![[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}
|
|
|
|
|
||
Vol. 280, Issue 2, 686-694, 1997
Departments of Neurology and Pharmacology and Physiology, University of Rochester, Rochester, New York (R.A.G.), Department of Molecular Biology and Pharmacology, Washington University, St. Louis, Missouri (D.F.C.), and Department of Neurology, University of Texas, Houston, Texas (J.A.F.)
 |
Abstract |
|---|
 Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Alkyl-substituted thiobutyrolactones increase or decrease
-aminobutyric acidA responses at or near the picrotoxin
site, but they are structurally similar to ethosuximide, which prompted us to determine the actions of thiobutyrolactones on voltage-dependent Ca++ currents. We measured Ca++ currents in
cultured neonatal rat dorsal root ganglion neurons in the absence and
presence of the anticonvulsant
-ethyl,-methyl--thiobutyrolactone (-EMTBL) and the
convulsant 
-ethyl,-methyl--thiobutyrolactone (-EMTBL).
Low-voltage-activated (T-type) currents were reduced in a
concentration-dependent manner, with a maximal reduction of 26% and
30% by -EMTBL and -EMTBL, respectively. -EMTBL reduced high-voltage-activated currents in a concentration- and
voltage-dependent manner: maximal responses were 7% when evoked from
80 mV, with more rapid current inactivation; 29% when evoked from
40 mV, with little effect on current inactivation. -EMTBL
increased high-voltage-activated currents 
20% at 10 to 300 µM, but
reduced currents at higher concentrations; the latter action was
similar to that of -EMTBL in its magnitude and voltage dependence.
Block of N-type channels with 
-conotoxin GVIA (10 µM) reduced the
effect of -EMTBL and eliminated its voltage dependence. The L-type
current component was also reduced by -EMTBL, with little effect on
P- or Q-type current components. The related compound,
-ethyl,-methyl--butyrolactone, had no effect on
Ca++ currents. We conclude that thiobutyrolactones affect
voltage-dependent Ca++ currents in a concentration- and
voltage-dependent manner, with greater potency on low-voltage-activated
channels. Both the ring structure and the position of its alkyl
substitutions determine the identity of the targeted Ca++
channel subtypes and the manner of regulation.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
The mechanisms of action of
antiseizure drugs have been studied intensively, relying on observed
drug effects in vitro to make inferences about actions
in vivo (Rogawski and Porter, 1990
; Macdonald and
Meldrum, 1995). For the most part, these drugs act on neuronal ion
channels or on the neurotransmitters that regulate their activity. For
example, phenytoin, carbamazepine and lamotrigine increase inactivation
of voltage-dependent Na+ channels (McLean and
Macdonald, 1983, 1986; Quandt, 1988; Cheung et al.,
1992), and barbiturates and benzodiazepines increase
GABAA-mediated inhibition (e.g., Twyman et
al., 1989; Macdonald and Meldrum, 1995). The role of
Ca++ channel blockade as an antiseizure mechanism is less
clear, however. The Ca++ channel blocker flunarizine and
the dihydropyridines have some antiseizure activity (Binnie, 1989;
Meyer, et al., 1990), but the effects of the latter are not
substantial. Barbiturates, at concentrations similar to those achieved
during treatment of status epilepticus, profoundly block
high-voltage-activated Ca++ currents (Gross and Macdonald,
1988a,b). The best case for the antiseizure effect of Ca++
channel blockade is ethosuximide, which blocks the
low-voltage-activated T-type channel, likely responsible for its
efficacy in primary generalized (absence) epilepsy (Coulter et
al., 1989).
Although newer pharmaceuticals are being developed with increasingly
specific and selective actions, some affect more than one ion channel
or neurotransmitter system; it may be difficult, therefore, to
determine which action is most desirable for clinical efficacy (or
which produces undesirable toxicity) and whether more than one
therapeutic action may be advantageous. The alkyl-substituted butyro-
and thiobutyrolactones are a novel group of compounds with either
anticonvulsant or convulsant activity, dependent on the location and
size of their alkyl groups. Compounds with small -substitutions are
anticonvulsant, whereas -substituted compounds are convulsant (Klunk
et al., 1982a,b; Holland et al., 1990). Both
groups act at or near the picrotoxin site of the GABAA
receptor complex (Weissman et al., 1984; Levine et
al., 1985; Canney, et al., 1991; Xu et al.,
1995), exhibiting either picrotoxin agonist (convulsant) or inverse
agonist (anticonvulsant) properties. These compounds do not appear to
gate the GABAA receptor directly, but the actions of some
are complex, with differing quantitative effects in the presence of
varying GABA concentrations (Yoon et al., 1993).
