Seizures can be triggered by stimulating non-cortical structures in the quaking mutant mouse

Epilepsy Res., 9 (1991) 19-31 Elsevier

19

EPIRES 00410

Seizures can be triggered by stimulating non-cortical structures in the quaking mutant mouse

Yves Gioannib, Henri Gioanni” and Nadia Mitrovic” ‘L~~~~~~~irede ~e~ruc~~~-A~~~m~e, Ins&t&des Newosciences ChrRS, UniversitPP. Brd4. Curie, Paris, and bU 97 INSERM, U&L;de Recherches SW ~Epi~eps~e,Paris (France) (Received 20 October 1990; revision received 15 February 1991; accepted 18 February 1991) Key words: Quaking mouse; Genetie madel; Seizure; Epilepsy; Electrical stimulation

Mutant Quaking mice (C57BL/6J) display convulsive tonic-clonic seizures that can be either spontaneous or triggered by manipulation of the animal or by auditory stimulation. Several abnormahties have been found (especially in the naradrenergic system) in the brainstem of this mutant strain. We first verified by el~aphysialo~cal recording that the cerebral cortex was not involved in the generatian or in the development of these fits. Then we showed that tonic-clonic seizures similar to those obtained in the freely moving animal were triggered by law-threshold (LT, S-50pA) or ~gh-threshold (HT, 5%15OpA) stimuli performed during head restraint, LT stimuli were mostly efficient in a number of into-b~bar and mesen~pha~c structures, including several reticular nuclei, the locus coeruleus, the nucleus sublet and the red nucleus, whereas HT stimuli were generally necessary to trigger fits by stimulating the nuclei pontis, the substantia n&a, the central gray area and the cerebellar nuclei. Seizures were also provoked at the diencephalic level with LT stimulation delivered in the medial tbalamic area, the nucleus reticularis thalami and some subthalamic regions (zona incerta, H field of Fore& In contrast, no fits were obtained by stimulating the cerebellar cortex and the inferior colliculus, the ventral and lateral groups of thalamic nuclei or the telencephalic regions (hippocampus, amygdala, caudate nucleus, putamen and cerebral cortex), with the exception of the globus pallidus.

Quaking (qk) mice, mutant of the C57BL16J strain, are characterized by a severe myelin deficiency, p~ncipally at the central nervous system leveP2. Clinically, they show a postural and action tremor. In addition, they exhibit tonic-clonic generalized seizures, either spontaneous, particularly during paradoxical sleep%, or triggered by manipulation of the animal or by auditory stimulaC~r~s~o~e~~e to: H. Gioanni, Lab. de Neuro~mie-Anatamie, Institut des Neurosciences CNRS, Universite P. & M. Curie, 9 Quai St. Bernard 75005 Paris. Tel.: (1) 44 27 32 33.

~2~~211~9l/$a3.Sa~

1991 Elsevier Science Publishers B.V.

tion. In the latter case, the tonic-clonic phase may be preceded by wild running activity, which may itself be preceded by a short tonic convulsive activity6. Myoclonic convulsions represent another type of seizure which can be t~ggered under particular conditions; however, these fits are rare and never spontaneous6 (see also Discussion). To date, there are no data indicating that the convulsive syndrome in the qk mice is directly linked to myelin deficiency. On the other hand, pharmacological and histological studies strongly suggest that the noradrenergic system, at the brainstem level, might play an essential role in the occurrence of seizures in this strain2tZ2’,

20 Only a few electrophysiological data are available in these mutants. They show that tonic-clonic seizures are associated with bilateral rapid discharges of low amplitude on the electrocorticogram (ECoG), resembling the ECoG in the waking state. Fu~hermore, there is no evidence of cortical paroxysmal activity6z4. All these results suggest that the seizures arise from structures located in the brainstem. Consequently, the aim of the present electrophysiological investigation was to identify the structures which could be implicated in the genesis and/or the propagation of the electrical seizures. To this end, various zones of the central nervous system were stimulated in waking mice under head restraint. The structures from which seizures were elicited and the corresponding stimulation thresholds were then determined.

