- Email: [email protected]
Electrophysiological effects of aconitine in rat hippocampal slices
Pergamon
002%3908(95)00153-0
Neurophormacology, Vol. 35, No. 1, pp. lS22, 1996 Copyright0 19% Ekvier Science Ltd Printed in Great Britain. All rights resewed ocm-3908% flS.oo + 0.00
Electrophysiological Effects of Aconitine in Rat Hippocampal Slices A. AMERI,‘* Q. SHIT J. ASCHOFF2 and T. PETERS’ ‘Department of Pharmacy and Pharmacology of Natural Compounds, University of Ulm, Helmholtzstr. 20, D-89081 Ulm, Germany and 2Department of Neurology, University of Ulm, Steinhovelstr. 9, D-89075 Ulm, Germany (Accepted 29 August 1995) Summary-The electrophysiological effects of aconitine were investigated in the rat hippocampal slice and compared with those of veratridine. Both alkaloids are known to bind at site 2 of sodium channels and to block its inactivation. Extracellular recordings revealed that aconitine and veratridine exert inhibitory effects on neuronal excitability. Aconitine slowly and reversibly decreased the population spike recorded in the CA1 pyramidal cell layer. The reduction of the spike amplitude was similar whether orthodromically or antidromically activated. The aconitine-induced inhibition did not differ from that of veratridine. However, following washout of aconitine, the amplitude of the antidromic spike was increased compared to the control amplitude. The veratridine-induced inhibition was only partially reversible. This inhibition was also observed during suppression of synaptic transmission by a low Ca*‘/high Mg*+-medium, indicating an inhibition of axonal conductance. The results show that in the absence of synaptic transmission the antidromic (alvear) spike is more sensitive to the inhibitory action of aconitine than the presynaptic fiber spike elicited by stimulation of the Schaffer collaterals. Furthermore, it is shown that aconitine acts in an activity-dependent manner, in that the latency of onset of the inhibition is prolonged when the stimulation frequency is decreased. Field excitatory postsynaptic potentials were also suppressed by aconitine, whereas excitatory postsynaptic currents recorded by the patch clamp technique were not influenced by aconitine when cells were held at - 60 mV. KeywordsSodium
channel, aconitine, hippocampus, veratridine.
in a series of diseases. There is evidence that a functional defect in the inactivation gate is involved in the pathophysiological mechanisms leading to epilepsy and migraine (Hansen and Lauritzen, 1984). Furthermore, recent studies have shown a Na+ overload under conditions of an energy deficit in neurons affected by ischemia (Prenen et al., 1988; Pauwels et al., 1991; Tegtmeier et al., 1992; Kiskin et al., 1993). These findings suggest that a modulation of sodium channels might be neuroprotective or might ameliorate damage in the diseases mentioned above. Voltage-dependent sodium channels have three separate receptor sites that modulate channel function. The alkaloids aconitine, veratridine and batrachotoxin bind to site 2 on the sodium channel (Catterall, 1980) and induce a persistent activation of the channel by blocking channel inactivation. Aconitine and veratridine are known to depolarize neuronal membranes, to shift the activation of the Na current to more hyperpolarized potentials, and to decrease the maximal inward current (Ulbricht and Flacke, 1965; Schmidt and Schmitt, 1974; Mozhayeva
The increase in sodium permeability resulting from depolarization of excitable cells is biphasic. These channels open to conduct sodium ions when the cells are depolarized (activation) and subsequently enter a closed state (inactivation) with a return of sodium permeability to the resting level during a maintained depolarization. The sodium channel can therefore exist in three functionally distinct states: resting, active and inactivated (Catterall, 1992). Both the resting and inactivated states are non-conducting. Thus, the inactivation of the sodium channel controls the return of the sodium permeability to the resting level. The loss of this control mechanism causes an increased influx of sodium ions, which in turn depolarizes the cell membrane and causes more sodium channels to open. Finally, due to this positive feedback, calcium-ions enter the cell via voltagedependent calcium channels leading to a rapid Ca*+ overload. In fact, functional alterations or mutations of the inactivation gate of the sodium channel are involved
*To whom correspondence
should be addressed.
et al., 1977; Cartneliet, 13
1991). Aconitine
is reported
to
A. Ameri et al.
14
induce a maintained Na current in myelinated nerve axons (Schmidt and Schmitt, 1974; Mozhayeva et al., 1977), but it does not do so in unmyelinated axons (Warashina, 1985). Recently, it was shown that aconitine and its derivates have antinociceptive properties (Suzuki et al., 1994). In the heart, aconitine was shown to prolong the Na influx during the action potential and to induce positive inotropic and arrhythmogenic effects (Honerjager and Meissner, 1983). The aim of the present study is to elucidate how aconitine influences neuronal excitability of a highly vulnerable brain region, the hippocampus. We took advantage of synchronously discharging action potentials by a large population of pyramidal neurons, which are recorded extracellularly as distinct ‘population spikes’ in the rat hippocampal slice. In order to compare the effects of aconitine with effects of drugs known to bind to site 2 of the sodium channel, we have also performed experiments with veratridine. METHODS Male Wistar rats (45-55 days) were anesthetized with ether and decapitated. The brain was quickly removed and chilled in ice-cold artificial cerebrospinal fluid (ACSF). After 1 min, the hippocampus of one hemisphere was removed and dissected free of surrounding tissue. Transverse slices of 400 pm thickness were cut transversely to the longitudinal axis of the hippocampus using a McIlwain tissue chopper. Immediately after cutting, one slice was transferred to the recording chamber where it was kept submerged and held down on a nylon net by a U-shaped piece of flattened platinum wire. The chamber temperature was slowly increased from room temperature to 28°C during a period of 30 min. Recordings were performed l-8 hr after dissection. The other slices were maintained in an incubation chamber at room temperature. The standard ACSF was continuously gassed with a mixture of 95% Od5% COz and contained (in mM): NaCl 124, KC1 3, NaHzP04 1.25, NaHCOs 26, CaClz 2.5, MgS04 2, glucose 15 at a pH of 7.4. In particular experiments, a modified ACSF was used in which no Ca*+ was added and the Mg*+ concentration was increased to 4 mM. The ACSF was perfused at 2 ml/ min. For extracellular recordings borosilicate glass capillaries (5-10 RM) were filled with 3 M NaCl. The recording electrode was connected via a chlorided silver wire to the probe of the amplifier. Recordings of stimulus-evoked population spikes were made from the stratum pyramidale of area CAl, recordings of field excitatory postsynaptic potentials (field EPSPs) were made from the stratum radiatum. For electrical stimulation, a concentric bipolar electrode with 0.25 mm o.d. (Rhodes Medical Instruments) was positioned in the Schaffer collateral commissural pathway (i.e. near the junction of CA1 and CA2 stratum radiatum), or in the alveus, for orthodromic and antidromic activation of CA1
pyramidal cells, respectively. Extracellular stimuli were rectangular current pulses of 200 psec in duration delivered every 15 set (in particular experiments every 30 or every 60 set). The stimulus strength ranged between 0.1 and 1.5 mA. Drug effects were investigated on population spikes elicited by using a half-maximal stimulus strength. The extracellular signals were recorded and amplified by a DP 301 amplifier (Warner Instruments, U.S.A.). Analog data were digitized and analysed with a TIDA system (HEKA electronic, Germany). For whole-cell patch clamp recordings only slices (300 pm) from male Wistar rats (l&23 days) were used. To obtain access to CA1 neurons, a part of the alveus and stratum oriens was removed by a saline jet from a micropipette with a large tip opening as described elsewhere (Garaschuk et af., 1992). Thereafter it was possible to obtain visually controlled high resistance electrical contact with the soma of CA1 neurons using the standard whole-cell patch clamp technique. Patch electrodes were pulled from filamented borosilicate glass capillaries of o.d. 1.5 mm (Clark Electromedical Instruments, U.K.). When filled with the internal solution (mM: CsF 125, Tris-Cl 10, HEPES 10, EGTA 10, pH 7.2), the tip resistance was 24 M. Excitatory postsynaptic currents (EPSCs) were elicited by stimulation of Schaffer collaterals and recorded using an Axopatch 200 A amplifier (Axon Instruments, U.S.A.). All drugs were applied by addition to the ACSF. Aconitine and veratridine were dissolved in 20% dimethyl sulfoxide (DMSO) to give stock solutions of 1 mM. This solution was diluted with ACSF to reach final concentrations of between 0.01 and 1 PM. The final DMSO concentration for the highest aconitine concentration tested (i.e. 1 PM) was 0.02%. Aconitine was purchased from Fluka (Germany), veratridine and tetrodotoxin (‘FIX) from Sigma (Germany). Values are given as mean + standard deviation (SD). Significance was assumed when P d 0.05. Statistical evaluation was performed using the Student’s r-test. The amplitude of the population spike was determined from the negative peak to a tangent drawn between the first and the second maximum positives. RESULTS The effects of aconitine on evoked field potentials in area CA1 of the rat hippocampus were investigated in 57 slices. Furthermore, in four CA1 neurons EPSCs were recorded by the patch clamp technique. Veratridine was examined in 16 slices. Recordings were only included in the data analysis if electrical stimulation at a maximal stimulus strength did not evoke a second population spike, and where the amplitude of the population spike was stable for a control period of at least 30 min. Aconitine (1 PM) slowly and reversibly decreased the amplitude of the orthodromically evoked population spike within 20 min [Fig. l(A)]. During the first l-2 hr of
Electrophysiological
effects of aconitine in rat hippocampal slices
15
ACONITINE (1 pM)
time
(min)
C
0
0
I
I
I
30
60
90
120
time (mill)
Fig. 1. Inhibitory action of aconitine (1 PM) on the orthodromic population spike recorded extracellularly in the CA1 pyramidal cell layer. (A) Population spikes were evoked by electrical stimulation of the Schaffer collateral pathway. The stimulus artefact is followed by the presynaptic fiber spike and the population spike. Each trace represents the average of 5 consecutive responses. (B) Time-course of the action of aconitine on the population spike. The filled bar indicates when aconitine was applied. The afferents were stimulated every 15 sec. The diagram depicts the measurements of the amplitude of the population spike in response to each stimulus. (C)Timecourse of the action of aconitine on the presynaptic fiber spike. During the 30 min application, the presynaptic fiber spike was resistant to an action of aconitine. Recordings in A and results in B and C were obtained from the same slice.