The structural similarities of the butyrolactone analogs to
succinimides, and to ethosuximide in particular, suggested that these
compounds may regulate voltage-dependent Ca++ channels as
well (fig. 1). If this were to be the case, then butyrolactone analogs may afford the opportunity to make inferences about the relative merits of GABA or Ca++ channel effects
vis-à-vis convulsant or anticonvulsant action. We therefore
sought to characterize the actions of a selected group of
alkyl-substituted butyrolactones on voltage-dependent Ca++
currents in cultured neurons. The effects of the anticonvulsant compound -EMTBL were compared with those of the convulsant -EMTBL and with -EMGBL, an anticonvulsant compound that blocks the action of -EMTBL on GABA responses in cultured hippocampal neurons (Holland et al., 1990). Our studies suggest that GABAA
receptor blockade is the major action of the convulsant compounds but
that blockade of Ca++ channels, particularly T-type
channels, may participate in the anticonvulsant actions of the
-substituted thiobutyrolactones.
|
| |
Methods |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Cell Culture
Primary DRG cultures were prepared from day 7 to 10 neonatal
Sprague-Dawley rats (Harlan) as described previously (Gross et al., 1990). The animals were sacrificed after CO2
narcosis by decapitation. The DRGs were dissected away from the spinal
cord and placed in ice-cold Hanks' buffer (pH 7.4 with HEPES; Sigma Chemical Company, St. Louis, MO). After enzymatic digestion with trypsin (1 mg/ml; Sigma) and mechanical trituration, the cells were
spun down and resuspended in plating medium [MEM containing bicarbonate (GIBCO, Grand Island, NY), 50 ng/ml neural growth factor
(Collaborative Biomedical Research, Bedford, MA), 5% equine serum and
5% fetal bovine serum (Hyclone Laboratories, Logan UT)]. Cells were
then plated on 35-mm dishes with collagen (Sigma) as substrate. The
medium was changed within several hours and replaced with MEM
containing 10% equine serum, without fetal bovine serum. Ara-C (1-5
µM; Sigma) was used, if needed, to inhibit the growth of nonneuronal
cells within the first week of longer term cultures. The cultures were
maintained at 37°C under a 95% air-5% CO2 atmosphere and fed on a twice-weekly basis with 50% exchanges of growth medium. These cultures were used for experiments as early as day 2, and for up
to 10 to 12 weeks. The vast majority of experiments were performed on
neurons in culture for 2 to 4 days.
Electrophysiology
Preparation of solutions.
Butyrolactones were dissolved in
external solution (below) on the day of the experiment. Nifedipine and
-conotoxin GVIA (both from Sigma) and -agatoxin IVA (Peptides
International, Louisville KY) were made fresh on the day of the
experiment. Nifedipine was dissolved in dimethyl sulfoxide (Sigma) at a
concentration of 10 mM and was diluted in external solution to a final
concentration of 10 µM. -Conotoxin GVIA and -agatoxin IVA were
stored frozen in water at a concentration of 10 mM and were used for
experiments within 2 to 3 months. On the day of the experiment, the
stock solutions were diluted in external solution.
Whole-cell patch-clamp recordings.
