Riluzole prevents hyperexcitability produced by the mast cell degranulating peptide and dendrotoxin I in the rat

European Journal of Pharmacology, 193 (1991) 223-229

223

© 1991 Elsevier Science Publishers B.V. 0014-2999/91/$03.50 A D O N I S 001429999100148C

EJP 51697

Riluzole prevents hyperexcitability produced by the mast cell degranulating peptide and dendrotoxin I in the rat J e a n - M a r i e S t u t z m a n n , G e o r g A n d r e e s Bt~hme, G a b r i e l G a n d o l f o 1, C l a u d e G o t t e s m a n n 1, J o s 6 p h i n e L a f f o r g u e , J e a n - C h a r l e s B l a n c h a r d , Pierre M. L a d u r o n a n d M i c h e l L a z d u n s k i 2 Rh6ne-Poulenc SantO, Centre de Recherches de Vitry, Unitd dWectrophysiologie, 94403 Vitry Sur Seine Cedex, France, i Laboratoire de Psychophysiologie, Facult~ des Sciences, 06034 Nice Cedex, France and 2 Centre de Biochimie, Centre National de la Recherche Seientifique, 06031 Nice Cedex, France

Received 18 July 1990, revised MS received 12 October 1990, accepted 6 November 1990

Using electroencephalographic (EEG) recordings in freely moving rats and extracellular neuronal firing-rate recordings in hippocampal slices, we examined the effects of riluzole (RP 54274), a compound with anti-glutamate properties, against the convulsive seizures and the cellular hyperexcitability produced by the mast-cell degranulating peptide (MCD), dendrotoxin I ( D T X i) and 4-aminopyridine (4-AP). I.c.v. administration of riluzole (10 nmol) prevented the seizures induced by MCD, and to a lesser extent those due to DIXi, whilst leaving 4-AP seizures unaffected. This effect was also present after oral administration of the compound (4 mg kg-t) and lasted for approximately 6 h. Electrophysiological recordings in vitro confirmed that riluzole dose dependently and reversibly abolished the sustained increase in firing rate induced by both MCD and D T X i in the hippocampus. These results indicate that the anti-epileptic spectrum of riluzole in this model has similarities with, but is not identical to, that of classical potassium channel openers, and differs from that of calcium channel blockers or other glutamate antagonists such as D(-)-2-amino-5-phosphono-valeric acid. However, since MCD releases glutamate, the preventive effect of riluzole in this model may involve direct or indirect interaction with glutamatergic processes. Mast-cell d e g r a n u l a t i n g peptide ( M C D ) ; D e n d r o t o x i n I ( D T X i); Electroencephalography; H i p p o c a m p a l slices; A n t i c o n v u l s a n t s ; Antiglutamate

I. Introduction

Riluzole (RP 54274) is a compound with a benzothiazole structure that possesses anticonvulsant properties in several models associated with glutamatergic neurotransmission. This drug prevents the convulsions induced by i.c.v, injection of glutamate or kainate and inhibits the tremors induced by harmaline, but is less potent than diazepam against seizures induced by glutamic acid decarboxylase inhibitors like 3-mercaptopropionic acid and isoniazid (Mizoule et al., 1985). Moreover, in contrast to barbiturates and benzodiazepines, riluzole was ineffective against the clonic convulsions induced by pentylenetetrazole, bicuculline or picrotoxin. This c o m p o u n d appears to interfere indi-

Correspondence to: J.-M. Stutzmann, Rh6ne-Poulenc Sant6, Centre de Recherches de Vitry, 13 Quai Jules Guesde, 94403 Vitry sur Seine Cedex, France.

rectly with glutamatergic neurotransmission (B~navid~s et al., 1985). Mast cell degranulating peptide ( M C D ) and dendrotoxin I (DTXi) are neurotoxins, purified from bee and snake venom respectively, which block a class of voltage-sensitive potassium channels, provoking seizures and convulsions after i.c.v, injection (Bidard et al., 1989). Their convulsive actions have been compared to the effect of another classical potassium channel blocker, 4-aminopyridine (4-AP) (Gandolfo et al., 1989a). In order to obtain more information about the anticonvulsant spectrum of riluzole, its effects on the electroencephalographic hyperexcitability produced by the three potassium channel blockers MCD, D T X i and 4-AP were examined. Moreover, since the hippocampus is considered to be the main target for the epileptogenic action of M C D (Bidard et al., 1987) whereas D T X i has both hippocampal and cortical epileptogenic foci (Bidard et al., 1989), the cellular interactions of riluzole with the two neurotoxins were examined using extracellular recording techniques with hippocampal slices maintained in vitro.

