-cell (with
[Na+]i = [Na+]o = 10 mM) and cell-attached
patch-clamp techniques. KB markedly slowed
INa inactivation, due to the development of
a slow-inactivating component (
slow 
50 ms) at
the expense of the normal, fast-inactivating component
(fast 2-3 ms). The effect was
concentration-dependent, with a half-maximally effective concentration
(K0.5) of 2.1 µM. KB also slowed the
recovery from inactivation and shifted the voltage-dependent
inactivation (
V0.5 = 
15 mV;
K0.5 
6.9 µM) and activation to
more negative potentials. Intracellular cell dialysis with 10 µM KB
had marginal or no effect on inactivation and did not prevent the
effect of extracellularly applied drug. In cell-attached patches,
extracellular KB prolonged Na+ channel opening. Amiodarone
(10 µM) and 10 µM 3,5,-diiodo-L-thyropropionic acid had
no effect on inactivation and did not prevent KB effects. 3,3',5-Triodo-L-thyronine (T3) also had no
effect on inactivation, but at 10 µM it increased
INa amplitude and partially prevented the
slowing of inactivation by KB. These data suggest the existence of a
binding site for KB and T3 on Na+ channels.
 |
Introduction |
Voltage-dependent
Na+ channels play an important role in the
initiation and conduction of electrical signals in excitable cells. A
dysfunction of Na+ channels, as may occur after
genetic mutation or as a consequence of drug action, is the basis for a
variety of cardiac arrhythmias (e.g., those related to the long QT and
the Brugada syndromes), skeletal muscle diseases, and epileptic
seizures (for reviews, see Fozzard and Hanck, 1996
; Catterall, 2000;
Balser, 2001; Goldin, 2001). On the other hand,
Na+ channels are targets on which a variety of
substances (hormones, neurotransmitters, and drugs) act to modify
cellular function. Therefore, drugs acting on Na+
channels are widely used as antiarrhythmic agents, muscle relaxants, or antiepileptics.
Among antiarrhythmic agents, amiodarone has proven to be the most
effective drug against ventricular arrhythmias (Kodama et al., 1999).
However, given the many undesirable side effects of the drug, there has
been a constant interest in synthesizing new molecules that can retain
its beneficial effects but be devoid of the side actions.
2-Methyl-3-(3,5- diiodo-4-carboxymethoxybenzyl)benzofuran (KB130015, hereafter
called KB) is a new drug whose structure is related to that of
amiodarone and thyroid hormones. KB is reported to antagonize the
action of 3,3',5-triiodo-L-thyronine
(T3) by binding onto thyroid hormone nuclear
receptors (Carlsson et al., 2002). Given the known actions of
amiodarone on many ion channels, the major target channels being
delayed rectifier K+ channels and
voltage-dependent Na+ channels (Carmeliet and
Mubagwa, 1998), KB may be expected have similar electrophysiological
effects. KB has been proposed as a potential antiarrhythmic agent,
possibly devoid of some side effects of its congener (Carlsson et al.,
2002). In the present study, we report marked acute effects of KB on
the inactivation of Na+ channels. These effects
were unexpected based on those known for amiodarone (Follmer et al.,
1987; Kohlhardt and Fichtner, 1988; Kodama et al., 1999; Maltsev et
al., 2001), but a few studies have reported similar effects with high
concentrations of thyroid hormones (Craelius et al., 1990; Harris et
al., 1991; Dudley and Baumgarten, 1993).
| |
Materials and Methods |
The measurements were performed on pig isolated single
ventricular myocytes. The study has been carried out in accordance with the Declaration of Helsinki and with the institutional guidelines for the care and use of laboratory animals.
Preparation of Pig Ventricular Myocytes.
The methods used
for the dissociation of pig cells have been described previously
(Stankovicova et al., 2000; Macianskiene et al., 2002). In short, a
piece of the left ventricular wall was excised with its supplying
artery, and cannulated and perfused for 30 min at 37°C and at
constant flow with an oxygenated Ca2+-free
Tyrode's solution, followed by a 20-min perfusion with a Ca2+-free Tyrode's solution containing 0.1 mg
ml
1 protease (type XIV; Sigma-Aldrich, St.
Louis, MO) and 1 mg ml1 collagenase (type A;
Roche Diagnostics, Mannheim, Germany). After a 15-min washing
perfusion with a 0.18 mM Ca2+ Tyrode's
solution, the tissue was removed from the perfusion and cut into small
pieces. Cells were dispersed by gentle mechanical agitation and were
stored at room temperature (21-22°C).
Electrophysiological Recordings in Myocytes.