Whole-cell voltage-clamp
recordings were obtained with the whole-cell variation of the
patch-clamp technique. Cells were bathed in a solution containing (in
mM): CaCl2, 5.0; choline Cl, 67; MgCl2, 0.8;
TEA, 100; glucose, 5.6; KCl, 5.3; HEPES, 10 (pH 7.3-7.4, 310-330
mOsm; all reagents from Sigma). Glass recording patch pipettes were
fashioned from Fisherbrand microhematocrit tubes with a Sutter
Instruments Brown-Flaming P-87 pipette puller. These electrodes had
resistances of 1.5 to 3.0 megohm when filled with a recording solution
consisting of the following (in mM): CsMeSO3, 140; HEPES,
10; ethyleneglycol-bis(-aminoethyl ether)-N,N,N
,N-tetraacetic acid, 10; ATP-Mg++, 5; GTP-Na, 0.1 (all reagents from
Sigma). The pH (7.3-7.4) was adjusted with 1 N CsOH after the addition
of ATP. The osmolality was 10 to 15% less than that of the bath
solution, 280 to 300 mOsm.
3 dB). Data were accepted only if space-clamp was adequate,
determined by well-controlled incremental current activation over a
series of voltage steps (usually, 65 to +10 mV), and if tail currents
deactivated rapidly. Furthermore, with 10-mV test pulses, we required
settling of the capacitance-charging transient within 2 msec.
Leak current was determined by a P/P4 or P/P6 protocol. This current
was digitally subtracted from the relevant inward current to obtain the
calcium current.
Butyrolactones, nifedipine, -agatoxin IVA and -conotoxin GVIA
were applied to the cell under study by pressure ejection (6-10 kPa)
from separate blunt-tipped (internal diameter, 12-15 µm) glass
micropipettes positioned ~30 µm from the cell. Applications of
these compounds were 10 to 15 sec in duration, just before currents
were evoked and the puffer pipettes were removed from the bath when not
in use. In some experiments, drug applications were accomplished by a
"U-tube" microperfusion system; the results were similar to those
obtained with puffer application. Diluents had no effect on evoked
currents. Control experiments showed that dimethyl sulfoxide
(nifedipine) at concentrations up to 1% had no effect on
Ca++ currents. In all experiments, the culture was perfused
continuously with bath (external) solution by a gravity-fed,
vacuum-removed system operating at about 0.3 ml/min; thus, drug
concentrations were minimized in the remainder of the culture dish.
Data Analysis
Data are expressed as means ± S.E.M., and statistical comparisons between group means were made with Student's two-tailed t test.
| |
Results |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Effects on low-voltage-activated whole-cell
Ca++ currents.
The effects of
thiobutyrolactones were tested first on low-voltage-activated (T-type)
Ca++ currents because of their structural similarity to
ethosuximide, an antiabsence seizure drug that reduces T-type currents
in thalamic and sensory neurons (fig. 2). T-type
currents, which require very negative Vh to remove
steady-state inactivation, can be studied in isolation by evoking
currents at relatively negative Vc (negative to 20 mV),
at which little if any high-voltage-activated currents are evoked. For
these experiments, currents were evoked from Vh = 90 mV
at a series of potentials, ranging from 65 to 5 mV (see figure).
Peak current magnitudes were measured, and the effect of each compound
was assessed on all currents uncontaminated with the more slowly
inactivating high-voltage-activated current components.
|
-EMTBL reduced T-type currents, with a
mean peak current reduction of 26 ± 3% (n = 12 ± SEM; P .001). There was no apparent effect on the
voltage dependence of current activation as evidenced in the
current-voltage plots (although unequivocal determination of the
potential at which the maximal T-type current occurred was difficult
because of contamination with high-voltage-activated current
components), nor any effect on current kinetics. The convulsant
compound -EMTBL had a similar action. Low-voltage-activated currents
were reduced with a maximal mean reduction of 30 ± 1%
(n = 4; P .001). By contrast, the anticonvulsant butyrolactone -EMGBL (500 µM) had no effect on T-type currents (see also below). These actions were compared with
those of 500 µM ethosuximide applied to neurons from the same culture
groups. Ethosuximide reduced T-type currents, but was less efficacious
than either thiobutyrolactone, producing a current reduction of 16 ± 4% (n = 8; P .002).