washout, a second population spike occurred in about half of the slices. In each of these slices, these additional spikes disappeared during prolonged washout. The presynaptic fiber spike (afferent volley) which represents the compound action potential generated in synchronously discharging axons was apparently unaffected
during the 30 min application of aconitine [Fig. l(B)]. In order to identify the small deflection preceding the postsynaptic population spike as the presynaptic fiber volley, 30 min after the start of the perfusion with aconitine the sodium channel blocker TTX (1 FM) was added. In every slice (n = 3) the spike was blocked by
16
A. Ameri et al.
-I orthodromic PS
antidromic PS
antidromic PS (after washout)
Fig. 2. Comparison of the inhibitory action of aconitine and veratridine on the orthodromic and antidromic population spike (PS). Both drugs were bath-applied at a concentration of 1 PM for 30 min. During this period the presynaptic fiber spike remained unchanged. The columns on the right represent the size of the antidromic population spike after washout of the drugs. The values after washout of aconitine and veratridine were significantly different (P < 0.01). Data are mean values +_ SD of at least 4-5 experiments.
fiber spike. The action of veratridine (1 PM) on the orthodromic population spike (n = 4) did not differ significantly from aconitine in this respect (Fig. 2). Both drugs reduced the antidromic population spike which is elicited by direct, alvear stimulation of CAl, in the same manner and time course as the orthodromic population spike (Figs 2 and 3). Immediately following application of aconitine, there was usually a slight and short-term increase in the amplitude of both the orthodromically and the antidromically evoked population spike (up to 15%). This increase was followed by a rapid decrease in the spike amplitude. The inhibitory effect of aconitine on the population spike was slowly reversible during a prolonged washout and showed a large overshoot (Fig. 3) which was more pronounced after the application of higher concentrations. The amount of the overshoot obtained after 4-5 hr washout of 1 PM aconitine was 137.2 + 11.4% of control (n = 5) thus significantly higher than that obtained after washout of 0.1 PM aconitine (116.4 f 6.9%, n = 3). During washout of veratridine only a partial recovery (68.7 f 7.1%, II = 3) was achieved (Fig. 2). Both the extent of inhibition and the latency of its onset was dependent on the concentration of aconitine used. The dose-response relationship of aconitine (Fig. 4) was examined by application of a single concentration to each slice. The ICse for the inhibitory effect of aconitine on the antidromic population spike was 0.03 ,uM. In response to antidromic stimulation synaptic activalTX,
indicating
that it was the presynaptic
tion of CA1 pyramidal cells via recurrent loops or by orthodromic stimulation of basal dendritic afferents might occur, and could contribute to the generation of the evoked potentials observed in these experiments. In order to eliminate the possible contribution of presynaptic mechanisms to the aconitine-induced action, synaptic transmission can be blocked by a low Ca2+/high Mg2+ medium (Andersen et al., 1978). The addition of aconitine (1 ,uM) to a low Ca2+/high Mg2+ medium attenuated the amplitude of the antidromic population spike by 92.0 + 1.8% (n = 4) and did not differ significantly from the inhibition caused by 1 PM aconitine added to normal ACSF (92.0 f 1.8%, II = 5). The aconitine-induced reduction of the antidromic population spike observed even in the absence of synaptic transmission suggests that it is mediated by a change in axonal excitability. The antidromic population spike results from synchronous activation of alvear fibers, whereas the presynaptic fiber volley which precedes the postsynaptic population spike in response to orthodromic stimulation is the compound action potential resulting from a synchronous discharging of the Schaffer collaterals (Andersen et al., 1978). As reported above (Fig. l), aconitine diminished the postsynaptic spike evoked by stimulation of the Schaffer collaterals (i.e. synaptic activation) while apparently sparing the presynaptic fiber volley. In order to investigate an effect of aconitine on the presynaptic fiber spike exclusively and in more detail, we again used a low Ca2+/high Mg2+ ACSF to block synaptic transmission. In these conditions, orthodromic stimulation of the Schaffer collaterals elicites only the presynaptic fiber spike but no postsynaptic spike, while the antidromic (aivear) population spike remains unchanged. Following suppression of the postsynaptic spike, we recorded the presynaptic fiber spike elicited by orthodromical stimulation of the Schaffer collaterals for a further 30 min control period before aconitine (1 PM) was applied. At this concentration, aconitine reduced the presynaptic fiber spike by only about 65.9 + 1.8% (n = 4). However, maximal inhibition was achieved 148.8 + 27.2 min after onset of the perfusion with aconitine. In contrast to this, the antidromically evoked (alvear) population spike was attenuated by aconitine (1 PM), added to a low Ca2+/high Mg2+ medium by about 92.1 f 1.8% (n =4) with a latency of 27.5 f 8.7 min after the start of the perfusion with aconitine. Figure 5 shows the time-course of the inhibition in two representative recordings, an antidromically evoked population spike [Fig. 5(A)] and an orthodromically evoked spike [Fig. 5(B)]. Both recordings were performed in a low Ca2+/high Mg2+ medium. However, during prolonged washout with the low Ca2+/ high Mg2+ medium, the amplitude of the antidromic population spike no longer reached the control value [Fig. 5(A)]. The maximal recovery of the antidromic spike recorded in the absence of extracellular Ca2+ was
Electrophysiological
A
CONTROL
17
effects of aconitine in rat hippocampal slices ACONITINE (1 pM)
WASH
I-- Ii---l
10ms
> E
2
time (min) Fig. 3. Inhibitory action of aconitine (1 PM) on the antidromic population spike. (A) Population spikes were elicited by electrical stimulation of alvear fibers. Each trace is the average of 5 consecutive events. (B) Timecourse of the action of aconitine on the antidromic population spike. Each point on the graph represents the measurement of the amplitude in response to a single stimulus. The time of application of aconitine is indicated by the filled bar. Note the marked overshoot during washout. Recordings in A and results in B were obtained from the same slice.