Fig. 1. Experimental set-up used during central stimulations performed in the head restraint condition. The head is immobilized in the stereotaxic position by means of a metal plate (p) cemented onto the skull. A cran~ectomy allows penetrations of the stim~ating bipolar electrode (e). The mouse is represented during a seizure triggered by a central stimulation.

METHODS Experiments were carried out on SO qk mutant mice of the C57BL/6J strain (homozygotes qMqk). The controls (8 animals) were heterozygotes of the same strain (qW+). The animals were 3-4 months old. surgical procedure

The mice were anesthetized either by intraperitoneal injection of pentobarbital (60 mg/kg) or by halothane-N,O inhalation. To prepare animals for the stimulating experiments performed in the restrained-head condition, the head was maintained in a stereotaxic position, and a small metal plate was cemented to the skull in order to immobilize the head during the experiments of stimulation. A craniectomy was then performed and the dura was removed to allow electrode penetrations in a restricted region of the brain (Fig. 1). The exposed cortex was finally protected by a layer of low fusion-point paraffin. Some animals were prepared to allow electrocortical (ECoG) and electromyographic (EMG) recordings either during spontaneous seizures occurring in freely moving mice, or during seizures provoked by central electrical stimulation in head restrained mice. For this purpose, 2 Teflon-coated silver-ball electrodes were placed on the frontal

and the occipital cortex on one side, through small holes made in the cranial bone. In addition, a pair of electromyographic electrodes (Teflon-coated steel wires) was placed in the neck muscles and another pair was set in the foreleg (biceps brachii). The electrocortical and electromyographic electrodes were then connected to microconnecto~ cemented onto the skull. A bilateral ablation of the motor cortex was performed in some animals. The motor cortex was delimited by micro-stimulations (monopolar electrode) delivered at 1 mm depth. Stimulations were trains (50 msec duration) of 0.25 msec square pulses (frequency 300 Hz) delivered at intensities of 1.5-50pA. Experimental procedure The ECoG and EMG activities were recorded in

either freely moving or in head-restrained animals, on an ink oscillograph (Siemens), andbccasionally on magnetic tapes. In the stimulation experiments, the head of the waking mouse was fastened to a holder, in painless containment, by way of a skull plate (Fig. 1). Different regions in the brainstem and in the forebrain were stereotaxically localized and then stimulated via a bipolar electrode made of 2 twisted Teflon-coated platinum wires (75 ym in diameter).

The penetrations were perpendicular to the cortical surface. Each stimulation consisted of a train (duration of 50 msec) of square pulses (duration of 250 ,us, frequency of 330 Hz) delivered at variable intensities (up to 150 yA in the mutant and 600 ,uA in the controls). These stimuli were given at 300 pm intervals along each electrode track. Two successive stimulations were separated by a minimal delay of 2 min. A small electrolytic lesion was made at the end of each electrode penetration to allow histological location of the stimulation sites. Histological control The mice were deeply anesthetized with ketamine (150 mg/kg) and intracardially perfused (physiological saline followed by 10% formaldehyde solution). Then the brains were cut at 60 pm and stained according to the cresyl-violet method. The different stimulating points were transcribed

onto drawings of frontal sections established from the atlas of Lehmann” and of Sidman et a1.31.Although shrinkage of brains due to fixation was generally weak, it was taken into account in the reconstruction of the different stimulating points. RESULTS Both spontaneous and triggered (by manipulation of the animal or by auditory stimuli produced, for example, by jingling keys) tonic-clonic convulsions were easily obtained in the freely moving mice. Moreover, in the head-restrained mice, it was possible to trigger seizures either by central electrical stimulation or by natural (tactile or auditory) stimulation. However, in this experimental condition, spontaneous seizures were present only in some animals.

Fig. 2. Examples of ECoG and EMG recorded during tonic-clonic seizures obtained in different experimental conditions. In A, the recordings correspond to a spontaneous fit which occurred in a freely moving animal. In B and C, seizures were triggered either by a natural somesthesic stimulation (B) or by the central electrical stimulation of the nucleus centralis caudalis pontis (C), in the same restrained-head animal. Upper trace: ECoG (antero-posterior recordings on the same side). Middle trace: EMG of the neck. Lower trace: EMG of a foreleg. The onset and the end of each seizure are indicated by arrows. In C, the artefacts present at the beginning of the fit correspond to the electrical stimulation. Note the absence of cortical paroxysms, the flattening of the ECoG and the strong electromyographic activity (tonic in the neck, tonic-clonic in the foreleg) during the fits.