Thiobutyrolactones reduced T-type currents in a concentration-dependent
manner but with slightly differing potencies and efficacies. T-type
currents were not significantly affected by 10 µM -EMTBL. At
greater concentrations, T-type currents were reduced 9 ± 2% (50 µM, n = 6), 18 ± 2% (100 µM,
n = 8), 25 ± 3% (300 µM, n = 5) and 26 ± 3% (500 µM, n = 13). Higher
concentrations did not have a greater effect. The
concentration-response plot (fig. 3) was used to
estimate by visual inspection the half-maximal concentration (EC50 ~75 µM). -EMTBL had similar
concentration-response parameters, slightly more efficacious but
slightly less potent. Currents were reduced 1.5% (50 µM,
n = 2), 2 ± 1% (100 µM, n = 5), 7 ± 1% (300 µM, n = 4), 30 ± 5%
(500 µM, n = 9) and 30 ± 1% (1000 µM,
n = 4), yielding an apparent EC50 ~350
µM.
|
Effects on high-voltage-activated whole-cell
Ca++ currents.
The effects of -EMTBL
and -EMTBL were assessed next on high-threshold-activated current
components. The thiobutyrolactones had complicated actions on these
current components, with both concentration- and voltage-dependent
effects. Across a wide concentration range (10-1000 µM), -EMTBL
reduced high-voltage-activated Ca++ current components in
all neurons tested, but did so in a voltage-dependent manner (fig.
4). By contrast, -EMTBL increased currents at
concentrations 300 µM, without a clear voltage dependence, and
reduced currents at greater concentrations in a voltage-dependent
manner similar to that of -EMTBL (fig. 5).
|
|
-EMTBL, with maximal peak current reductions at a
concentration of 500 µM. When currents were evoked from
Vh = 80 mV at +10 mV, there was no apparent effect of 10 µM -EMTBL. At higher concentrations, peak current magnitude was
reduced 2 ± 1% (50 µM, n = 8), 2 ± 1%
(100 µM, n = 11), 4 ± 1% (300 µM, n = 9, P .001), 7 ± 1% (500 µM,
n = 45, P .001) and 6 ± 1% (1000 µM,
n = 4, P .005). Although the reduction in peak
current was modest, currents inactivated more rapidly (see fig. 4).
When currents were evoked from Vh = 40 mV at +10 mV, the
peak reductions were greater, 4 ± 1% (50 µM, P .01),
8 ± 1% (100 µM, P .001), 16 ± 2% (300 µM,
P .001), 28 ± 1% (500 µM, P .001) and 30 ± 4% (1000 µM, P .001). When currents were evoked from
these more positive potentials in the presence of -EMTBL, the rate of current inactivation was slightly greater than control, but not
nearly so rapid as when evoked from 80 mV. The apparent
EC50 was ~300 µM.
In the presence of -EMTBL, currents were also affected in a
concentration- and voltage-dependent manner, but an additional effect
was seen, enhancement of Ca++ currents at relatively low
concentrations (fig. 5). As with -EMTBL, there was no effect of 10 µM -EMTBL. When currents were evoked from Vh = 80 mV
at +10 mV, peak current magnitudes were increased 10 ± 4% (50 µM, n = 9, P .02), 13 ± 5%
(100 µM, n = 7, P .05) and 7 ± 5% (300 µM, n = 4); at higher concentrations, peak current magnitudes were reduced 7 ± 2% (500 µM,
n = 10, P .02) and 7 ± 2% (1000 µM,
n = 6, P .005). When currents were evoked from Vh = 40 mV at +10 mV, currents were increased
10 ± 4% (50 µM, P .02), 10 ± 5% (100 µM,
P .05) and 5 ± 9% (300 µM); at higher concentrations,
peak current magnitudes were reduced 23 ± 4% (300 µM, P .001) and 29 ± 3% (1000 µM, P .001).
The current enhancements seen with -EMTBL were not voltage
dependent, i.e., the percent increase was similar when
currents were evoked from either 80 mV or 40 mV. Current reductions
seen in the presence of -EMTBL, by contrast, were voltage dependent:
lesser peak current reductions and faster current inactivation were
evident in currents evoked from Vh = 80 mV; greater peak
current reductions and more normal current inactivation were evident in
currents evoked from Vh = 40 mV. For current enhancement,
the EC50 could not be determined unequivocally, because we
could not determine, at any given concentration, whether -EMTBL was
acting solely to increase currents. Assuming that the maximal
enhancement was 13% at 100 µM, and given that 10 µM had no effect,
a rough estimate of EC50 is 50 µM. With the same
limitation, the EC50 for current reduction was ~400 µM.