224 2. Materials and methods

2.1. Drugs M C D and D T X i were purified from bee (Apis mellifera) and m a m b a snake (Dendroaspis polylepis polylepis) venoms respectively, according to previously described techniques (Rehm et al., 1988; Taylor et al., 1984). Stock solutions of the peptides were prepared in physiological saline (i.c.v. studies) or in artificial cerebrospinal fluid (brain slice studies, composition see below) and frozen in small aliquots until required. 4-AP was obtained from Sigma (France). Riluzole (2-amino6-trifluoromethoxybenzothiazole) was synthesized in the Chemistry Department of Rh6ne-Poulenc Sant&

2.2. EEG studies Adult male Wistar and Sprague-Dawley rats (Charles River, France) weighing 270-300 g, anaesthetised with pentobarbital, were equipped with chronic cortical electrodes for E E G recording. Silver balls of I m m diameter were placed on the frontal and occipital cortex (area 10 and 17 from Krieg, 1946, respectively) for i.c.v, studies. The animals to be used in oral route studies were implanted with screw electrodes in the skull overlying these areas. The electromyogram (EMG) was recorded from the dorsal neck musculature by means of stainless-steel or silver wires. A bevel-edged cannula was inserted into the lateral ventricle for i.c.v, injections (coordinates: 1 m m posterior to bregma, 1.3 m m lateral to the midline and 5 m m below the surface of the skull). Placement of the cannula in the ventricle was verified both by extraction of cerebrospinal fluid (CSF) and dissection at the end of the experiments. The cortical and muscular recording electrodes were soldered to a connector and secured, together with the tip of the cannula, to the skull of the animal with dental cement (Atlantic Codental, France). After surgery, the animals were allowed a minimum of 7 days of recovery during which postoperative care, including prophylactic antibiotic therapy with benzathine-benzylpenicilline (Specia, France), was conducted as previously described (Bidard et al., 1987; Stutzmann et al., 1987). All animals were individually housed in an environment with controlled temperature and light (12 h l i g h t / d a r k schedule) with free access to commercial rat chow and water. Before each experiment the rats were acclimatized to their cage for one h. The p.o. and i.c.v, effects of riluzole were assessed as follows: for i.c.v, route studies, EEG recordings were performed on three groups of rats: (i) the control rats (n = 5) received an i.c.v, injection of the potassium channel blockers alone (injected volume = 10 /~1 over 2 min, the injection needle remaining in the cannula 2 further minutes to avoid back flow of the convulsant);

(ii) the sham rats (n = 6) received an i.c.v, vehicle injection (10 /d of 1% dimethylsulfoxide in isotonic saline solution) 15-20 min prior to convulsant administration; (iii) the experimental group (n = 6) received an i.c.v. (10 /d) injection of riluzole (10 nmol) 15-20 min prior to convulsant administration or during the potassium channel blocker-induced seizures. The criteria for a positive assessment of epileptiform activity in the E E G were as previously described (Bidard et al., 1987). For p.o. studies, EEG recordings were performed on one group of vehicle-treated rats (control) and three groups of rats receiving riluzole (4 mg kg-~ p.o.) at different times before i.c.v, injection of the convulsants in order to estimate its duration of action. Vehicle (saline) or riluzole was given to gently hand-restrained animals through an oesophageal cannula in a volume of 5 ml kg -~ body weight. The control rats (n = 12) received an i.c.v, injection (4 /al) of 0.10 nmol M C D 1 h after the vehicle. The three experimental groups received the same dose of toxin injected 1 (n = 8), 6 (n = 6) or 18 h (n = 6) after treatment with riluzole. The epileptogenic events were qualitatively evaluated according to the following rating scale: stage 0 = desynchronized E E G without epileptic events; stage 1: isolated spike and wave complexes; stage 2: paroxystic bursts of epileptiform grapho-elements; stage 3 = generalized seizures.