22 Electroclinical recordings of seizures obtained in the freely moving and in head restraint conditions

The fits triggered by electrical stimulation in the head-restrained animals presented electroclinical features similar to those observed in the freely moving animals. Under both conditions the fits were of the tonic-clonic type. Fig. 2 shows typical examples of ECoG and EMG recorded during a seizure occurring either in a freely moving animal (A) or in a head-restrained animal (B, C). The seizure illustrated in Fig. 2A occurred spontaneously whereas the fits shown in Fig. 213 and C were triggered by somesthetic and central electrical stimulation, respectively. The seizures were never accompanied by cortical paroxysms. Moreover, during the fits, the ECoG was close to a waking state activity, so that when the animal was drowsy, as shown in these examples, the ECoG was flattened from onset and during most of the seizure. On the other hand, the EMGs revealed a strong tonic activity of the neck muscles, and both tonic and clonic activities in the foreleg muscles. The clonic activity was generally present at the beginning and end of the seizures. Experiments with central electrical stimuli

In order to detect the regions from which seizures can be electrically triggered, stimuli were delivered in most parts of the brain (forebrain and brainstem), both in controls and in quaking mutant mice. The experiments were performed in waking, head-restrained animals. Control electrical stimulation

Control experiments were performed by stimulating (with the same parameters as those used in mutant mice) the bulbo-pontine and cerebellar regions (3 animals), the mesencephalic area including the grisea centralis (2 animals) and the thalamic and telencephalic structures (3 animals). Seizures were never provoked by these central stimulations, even with an intensity of 6OOpA. The stimuli were never accompanied by vocalizations or by any other sign of pain. Ele&~~~alstimulation in quaking mutant mice The structures investigated in the mutant mice

were divided into 3 groups according to their abili-

ty to trigger a seizure when they were stimulated either by a ‘low-threshold (LT) stimulation’ (stimulation intensity did not exceed 50 PA) or by a ‘high-threshold (I-IT) stimulation’ (stimulation intensities ranging from 55-150 PA). The third group included those structures from which seizures could not be elicited using stimulation intensities extending up to 150 ,uA. The mean thresholds were 36.4 & 14.9 yA (n = 223) for the LT values, and 104.9 It: 31.9 PA (n = 370) for the HT values, All the seizures triggered by stimulating the different structures were of the same type: a clonic phase, p~ncipally involving the forelegs, was followed by a tonic generalized seizure which often ended with a few myoclonic jerks of the forelegs. The fit duration varied from 8 to 35 set and depended principally on animals. No correlation has been found between the fits duration and the stimulated structures. When stimulations were delivered to the grisea centralis, a stereotyped reaction was frequently observed, consisting of a complete immobilization of the animal which began after the seizure and lasted for several seconds. Data obtained for the main structures from which fits were elicited are shown in Table I. Both the percentage of LT. HT and ‘no response’ points and the mean threshold of LT + HT stimuli are reported. The structures can be divided into 2 groups, the ‘LT structures’ which include the locus coeruleus, the bulbo-pontine reticular formation and the medial thalamic nuclei, and the ‘HT structures’ which include the substantia grisea centralis, the hypothalamic area and the pons. The lowest mean thresholds correspond to the higher percentages of LT stimuli, and the higher mean thresholds to the higher percentages of HT stimuli. The mean threshold of each structure of the first group is significantly different from the mean threshold of each structure of the second group. The structures will be examined from the posterior to the anterior levels of the brain. Ponto-bulbar and mesencephalic regions

From a quantitative point of view, the pontobulbar and mesencephalic regions appeared to be the most prominent zones involved in the genesis of seizures, since fits were elicited by stimulating

23 TABLE 1 ~~~~~tive

da& obtainedfor the main structuresin whichseizures were e~ec~~u~~y triggered

Both the percentage of ‘low-threshold’ (LT), ‘high-threshold’ (HT) and ‘no response’ stimuli and the mean thresholds (LT + NT stimuli) are reported. The structures are ordered from the lowest to the highest mean threshold. The bulbo-pontine KJ? includes the nuclei Pv, Gi, PGi, and CCp, the Pons camprises the NP, RTP, Cop, Lem, Pe, BP and the medial thalamic nuclei represent the MD, CM, PV, I&, and the NCM,

Number of animals Number of s~rnu~a~on points %LT %HT % No response Mean threshold &A) SD.