At the higher concentrations tested, current-voltage plots showed that
the maximal high-voltage-activated currents occurred at similar
potentials in the absence and presence of thiobutyrolactones and that
ECa did not appear shifted (see also fig. 2).
We next examined whether -EMGBL, an anticonvulsant compound that
blocks the effects of -EMTBL on GABAA receptors,
affected Ca++ currents in a similar manner. -EMGBL was
tested in the absence and presence of -EMTBL, with both compounds
applied at a concentration of 500 µM (fig. 6). When
applied alone, -EMGBL had no effect on high-voltage-activated
Ca++ currents, evoked from Vh of 80 and 40
mV; this lack of effect was observed across the entire physiological
range of current activation, as shown in the current-voltage plot
(Vh = 90 mV). In the next series of experiments,
-EMGBL was applied first to verify its lack of effect. Next,
-EMGBL and -EMTBL were applied simultaneously from another puffer
pipette; compared with the effect of -EMTBL alone, obtained at the
end of the experiment, -EMGBL failed to alter the action of
-EMTBL. In a series of control experiments, -EMGBL was applied
continuously from one puffer pipette, with a second pipette used to
apply -EMTBL. Preapplication of -EMGBL in this manner still
failed to affect the action of -EMTBL (data not shown).
|
; Regan et al., 1991; Mintz
et al., 1992; Zhang et al., 1993). Because
earlier experiments had already studied low-voltage-activated (T-type)
currents in isolation, these experiments on high-voltage-activated
current components studied the effect of -EMTBL before and after the
application of biological toxins that block select channel types (fig.
7). High-voltage-activated currents were evoked from
Vh of 80 and 40 mV in the absence and presence of 500 µM -EMTBL to obtain a control response. Digital subtraction of the
traces obtained in the presence of -EMTBL from those obtained in its
absence permitted the direct visualization of the drug effect (traces
far right in fig. 7). Next, the selective and virtually irreversible
N-type channel blocker -conotoxin GVIA (10 µM) was applied for 10 to 30 sec, and the effect of -EMTBL was reassessed. P- and Q-type
channels were blocked by use of a high (500 nM) concentration of
-agatoxin IVA (Zhang et al., 1993); even at this
concentration, full block of these channels, as determined by the
attainment of a new stable base-line current, took 5 to 8 min. For most
experiments, therefore, the toxins were applied simultaneously:
reapplication of -EMTBL shortly after -conotoxin GVIA allowed the
determination of the effect of the compound on the residual L-, P- and
Q-type current components; and reapplication after a stable base line
was obtained allowed the determination of -EMTBL's action after
block of N-, P- and Q-type channels. In the latter case, the L-type
current component was assumed to constitute most if not all of the
remaining current.
|
-EMTBL blocked
high-voltage-activated Ca++ current components
nonselectively, but that its voltage-dependent effect was selective for
N-type channels. As seen in the derived difference currents, the effect
of -EMTBL, when applied at 80 mV, was slow to develop, requiring
most of the 100-msec voltage step to achieve maximal block. When
applied at 40 mV, the effect was greatest at the current peak, with a
parallel reduction in current persisting throughout the remainder of
the voltage step. After application of -conotoxin GVIA, the rapidly
inactivating N-type current component was lost. The subsequent
application of -EMTBL showed that the effect of this compound was
reduced by 51 ± 7% (n = 9). An interesting
finding was that the voltage-dependent effect was lost, with equal
current reductions from both Vh values; the current
reductions developed rapidly, with a parallel current reduction
throughout the voltage step. Elimination of P- and Q-type current
components by continued applications of the toxins reduced the control
currents further, with little current inactivation evident during the
voltage step. Reapplication of -EMTBL produced a similar current
reduction after block of P- and Q-type currents: in 5 of 8 neurons, the
effect of -EMTBL was the same, whether -agatoxin IVA was applied
alone or in combination with -conotoxin GVIA; in the remaining 3 neurons, -agatoxin IVA reduced the -EMTBL effect by no more than
10%. These data show that the greatest effect of -EMTBL on
high-voltage-activated Ca++ current components was caused
by an action on N- and L-type channels, with the former occurring in a
voltage-dependent manner.