2.3. Brain slice studies H i p p o c a m p a l slices were obtained from male Sprague-Dawley rats weighing 150-190 g (Charles River, France) and were prepared for electrophysiological recordings in vitro in a submersion-type chamber as described before (BiShme et al., 1988). The composition of the artificial cerebrospinal fluid (ACSF) was as follows in m M : NaC1 124, KC1 5, MgSO 4 2, CaC12 2, N a H C O 3 26, K H z P O 4 1.25 and glucose 10. The temperature was maintained constant at 32_+0.2°C and the flow rate was held between 2.5-3 ml min-1. Extracellular recordings of the action potential discharge of single hippocampal neurons were performed using metal microelectrodes (Frederick Haer, Brunswick, U.S.A.) lowered under visual guidance into the pyramidal cell layer of the CA1 area. The signal was amplified (Dagan, Mineapolis, U.S.A.) and a measurement of the firing frequency was obtained using a spike detector coupled to an integrating device (time constant: 1 s) the output of which was displayed on a chart recorder throughout the experiment. Once a stable firing rate was obtained, the peptidic toxins to be tested were applied to the slices in the following concentrations (0.5 and 1 /~M for D T X i and MCD, respectively) by switching from control ACSF to ACSF + M C D or DTX~ without interruption of flow (the flow rate allowed the bathing medium to reach equilibrium within 4-5 min).

225

were quantified by their frequency of occurrence (i.c.v. studies, table 1) or by their strength (p.o. studies, see below). These seizures increased in frequency in a dosedependent manner, becoming continuous at the highest doses of the potassium channel blockers in control and sham rats. Riluzole was first given alone at the dose of 10 nmol i.c.v., to verify that it had no action on the electrophysiological recordings and behaviour of animals not treated with convulsant (fig. 1). Figure 1 and table 1 show that a dose of 10 nmol of riluzole significantly

3. R e s u l t s

3.1. EEG studies

As previously described (Bidard et al., 1987; 1989) the i.c.v, injection of MCD, DTX~ and 4-AP induces desynchronization of the background EEG, interrupted by epileptic discharges (composed of bursts of spike and wave complexes accompanied by clonic movements) or generalized convulsive seizures (fig. 1) which No aru~