12.2% (n = 4). This stands in contrast to the recordings performed in normal ACSF showing a marked overshoot (137.2 + 11.4%, n = 5) of the antidromic population spike amplitude [Fig. 3(B)] during washout. The discrepancies in the extent of the overshoot observed during washout of aconitine by standard ACSF and by low Ca +-ACSF, raises the question of whether it is due to a blockade of inactivation, or if it is a consequence of the Ca2+ influx caused by these drugs (Ashton et al., 1990). In order to allow a long-lasting CaZt influx we applied aconitine (1 ,uM) for a period of 2 hr. In all of the slices tested (n = 3) the inhibitory effect of aconitine was reversible. The overshoot of the antidromic spike amplitude did not differ from the results obtained during washout of aconitine following a 30 min application. The inhibitory effect of aconitine (0.03 PM) on the antidromically evoked population spike recorded in the 35.0 f
CA1 region of the hippocampus is decelerated by decreasing the stimulation frequency (Fig. 6), consistent with its activity-dependent action in other preparations (Honejager and Meissner, 1983). A reduction of the stimulation frequency from 4 stimulus pulses per min [Fig. 6(A)] to 2 pulses per min [Fig. 6(B)], and to 1 pulse per min [Fig. 6(C)] increased the time for the maximal aconitine-induced inhibition to be achieved. In contrast to the longer duration, the maximal percentage of inhibition of the population spike by aconitine is unaltered. It is emphasized that the dependence on the stimulation frequency can be overcome by aconitine used at maximally effective concentrations (0.3 and 1 PM). Since the postsynaptic population spike was completely blocked by aconitine (Fig. l), recordings of the field EPSP were performed (Fig. 7) to examine the effect of aconitine on synaptic transmission. Aconitine (1 PM) depressed the field EPSP recorded in the CA1 stratum
18
A. Ameri et al. radiatum by 84.4 + 47.0% (n = 5). Maximal inhibition of the field EPSP was achieved 44 + 6.5 min following
onset of the application of aconitine. The effect on the field EPSP was only partially reversible during washout of aconitine. In order to confirm the results of the field data at a cellular level, patch clamp experiments on single pyramidal neurons in the hippocampal slice were performed. The advantage of patch clamping in slices is that the synaptic connections are preserved. The peak amplitude and duration of EPSCs recorded in response to electrical stimulation of the Schaffer collaterals (Fig. 8) remained unaltered by aconitine (1 PM) (n = 4).
I
I
I
I
-0
-7
-6
-5
concentration (log; M)
Fig. 4. Concentration-response curve for aconitine in rat hippocampal slices. For each curve, data were obtained from measurements performed in 4 and 5 different slices. Aconitine was applied at each concentration to a single slice. Slices were stimulated antidromically every 15 sec. The amplitude of the population spike (PS) was normalized with respect to control and plotted as a function of the logarithm of the aconitine concentration. Data points represent mean values + SD. In cases where data points are lacking error bars, the standard deviation is smaller than the size of the symbol.
DISCUSSION The major finding of the present study is an inhibitory effect of aconitine and veratridine on the excitability of rat hippocampal slices. At a concentration range of O.Ol1 PM, aconitine caused a slow and reversible decrease of the amplitude of the population spike recorded in the CA1 stratum pyramidale evoked by both orthodromic (Fig. 1) and antidromic (Fig. 3) stimulation. The inhibition of the orthodromic and antidromic population spike induced by 1 PM aconitine does not differ from the inhibition induced by 1 PM veratridine (Fig. 2). Immediately upon application of both drugs, the amplitude of the population spike was slightly increased. This initial enhancement of the population spike has also been observed by others (Ashton et al., 1990) following
A
120
time (min)
0
120
240 time (min)
360
Fig. 5. Time-course of the inhibitory effect of aconitine (1 PM) in the absence of synaptic transmission. (A) The slice was stimulated antidromically every 15 sec. (B) The slice was stimulated orthodromically. Since the recording was performed in a low Ca*“/high Mg*+-medium the population spike was suppressed. The data points represent measurements of the presynaptic fiber spike which is the compound action potential of the Schaffer collaterals. Note the longer time of application of aconitine (filled bars) in B.
Electrophysiological
{*50;,
,
0
,lT”,
120
,
,“p”.
240
,
,‘p”,
360
,
effects of aconitine in rat hippocampal slices
,4i
480
time (min)
Fig. 6. Dependence of the latency of onset of the aconitineinduced effect on the stimulation frequency. The slices were stimulated antidromically with 4 stimulus pulses per min (A), 2 pulses per min (B), and 1 pulse per min (C), respectively. In each experiment, aconitine was applied at a concentration of 0.03 PM at the time marked by the bar above the graph. Note the longer duration of drug-application and the later onset of inhibition with decreasing stimulation frequency.
blockade of sodium channel inactivation by veratridine. These authors report that veratridine both decreases the extracellular concentration of Na+ and Ca2+ and increases the extracellular concentration of K+, accompanied by a persistent depolarization. The veratridine-induced Na+ influx is thought to be the result of the’ binding of veratridine to site 2 of the Na+ channel (Catterall, 1980), the common binding site for veratridine and aconitine. These findings suggest that the initial increase in spike amplitude reflects blockade of the inactivation of voltagedependent sodium channels. Due to the prolonged activation, the membranes are depolarized by sustained sodium influx (Schmidt and Schmitt, 1974; Mozhayeva et al., 1977). Nodal membranes of myelinated nerve fibres of Xenopus Zuevis are depolarized by 10-15 mV, leading finally to complete inexcitability (Schmidt and Schmitt, 1974). Furthermore, there: is evidence that aconitine and veratridine, in addition to their action on inactivation, shift the voltage-dependence for the activation of the Na+ current in a hyperpolarizing direction and reduce the maximal inward current (Ulbricht and Flacke, 1965;
19
Schmidt and Schmitt, 1974; Mozhayeva et al., 1977; Carmeliet, 1991). These findings explain the suppression of the population spike by aconitine and veratridine observed in the present study. Furthermore, we observed that during the first 2 hr of washout of aconitine about half of the slices developed a second population spike, which might be a result of a shift of potassium from the intracellular to the extracellular space (Ashton et al., 1990). Alternatively, the second transient, population spike might reflect a shift of the activation parameters in a hyperpolarizing direction (Ulbricht and Flacke, 1965; Schmidt and Schmitt, 1974; Mozhayeva et al., 1977; Carmeliet, 1991) so that less depolarization is required to activate the Na+ channel; the threshold will be attained earlier and more easily, which explains the repetitive activity. The antidromic (alvear) population spike is attenuated to the same extent and time course as the orthodromically (synaptically) evoked population spike by aconitine (1 ,uM). The antidromic population spike is produced by electrical activation of the axons of the neurons. Since this potential is unaffected by drugs that modify synaptic transmission (Dunwiddie, 1986), the present results indicate that aconitine inhibits preferentially axonal conduction. During washout of aconitine, the amplitude of the population spike was significantly increased compared to control. This overshoot was concentration-dependent and was also observed when aconitine was applied for 2 hr in order to allow a sustained Ca2+ influx. When aconitine was applied in the absence of Ca2+, however, there was neither an overshoot nor a full recovery of the spike amplitude. Veratridine failed to produce an overshoot of the spike amplitude during washout (Fig. 2). In a previous study, veratridine, in contrast to aconitine, is reported to cause a reduction in the maximal intensity of the inward current, which is persistent during washout (Warashina, 1985). These findings imply that both drugs have a different mode of interaction with Na+ channels, which is in accordance with the results of the present study. Further experiments are required to clarify the mechanisms underlying the increase in spike amplitude after washout of aconitine. The most intriguing finding of the present study is the difference in the extent of the aconitine-induced inhibition of the fiber spike amplitude evoked by orthodromic (Schaffer collaterals) and antidromic (alvear) stimulation (Fig. 5). By use of TTX, we have identified the negative deflection preceding the orthodromic population spike as the presynaptic fiber spike which represents the compound action potential of the Schaffer collaterals. While a previous study failed to show an effect of aconitine in concentrations of up to 1 PM on the presynaptic fiber spike (Worley and Baraban, 1987) the present study clearly demonstrated that both alvear fibers and Schaffer collaterals are sensitive to aconitine. These authors applied aconitine for 30 min, a period which in the
20
A. Ameri et al.
A
CONTROL
WASH
ACONITINE(lpM)
B
0 0
I 60
I I 120 180 time (min)
1 240
300
Fig. 7. Inhibitory action of aconitine (1 PM) on the field excitatory postsynaptic potential (EPSP). (A) Field EPSPs recorded extracellularly in the apical dendritic region of area CAL Schaffer collaterals were stimulated every 15 sec. Each curve is the average of 4 subsequent responses. (B) Time-course of the peak amplitude of the field EPSP. Each point represents the mean of four consecutive traces. The filled bar indicates when aconitine was applied. Recordings in A and results in B were obtained from the same slice.