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This study investigated the regulation of voltage-dependent
Ca++ currents in cultured DRG neurons by alkyl-substituted
thiobutyrolactones. The results show for the first time that these
compounds affected both low-voltage- and high-voltage-activated
Ca++ currents in a concentration-dependent manner and in
the presumed active range of 200 to 400 µM (Canney et al.,
1991). A particularly novel result was that -EMTBL only reduced
low-voltage-activated (T-type) currents, but had a biphasic effect on
high-voltage-activated currents, producing an increase at low
concentrations (300 µM) and a decrease at higher concentrations. By
contrast, -EMTBL only reduced low-voltage- and
high-voltage-activated currents and -EMGBL had no effect on
Ca++ currents. The predominant effects of the
thiobutyrolactones were on the T-, N- and L-type current components,
with little effect on the P- and Q-type current components.
Furthermore, thiobutyrolactone-induced reductions in the N-type current
component were voltage dependent.
Butyrolactone structure and action on Ca++
channels.
Thiobutyrolactone actions were concentration dependent
and exhibited some selectivity for Ca++ channel subtypes.
These findings stand in contrast to the lack of effect of the
-butyrolactones and suggest that Ca++ channel subtypes
represent specific targets for the thiobutyrolactones. The inability of
-EMGBL to affect low-voltage- or high-voltage-activated currents, or
to block the actions of -EMTBL, for example, illustrates this
difference in binding sites. (This last result highlights a clear
difference in the effects of these compounds, at similar concentrations, on the GABAA receptor, at which -EMGBL
acts as an "antagonist" to -EMTBL [Holland et al.,
1990].) It is possible, however, that -EMGBL may bind at the
thiobutyrolactone site on Ca++ channels, but with a
markedly lower affinity or with less potency. Additional experiments
will be required to distinguish between these possibilities, but the
present results make clear that the identity of the ring heteroatom is
of primary importance in determining actions on Ca++
channels. Studies with additional analogs may allow firmer conclusions regarding the importance of the ring structure in targeting
Ca++ channels of different subtypes. Ethosuximide, for
example, which differs from -EMTBL in its ring heteroatom and in the
number of ring carbonyl groups, targets only T-type channels.
-EMTBL on T-type currents was
slightly more efficacious but less potent than for -EMTBL. Thus, for
low-voltage-activated currents, the position of the alkyl substitutions
of thiobutyrolactiones appears to alter apparent affinity. As for
high-voltage-activated currents, an interesting and novel result of the
present study was that the anticonvulsant -EMTBL only reduced
high-voltage-activated Ca++ currents, whereas the
convulsant -EMTBL had biphasic actions. The current enhancement,
seen at relatively low concentrations, was voltage independent, but
the current reduction at higher concentrations was voltage dependent,
similar to the effect of -EMTBL. This suggests that -EMTBL may be
acting either at two different sites on high-voltage-activated channels
or may have differing actions on separate channel types; additional
experiments will be required for clarification. For
high-voltage-activated channels, then, the position of the alkyl
substitution confers differing properties and imparts to
-substituted compounds an unusual characteristic, Ca++
current enhancement.
The structure of a given compound may determine its action on
particular Ca++ channel subtypes. Unlike many
Ca++ channel blockers, which show preferences for either
single-channel subtypes, such as -conotoxin GVIA for the N-type
channel (for example, Regan et al., 1991), or a relative
preference for either low-voltage- or high-voltage-activated channels,
such as Ni++ or Cd++, respectively (Fox
et al., 1987), the thiobutyrolactones affect both
low-voltage-activated (T-type) and high-voltage-activated Ca++ current components. Of the latter, both N- and L-type
current components were affected, with little apparent effect on the P- and Q-type current components. This last finding must remain tentative, however, because DRG neurons express relatively little P- and Q-type
current. In experiments with 500 nM -agatoxin IVA, for example, the
high-voltage-activated currents were reduced by only 10 to 15% by this
toxin, similar to the findings of others who used this cell type (Mintz
et al., 1992; Regan et al., 1991). Thus, it may
have been difficult to determine with certainty whether and to what
extent these current components are affected by thiobutyrolactones. Additional experiments will be required, with either expressed channels
or cells, such as cerebellar Purkinje or granule cells, which express a
greater proportion of P- and Q-type channels.