Oe

PAl. {IO nmoB

.,

,

MCD 0,1 n m a l

DTXI 0 . 0 8 nmol

~

_

~?~'~

.~_j

.,

J

~

_

Ril. + MCD l 0 nmol

Rfl. + MCD

0.I rm~l

t0nmol

Ril.+ DTXx 10 nmol

DTXI

0 . 0 8 nmol

0 . 1 2 nmol

4-AP

I , ~

MCD + Ril. 0.1 nmol

20nmol

Rid. + D'I'~ I 0 nmol

0 . 1 2 nmol

Rfl. + 4 - A P

I00 nrnol

t

0A5nmol

I0 nmol

~'"f~

I00 nmol

-.,l~,

Fig. 1. Effect of i.c.v, injections of riluzole on potassium channel blocker-induced seizures. From top to bottom. Control: riluzole given alone (fight) did not modify the EEG recordings (slow waves alternating with waking desynchronization) made 20 rain after an i.c.v, injection of 10 nmol, when compared to those from the same rat prior to treatment (left). M C D (from left to right): 15-20 rain after vehicle administration, an i.c.v, injection of 0.1 nmol of M C D induced seizures (sham rat:recordings performed 20 min after M C D injection); when 10 nmol of riluzole were injected i.c.v. 15-20 rain prior to M C D administration (at a dose of 0.1 nmol), all seizures were prevented (recordings at 18 min). A normally lethal dose of 0.45 nmol M C D was needed after pretreatment with riluzole for the induction of seizures, which were no stronger than those induced in sham rats (recordings at 16 min). When riluzole (even at a dose of 20 nmol) was given after the M C D - i n d u c e d seizures had started, it was unable to disrupt the seizures (recordings at 14 min). D T X i (from left to right): s h a m rat i.c.v, injected with 0.08 nmol of D T X i showed seizures (recording performed 20 min after D T X i injection), but the same dose of D T X i was unable to produce seizures when 10 nmol of riluzole was injected i.c.v. 15-20 rain prior to D T X i administration (recording at 25 rain). For a dose of 0.12 nmol of DTXi, the seizure induced was less intense (more spaced spiked waves) after a prior injection of riluzole (10 nmol) than that induced in s h a m rats (recordings at 17 and 15 rain respectively). 4-AP: sham rat i.c.v, injected with 100 nmol of 4-AP showed seizures; 10 nmol riluzole i.c.v, injected 15-20 rain prior to 4-AP administration was unable to prevent seizures (recording at 16 and 24 min respectively). Abbreviations. Ril.: riluzole; M C D : mast cell-degranulating peptide; DTXi: dendrotoxin I; 4-AP: 4-aminopyridine; FF: bihemispheric frontal cortex; F: monopolar frontal cortex; O: monopolar occipital cortex; c: contralatcral structure (with respect to the injection side); i: ipsilateral structure; EMG: dorsal neck electromyogram. Calibration: 100/~V (same rat in the two highest panels); 1 s (common time scale for all panels).

226 100

increased the threshold for seizures induced by MCD. A minimum lethal dose (0.30 nmol) or even a higher dose (0.45 nmol) of M C D produced only weak paroxysmal seizures, similar to those induced by a dose of 0.10 nmol of M C D in control or sham rats. This protective effect was only preventive and was not seen when riluzole, even at a 2-fold higher dose (20 nmol), was administered after MCD-induced seizures had started (fig. 1). The protective effect of riluzole against DTXi-induced seizures was less potent but was still observable. As shown in table 1, the threshold of seizure occurrence was slightly, but significantly (P < 0.05), increased from 0.08 to 0.12 nmol DTXi; the dose of 0.08 nmol being unable to elicit a seizure and 0.12 nmol producing a less severe seizure with more spaced spike-and-wave complexes (fig. 1). Moreover, no continuous seizures were observed with doses of D T X i up to 0.15 nmol (table 1). On the contrary, the 4-AP-induced seizures were not significantly prevented by riluzole (fig. 1 and table 1). The protective effect of riluzole against MCD-seizures was also very potent when administered p.o., indicating that the compound has good bioavailability and brain permeation. Under control conditions, 11 out of 12 rats showed stage 2 or 3 seizures. When administered 1 h before the toxin injection, riluzole (4 mg kg -1 p.o.) slowed the background E E G activity, and 5 out of 8 animals showed stage 0 or 1 seizures. When administered 6 h before MCD, a marked protection still remained in 4 out of 6 rats according to this criterion, indicating that riluzole has a relatively long duration of action. However, when 18 h separated riluzole administration and convulsant treatment, most (4 out of 6) of the animals showed stage 2 or 3 seizures (fig. 2).

_.m o 80

E o

60 40

®

20

0

MCD: RILUZOLE:

Time:

0.1 0

0.1 0.1 0.1 4 4 4 1 6 18

control

n m o l i.c.v.

m g / k 9 p.o. hours

Fig. 2. Time course of the anti-MCD effect of orally administered riluzole. When given 1 h before i.c.v, administration of MCD, riluzole exerted a powerful anti-epileptic action as shown by the significant reduction (P < 0.02, Fisher's exact test) in the proportion of animals showing severe (stage 2 or 3) seizures (narrow right-hatched bar graph). A similarly significant effect was still observable when the compound had been given 6 h before the toxin (wide left-hatched bar graph), but no significant protection could be seen 18 h after riluzole treatment although a lower proportion of animals showed stage 2 or 3 seizures in this group (cross-hatched bar graph) than in the control group (open bar graph).