present study also failed to exert an inhibition presynaptic fiber spike. However, inhibition
of the
of the presynaptic fiber spike is obtained after long-term application of aconitine [Fig. 5(B)]. At a concentration of 1 ,uM, which caused maximal inhibition of the alvear spike, aconitine reduced the amplitude of the presynaptic fiber spike by about 65%. The difference in sensitivity to aconitine might reflect differences in myelination. The Schaffer collaterals are not myelinated whilst the alvear fibers are (Andersen et al., 1978). It is well known that sodium channels are highly concentrated in the membrane at the nodes of Ranvier as compared to unmyelinated axons. The high density of sodium channels in the nodes of Ranvier is the basis of the low threshold and generation of action potentials at this site (Hille, 1992). Differences in the action of aconitine on myelinated and nonmyelinated nerve fibers, have been reported previously by others (Schmidt and Schmitt, 1974; Mozhayeva et al., 1977; Warashina, 1985) suggesting different modes of interactions between toxin molecules and open
Na+ channels. In unmyelinated
axons, aconitine did not induce a Na+ current which was maintained, although it reduced the current to a large extent (Warashina, 1985). This contrasts with the effect of aconitine on myelinated nerve axons, in which a Na+ current which is maintained and a shift of the activation in a hyperpolarizing direction was observed (Schmidt and Schmitt, 1974; Mozhayeva et aZ., 1977). Alternatively, aconitine might interact differentially with various Na+ channel subtypes. Indeed, Westenbroek et cd., (1989) have shown a differential localization of Rr and Rn Na+ channel subtypes in the hippocampus pyramidal cell layer and on Schaffer collaterals, respectively, Whether the differences observed in the present study are due to myelination and non-myelination, or to different subtypes of the voltagedependent sodium channel cannot be decided at present. An activity-dependent effect of aconitine was obvious in our study since the inhibitory action of aconitine (0.03 PM) depends on the stimulation frequency (Fig. 6). While a reduction of the stimulation frequency leads
Electrophysiological
21
effects of aconitine in rat hippocampal slices
A ACONlTINE (1 pM)
CONTROL
WASH
B
01
0
I
I
60
120
I 180
time (min) Fig. 8. EPSCs record’ed in a CA1 pyramidal cell evoked by stimulation of the Schaffer collaterals. The CA1 neuron was held at - 60 mV. (A) Application of aconitine (1 PM) did not irdluence the EPSC. Each curve is the mean of 3 sequential traces. (B) Time-course of the peak amplitude of the EPSC. The filled bar indicates when aconitine (1 PM) was applied. The Schaffer collaterals were stimulated every 10 sec.
in the rate of the aconitine-induced inhibition of the population spike, the extent of the inhibition remains unaffected. At maximally effective concentrations (i.e. 0.3 and 1 PM), aconitine failed to show frequency-dependence. This suggests that higher concentrations of aconitine compensate for fewer stimulus pulses. It is likely that by using higher concentrations the probability of finding open sodium channels is still high enough to elicit maximal blockade of inactivation. There was no effect d aconitine on the excitatory postsynaptic currents (EPSCs) recorded in the CA1 neurons by the patch clamp technique (Fig. 8). However, as the membrane potential was clamped at -60 mV, we received no information on aconitine-induced changes in the membrane potential which in turn could influence synaptic transmission. In contrast, field EPSPs recorded in the CA1 stratum radiatum were depressed by aconitine. The field EPSP is thought to reflect synaptic currents in the dendrites as a result of the action of the neurotransmitter (Dunwisddie, 1986). The time-course of the inhibitory action of aconitine on the field EPSPs is comparable with the suppression of the population spike
to
a decrease
amplitude. However, the inhibition of the field EPSPs was only partially reversible. In conclusion, the present results show that aconitine has, in addition to its inhibitory action on axonal conductance, an inhibitory effect on excitatory synaptic transmission. It is not yet clear if the various effects of the toxin described in the present study are derived from a unique mechanism comprising a very specific interaction between a toxin molecule and a Na+ channel, or if they are induced by different interactions. Further experiments are required to elucidate the effects of aconitine on the membrane potential of CA1 neurons and the mechanisms involved in the overshoot of the spike amplitude after washout. Acknowledgement-The authors Strack for computer support.
are grateful
to Sebastian
REFERENCES Andersen P., Silvenius H., Sundberg S. H., Sveen 0. and Wigstrom H. (1978) Functional characteristic of myelinated fibers in the hippocampal cortex. Brain Res. 144: 11-18. Ashton D., Willems R., Marrannes R. and Janssen P. A. J.
22
A. Amel ri et al.
(1990) Extracellular ions during veratridine-induced neurotoxicity in hippocampal slices: neuroprotective effects of flunarizine and tetrodotoxin. Brain Rex 528: 212-222. Carmeliet E. (1991) Ion channels agonists: expectations for therapy. Eur. Heart J. 12: 30-37. Catterall W. A. (1980) Neurotoxins that act on voltagesensitive sodium channels in excitable membranes. A. Rev. Pharm. Toxic. 20: 15-43. Catterall W. A. (1992) Cellular and molecular biology of voltage-gated sodium channels. Physiol. Rev. 72: 15-48. Dunwiddie T. V. (1986) The use of in vitro brain slices in neuropharmacology. In: Electrophysiological Techniques in Pharmacology. Alan R. Liss, New York. Garaschuk O., Kovalschuk Y. and Krishtal 0. (1992) Adenosine-dependent enhancement by methylxanthines of excitatory synaptic transmission in hippocampus of rats. Neurosci. Lett. 135: 10-12. Hansen A. J. and Lauritzen M. (1984) The role of spreading depression in acute brain disorders. Ann. Acad. Bras. Cien. 54: 457479. Hille B. (1992) Ionic channels of excitable membranes. Sinauer Associates, Sunderland. Honerjager P. and Meissner A. (1983) The positive inotropic Naunyn-Schmiedeberg’s Archs. effect of aconitine. Pharmac. 322: 49-58. Kiskin N. I., Chizhmakov I. V., Tsyndrenko A. Y., Krishtal 0. A. and Tegtmeier F. (1993) R56865 and flunarizine as Na+channel blockers in isolated purkinje neurons of rat cerebellum. Neuroscience 54: 575-585. Mozhayeva G. N., Naumov A. P., Negulayev Y. A. and Nosyreva E. D. (1977) The permeability of aconitinemodified sodium channels to univalent cations in myelinated nerve. Biochim. Biophys. Acta 466: 461473.
Pauwels P. J., Leysen J. I. and Janssen P. A. J. (1991) Ca*+ and Na+ channels involved in neuronal cell death: protection by flunarizine. Life Sci. 48: 1881-1893. Prenen G. H. M., Go K. G., Postema F., Zuiderveen F. and Korf J. (1988) Cerebral shifts in hypoxic-ischemic brain damage are prevented by the sodium channel blocker tetrodotoxin. Expl. New. 99: 118-132. Schmidt H. and Schmitt 0. (1974) Effect of aconitine on the sodium permeability of the node of Ranvier. Plugers Arch. 349: 133-148. Suzuki Y., Oyama T., Ishige A., Bono T., Asami A., Ikeda Y., Noguchi M. and Omiya Y. (1994) Antinociceptive mechanism of the aconitine alkaloids mesaconitine and benzoylmesaconine. Planta Med. 60: 391-394. Tegtmeier F., Xie Y., Scheller D., Gleitz J., Osikowska-Evers B. and Peters T. (1992) The role of sodium channels in cerebral ischemia. Wiss Zeitschrifr der Humboldt-Universitat zu Berlin, R. Medizin 41: 45-53. Ulrbricht W. and Flacke W. (1965) The effect of veratridine on excitable membrane of nerve and muscle. Ergeb. Physiol. 61: 18-71. Warashina A. (1985) Frequency-dependent effects of aconitine and veratridine on membrane currents in the crayfish giant axon. Japan J. Physiol. 35: 463-482. Westenbroek R. E., Merrick D. K. and Catterall W. A. (1989) Differential subcellular localization of RI and RI1 Nat channel subtypes in central neurons. Neuron 3: 695-704. Worley P. F. and Baraban J. M. (1987) Site of anticonvulsant action on sodium channels: autoradiographic and electrophysiological studies in rat brain. Proc. Natn. Acad. Sci. U.S.A. 84: 3051-3055.