Voltage dependence of thiobutyrolactone action.
The
thiobutyrolactone effects on T- and L-type currents did not appear to
be voltage dependent as did the -conotoxin GVIA-sensitive effects on
the N-type current component. The mechanism of this voltage-dependent
effect on N-type channels is uncertain, although the present results
suggest possible alternative explanations. The finding that
thiobutyrolactones, applied at Vh = 80 mV, had little
effect on peak currents, but increased the rate of apparent current
inactivation at +10 mV, suggests that the major effect may have been on
open channels. This hypothesis predicts that Ca++ channels
needed to be in the open state for the thiobutyrolactones to effect a
reduction in current, either by an inhibitory effect on channel
activity or by blocking open channels. The finding that peak currents
were substantially reduced by thiobutyrolactones applied at
Vh = 40 mV, a potential at which high-voltage-activated channels were not open, counters this hypothesis and suggests, rather,
that closed channels can be affected by these compounds. Further, the
fact that inactivation rates of currents evoked from 40 mV were not
substantially affected by thiobutyrolactones suggests that the open
state of high-voltage-activated channels may not be the only channel
conformation to be affected by the thiobutyrolactones. Indeed, if
"open channel block" were the only or predominant mechanism of
action, a shift in the apparent ECa would be expected,
which is not evident in the present results.
40 mV. Our findings
that thiobutyrolactone effects on high-voltage-activated currents were relatively low in affinity and that onset and reversal of effects was
rapid support this view. Alternatively, thiobutyrolactones may bind in
a voltage-independent fashion, but may change the voltage-dependent
properties of the target channel, resulting in an increase in
steady-state inactivation and/or inactivation at depolarized
potentials. Clearly, additional experiments are needed, perhaps with
expressed N-type channels, for a fuller explanation of the mechanism of
action of these compounds and their voltage dependence on this channel
subtype.
Anticonvulsant action and the block of Ca++
channels.
Butyrolactone analogs afford an opportunity to place in
perspective the role of Ca++ channel blockade as an
anticonvulsant mechanism of action. Some antiseizure drugs currently in
use have the ability to block Ca++ currents, although in
most cases it is not clear whether this action is either necessary or
sufficient or even contributes to clinical efficacy. Phenytoin, for
example, reduces synaptosomal Ca++ uptake, but does so at
concentrations above the therapeutic range (Sohn and Ferrendelli,
1976). Similarly, the barbiturates markedly reduce
high-voltage-activated Ca++ currents, but at concentrations
that are relevant only in the treatment of status epilepticus, when
anesthetic concentrations are achieved (Gross and Macdonald, 1988a,b).
The Ca++ channel block is substantial at these
concentrations, however, and may account for the cardiovascular
suppression commonly seen when patients are under barbiturate
anesthesia. The reduction of T-type currents by ethosuximide is the
best example of an antiseizure agent that acts on Ca++
channels at therapeutic concentrations; to date, T-type current reduction is probably the most likely mechanism for ethosuximide's efficacy in treating primary generalized (absence) seizures (Coulter et al., 1989). Valproic acid may also reduce T-type
currents, although whether this contributes to its broad antiseizure
efficacy is not certain (Kelly et al., 1990).
). The effects on high-voltage-activated Ca++
currents are more complicated. It is worth noting that -EMTBL, a
convulsant compound, not only blocks T-type Ca++ currents
but also increases high-voltage-activated Ca++ currents, at
concentrations that reduce GABAA-mediated inhibition. At
high concentrations, -EMTBL blocks both low-voltage and
high-voltage-activated Ca++ currents, but also
increases GABAA-mediated inhibition (Holland et al., 1995). It would seem likely that, at low
concentrations, the block of GABAA inhibition by -EMTBL
is by far the more significant clinical (convulsant) effect than its
potentially anticonvulsant reduction of T-type Ca++
currents. It is also possible that enhancement of
high-voltage-activated currents may be contributing to the convulsant
actions of this compound. As for -EMTBL, its anticonvulsant action
is likely caused by enhancement of GABAA receptor activity,
but may be caused in part by Ca++ current reduction,
particularly of T-type currents. It will be interesting to determine
whether -EMTBL, like ethosuximide, is effective for control of
absence seizures.
| |
Acknowledgments |
|---|
The authors thank Robert W. Subiaga, Jr. and Mark J. Gallagher for expert technical assistance.
| |
Footnotes |
|---|
Accepted for publication October 18, 1996.