campal CA1 neurons, which were recorded for up to 5 h. U p o n delivery of the neurotoxins to the recording chamber, the firing rate of all 9 neurons slowly increased and reached a sustained plateau generally above 20-30 Hz. Cells undergoing this sustained activity were exposed, via the superfusion bath, to various (at least two) concentrations of riluzole ranging from 1 to 30 /~M, added cumulatively. As shown in fig. 3, a concentration-dependent reduction of the M C D - or DTXiactivated firing rate levels was seen as soon as the compound reached equilibrium in the recording chamber. In the presence of 1 /~M riluzole, the excitations induced by M C D and DTX~ were reduced to 70 _+ 12% (mean + S.E., n = 5 trials) and 78 + 27% (n = 8 trials) respectively, of the pre-drug, toxin-induced sustained firing levels taken as 100%. In the presence of 3 /~M riluzole, firing fell to 31 + 5% (n = 5 trials) and 58 + 9%

3.2. Brain slice studies

The potassium channel blockers, M C D (1 /~M) and D T X i (0.5 ~M), were applied to a total of 9 hippoTABLE 1

Actions of riluzole upon potassium channel blocker-induced seizures. For the sake of clarity, only the mean frequency of induced seizure occurrences during 10 min after the first generalized crisis is given (in rain-1 ). A value of 1 indicates continuous seizures and absence of seizures is indicated by 0. There was no significant difference in seizure frequency between control, sham and experimental rats for which there was failure of seizure prevention for any given dose of potassium channel blocker, but all cases of prevention were significant (P < 0.01-0.05, Duncan's test). All animals received a single dose of convulsant except 3 rats out of 6 which were injected with increasing doses (0.45 nmol) of MCD when they were protected by riluzole from a normally lethal dose (0.30 nmol). Riluzole was unable to antagonize 4-AP-induced seizures while it significantly raised the threshold for DTXi-induced seizures and potently antagonized seizures induced by MCD. Abbreviations: see fig. 1. Number of rats in parentheses. Frequency of induced seizure occurrence/min MCD (nmol)

Control Sham Riluzole (10 nmol)

DTX i (nmol)

4-AP (nmol)

0.10

0.15

0.30 c

0.45 c

0.08

0.12

0.15 ~

100

200 c

(5) (6)

0.2 0.1

0.5 0.3

0.8

1

0.4 0.2

0.6 0.7

1

0.5 0.6

1

(6)

0 a

0 a

0.2 a

0.2 b

0 a

0.4

0.8

0.3

1

Duncan's test a p < 0.05, b p < 0.01; c Lethal doses.

227

®

~

RILUZOLE 1/zM RILUZOLE3~M 80

i

l

MCD 1/~M /

Illllllll

G)

~ D T X. . . . . . . 0,5#M

RILUZOLE 1/zM

RILUZOLE5/~M

RILUZOLE1OHM

o

~5 m i n . f

Fig. 3. Effect of riluzole and MCD- and DTXi-induced neuronal firing in hippocampal slices in vitro. Horizontal lanes A and B are action-potential discharge frequency recordings taken from 2 different hippocampal CAl-neurones. (A) Reversible effect of riluzole (1 and 3 /~M bath-applied during the time indicated by solid bars above tracings) on the sustained firing induced by 1 /zM MCD previously applied (not shown). Note the concentration-dependent reduction of firing and its reversibility on renewed application of the toxin. (B) Same effect as in (A) on DTXi-induced firing. Another cell was superfused with ACSF containing 0.5 p,M DTX i during the time indicated by the dots. Note that after each riluzole application, the spike-discharge frequency was concentration dependently reduced to pre-toxin firing levels.

(n = 6 trials) of pre-drug levels. After application of 10 /~M of the compound, the firing rate was reduced to 7.5 + 2.5% (n = 2) and 46 + 5% (n = 5) of the firing rate levels induced by MCD and DTX~ respectively. One cell, having been excited by DTX~, was treated with a concentration of 30 /~M riluzole which resulted in a decrease of the firing level down to 6% of pre-drug values. These inhibitory effects of riluzole appeared to be reversible upon washout of the drug by prolonged superfusion with toxin-free ACSF, or by a renewed application of 1 /LM MCD or 0.5/~M DTXp Assuming linearity of the relation in the concentration range studied, half-maximal effective concentrations (ECs0) of 1.8 ~M and 4.6 ~M were calculated for the effect of riluzole against MCD and DTX, respectively.