002%3908(95)00153-0
Neurophormacology, Vol. 35, No. 1, pp. lS22, 1996 Copyright0 19% Ekvier Science Ltd Printed in Great Britain. All rights resewed ocm-3908% flS.oo + 0.00
Electrophysiological Effects of Aconitine in Rat Hippocampal Slices A. AMERI,‘* Q. SHIT J. ASCHOFF2 and T. PETERS’ ‘Department of Pharmacy and Pharmacology of Natural Compounds, University of Ulm, Helmholtzstr. 20, D-89081 Ulm, Germany and 2Department of Neurology, University of Ulm, Steinhovelstr. 9, D-89075 Ulm, Germany (Accepted 29 August 1995) Summary-The electrophysiological effects of aconitine were investigated in the rat hippocampal slice and compared with those of veratridine. Both alkaloids are known to bind at site 2 of sodium channels and to block its inactivation. Extracellular recordings revealed that aconitine and veratridine exert inhibitory effects on neuronal excitability. Aconitine slowly and reversibly decreased the population spike recorded in the CA1 pyramidal cell layer. The reduction of the spike amplitude was similar whether orthodromically or antidromically activated. The aconitine-induced inhibition did not differ from that of veratridine. However, following washout of aconitine, the amplitude of the antidromic spike was increased compared to the control amplitude. The veratridine-induced inhibition was only partially reversible. This inhibition was also observed during suppression of synaptic transmission by a low Ca*‘/high Mg*+-medium, indicating an inhibition of axonal conductance. The results show that in the absence of synaptic transmission the antidromic (alvear) spike is more sensitive to the inhibitory action of aconitine than the presynaptic fiber spike elicited by stimulation of the Schaffer collaterals. Furthermore, it is shown that aconitine acts in an activity-dependent manner, in that the latency of onset of the inhibition is prolonged when the stimulation frequency is decreased. Field excitatory postsynaptic potentials were also suppressed by aconitine, whereas excitatory postsynaptic currents recorded by the patch clamp technique were not influenced by aconitine when cells were held at - 60 mV. KeywordsSodium
channel, aconitine, hippocampus, veratridine.
in a series of diseases. There is evidence that a functional defect in the inactivation gate is involved in the pathophysiological mechanisms leading to epilepsy and migraine (Hansen and Lauritzen, 1984). Furthermore, recent studies have shown a Na+ overload under conditions of an energy deficit in neurons affected by ischemia (Prenen et al., 1988; Pauwels et al., 1991; Tegtmeier et al., 1992; Kiskin et al., 1993). These findings suggest that a modulation of sodium channels might be neuroprotective or might ameliorate damage in the diseases mentioned above. Voltage-dependent sodium channels have three separate receptor sites that modulate channel function. The alkaloids aconitine, veratridine and batrachotoxin bind to site 2 on the sodium channel (Catterall, 1980) and induce a persistent activation of the channel by blocking channel inactivation. Aconitine and veratridine are known to depolarize neuronal membranes, to shift the activation of the Na current to more hyperpolarized potentials, and to decrease the maximal inward current (Ulbricht and Flacke, 1965; Schmidt and Schmitt, 1974; Mozhayeva
The increase in sodium permeability resulting from depolarization of excitable cells is biphasic. These channels open to conduct sodium ions when the cells are depolarized (activation) and subsequently enter a closed state (inactivation) with a return of sodium permeability to the resting level during a maintained depolarization. The sodium channel can therefore exist in three functionally distinct states: resting, active and inactivated (Catterall, 1992). Both the resting and inactivated states are non-conducting. Thus, the inactivation of the sodium channel controls the return of the sodium permeability to the resting level. The loss of this control mechanism causes an increased influx of sodium ions, which in turn depolarizes the cell membrane and causes more sodium channels to open. Finally, due to this positive feedback, calcium-ions enter the cell via voltagedependent calcium channels leading to a rapid Ca*+ overload. In fact, functional alterations or mutations of the inactivation gate of the sodium channel are involved
*To whom correspondence
should be addressed.
et al., 1977; Cartneliet, 13
1991). Aconitine
is reported
to
A. Ameri et al.
14
induce a maintained Na current in myelinated nerve axons (Schmidt and Schmitt, 1974; Mozhayeva et al., 1977), but it does not do so in unmyelinated axons (Warashina, 1985). Recently, it was shown that aconitine and its derivates have antinociceptive properties (Suzuki et al., 1994). In the heart, aconitine was shown to prolong the Na influx during the action potential and to induce positive inotropic and arrhythmogenic effects (Honerjager and Meissner, 1983). The aim of the present study is to elucidate how aconitine influences neuronal excitability of a highly vulnerable brain region, the hippocampus. We took advantage of synchronously discharging action potentials by a large population of pyramidal neurons, which are recorded extracellularly as distinct ‘population spikes’ in the rat hippocampal slice. In order to compare the effects of aconitine with effects of drugs known to bind to site 2 of the sodium channel, we have also performed experiments with veratridine. METHODS Male Wistar rats (45-55 days) were anesthetized with ether and decapitated. The brain was quickly removed and chilled in ice-cold artificial cerebrospinal fluid (ACSF). After 1 min, the hippocampus of one hemisphere was removed and dissected free of surrounding tissue. Transverse slices of 400 pm thickness were cut transversely to the longitudinal axis of the hippocampus using a McIlwain tissue chopper. Immediately after cutting, one slice was transferred to the recording chamber where it was kept submerged and held down on a nylon net by a U-shaped piece of flattened platinum wire. The chamber temperature was slowly increased from room temperature to 28°C during a period of 30 min. Recordings were performed l-8 hr after dissection. The other slices were maintained in an incubation chamber at room temperature. The standard ACSF was continuously gassed with a mixture of 95% Od5% COz and contained (in mM): NaCl 124, KC1 3, NaHzP04 1.25, NaHCOs 26, CaClz 2.5, MgS04 2, glucose 15 at a pH of 7.4. In particular experiments, a modified ACSF was used in which no Ca*+ was added and the Mg*+ concentration was increased to 4 mM. The ACSF was perfused at 2 ml/ min. For extracellular recordings borosilicate glass capillaries (5-10 RM) were filled with 3 M NaCl. The recording electrode was connected via a chlorided silver wire to the probe of the amplifier. Recordings of stimulus-evoked population spikes were made from the stratum pyramidale of area CAl, recordings of field excitatory postsynaptic potentials (field EPSPs) were made from the stratum radiatum. For electrical stimulation, a concentric bipolar electrode with 0.25 mm o.d. (Rhodes Medical Instruments) was positioned in the Schaffer collateral commissural pathway (i.e. near the junction of CA1 and CA2 stratum radiatum), or in the alveus, for orthodromic and antidromic activation of CA1
pyramidal cells, respectively. Extracellular stimuli were rectangular current pulses of 200 psec in duration delivered every 15 set (in particular experiments every 30 or every 60 set). The stimulus strength ranged between 0.1 and 1.5 mA. Drug effects were investigated on population spikes elicited by using a half-maximal stimulus strength. The extracellular signals were recorded and amplified by a DP 301 amplifier (Warner Instruments, U.S.A.). Analog data were digitized and analysed with a TIDA system (HEKA electronic, Germany). For whole-cell patch clamp recordings only slices (300 pm) from male Wistar rats (l&23 days) were used. To obtain access to CA1 neurons, a part of the alveus and stratum oriens was removed by a saline jet from a micropipette with a large tip opening as described elsewhere (Garaschuk et af., 1992). Thereafter it was possible to obtain visually controlled high resistance electrical contact with the soma of CA1 neurons using the standard whole-cell patch clamp technique. Patch electrodes were pulled from filamented borosilicate glass capillaries of o.d. 1.5 mm (Clark Electromedical Instruments, U.K.). When filled with the internal solution (mM: CsF 125, Tris-Cl 10, HEPES 10, EGTA 10, pH 7.2), the tip resistance was 24 M. Excitatory postsynaptic currents (EPSCs) were elicited by stimulation of Schaffer collaterals and recorded using an Axopatch 200 A amplifier (Axon Instruments, U.S.A.). All drugs were applied by addition to the ACSF. Aconitine and veratridine were dissolved in 20% dimethyl sulfoxide (DMSO) to give stock solutions of 1 mM. This solution was diluted with ACSF to reach final concentrations of between 0.01 and 1 PM. The final DMSO concentration for the highest aconitine concentration tested (i.e. 1 PM) was 0.02%. Aconitine was purchased from Fluka (Germany), veratridine and tetrodotoxin (‘FIX) from Sigma (Germany). Values are given as mean + standard deviation (SD). Significance was assumed when P d 0.05. Statistical evaluation was performed using the Student’s r-test. The amplitude of the population spike was determined from the negative peak to a tangent drawn between the first and the second maximum positives. RESULTS The effects of aconitine on evoked field potentials in area CA1 of the rat hippocampus were investigated in 57 slices. Furthermore, in four CA1 neurons EPSCs were recorded by the patch clamp technique. Veratridine was examined in 16 slices. Recordings were only included in the data analysis if electrical stimulation at a maximal stimulus strength did not evoke a second population spike, and where the amplitude of the population spike was stable for a control period of at least 30 min. Aconitine (1 PM) slowly and reversibly decreased the amplitude of the orthodromically evoked population spike within 20 min [Fig. l(A)]. During the first l-2 hr of
Electrophysiological
effects of aconitine in rat hippocampal slices
15
ACONITINE (1 pM)
time
(min)
C
0
0
I
I
I
30
60
90
120
time (mill)
Fig. 1. Inhibitory action of aconitine (1 PM) on the orthodromic population spike recorded extracellularly in the CA1 pyramidal cell layer. (A) Population spikes were evoked by electrical stimulation of the Schaffer collateral pathway. The stimulus artefact is followed by the presynaptic fiber spike and the population spike. Each trace represents the average of 5 consecutive responses. (B) Time-course of the action of aconitine on the population spike. The filled bar indicates when aconitine was applied. The afferents were stimulated every 15 sec. The diagram depicts the measurements of the amplitude of the population spike in response to each stimulus. (C)Timecourse of the action of aconitine on the presynaptic fiber spike. During the 30 min application, the presynaptic fiber spike was resistant to an action of aconitine. Recordings in A and results in B and C were obtained from the same slice.