Received for publication June 3, 1996.
1 Supported in part by National Institutes of Health grants NS19613 (to R.A.G.) and NS14834 (to D.F.C. and J.A.F.).
Send reprint requests to: Robert A. Gross, M.D., Ph.D., Departments of Neurology and Pharmacology and Physiology, University of Rochester, Box 711, 601 Elmwood Avenue, Rochester, NY 14642.
| |
Abbreviations |
|---|
-EMGBL, -ethyl,-methyl--butyrolactone;
-EMTBL, -ethyl,-methyl--thiobutyrolactone;
-EMTBL, -ethyl,-methyl--thiobutyrolactone;
ATP, adenosine
triphosphate;
DRG, dorsal root ganglion;
GABAA, -aminobutyric acid;
HEPES, N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid.
| |
References |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
-butyrolactones and -thiobutyrolactones: ligands for the picrotoxin receptor.
J. Med. Chem.
34: 1460-1467, 1991[Medline].-butyrolactones and -thiobutyrolactones in cultured hippocampal cells.
J. Neurosci.
10: 1719-1727, 1990[Abstract].-aminobutyric acid type A receptor/ionophore by alkyl-substituted -butyrolactones.
Mol. Pharmacol.
47: 1217-1223, 1995[Abstract].-, -, and -,-substituted -butyrolactones.
Mol. Pharmacol.
22: 438-443, 1982a[Abstract].-alkyl-substituted -butyrolactones.
Mol. Pharmacol.
22: 431-437, 1982b[Abstract].-butyrolactones bind at the picrotoxinin/t-butyl-bicyclophosphorothionate (TBPS) receptor.
Biochem. Pharmacol.
34: 4187-4190, 1985[Medline].-AGA-IVA.
Nature
355: 827-829, 1992[Medline].-aminobutyric acid receptor channels by diazepam and phenobarbital.
Ann. Neurol.
25: 213-220, 1989[Medline].-butyrolactones inhibit [35S]TBPS binding to a GABA-linked chloride ionophore.
Eur. J. Pharmacol.
105: 195-196, 1984[Medline].This article has been cited by other articles:
|
|
 |
|
  S. M. Todorovic, E. Perez-Reyes, and C. J. Lingle Anticonvulsants But Not General Anesthetics Have Differential Blocking Effects on Different T-Type Current Variants Mol. Pharmacol., July 1, 2000; 58(1): 98 - 108. [Abstract] [Full Text]
|
|
|
|
 |
|
 P. Salvati, R. Maj, C. Caccia, M. A. Cervini, M. G. Fornaretto, E. Lamberti, P. Pevarello, G. A. Skeen, H. S. White, H. H. Wolf, et al. Biochemical and Electrophysiological Studies on the Mechanism of Action of PNU-151774E, A Novel Antiepileptic Compound J. Pharmacol. Exp. Ther., March 1, 1999; 288(3): 1151 - 1159. [Abstract] [Full Text]
|
|
|
|
 |
|
 A.-M. Autere, K. Lamsa, K. Kaila, and T. Taira Synaptic Activation of GABAA Receptors Induces Neuronal Uptake of Ca2+ in Adult Rat Hippocampal Slices J Neurophysiol, February 1, 1999; 81(2): 811 - 816. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Todorovic and C. J. Lingle Pharmacological Properties of T-Type Ca2+ Current in Adult Rat Sensory Neurons: Effects of Anticonvulsant and Anesthetic Agents J Neurophysiol, January 1, 1998; 79(1): 240 - 252. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
) and 
) over the
stated ranges of concentrations. See text for details.
) or 
) at +10 mV in the absence or presence of 500 µM
) or 