4. Discussion

The present study demonstrates the riluzole is a powerful antagonist of the hyperexcitability induced by the potassium channel blocking neurotoxin, MCD. Riluzole also protects against DTXi-induced seizures, although to a slightly lesser degree than MCD-induced seizures, and does not affect 4-AP induced seizures. Previous research (Bidard et al., 1987) has shown that molecules such as naloxone, morphine, diazepam and progabide do not prevent MCD-induced seizures. These seizures are, however, prevented by the potassium channel openers, cromakalim (BRL 34915), RP 49356 and RP 61419 (Gandolfo et al., 1989b), but none of these compounds are able to prevent DTX~- and 4-APinduced seizures (Gandolfo et al., 1989c). On the other hand, L-type calcium channel blockers, such as ( + ) P N 200-110, desmethoxyverapamil or fluspirilene, have been shown to be potent antagonists of seizures induced by

all three potassium channel blockers (Gandolfo et al., 1989a). Given the spectrum of the protective effect of rihizole now described, it appears that this drug does not behave identically to potassium channel openers since, in contrast to the above mentioned members of this class of drugs, riluzole was also effective against DTXiinduced seizures. Moreover, riluzole does not behave as a calcium channel blocker, which would have completely prevented the seizures induced by 4-AP in addition to those induced by M C D and DTX i. Furthermore, this absence of activity against 4-AP seizures indicates that riluzole does not behave like the classical NMDA-selective anti-glutamate, D ( - ) - 2 - a m i n o - 5 phosphono-valeric acid, which has been shown to prevent only 4-AP-, but not MCD- and DTXi-induced seizures (Gandolfo et al., 1989a). Riluzole appears, however, to interfere with glutamatergic neurotransmission (Brnavid~s et al., 1985) as indicated by the results of a number of biochemical and electrophysiological studies. For example, riluzole decreases the spontaneous release of glutamic acid (Chrramy, personal communication) as well as the glutamic acid-induced release of striatal dopamine in vivo (Baruch et al., 1988). In vitro assays have shown that riluzole counteracts the glutamic acid-induced release of aspartic acid from cerebellar cells (Drejer et al., 1986; Hubert and Doble, 1989), and the NMDA-evoked release of acetylcholine from striatum or olfactory tubercules (Brnavidrs et al., 1985). Extracellular recordings in vivo have shown that riluzole inhibits the responses evoked by excitatory amino acids micro-iontophoretically applied on motoneurones of the facial nucleus (Girdlestone et al., 1989). Recent evidence from biochemical studies with MCD suggests that the protective effect of riluzole against