washout, a second population spike occurred in about half of the slices. In each of these slices, these additional spikes disappeared during prolonged washout. The presynaptic fiber spike (afferent volley) which represents the compound action potential generated in synchronously discharging axons was apparently unaffected
during the 30 min application of aconitine [Fig. l(B)]. In order to identify the small deflection preceding the postsynaptic population spike as the presynaptic fiber volley, 30 min after the start of the perfusion with aconitine the sodium channel blocker TTX (1 FM) was added. In every slice (n = 3) the spike was blocked by
16
A. Ameri et al.
-I orthodromic PS
antidromic PS
antidromic PS (after washout)
Fig. 2. Comparison of the inhibitory action of aconitine and veratridine on the orthodromic and antidromic population spike (PS). Both drugs were bath-applied at a concentration of 1 PM for 30 min. During this period the presynaptic fiber spike remained unchanged. The columns on the right represent the size of the antidromic population spike after washout of the drugs. The values after washout of aconitine and veratridine were significantly different (P < 0.01). Data are mean values +_ SD of at least 4-5 experiments.
fiber spike. The action of veratridine (1 PM) on the orthodromic population spike (n = 4) did not differ significantly from aconitine in this respect (Fig. 2). Both drugs reduced the antidromic population spike which is elicited by direct, alvear stimulation of CAl, in the same manner and time course as the orthodromic population spike (Figs 2 and 3). Immediately following application of aconitine, there was usually a slight and short-term increase in the amplitude of both the orthodromically and the antidromically evoked population spike (up to 15%). This increase was followed by a rapid decrease in the spike amplitude. The inhibitory effect of aconitine on the population spike was slowly reversible during a prolonged washout and showed a large overshoot (Fig. 3) which was more pronounced after the application of higher concentrations. The amount of the overshoot obtained after 4-5 hr washout of 1 PM aconitine was 137.2 + 11.4% of control (n = 5) thus significantly higher than that obtained after washout of 0.1 PM aconitine (116.4 f 6.9%, n = 3). During washout of veratridine only a partial recovery (68.7 f 7.1%, II = 3) was achieved (Fig. 2). Both the extent of inhibition and the latency of its onset was dependent on the concentration of aconitine used. The dose-response relationship of aconitine (Fig. 4) was examined by application of a single concentration to each slice. The ICse for the inhibitory effect of aconitine on the antidromic population spike was 0.03 ,uM. In response to antidromic stimulation synaptic activalTX,
indicating
that it was the presynaptic
tion of CA1 pyramidal cells via recurrent loops or by orthodromic stimulation of basal dendritic afferents might occur, and could contribute to the generation of the evoked potentials observed in these experiments. In order to eliminate the possible contribution of presynaptic mechanisms to the aconitine-induced action, synaptic transmission can be blocked by a low Ca2+/high Mg2+ medium (Andersen et al., 1978). The addition of aconitine (1 ,uM) to a low Ca2+/high Mg2+ medium attenuated the amplitude of the antidromic population spike by 92.0 + 1.8% (n = 4) and did not differ significantly from the inhibition caused by 1 PM aconitine added to normal ACSF (92.0 f 1.8%, II = 5). The aconitine-induced reduction of the antidromic population spike observed even in the absence of synaptic transmission suggests that it is mediated by a change in axonal excitability. The antidromic population spike results from synchronous activation of alvear fibers, whereas the presynaptic fiber volley which precedes the postsynaptic population spike in response to orthodromic stimulation is the compound action potential resulting from a synchronous discharging of the Schaffer collaterals (Andersen et al., 1978). As reported above (Fig. l), aconitine diminished the postsynaptic spike evoked by stimulation of the Schaffer collaterals (i.e. synaptic activation) while apparently sparing the presynaptic fiber volley. In order to investigate an effect of aconitine on the presynaptic fiber spike exclusively and in more detail, we again used a low Ca2+/high Mg2+ ACSF to block synaptic transmission. In these conditions, orthodromic stimulation of the Schaffer collaterals elicites only the presynaptic fiber spike but no postsynaptic spike, while the antidromic (aivear) population spike remains unchanged. Following suppression of the postsynaptic spike, we recorded the presynaptic fiber spike elicited by orthodromical stimulation of the Schaffer collaterals for a further 30 min control period before aconitine (1 PM) was applied. At this concentration, aconitine reduced the presynaptic fiber spike by only about 65.9 + 1.8% (n = 4). However, maximal inhibition was achieved 148.8 + 27.2 min after onset of the perfusion with aconitine. In contrast to this, the antidromically evoked (alvear) population spike was attenuated by aconitine (1 PM), added to a low Ca2+/high Mg2+ medium by about 92.1 f 1.8% (n =4) with a latency of 27.5 f 8.7 min after the start of the perfusion with aconitine. Figure 5 shows the time-course of the inhibition in two representative recordings, an antidromically evoked population spike [Fig. 5(A)] and an orthodromically evoked spike [Fig. 5(B)]. Both recordings were performed in a low Ca2+/high Mg2+ medium. However, during prolonged washout with the low Ca2+/ high Mg2+ medium, the amplitude of the antidromic population spike no longer reached the control value [Fig. 5(A)]. The maximal recovery of the antidromic spike recorded in the absence of extracellular Ca2+ was
Electrophysiological
A
CONTROL
17
effects of aconitine in rat hippocampal slices ACONITINE (1 pM)
WASH
I-- Ii---l
10ms
> E
2
time (min) Fig. 3. Inhibitory action of aconitine (1 PM) on the antidromic population spike. (A) Population spikes were elicited by electrical stimulation of alvear fibers. Each trace is the average of 5 consecutive events. (B) Timecourse of the action of aconitine on the antidromic population spike. Each point on the graph represents the measurement of the amplitude in response to a single stimulus. The time of application of aconitine is indicated by the filled bar. Note the marked overshoot during washout. Recordings in A and results in B were obtained from the same slice.