Membrane
currents were measured using whole-cell or cell-attached configurations
of the patch-clamp technique (Hamill et al., 1981). Heat-polished
borosilicate glass electrodes (horizontal puller; Zeitz Instrumente,
Munich, Germany), with tip resistances of 1 to 1.5 M
when filled
with the internal solution, were connected to an Axopatch 200A or 200B
amplifier (Axon Instruments, Union City, CA), and an analog-to-digital
interface controlled by the pClamp software (Axon Instruments) was used
to generate command pulses and acquire data. All experiments were
carried out at room temperature.
Capacitance and series resistance were partially compensated.
Whole-cell currents were generated by step depolarizations given from a
holding potential (VH) of 80 or
120 mV. Repetitive depolarizations to 30 mV were given every 1 s when changing from one solution to another. Upon reaching steady
state in a given solution, depolarizations to various levels were given
every 5 s to obtain current-voltage relationships or inactivation
curves, or double pulses with varying intervals were given to 30 mV
to obtain the recovery from inactivation. For steady-state
inactivation, prepulses lasting 1 s were given to potentials
between 120 and +40 mV (in 5-mV steps) before depolarizing to a test
potential of 30 mV, and Na+ current
(INa) after a given prepulse was
normalized relative to the maximum current (induced by depolarizing
directly from 120 mV). Single channel currents were recorded with
cell-attached electrodes coated with Sylgard 184 (Dow Corning,
Wiesbaden, Germany), whereas the intracellular potential was set close
to zero using a 150 mM [K+] extracellular
solution (see composition below). The pipette potential was held at
+120 mV (relative to ground), and patch depolarization steps lasting 60 ms (for control measurements) or 400 ms (for measurements with KB) were
given every 3 s. Current signals were filtered at either 2 kHz
(for whole-cell currents) or at 5 to 10 kHz (for single channel
currents) and were digitized at 10 or 20 kHz, respectively.
Using low [Na+]o and
working at room temperature should reduce
INa amplitude, making possible a study
of this current under whole-cell patch clamp. Despite possibly
persisting limitations in the quality of voltage control, the drug
effects on the kinetics of whole-cell
INa inactivation (see
Results) were independent of the changes in peak
INa amplitude or access resistance,
indicating that they could not be attributed to a deteriorating voltage
clamp. In addition, qualitatively similar results could be derived from experiments in the cell-attached patch-clamp mode, where voltage control is ideal.
Data were analyzed using Clampfit (Axon Instruments), Winascd (Prof. G. Droogmans, University of Leuven, Leuven, Belgium) or Origin (Microcal,
Northampton, MA). Normalized inactivation (or availability) curves were
fitted using one single Boltzmann distribution function:
where V0.5 is the potential of
half-maximum inactivation and k is the slope factor of the
distribution curve. Concentration-effect relationships were fitted by
the Hill equation:
where D is the drug concentration,
K0.5 is the half-maximum effective
concentration, and nH is the Hill
coefficient. The time course of INa
inactivation was fitted by one exponential or, when necessary, by a sum
of two exponentials:
where Af and
As are the amplitudes of the fast and
slow decaying components, respectively, and
fast and slow are
their respective time constants. Average data are given as mean ± S.E.M. Statistical comparison was made using a two-tailed t test.
Solutions and Drugs.
The composition of the standard
Tyrode's solution used during cell isolation was 150 mM NaCl, 5.4 mM
KCl, 1.8 mM CaCl2, 0.9 mM
MgCl2, 0.33 mM
NaH2PO4, 10 mM HEPES, and
10 mM glucose; pH was adjusted to 7.4 with NaOH. The solution was
bubbled with 100% O2. During measurements of
whole-cell INa, the myocytes were
superfused with a K+-free Tyrode's solution
containing 10 mM NaCl, 138 mM CsCl, 2.6 mM MgCl2,
0.1 mM CaCl2, 10 mM HEPES, and 10 mM glucose (pH
adjusted to 7.4 with CsOH). Low
[Na+]o was used to
decrease INa to levels (<10 nA) that
are compatible with maintained voltage control. Low
[Ca2+]o was used to
suppress ICa, whereas Mg2+
was increased to prevent opening of nonselective channels by the low
[Ca2+]o (Mubagwa et al.,
1997; Macianskiene et al., 2001). The internal solution contained 10 mM
NaCl, 120 mM Cs-glutamate, 20 mM tetraethylammonium-Cl, 5 mM
MgATP, 0.1 mM Na2GTP, 1 mM EGTA, and 5 mM HEPES
(pH adjusted to 7.25 with CsOH). During measurements of unitary
Na+ currents, the pipette was filled with a
solution containing 150 mM Na-glutamate, 2.7 mM
MgCl2, and 10 mM HEPES (pH was adjusted to 7.4 with NaOH), and the cells were superfused with a depolarizing solution
containing150 mM KCl, 10 mM glucose, and 10 mM HEPES (pH adjusted to
7.4 with KOH).