228 t o x i n - i n d u c e d h y p e r - e x c i t a b i l i t y c o u ld result f r o m its a n t i - g l u t a m a t e properties. Indeed, in vivo ' p u s h - p u l l ' p e r f u s i o n studies have s h o w n that M C D p r o d u c e s a sustained release of excitatory a m i n o acids in the CA1 region of the h i p p o c a m p u s d u r i n g e p i l e p t o g e n i c activity ( A n i k s z t e j n et al., 1990). As a c o n s e q u e n c e , a l t h o u g h M C D and D T X i are b o t h p r i m a r i l y specific blockers of a class of v o l t a g e - d e p e n d e n t p o t a s s i u m c h a n n e l s (Benoit a nd D u b o i s , 1986; Stansfeld et al., 1987), the expression of their e p i l e p t o g e n i c action m a y involve a c a s c a d e of e ve nt i n c l u d i n g the release of g l u t a m a t e , the latter b ei n g the substrate of the p r o t e c t i v e effects of riluzole. T h e activity of riluzole o b s e r v e d on w h o l e - b r a i n E E G is in g o o d a g r e e m e n t with its effect in excised hippoc a m p a l slices. Since the h i p p o c a m p a l CA1 region has b e e n shown to be a theta g e n e r a t o r (Bland et al., 1975) possibly m o d u l a t e d by a septal p a c e m a k e r (Petsche et al., 1962), the sustained increase in n e u r o n a l firing-rate i n d u c e d by b o t h M C D a n d DTX~ at the single-cell level is very likely to be the cellular c o u n t e r p a r t of the q u a s i - p e r m a n e n t theta r h y t h m o b s e r v e d in vivo (Bidard et al., 1987). H o w e v e r , in c o n t r a s t to the results of E E G studies, riluzole exerted its p r o t e c t i v e effect in brain slices even w h e n t o x i n - i n d u c e d firing had developed, a nd its relative p o t e n c y to w a r d s D T X i as c o m p a r e d to M C D was m u c h stronger in the slice e x p e r i m e n t s than in the E E G studies. T h e s e results are p r o b a b l y related to the previously r e p o r t e d dual cortical a n d hippoc a m p a l origin of the seizures r e c o r d e d after DTX~ injection, whereas M C D - i n d u c e d seizures are of p u r e l y hipp o c a m p a l origin (B i d a r d et al., 1987; 1989). T h a t the effects of riluzole against the t o x i n - i n d u c e d h y p e r e x c i t a bility are p a r t i c u l a r l y p r o m i n e n t in the h i p p o c a m p u s , is consistent with p r e v i o u s research showing that this drug exhibits p r o t e c t i v e activity against n e u r o n a l cell loss in this structure (as d e t e r m i n e d by q u a n t i t a t i v e a u t o r ad i o g r a p h y of cholinergic receptors) a n d s u b s e q u e n t m e m ory deficit, c o n s e c u t i v e to transient cerebral i s c h a em i a in gerbils ( M a l g o u r i s et al., 1989). In conclusion, the p r e s e n t study shows that the spect r u m of riluzole's p r o t e c t i v e activity against the hyperexcitability p r o d u c e d by p o t a s s i u m c h a n n e l blockers has similarities with, b u t is n o t identical to, that of p o t a s s i u m c h a n n e l openers, a n d differs f r o m that of c a l c i u m c h a n n e l blockers a n d o t h e r a n t i - g l u t a m a t e c o m pounds. T h e a n t i - e p i l e p t i c action of riluzole in this m o d e l m ay h o w e v e r be the c o n s e q u e n c e of direct or i n d i r e c t effects on g l u t a m a t e r g i c systems.

Acknowledgements The authors wish to thank J. Pratt for fruitful discussions and K. Birmingham for style editing of the manuscript.