12.2% (n = 4). This stands in contrast to the recordings performed in normal ACSF showing a marked overshoot (137.2 + 11.4%, n = 5) of the antidromic population spike amplitude [Fig. 3(B)] during washout. The discrepancies in the extent of the overshoot observed during washout of aconitine by standard ACSF and by low Ca +-ACSF, raises the question of whether it is due to a blockade of inactivation, or if it is a consequence of the Ca2+ influx caused by these drugs (Ashton et al., 1990). In order to allow a long-lasting CaZt influx we applied aconitine (1 ,uM) for a period of 2 hr. In all of the slices tested (n = 3) the inhibitory effect of aconitine was reversible. The overshoot of the antidromic spike amplitude did not differ from the results obtained during washout of aconitine following a 30 min application. The inhibitory effect of aconitine (0.03 PM) on the antidromically evoked population spike recorded in the 35.0 f
CA1 region of the hippocampus is decelerated by decreasing the stimulation frequency (Fig. 6), consistent with its activity-dependent action in other preparations (Honejager and Meissner, 1983). A reduction of the stimulation frequency from 4 stimulus pulses per min [Fig. 6(A)] to 2 pulses per min [Fig. 6(B)], and to 1 pulse per min [Fig. 6(C)] increased the time for the maximal aconitine-induced inhibition to be achieved. In contrast to the longer duration, the maximal percentage of inhibition of the population spike by aconitine is unaltered. It is emphasized that the dependence on the stimulation frequency can be overcome by aconitine used at maximally effective concentrations (0.3 and 1 PM). Since the postsynaptic population spike was completely blocked by aconitine (Fig. l), recordings of the field EPSP were performed (Fig. 7) to examine the effect of aconitine on synaptic transmission. Aconitine (1 PM) depressed the field EPSP recorded in the CA1 stratum
18
A. Ameri et al. radiatum by 84.4 + 47.0% (n = 5). Maximal inhibition of the field EPSP was achieved 44 + 6.5 min following
onset of the application of aconitine. The effect on the field EPSP was only partially reversible during washout of aconitine. In order to confirm the results of the field data at a cellular level, patch clamp experiments on single pyramidal neurons in the hippocampal slice were performed. The advantage of patch clamping in slices is that the synaptic connections are preserved. The peak amplitude and duration of EPSCs recorded in response to electrical stimulation of the Schaffer collaterals (Fig. 8) remained unaltered by aconitine (1 PM) (n = 4).
I
I
I
I
-0
-7
-6
-5
concentration (log; M)
Fig. 4. Concentration-response curve for aconitine in rat hippocampal slices. For each curve, data were obtained from measurements performed in 4 and 5 different slices. Aconitine was applied at each concentration to a single slice. Slices were stimulated antidromically every 15 sec. The amplitude of the population spike (PS) was normalized with respect to control and plotted as a function of the logarithm of the aconitine concentration. Data points represent mean values + SD. In cases where data points are lacking error bars, the standard deviation is smaller than the size of the symbol.
DISCUSSION The major finding of the present study is an inhibitory effect of aconitine and veratridine on the excitability of rat hippocampal slices. At a concentration range of O.Ol1 PM, aconitine caused a slow and reversible decrease of the amplitude of the population spike recorded in the CA1 stratum pyramidale evoked by both orthodromic (Fig. 1) and antidromic (Fig. 3) stimulation. The inhibition of the orthodromic and antidromic population spike induced by 1 PM aconitine does not differ from the inhibition induced by 1 PM veratridine (Fig. 2). Immediately upon application of both drugs, the amplitude of the population spike was slightly increased. This initial enhancement of the population spike has also been observed by others (Ashton et al., 1990) following
A
120
time (min)
0
120
240 time (min)
360
Fig. 5. Time-course of the inhibitory effect of aconitine (1 PM) in the absence of synaptic transmission. (A) The slice was stimulated antidromically every 15 sec. (B) The slice was stimulated orthodromically. Since the recording was performed in a low Ca*“/high Mg*+-medium the population spike was suppressed. The data points represent measurements of the presynaptic fiber spike which is the compound action potential of the Schaffer collaterals. Note the longer time of application of aconitine (filled bars) in B.
Electrophysiological
{*50;,
,
0
,lT”,
120
,
,“p”.
240
,
,‘p”,
360
,
effects of aconitine in rat hippocampal slices
,4i
480
time (min)
Fig. 6. Dependence of the latency of onset of the aconitineinduced effect on the stimulation frequency. The slices were stimulated antidromically with 4 stimulus pulses per min (A), 2 pulses per min (B), and 1 pulse per min (C), respectively. In each experiment, aconitine was applied at a concentration of 0.03 PM at the time marked by the bar above the graph. Note the longer duration of drug-application and the later onset of inhibition with decreasing stimulation frequency.
blockade of sodium channel inactivation by veratridine. These authors report that veratridine both decreases the extracellular concentration of Na+ and Ca2+ and increases the extracellular concentration of K+, accompanied by a persistent depolarization. The veratridine-induced Na+ influx is thought to be the result of the’ binding of veratridine to site 2 of the Na+ channel (Catterall, 1980), the common binding site for veratridine and aconitine. These findings suggest that the initial increase in spike amplitude reflects blockade of the inactivation of voltagedependent sodium channels. Due to the prolonged activation, the membranes are depolarized by sustained sodium influx (Schmidt and Schmitt, 1974; Mozhayeva et al., 1977). Nodal membranes of myelinated nerve fibres of Xenopus Zuevis are depolarized by 10-15 mV, leading finally to complete inexcitability (Schmidt and Schmitt, 1974). Furthermore, there: is evidence that aconitine and veratridine, in addition to their action on inactivation, shift the voltage-dependence for the activation of the Na+ current in a hyperpolarizing direction and reduce the maximal inward current (Ulbricht and Flacke, 1965;
19
Schmidt and Schmitt, 1974; Mozhayeva et al., 1977; Carmeliet, 1991). These findings explain the suppression of the population spike by aconitine and veratridine observed in the present study. Furthermore, we observed that during the first 2 hr of washout of aconitine about half of the slices developed a second population spike, which might be a result of a shift of potassium from the intracellular to the extracellular space (Ashton et al., 1990). Alternatively, the second transient, population spike might reflect a shift of the activation parameters in a hyperpolarizing direction (Ulbricht and Flacke, 1965; Schmidt and Schmitt, 1974; Mozhayeva et al., 1977; Carmeliet, 1991) so that less depolarization is required to activate the Na+ channel; the threshold will be attained earlier and more easily, which explains the repetitive activity. The antidromic (alvear) population spike is attenuated to the same extent and time course as the orthodromically (synaptically) evoked population spike by aconitine (1 ,uM). The antidromic population spike is produced by electrical activation of the axons of the neurons. Since this potential is unaffected by drugs that modify synaptic transmission (Dunwiddie, 1986), the present results indicate that aconitine inhibits preferentially axonal conduction. During washout of aconitine, the amplitude of the population spike was significantly increased compared to control. This overshoot was concentration-dependent and was also observed when aconitine was applied for 2 hr in order to allow a sustained Ca2+ influx. When aconitine was applied in the absence of Ca2+, however, there was neither an overshoot nor a full recovery of the spike amplitude. Veratridine failed to produce an overshoot of the spike amplitude during washout (Fig. 2). In a previous study, veratridine, in contrast to aconitine, is reported to cause a reduction in the maximal intensity of the inward current, which is persistent during washout (Warashina, 1985). These findings imply that both drugs have a different mode of interaction with Na+ channels, which is in accordance with the results of the present study. Further experiments are required to clarify the mechanisms underlying the increase in spike amplitude after washout of aconitine. The most intriguing finding of the present study is the difference in the extent of the aconitine-induced inhibition of the fiber spike amplitude evoked by orthodromic (Schaffer collaterals) and antidromic (alvear) stimulation (Fig. 5). By use of TTX, we have identified the negative deflection preceding the orthodromic population spike as the presynaptic fiber spike which represents the compound action potential of the Schaffer collaterals. While a previous study failed to show an effect of aconitine in concentrations of up to 1 PM on the presynaptic fiber spike (Worley and Baraban, 1987) the present study clearly demonstrated that both alvear fibers and Schaffer collaterals are sensitive to aconitine. These authors applied aconitine for 30 min, a period which in the
20
A. Ameri et al.
A
CONTROL
WASH
ACONITINE(lpM)
B
0 0
I 60
I I 120 180 time (min)
1 240
300
Fig. 7. Inhibitory action of aconitine (1 PM) on the field excitatory postsynaptic potential (EPSP). (A) Field EPSPs recorded extracellularly in the apical dendritic region of area CAL Schaffer collaterals were stimulated every 15 sec. Each curve is the average of 4 subsequent responses. (B) Time-course of the peak amplitude of the field EPSP. Each point represents the mean of four consecutive traces. The filled bar indicates when aconitine was applied. Recordings in A and results in B were obtained from the same slice.