KB (free acid) was from Karo-Bio AB (Huddinge, Sweden). All other drugs
and products were from Sigma Chemie (Bornem, Belgium). Stock solutions
(10-100 mM) of KB, amiodarone, or
3,5-diiodo-L-thyropropionic acid (DITPA) were prepared in
DMSO. Stock solutions of T3 were prepared in
either DMSO or NaOH. The highest DMSO concentration (0.1%, v/v) was
without any effect of its own on either the
INa amplitude (98 ± 3% of its
basal level after 10 min) or the inactivation kinetics (2.0 ± 0.2 versus 1.9 ± 0.2 ms; basal versus vehicle; ncells = 6; Ogura et al., 1995).
| |
Results |
KB Slows Na+ Channel Inactivation.
Figure
1 illustrates typical effects of KB on
INa recorded during step
depolarizations from a VH of 120 mV.
Under control conditions (i.e., in the absence of drug; Fig. 1A, left),
INa inactivated rapidly after reaching
the peak level. After 5 min of exposure of the same cell to 10 µM KB,
there was a marked slowing of inactivation (Fig. 1A, right).
Superimposing INa recordings during
long pulses at 30 mV in the absence or in the presence of KB (Fig.
1B) clearly shows that the INa decay
in KB was remarkably slow but complete, with no maintained current. The
inactivation of control INa in most
cells could be fitted satisfactorily with one exponential, with a time
constant () of 2.1 ± 0.7 ms
(ncells = 45) at 30 mV, but in 13%
of the cells two exponentials (fast = 2.1 ± 0.2 ms, slow = 20.8 ± 2.0 ms;
ncells = 7) were needed. In the latter
cases, the slow component accounted for a minor part (8 ± 3.9%)
of total INa. With 10 µM KB, the
INa inactivation was dominated by a
slowly decaying component (slow = 47.8 ± 3.6 ms, 77 ± 6% of total INa;
ncells = 7), but a fast component,
with time constant (fast = 3.4 ± 0.5 ms)
similar to that of control was still present. Similar effects were
obtained at other potentials, and the time constants in different cells
are summarized in the inset of Fig. 1B.
INa-voltage relationships from traces
in Fig. 1A are shown in Fig. 1C: the threshold of
INa activation and the potential of
maximum INa were more negative in KB
(filled symbols), suggesting a negative shift of the voltage-dependent
activation. The INa reversal potential
was close to 0 mV (consistent with the experimental conditions used:
[Na+]i = [Na+]o = 10 mM) and was
unchanged by the drug.
View larger version (35K):
[in this window]
[in a new window]
|
Fig. 1.
Changes of INa by KB. A,
INa traces elicited by depolarizations to
various potentials in the absence (left) and in the presence (right) of
10 µM KB. The horizontal dotted line indicates zero current level. B,
superimposed INa traces induced at 30 mV
in control ( ) and in the presence ( ) of 10 µM KB. Inset, time
constants of inactivation at various potentials
(ncells = 5), in control, and in 10 µM KB. C, INa-voltage relations in control
() and in the presence () of 10 µM KB. Same cell as in A. VH of 120 mV.
|
|
KB was usually applied during repetitive depolarizations. Its effects
developed progressively over time, and usually reached steady state
within 5 to 10 min. In three experiments, applying KB for 10 min while
continuously holding the potential at 120 mV (without depolarizing
steps) caused typical effects on INa induced by the first depolarizing pulse (data not shown). The current
amplitude or time course of inactivation did not change significantly
with additional depolarizing pulses (with 1-s interpulse interval),
implying that there was no marked use dependence in the drug action at
1 Hz.
Concentration Dependence of KB Effects.
Figure
2A shows traces obtained from another
cell, successively exposed to control solution, and to 0.3, 3, and 100 µM KB. The slowing of INa depended
on the KB concentration. Fitting the INa decay at 30 mV by a sum of two
exponentials indicated that the major effect of KB consisted of
inducing a slowly decaying INa
component and that changing the KB concentration mainly affected the
relative amplitude on the fast versus slow components, with less marked
effect on their time constants. The slowly decaying INa component increased with drug
concentration, at the expense of the fast-inactivating component. To
quantitatively examine the concentration dependence of the KB effect,
the relative magnitude of the slowly inactivating component
(As; total
INa = As+
Af = 1) was plotted as a function of the
drug concentration (Fig. 2B). The relationship could be fitted by the
Hill equation (see Materials and Methods), with a
K0.5 of 2.1 µM and a Hill
coefficient of 1.1.
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 2.