Locus coerukw

Bulbo-pontine RF

Medial thalamic Grkea centralis nuclei

7 19 58 31.5 10.5 54.58 44.48

13 121 43 25.5 31.5 57.76 47.42

4 19 52.6 31.6 15.8 65 33.01

P 2.96

P 2.32

P 1.64

P 1.56

10 75 25.5 42.5 32 80.78 42.67

Hypothalamic area

Pons

6 22 4.5 50 45.5 87.oFI 27.17

15 101 15.8 63.4 20.8 92.87 39.79

P 1.84

P -1.04

Fig. 3. Schematic repre~ntat~on of data obtained by central stimulation. Ali the stimuiation points are reported on drawings of frontal sections of the brainstem (A, B) and of the forebrain (E, C). The drawings are presented caudaI to rostra1 and the corresponding stereotaxic levels are mentioned. According to their ability to provoke a seizure, the stimulation points are designated: by filled circles (seizures triggered at low threshold, S-50pA), by filled triangles (seizures triggered at high threshold, 55-150fiA) or by empty triangles (no seizure triggered with stimulations extending up to 150pA).

A 0.10

A 0.50

AD, &g-J, aha,

ahp, AM, apl, ar, AV, BP, cc, CCP, Cd, Cei, Cel, Cem, Cf. CL Ci, Cing,

n. anterodorsalis thalami amygdala area hypothalamica anterior area hypothalamica posterior n. anteromedialis thalami area preoptica lateralis n. infundibularis n. anteroventralis thalami brachium ponti corpus callosum n. centralis caudalis pontis n. caudatus n. interpositus cerebelli n. lateralis cerebelli n. medialis cerebelli commissura fornicis colliculus inferior capsula interna cingulum

CL Cl, CM, Cmd, Cmv,

COP, cs, CT, CUT dCR, dmh, DTg, RP, FD, W FIR, Rx, G VII, GC,

n. centralis lateralis thalami claustrum n. centralis medialis thalami n. centralis medullae oblongatae, pars dorsalis n. centralis medullae oblongatae, pars ventralis n. centralis oralis pontis colliculus superior n. corporis trapezoidei n. cuneatus lateralis decussatio corpus restiformis n. dorsomedialis hypothalami n. dorsalis tegmenti n. entopeduncularis fascia dentata fimbria hippocampi fibrae ponti columna fomicis genu facialis substantia grisea centralis mesencephali

A 1.30

-

GCP, Gi, GLd, GLv, GM, GP, H, Hbl, Hbm, Hip, HY, HYP, IC, IP, L, LA, LC, lem, LMd, LP, MV,

substantia grisea centralis pontis n. gigantocellularis . n. corporis geniculati dorsalis thalamt n. corporis geniculati lateralis thalami n. corporis geniculati medialis thalami globus pallidus H field of Fore1 n. lateralis habenulae n. medialis habenulae hippocampus n. hypoglossi hypophysis n. interstitialis (Cajal) n. interpeduncularis n. lateralis thalami n. lateralis thalami, pars anterior n. locus coeruleus lemniscus medialis n. dorsalis lemnisci lateralis n. lateralis thalami, pars posterior n. motorius nervus trigemini

A 1.75

A 2.05

A 2.65

A 3.15

n. medialis dorsalis thalami n. tractus mesencephali nervus trigemini medial forebrain bundle (fasciculus medialis -, telencephali) MV n. medialis ventralis thalami n VII, n. facialis n. cuneiformis NC, NCf, n. cuneiformis NCP, n. commissurae posterioris NCs, n. centralis superior n. accessorius medialis (Darkschewitsch) ND, N Mam, n. mamillaris medialis N Mam p, n. mamillaris medialis, pars posterior n. pontis NP, n. tuber NR NSI, n. lateralis septi n. olivaris inferior GI, n. parvocellularis compactus p, n. parabrachialis PB, n. paracentralis PC, pedunculus cerebri Pe,