References Aniksztejn, L., C. Charriaut-Marlangue, M.P. Roisin and Y. Ben-Ari, 1990, Long term potentiation in the rat hippocampus induced by the mast cell degranulating peptide: analysis of the release of endogenous excitatory amino acids and proteins, Neuroscience 35, 63. Baruch, P., F. Artaud, G. Godeheu, L. Barbeito, J. Glowinski and A. Ch6ramy, 1988, Substance P and neurokinin A regulate by different mechanisms dopamine release from dendrites and nerve terminals of the nigrostriatal dopaminergic neurons, Neuroscience 25, 889. B6navid6s, J., J.C. Camelin, N. Mitrani, F. Flamand, A. Uzan, J.J. Legrand, C. Gueremy and G. Le Fur, 1985, 2-amino-6-trifluoromethoxybenzothiazole, a possible antagonist of excitatory amino acid neurotransmission. II. Biochemical properties, Neuropharmacology 24, 1085. Benoit, E. and J.M. Dubois, 1986, Toxin I from the snake Dendroaspis polylepis polylepis: a highly specific blocker of one type of potassium channel in myelinated nerve-fiber, Brain Res. 377, 374. Bidard, J.N., G. Gandolfo, C. Mourre, C. Gottesmann and M. Lazdunski, 1987, The brain response to the bee venom peptide MCD. Activation and desensitization of a hippocampal target, Brain Res. 418, 235. Bidard, J.N., C. Mourre, G. Gandolfo, H. Schweitz, C. Widmann, C. Gottesmann and M. Lazdunski, 1989, Analogies and differences in the mode of action and properties of binding sites (localization and mutual interactions) of two K + channels toxins, MCD peptide and dentrotoxin I, Brain Res. 495, 45. Bland, B.H., P. Andersen and T. Ganes, 1975, Two generators of theta activity in rabbits, Brain Res. 94, 199. B~3hme, G.A., J.M. Stutzmann and J.C. Blanchard, 1988, Excitatory effects of cholecystokinin in rat hippocampus: pharmacological response compatible with "central"- or B-type CCK-receptors, Brain Res. 451, 309. Drejer, J., T. Honore, E. Meier and A. Schousboe, 1986, Pharmacologically distinct glutamate receptors on cerebellar granule cells, Life Sci. 38, 2077. Gandolfo, G., C. Gottesmann, J.N. Bidard and M. Lazdunski, 1989a, Ca 2+ channel blockers prevent seizures induced by a class of K ÷ channel inhibitors, European J. Pharmacol. 160, 173. Gandolfo, G., C. Gottesmann, J.N. Bidard and M. Lazdunski, 1989b, K ÷ channel openers prevent epilepsy induced by the bee venom peptide MCD, European. J. Pharmacol. 159, 329. Gandolfo, G., C. Gottesmann, J.N. Bidard and M. Lazdunski, 1989c, Subtypes of K + channels differentiated by the effect of K + channel openers upon K ÷ channel blocker-induced seizures, Brain Res. 495, 189. Girdlestone, D., A. Dupuy, A. Coston, L. Roy-Contacin and D. Escande, 1989, Riluzole antagonises excitatory amino acid-evoked firing in rat facial motoneurones in vivo, Br. J. Pharmacol. 97, 583 P. Hubert, J.P. and A. Doble, 1989, Ibotenic acid stimulates D[3H]aspartate release from cultured cerebellar granule cells, Neurosci. Lett. 96, 345. Krieg, W.J.S., 1946, Connections of the cerebral cortex. I-The albino rat. A-Topography of the cortical area, J. Comp. Neurol. 84, 221. Malgouris, C., F. Bardot, M. Daniel, F. Pellis, J. Rataud, A. Uzan, J.C. Blanchard and P.M. Laduron, 1989, Riluzole, a novel antiglutamate, prevents memory loss and hippocampal neuronal damage in ischaemic gerbils, J. Neurosci. 9, 3720. Mizoule, J., B. Meldrum, M. Mazadier, M. Croucher, C. Ollat, A. Uzan, J.J. Legrand, C. Gueremy and G. Le Fur, 1985, 2-Amino6-trifluoromethoxy benzothiazole, a possible antagonist of excitatory amino acid neurotransmission I. Anti-convulsant properties, Neuropharmacology 24, 767. Petsche, H., C. Stumpf and Gogolak, 1962, The significance of the

229 rabbit's septum as a relay station between the midbrain and the hippocampus. I. The control of hippocampus arousal activity by the septum cells, Electroenceph. Clin. Neurophysiol. 14, 202. Rehm, H., J.N. Bidard, H. Schweitz and M. Lazdunski, 1988, The receptor site for the bee venom mast-cell degranulating peptide. Affinity labelling and evidence for a c o m m o n molecular target for mast-cell degranulating peptide and dendrotoxin I, a snake toxin active on K + channels, Biochemistry 27, 1827. Stansfeld, C.E., S.J. Marsh, D.N. Parcej, J.O. Dolly and D.A. Brown, 1987, Mast cell degranulating peptide and dendrotoxin selectively

inhibit a fast-activating potassium current and bind to c o m m o n neuronal proteins, Neuroscience 23, 893. Stutzmann, J.M., G.A. B~3hme, M. Cochon, M. Roux and J.C. Blanchard, 1987, Proconflict and electrocorticographic effects of drugs modulating GABAergic neurotransmission, Psychopharmacology 91, 74. Taylor, J.W., J.N. Bidard and M. Lazdunski, 1984, The characterization of high affinity binding sites in rat brain for the mast cell degranulating peptide from the bee venom using the purified monoiodinated peptide. J. Biol. Chem. 259, 13957.