Concentration dependence of the slowing of
INa inactivation by KB. A,
INa traces elicited by depolarizations to
various potentials in the absence (control) and in the presence of 0.3, 3, or 100 µM KB. Horizontal dotted line indicates zero current level.
VH of 120 mV. B, relationship between
relative amplitude of the slow component induced by KB and drug
concentration. Data were fitted to the Hill equation (see
Materials and Methods). As and
Af are the magnitudes of the slow and fast components of
INa, respectively,
K0.5 is the half-maximum effective
concentration, and nHill is the Hill
coefficient.
|
|
KB Effects on Na+ Current Amplitude Depend on Holding
Potential.
The above-mentioned data (Fig. 1C) indicate that the KB
effect on the INa amplitude was marked
at potentials close to the activation threshold. The effect also
depended on VH. Figure
3A shows superimposed
INa recorded under control conditions
and in the presence of 10 µM KB, during steps to 30 mV from a
VH of 80 mV, either without prepulse
(Fig. 3A, left) or with a 1-s prepulse to 120 mV (Fig. 3A, right).
Although the most prominent drug effect was a slowing of inactivation,
there was in addition a marked decrease of peak
INa amplitude, the decrease being
relatively more marked when depolarizing directly from
VH of 80 mV. To avoid any drug
effect, due to holding at VH of 80
mV, that may not be removed by the 1-s hyperpolarization to 120 mV,
in further experiments a continuous VH
of either 80 or 120 mV was used. Figure 3, B and C, show pooled
current-voltage relationships from various cells in which
INa was recorded at various
potentials, in control conditions (unfilled symbols) and in the
presence of 10 µM KB (filled symbols). Control
INa was smaller with a
VH of 80 mV than with a
VH of 120 mV (compare unfilled
circles in Fig. 3, B and C), as expected from less channel availability
at 80 mV. KB decreased INa at all
potentials when using a VH of 80 mV
(at 30 mV, decrease from 10.4 ± 1.1 to 5.6 ± 0.9 pA/pF; P < 0.01, ncells = 5), whereas it increased INa at
threshold potentials and had marginal or no significant effect at
potentials more than 40 mV when using a
VH of 120 mV (at 30 mV, 24.6 ± 2.4 versus 20.9 ± 1.5 pA/pF, in control and in KB, respectively;
P > 0.05, ncells = 6).
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 3.
Dependence on the holding potential of KB effects on
INa amplitude. A, superimposed
INa traces, elicited by depolarizations to
30 mV in the absence () and in the presence () of 10 µM KB.
VH of 80 mV. Insets, voltage-pulse
protocol. Horizontal dotted line indicates zero current level. Left,
depolarization without prestep. Right, depolarization after 1-s prestep
to 120 mV. B and C, INa-voltage relations
in control () and in the presence () of 10 µM KB.
VH of 80 mV (B) or 120 mV (C).
|
|
Shift of Voltage-Dependent Inactivation.
The dependence of KB
effect on the prestep or VH suggests a
change of the voltage-dependent inactivation. Figure
4A shows INa induced at 30 mV after 1-s
prepulses to different levels (VH = 120 mV), in the absence (Fig. 4A, left) and in the presence (Fig. 4A,
right) of 10 µM KB. The steady-state availability (or inactivation)
curve obtained from such tracings in 14 cells was concentration
dependently shifted to more negative potentials in the presence of KB
(Fig. 4B). The change in potential of half-maximum inactivation
(V0.5) is given as a function of
the KB concentration in Fig. 4C. Assuming that 100 µM KB caused
maximal effect (V0.5 of 15 ± 3.5 mV), the average relationship
(ncells = 3-7 for each concentration)
could be fitted by the Hill equation with a
K0.5 of 6.9 µM KB
(nH = 1.2).
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 4.
Effect of KB on steady-state inactivation of
INa. A, current traces obtained at 30 mV
after 1-s prepulses to various levels in control conditions (left) and
in the presence of 10 µM KB (right). Horizontal dotted line indicates
zero current level. VH of 120 mV.
Prepulses in the traces shown are to 105, 95 (identical with trace
following 105), 85, 75, 65, and 55 mV. B, availability (or
inactivation) curves in the absence () and in the presence of 10 µM () or 100 µM ( ) KB. Data obtained by normalizing currents
measured at 30 mV after 1-s conditioning steps to various potentials.
The fitting lines are Boltzmann distribution functions (see
Materials and Methods) drawn using the mean of
parameters obtained in each experiment (control:
V0.5 = 79 ± 1.5 mV, slope = 4.8 ± 0.2; 10 µM KB: V0.5 = 84 ± 2.7 mV, slope = 4.5 ± 0.3; 100 µM KB:
V0.5 = 94 ± 2.4 mV, slope = 6.0 ± 0.4; P < 0.05 for
V0.5 in KB versus control;
P > 0.05 for slopes in KB versus control). C,
relationship between magnitude of shift in potential of half-maximum
inactivation (V0.5) and drug
concentration.
|
|
Slowing of Recovery from Inactivation.