MD,

MeV,

26 K PG, PGd, PGI, Pf-f, PMd, PMv, PO, PP, Pt, Put, PV, Pv, Rd, Re, Ret, Rh, Ro, RP, RTP, S, SV, Sm

n. parafascicularis thalami n. paragigantocellularis n. paragigantocellularis dorsalis n. paragigantocellularis lateralis n. prepositus hypoglossi n. premamillaris dorsalis n. premamillaris ventralis complexus posterior n. peripeduncularis n. pretectalis thalami putamen n. paraventricularis thalami n. parvocellularis n. raphe dorsalis n. reuniens thalami n. reticularis thalami n. rhomboideus thalami n. Rolleri n. reticularis paramedianus n. reticularis tegmenti pontis n. tractus solitarii n. sensorius principalis n. trigemini stria medullaris thalami

many structures, with low-threshold or highthreshold stimulations, at these levels. Fig. 3A,B shows a map of all the stimulation points located in these regions. The ponto-bulbar and the mesencephalic reticular nuclei constitute the most extensive zone from which fits were triggered with LT stimuli. These nuclei principally include the nuclei gigantocellularis (Gi), paragigantocellularis (PGi), parvocellularis (Pv), and the nucleus centralis caudalis pontis (CCp). Another region where many LT and HT stimuli were effective includes the locus coeruleus (LC), the nucleus subcoeruleus (Sub C) and the zone anterior to these nuclei. Furthermore, it can be noted that seizures were also provoked at low or high threshold, although less frequently, by stimulating the nucleus olivaris inferior (01) and the nuclei vestibularis (Ve). Finally, at the mesencephalic level, fits were provoked, at low or high threshold, by stimulations located in the nuclei ruber (NR) and cuneiformis (NCf), and more ventrally by stimulating the region including the lemniscus medialis (lem) and the pedunculus cerebri (Pe). In a second group of structures, fits were mostly induced by HT stimulation. These structures include, from the caudal to rostra1 levels, the nuclei centralis medullae oblongata (Cmd, Cmv), raphe

SN, SNc, SNr, Spa V, St, Sub,

Sub C, tc, to, v III, VA, Ve, Vel, Vem, Ves, VL V lat, vmh, VPL, VPm, VTg, Zf,

substantia nigra substantia nigra, pars compacta substantia nigra, pars reticularis n. tractus spinalis n. trigemi oralis stria terminalis subiculum n. subcoeruleus tractus corticospinalis tractus opticus ventriculus tertius n. ventralis thalami, pars anterior n vestibularis n. vestibularis lateralis n. vestibularis mediahs n. vestibularis spinalis n. ventralis thalami, pars lateralis ventriculus lateralis n. ventromedialis hypothalami n. ventralis posterolateralis thalami n. ventralis posteromedialis thalami n. ventralis tegmenti zona incerta

magnus, spinalis and motorius trigemini (SV, MV), pontis (NP), centralis oralis pontis (Cop), reticularis tegmenti pontis (RTP), the substantia grisea centralis and the substantia nigra (SNr, SNc). On the other hand, it is interesting to note that seizures were often provoked by HT stimulation of the nuclei cerebelli (Cei, Cep, Cem), but were never obtained by stimulating the cerebellar cortex, even with stimulation intensities extending up to 300 ,uA. In Fig. 3A, the cerebellar cortex is not completely represented, but was extensively investigated in our experiments. Finally, fits were generally not produced by stimulation of the inferior colliculus (CI), whereas the results appear more complex for the superior colliculus (CS) since a number of seizures were provoked by LT and HT stimulations of the medial part, but not of the lateral part of this structure. Diencephalic region

Fits were also triggered by LT and HT stimuli delivered in the diencephalon. However, the LT convulsive area appeared to be more restricted in the diencephalon as compared to the brainstem. Although seizures were triggered by the stimulation of different nuclei or bundles, they were most-