present study also failed to exert an inhibition presynaptic fiber spike. However, inhibition
of the
of the presynaptic fiber spike is obtained after long-term application of aconitine [Fig. 5(B)]. At a concentration of 1 ,uM, which caused maximal inhibition of the alvear spike, aconitine reduced the amplitude of the presynaptic fiber spike by about 65%. The difference in sensitivity to aconitine might reflect differences in myelination. The Schaffer collaterals are not myelinated whilst the alvear fibers are (Andersen et al., 1978). It is well known that sodium channels are highly concentrated in the membrane at the nodes of Ranvier as compared to unmyelinated axons. The high density of sodium channels in the nodes of Ranvier is the basis of the low threshold and generation of action potentials at this site (Hille, 1992). Differences in the action of aconitine on myelinated and nonmyelinated nerve fibers, have been reported previously by others (Schmidt and Schmitt, 1974; Mozhayeva et al., 1977; Warashina, 1985) suggesting different modes of interactions between toxin molecules and open
Na+ channels. In unmyelinated
axons, aconitine did not induce a Na+ current which was maintained, although it reduced the current to a large extent (Warashina, 1985). This contrasts with the effect of aconitine on myelinated nerve axons, in which a Na+ current which is maintained and a shift of the activation in a hyperpolarizing direction was observed (Schmidt and Schmitt, 1974; Mozhayeva et aZ., 1977). Alternatively, aconitine might interact differentially with various Na+ channel subtypes. Indeed, Westenbroek et cd., (1989) have shown a differential localization of Rr and Rn Na+ channel subtypes in the hippocampus pyramidal cell layer and on Schaffer collaterals, respectively, Whether the differences observed in the present study are due to myelination and non-myelination, or to different subtypes of the voltagedependent sodium channel cannot be decided at present. An activity-dependent effect of aconitine was obvious in our study since the inhibitory action of aconitine (0.03 PM) depends on the stimulation frequency (Fig. 6). While a reduction of the stimulation frequency leads
Electrophysiological
21
effects of aconitine in rat hippocampal slices
A ACONlTINE (1 pM)
CONTROL
WASH
B
01
0
I
I
60
120
I 180
time (min) Fig. 8. EPSCs record’ed in a CA1 pyramidal cell evoked by stimulation of the Schaffer collaterals. The CA1 neuron was held at - 60 mV. (A) Application of aconitine (1 PM) did not irdluence the EPSC. Each curve is the mean of 3 sequential traces. (B) Time-course of the peak amplitude of the EPSC. The filled bar indicates when aconitine (1 PM) was applied. The Schaffer collaterals were stimulated every 10 sec.
in the rate of the aconitine-induced inhibition of the population spike, the extent of the inhibition remains unaffected. At maximally effective concentrations (i.e. 0.3 and 1 PM), aconitine failed to show frequency-dependence. This suggests that higher concentrations of aconitine compensate for fewer stimulus pulses. It is likely that by using higher concentrations the probability of finding open sodium channels is still high enough to elicit maximal blockade of inactivation. There was no effect d aconitine on the excitatory postsynaptic currents (EPSCs) recorded in the CA1 neurons by the patch clamp technique (Fig. 8). However, as the membrane potential was clamped at -60 mV, we received no information on aconitine-induced changes in the membrane potential which in turn could influence synaptic transmission. In contrast, field EPSPs recorded in the CA1 stratum radiatum were depressed by aconitine. The field EPSP is thought to reflect synaptic currents in the dendrites as a result of the action of the neurotransmitter (Dunwisddie, 1986). The time-course of the inhibitory action of aconitine on the field EPSPs is comparable with the suppression of the population spike
to
a decrease
amplitude. However, the inhibition of the field EPSPs was only partially reversible. In conclusion, the present results show that aconitine has, in addition to its inhibitory action on axonal conductance, an inhibitory effect on excitatory synaptic transmission. It is not yet clear if the various effects of the toxin described in the present study are derived from a unique mechanism comprising a very specific interaction between a toxin molecule and a Na+ channel, or if they are induced by different interactions. Further experiments are required to elucidate the effects of aconitine on the membrane potential of CA1 neurons and the mechanisms involved in the overshoot of the spike amplitude after washout. Acknowledgement-The authors Strack for computer support.
are grateful
to Sebastian
REFERENCES Andersen P., Silvenius H., Sundberg S. H., Sveen 0. and Wigstrom H. (1978) Functional characteristic of myelinated fibers in the hippocampal cortex. Brain Res. 144: 11-18. Ashton D., Willems R., Marrannes R. and Janssen P. A. J.
22
A. Amel ri et al.
(1990) Extracellular ions during veratridine-induced neurotoxicity in hippocampal slices: neuroprotective effects of flunarizine and tetrodotoxin. Brain Rex 528: 212-222. Carmeliet E. (1991) Ion channels agonists: expectations for therapy. Eur. Heart J. 12: 30-37. Catterall W. A. (1980) Neurotoxins that act on voltagesensitive sodium channels in excitable membranes. A. Rev. Pharm. Toxic. 20: 15-43. Catterall W. A. (1992) Cellular and molecular biology of voltage-gated sodium channels. Physiol. Rev. 72: 15-48. Dunwiddie T. V. (1986) The use of in vitro brain slices in neuropharmacology. In: Electrophysiological Techniques in Pharmacology. Alan R. Liss, New York. Garaschuk O., Kovalschuk Y. and Krishtal 0. (1992) Adenosine-dependent enhancement by methylxanthines of excitatory synaptic transmission in hippocampus of rats. Neurosci. Lett. 135: 10-12. Hansen A. J. and Lauritzen M. (1984) The role of spreading depression in acute brain disorders. Ann. Acad. Bras. Cien. 54: 457479. Hille B. (1992) Ionic channels of excitable membranes. Sinauer Associates, Sunderland. Honerjager P. and Meissner A. (1983) The positive inotropic Naunyn-Schmiedeberg’s Archs. effect of aconitine. Pharmac. 322: 49-58. Kiskin N. I., Chizhmakov I. V., Tsyndrenko A. Y., Krishtal 0. A. and Tegtmeier F. (1993) R56865 and flunarizine as Na+channel blockers in isolated purkinje neurons of rat cerebellum. Neuroscience 54: 575-585. Mozhayeva G. N., Naumov A. P., Negulayev Y. A. and Nosyreva E. D. (1977) The permeability of aconitinemodified sodium channels to univalent cations in myelinated nerve. Biochim. Biophys. Acta 466: 461473.
Pauwels P. J., Leysen J. I. and Janssen P. A. J. (1991) Ca*+ and Na+ channels involved in neuronal cell death: protection by flunarizine. Life Sci. 48: 1881-1893. Prenen G. H. M., Go K. G., Postema F., Zuiderveen F. and Korf J. (1988) Cerebral shifts in hypoxic-ischemic brain damage are prevented by the sodium channel blocker tetrodotoxin. Expl. New. 99: 118-132. Schmidt H. and Schmitt 0. (1974) Effect of aconitine on the sodium permeability of the node of Ranvier. Plugers Arch. 349: 133-148. Suzuki Y., Oyama T., Ishige A., Bono T., Asami A., Ikeda Y., Noguchi M. and Omiya Y. (1994) Antinociceptive mechanism of the aconitine alkaloids mesaconitine and benzoylmesaconine. Planta Med. 60: 391-394. Tegtmeier F., Xie Y., Scheller D., Gleitz J., Osikowska-Evers B. and Peters T. (1992) The role of sodium channels in cerebral ischemia. Wiss Zeitschrifr der Humboldt-Universitat zu Berlin, R. Medizin 41: 45-53. Ulrbricht W. and Flacke W. (1965) The effect of veratridine on excitable membrane of nerve and muscle. Ergeb. Physiol. 61: 18-71. Warashina A. (1985) Frequency-dependent effects of aconitine and veratridine on membrane currents in the crayfish giant axon. Japan J. Physiol. 35: 463-482. Westenbroek R. E., Merrick D. K. and Catterall W. A. (1989) Differential subcellular localization of RI and RI1 Nat channel subtypes in central neurons. Neuron 3: 695-704. Worley P. F. and Baraban J. M. (1987) Site of anticonvulsant action on sodium channels: autoradiographic and electrophysiological studies in rat brain. Proc. Natn. Acad. Sci. U.S.A. 84: 3051-3055.