We also examined the
effect of KB on the recovery from inactivation. Figure
5A shows
INa traces from a typical experiment in which double pulses were given at 30 mV, with increasing
interpulse interval at VH of 120 mV,
before and during treatment with 10 µM KB. The first pulse was 1 s in duration (of which only the initial 40 ms are displayed in traces
of Fig. 5A) and allowed complete inactivation of
INa, even in the presence of KB. The second pulse allowed a recovery of INa
as a function of the interpulse interval. The time course of the
relative recovery measured in five cells
(INa in the second pulse as percentage
of INa in the first pulse) is shown in
Fig. 5B and illustrates that the removal of inactivation was slowed in
the presence of KB. Two exponential components were sufficient to
account for the recovery process. In control conditions, a fast
component with a time constant (fast) of
25 ± 2.8 ms accounted for 50 ± 2.2% of the total recovery, whereas a slow component (50 ± 2.3%) had a time constant
(slow) of 117 ± 20.2 ms. In the presence
of 10 µM KB, the two components were slowed
(fast = 55 ± 13.2 ms,
slow = 318 ± 123.8 ms; Fig. 5C, left).
To assess the change in the relative contributions of both components,
the recovery in KB was fitted using the same time constants as in
control. The slow recovery component was predominant in the presence of
KB (Fig. 5C, right). In addition, it was also noticed that a small
INa recovered after brief interpulse intervals (e.g., traces labeled "a" in Fig. 5A, right, and inset) inactivated faster than a larger INa
recovered after longer intervals (e.g., traces labeled "b" in Fig.
5A, right, and inset). This result is also consistent with a slower
recovery of slowly inactivating Na+ channels.
View larger version (49K):
[in this window]
[in a new window]
|
Fig. 5.
Effect of KB on the recovery from inactivation. A,
superimposed currents, obtained by a double-pulse protocol, with
varying interval between the two pulses. VH
of 120 mV. Test potential, 30 mV. Pulses given every 5 s.
Horizontal dotted line indicates zero current level. Left, in the
absence of drug. Right, in the presence of 10 µM KB. Left inset,
voltage pulse protocol. Right inset, traces after 20-ms (a) and 500-ms
(b) interval. Note fast inactivation of the trace after shorter
interval. B, amplitude of INa during the
second pulse expressed relative to the current during the first pulse
and plotted as a function of the interval between the pulses. Data from
five cells, calculated from tracings such as those in A. Inset,
semilogarithmic plot of 100 minus percentage of recovery. C, time
constants and amplitudes of the biexponential fitting of inactivation.
Unfilled columns, control conditions. Hatched columns, 10 µM KB. *,
P < 0.05 for KB versus control.
|
|
Effect of Tetrodotoxin and Interaction with Other Drugs Related to
Thyroid Hormones.
INa in control
conditions and in the presence of KB could be largely suppressed by 10 µM tetrodotoxin (ncells = 3; data
not shown), suggesting that sensitivity to this toxin was not
suppressed by KB.
Due to their structural resemblance, we also examined a possible
interaction between KB and amiodarone. Amiodarone (10 µM) itself had
no effect on INa induced from a
VH of 120 mV and did not slow the
inactivation process (Fig. 6A). In five
cells, INa at 30 mV inactivated with
of 2.1 ± 0.2 and 1.8 ± 0.2 ms before and after 10 min
in the presence of 10 µM amiodarone. However, the drug shifted the
inactivation curve to more negative potentials and slowed the recovery
from inactivation of INa
(ncells = 4; data not shown).
Application of KB on top of amiodarone still induced its usual slowing
of inactivation (Fig. 6A, right), enhanced the negative shift of the
inactivation curve, and further slowed the recovery from inactivation.
Because, on the one hand, KB like amiodarone may share some receptors
with thyroid hormones, and on the other hand these hormones have been
reported to slow INa inactivation
(Craelius et al., 1990; Harris et al., 1991; Dudley and Baumgarten,
1993), we also examined the effect of T3 and of its analog DITPA and their interaction with KB.