27 ly provoked by stimulating several small nuclei placed in a medial thalamic area inclu~ng the intralaminar and the midline regions (Fig. 3B,C). The nuclei from which seizures were elicited at low-threshold principally belong to the following, often called ‘non-specific’, nuclei: reticularis thalami (Ret), centralis mediaiis thalami (CM), paracentralis (PC), medialis dorsalis thalami (MD), reuniens thalami (Re), paraventricularis thalami (PV) and anteroventralis thalami (AV). Furthermore, some fits were obtained with HT stimuli localized in the nuclei anteromedialis thalami (AM), and ventralis anterior thalami (VA). In the dorsal thalamus, many seizures were elicited by HT stimuli of the nucleus corporis geniculati medialis (GM). Despite the small number of stimuli placed in the nuclei corporis geniculati ventralis (GLv) and dorsalis (GLd), it seems that seizures can be triggered at LT or HT by stimulating the former but not the latter nucleus. At this same level, convulsive seizures were induced at either LT or HT by stimulating the nuclei pretectalis thalami (Pt). In the ventral thalamus, a large proportion of fits were also provoked by LT or HT stimulations of the zona incerta (ZI), and the area tegmentalis (H field of Forel) . It should be emphasized that practically no fits were triggered by stimulating either the nuclei ventralis posterior (VP) and ventralis lateralis (VL) thalami or the lateral complex including the nuclei lateralis (L), lateralis anterior (LA) and lateralis posterior (LP) thalami. In the hypothalamic region, seizures were elicited with HT stimuli in the area hypothalamica posterior (ahp), in the nucleus arcuatus hypothalami or nucleus infundibularis (ar), and more rostrally in the nuclei dorsomedialis (dmh) and ventromedialis (vmh) hypothalami. Finally, several fits were elicited by stimulating some well identified bundles such as the capsula interna (Ci) either at low or high threshold, and the stria terminalis (St), at high threshold.

With the exception of some seizures provoked by HT stimuli located in the globus pallidus (GP), no seizures were obtained by stimulating the fol-

lowing other telencephalic structures: hippocampus (Hip), amygdala (Agd), nuclei caudatus (Cd), putamen (Put) and neocortex (Fig. 3B,C). Furthermore, the bilateral ablation of the motor cortex (delimited by microstimulation) did not suppress the s~ntaneous seizures in the freely moving animal.

DISCUSSION With regard to the data obtained by central electrical stimulation, thresholds were divided into 3 categories (LT, 5-50 PA, HT: 55-150 PA and no response). The limit between LT and HT stimuli was fixed at 50 PA because the structures from which seizures were triggered with the lowest intensities had thresholds below 50 ,uA. Moreover, we verified that in the main structures in which fits were elicited, the LT and HT values were significantly different. The highest threshold value was set at 150pA in order to limit the extention of stimulated sites. It can be seen in several tracks (Fig. 3, for example at the levels A 1.75, P 0.5, P 2.96) that a 300 pm displacement of the electrode was sufficient to switch from a negative point to a LT point. This indicates of the possible spreading of current, since a stimulus of 150 PA was unable to involve a LT site located 300 ,um from the tip of the electrode. The stimulations were delivered at 2-min intervals. Thus, due to postictal refractoriness, the seizure thresholds determined after the induction of a first seizure or several seizures could be increased. Nevertheless, we verified that, once a seizure has been triggered (and its threshold determined) by stimulating a structure, successive stimulations delivered at 2-min intervals in the same structure provoked seizures with the same threshold. Moreover, in several cases, successive seizures were triggered with LT stimulations delivered in different structures along the same electrode track (with the same stimulation intervals). Therefore, the presence of a seizure induced from a structure did not appreciably increase the threshold for the following seizures, provoked by stimulating either the same structure or neighboring ones. We have also observed that it was possible, in quiet animals,