T3 at physiological concentrations (10 pM-1 nM;
ncells = 3) did not affect
INa. Even at higher concentrations (10 nM, ncells = 2; or 10 µM,
ncells = 6), T3
failed to slow INa inactivation under
our experimental conditions (Fig. 6B). However, at 10 µM
T3 increased INa
peak amplitude (Fig. 6B, middle; Huang et al., 1999). In addition, we
noticed that in the presence of 10 µM T3 the
effect of 10 µM KB (applied on top of T3) on
inactivation was markedly reduced (Fig. 6B, right). DITPA at 10 µM
(ncells = 2) did not affect
INa or prevent the KB effect (data not
shown). The above-mentioned results with amiodarone and
T3 are summarized in Fig. 6C, which presents the
relative magnitude of the slowly inactivating
INa component in the different
experimental conditions.
View larger version (36K):
[in this window]
[in a new window]
|
Fig. 6.
Interaction between KB and amiodarone or
T3. A, INa traces at various
potentials in control conditions, in the presence of 10 µM
amiodarone, and in the presence of 10 µM KB added on top of
amiodarone. Horizontal dotted line indicates zero current level. B,
INa traces at 30 mV in control conditions,
in the presence of 10 µM T3, and in the presence of 10 µM KB added on top of T3. C, pooled data on the effects
of KB and other drugs on inactivation kinetics. Amplitudes of slow
component of INa inactivation in various
experimental conditions. Unfilled columns: in the absence of KB (but in
presence of 10 µM of the drugs indicated below the columns; control,
no drug). Hatched column: in the presence of 10 µM KB (added on top
of the drugs indicated below the column). *, P < 0.0001 for amplitudes with 10 µM KB in T3 versus in
control. VH of 120 mV.
|
|
Preferential Binding of KB from Extracellular Side.
The
removal of KB for up to 30 min was not associated with a washout of its
effect (ncells = 3). Because the drug
is lipophylic and could reach an intracellular site of action even when
given extracellularly, we tested the effect of adding KB to the
cell-dialyzing pipette solution. In 11 cells, intracellular dialysis
with 10 µM KB had no apparent effect on
INa, which inactivated as fast as in
untreated cells ( = 2.5 ± 0.6 ms, after 12-15 min of
dialysis with KB; V0.5 of inactivation
of 79 ± 1.4 mV; P > 0.05 versus 80 ± 1.2 mV in 26 control cells). In addition, these cells still responded
to extracellularly applied KB (data not shown). In four other cells
there was a modest slowing of INa
inactivation while dialyzing with 10 µM KB
(fast = 1.7 ± 0.1 ms,
slow = 32.1 ± 2.3 ms; and
As = 13 ± 3.4% of total
INa). However, even in these cells the
slowing of inactivation became most marked when the same drug
concentration was further applied externally. Taken together, these
results do not allow excluding access of KB to its binding site from
the intracellular medium, but indicate that the drug is more effective
when applied externally.
To further examine the membrane side from which the drug reached its
binding site, we measured single Na+ channel
activity in cell-attached patches. Figure
7 compares recordings obtained without
and with KB added to the pipette solution. When KB was not included in
the pipette solution (Fig. 7A), multiple Na+
channel openings were clustered in the first 10 ms of the voltage step
and hardly seen at potentials negative to 70 mV. In contrast, when 10 µM KB was present in the pipette (i.e., on the extracellular side of
the patch; Fig. 7B), the channel openings were present for a longer
time (during the first 100 ms of the voltage step; notice difference in
time scale between A and B) but also decayed slowly with time. In
addition, channel openings were frequently observed at potentials more
negative than 70 mV. Typical ensemble-average currents obtained from
such recordings in another experiment with 10 µM KB in the pipette
are illustrated in Fig. 7C and confirmed that the channel inactivation
was markedly slowed by the application of KB on the external side of
the patch. Under these conditions, the inactivation time constant was
28 ± 6.6 ms (ncells = 3) at 30
mV, a value that is of the same order of magnitude as the dominant slow
time inactivation of whole-cell INa
(Fig. 1B). The current-voltage relationship obtained using ensemble
averages has the shape expected for
INa and reversed at an extrapolated potential of >+50 mV (consistent with
[Na+]pipette = 150 mM).
When KB was applied in the bath, no or less marked effect on the patch
channels was obtained (data not shown). These results are also
consistent with an easier access of the drug from the extracellular
side.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 7.
Effect of externally added KB on single
Na+ channel currents. A and B, Na+ channel
activity recorded in the cell-attached patch-clamp mode without (A) or
with (B) 10 µM KB included in the pipette solution.
VH of approximately 120 mV (pipette
potential set to +120 mV while zeroing intracellular potential with
high-K+ bath solution; see Materials and
Methods). Artifactual upward deflections at the beginning of
the traces indicate start of depolarization. Horizontal dotted line
indicates zero current level. C, ensemble-averages of the single
Na+ channel currents recorded with KB (10 µM) added to
the pipette solution. Each trace is an average of 29 original traces.
Notice the slow decay of the currents when KB was present in the
pipette. D, INa-voltage relationship, using
peak current amplitudes in C. Different experiments in A, B, and C.
|
|
| |
Discussion |
Effect on Inactivation.
The present study shows that KB, a
drug related to amiodarone and to thyroid hormones, has marked effect
on INa. The most prominent KB effect
was a slowing of INa inactivation, but
the drug also had an effect on voltage-dependent activation and caused a change of INa peak amplitude,
depending on the voltage protocol.
The effect on the kinetics of INa
inactivation was unexpected given the known actions of amiodarone,
which either does not change the kinetics of fast inactivation (Follmer
et al., 1987; Kohlhardt and Fichtner, 1988) or accelerates the decay of
late currents (Maltsev et al., 2001). It resembles that of many drugs, including animal or plant toxins, that interact with specific sites on
the voltage-dependent Na+ channel and modify
inactivation (Narahashi, 1996; Catterall, 2000; Goldin, 2001). At the
molecular level, inactivation of Na+ channels and
its coupling to activation involve many parts of the channel protein,
comprising intracellular loops (e.g., between domain DIII and DIV of
the 
-subunit) but also intracellular, mid-, or even extracellular
portions of certain transmembrane segments (Catterall, 2000; Goldin,
2001). Thus, inactivation can be modified by agents that act from
either side of the membrane. For example, the binding site of
insecticides such as DDT or pyrethroids can be accessed from the
extracellular side of the membrane to change inactivation (Tatebayashi
and Narahashi, 1994; Narahashi, 1996; Spencer et al., 2001). In the
present study, the lack of a pronounced effect of intracellularly
dialyzed KB in whole-cell experiments and the marked increase of
channel activity by external KB in cell-attached patches are suggestive
of a preferential access of KB from the external side.
The effect of KB on INa peak amplitude
depended on both VH and test
potential. The marked decrease of INa
induced from 80 mV (Fig. 3, A and B), with less pronounced, no, or
opposite effect when inducing INa from
120 mV (Fig. 3C) is consistent with a shift of the
INa availability curve (Fig. 4B).
Similar decreases of INa amplitude and
shifts of its inactivation curve are usually obtained by local
anesthetic or class 1 antiarrhythmic drugs (Carmeliet and Mubagwa,
1998), but also by many other drugs that slow inactivation. Amiodarone
also causes such an effect, which may contribute to its beneficial
action against cardiac arrhythmias. The shift in inactivation curve by
KB occurred at higher concentrations
(K0.5 6.9 µM) compared with the
effect on kinetics (K0.5 = 2.1 µM), making it plausible that different sites could be involved in the two
effects. For example, binding at a local anesthetic site, accessible
from the inside, could be invoked to account for the effect on
voltage-dependent inactivation. However, the absence of a difference in
V0.5 of inactivation curve between
cells internally dialyzed with KB and control cells, suggests that the
shift of inactivation curve was not due to an internally accessible
binding site.
KB applied while channels were rested at the holding potential, without
any depolarizing pulse, produced its maximum effect on the first pulse.
This indicates that the drug interacted either with the rested state
and/or that there is very fast binding to open channels. At high drug
concentrations, the largest component of the total
INa inactivated with a slow time
constant, suggesting that under these conditions nearly all channels
interacted with the drug. If such a KB binding were to occur only in
the closed state, a significant increase in
INa amplitude should be obtained at
all potentials in the presence of the drug due to less inactivation at
the time of peak current. However, INa
was increased only at potentials less than 50 mV while using
VH of 120 mV. This increase can be
accounted for by a change of voltage-dependent activation. The lack of
an INa increase at potentials of
maximum activation (more than 40 mV) therefore suggests that to a
certain extent binding also involves the open state.
Relation to Thyroid Hormones.
Given the reported slowing of
INa inactivation by
T3 (Craelius et al., 1990; Harris et al., 1991;
Dudley and Baumgarten, 1993) and the structural similarity between this
hormone and KB, we examined the effect of T3,
with the hypothesis that any T3-induced slowing
of inactivation was likely to involve the same membrane receptor as for
KB. We were surprised to find that we could not reproduce the actions
of T3 on INa
inactivation reported by others, even at very high concentrations. The
reason for this discrepancy with previous studies is not clear. But the
same authors later obtained much less marked effect on the kinetics of
inactivation in another study (Huang et al., 1999) and proposed that
the T3 effects could be influenced by the cell
conditions. (Nevertheless, our study confirms that
T3 increases INa
peak amplitude; Huang et al., 1999.) We then tested whether
T3 was occupying the receptor involved in the KB
action, but unable to induce the pharmacological effect. The decrease
of KB effect when added on top of T3 (Fig. 6, B
and C) is in favor of this possibility, although a definitive proof
would require more extensive and specific experimental techniques.
KB has been shown to interact with nuclear thyroid hormone receptors
(Carlsson et al., 2002). Nuclear receptors mediate genomic effects that
develop over hours or days. It is unlikely that the acute effects
observed in the present study are mediated by such receptors.
Extranuclear receptors, which are also known to exist also for acute,
nongenomic effects of thyroid hormones, are likely to be involved. The
identity of the extranuclear receptors is still unknown. The lack of
response in most cells when including KB in the cytosol-dialyzing
pipette solution and the marked effect of KB when the drug was applied
to the external side of the patch are suggestive of an action site more
easily accessible from the extracellular side. However, access to this
site from an intracellular site cannot be completely excluded by the
present experiments because in a few cases a small effect of KB applied
via the pipette was obtained, and the ability of KB to be exchanged
between pipette solution and cytosol is not known. At present we also
cannot exclude the possibility of intermediate structures coupling the
receptor for KB to the effector, i.e., the Na+
channel. However, given the similarity with other
inactivation-modifying drugs, it is possible that the
Na+ channel -subunit constitutes itself a
membrane receptor for KB. Surprisingly, amiodarone, which is
structurally closer to KB, did not antagonize the slowing of
inactivation. Because only one concentration (10 µM) was tested, we
cannot exclude the possibility that this negative result is due to a
low-affinity interaction of amiodarone with the site responsible for
slowing inactivation or the possibility that amiodarone is unable to
induce the necessary conformation change.
Practical Importance.
The effect of KB on
INa was not specific to the pig cells
or to the special extracellular and cell-dialyzing solutions used to
optimize INa measurements. A similar
effect could be obtained in ventricular myocytes of other species
(guinea pig, rabbit, rat, and human) during superfusion with
extracellular solutions containing 150 mM
[Na+]o, or while
recording with the perforated patch method (R. Macianskiene, S. Viappiani, K. Mubgwa, unpublished data).
KB action may be of importance for pathophysiological conditions such
as arrhythmias and heart failure. KB has been proposed to have a
protective effect during reperfusion after ischemia (Carlsson et al.,
2002). However, because increased
[Na+]i is known to play a
role in the deleterious effect of ischemia and reperfusion by promoting
[Ca2+]i overload, the
slowing of INa inactivation by KB, as
described in the present study, will tend by itself to overload the
cells with Na+ and to enhance cell damage. In
addition, slowed inactivation is generally associated with the long QT
syndrome, with increased propensity to "torsades de pointes"
arrhythmias and to fibrillation. Thus, a protective action based on
this effect on Na+ channel inactivation is
unlikely. However, the "local anesthetic" action (shift in the
inactivation curve), which implies that fewer channels are available to
open from the resting potential, could partly offset the deleterious
effect of slowed inactivation.
The slowing of Na+ channel inactivation may be of
interest for inotropy. Moderate increases in
[Na+]i are proposed to be
the basis of the beneficial effect of cardiac glycosides in supporting
cardiac contraction in heart failure. Whereas the cardiac glycosides
increase [Na+]i by
inhibiting the
Na+-K+-ATPase, such an
increase can also be achieved by increasing Na+
influx. There is increasing interest in a new class of positive inotropic agents (e.g., DPI-201-106, BDF-9148, BDF-9196, and
LY-366634) that act by altering Na+ channel
function, because their action is preserved in failing myocardium
(whereas the effect of cAMP-increasing agents tend to be limited;
Muller-Ehmsen et al., 1997, 1998). For all positive inotropic agents,
the beneficial use in heart failure is limited by their potential
arrhythmogenic action. In this respect, it has been shown that many of
the Na+ channel-modifying agents increase the
action potential duration (Amos and Ravens, 1994) and the QT interval
on the electrocardiogram. It will be of interest to carefully examine
whether KB lacks this side effect on action potential duration
(Carlsson et al., 2002) and QT while probably retaining the positive
inotropic action.
We thank Virginie Bito for supplying ventricular tissue and some
of the isolated myocytes, Dr. F. Rega for supplying ventricular tissue,
and Prof. G. Droogmans for generously allowing us to use the Winascd software.
This study was supported by grants from Fonds voor
Wetenschappelijk Onderzoek, the Flemish Foundation for Science.
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Abstract
Introduction
![<UP>availability</UP>=1/{1+<UP>exp</UP>[(V−V<SUB>0.5</SUB>)/k]}](/content/vol304/issue1/fulltext/130/img001.gif)
![<UP>relative effect</UP>=1/[1+(K<SUB>0.5</SUB>/D)<SUP>n</SUP>]](/content/vol304/issue1/fulltext/130/img002.gif)