28

to provoke seizures at 2-min intervals for a period of 1 h by tactile stimulation (handling). Thus, the postictal refractoriness is very short in this strain. This observation is in agreement with the results of Caboche et al5 and Mitrovic et a1.26who used also short intervals of tactile stimulation (45 set) to test the seizure susceptibility of qk mice. The results show that the structures from which seizures were elicited by stimuli of low intensity are essentially located in the brainstem and in the medial part of the thalamus. Seizure induction by electrical stimulation of a specific brain region does not necessarily mean that this region is involved in epileptogenesis. Nevertheless, several observations suggest that the structures from which seizures were elicited in our experiments are involved at least in the development of seizures: (1) the electrically triggered seizures were clinically similar to the spontaneous seizures; (2) the ECoG activity was also identical in these two conditions; (3) in the case of the locus coeruleus, as discussed later, the results concerning experiments of lesion23 are in perfect agreement with our results of stimulation; (4) it may be interesting to recall that, in human epilepsy, central electrical stimulations are used to trigger seizures and to determine their origin’. Previous studies have reported that electrical stimulation of the reticular formation (RF), with a relatively high intensity and a long duration, can elicit convulsions without any epileptiform activity in the cortex of rabbits, cats and rats2,16.In the rabbit, the ‘low-threshold convulsive area’ was found to be limited to a part of the mesencephalic RF2. In our study, seizures were also elicited in the qk mice, by stimulating a similar region with LT intensities, but fits were additionally obtained by stimulating the bulbo-pontine RF, especially in a caudal part including the nuclei gigantocellularis, centralis caudalis pontis and parvocellularis. No seizures were triggered in the control mice with our stimulations. Since we used short duration and low intensity stimuli, it could be possible that stronger stimulations would have provoked convulsions. However, it should be noted that the convulsions described in normal animals2g16differ markedly from the seizures of the qk mice with regard to their respective clinical aspect.

In a number of recent investigations, it has been demonstrated that the RF is implicated in the initiation and propagation of the seizures in several models of generalized epilepsy (kindling, electroshock, alcohol withdrawal or administration of convulsant drugs) including some genetic models (genetically epilepsy-prone rat (GEPR), photomyoclonus), (see Faingold12 for a review). Our results suggest that, in the qk mouse, the bulbo-ponto-mesencephalic RF also plays a major role in the triggering of tonic-clonic seizures. This point is strengthened by the fact that, as in the seizures obtained by stimulation of the brainstem in normal animals2,‘6, the cortical EEG of qk mice is devoid of any epileptic-like activity during the fits. This lack of cortical epileptiform activity seems to be a general feature of the audiogenic seizures (AGS), since it has been described in several strains of m&25, in the ratis, rabbit2’ and cat35. Cortical EEG spikes have been observed by Chauvel et a1.6 in the qk mouse during another type of seizure, mostly of the myoclonic type. However, these myoclonic seizures were triggered only rarely and after creating an intense stresscondition (noise, manipulation of the cage) during at least 10 min, so that the mice exhibited many tonic-clonic seizures before the appearance of myoclonic fits and of the associated cortical spikes6 (and personal communication). These myoclonic fits were neither observed in natural conditions (spontaneously or after stimulation) nor with our central stimulations. This phenomenon can be compared to the observations of Krushinski18 regarding the myoclonic convulsive seizures in the rat susceptible to AGS. After repeated auditory stimuli, besides audiogenic seizures, these animals developed myoclonic convulsions with cortical EEG spikes. More recently, Tacke et a1.33and Naritoku et a1.28reported that, in the GEPR, repeated induction of maximal AGS produced spikewave discharges at the cortical level. In these cases, the seizures were shown to originate in the brainstem and to spread secondarily to the cortex. Finally, in Wistar rats, Marescaux et a1.20have recently shown that daily auditory stimuli provoked the progressive development of cortical EEG spikes and spike and waves. In the qk mouse, it may also be so that the myoclonic seizures result

29 from the spread of paroxysmal activity originating in sub-cortical structures due to repetitive strong stimulations. The fact that seizures were triggered by stimuli delivered at low intensity in the locus coeruleus and the nearby zone is interesting to compare with pharmacological results obtained in the qk mouse in relation to the noradrenergic system. Seizures are suppressed by the aZadrenoceptor antagonist yohimbine, and this effect is reversed by the a2adrenoceptor agonist clonidine7T8. Furthermore, a series of neuropharmacological experiments have demonstrated an abnormal noradrenergic neurotransmission in the brainstem of qk mice compared to the control of the same strain. These abnormalities mainly consist of a higher concentration of endogenous noradrenaline2’, an enhanced number of a-adrenoceptor binding sites24 and a higher number of noradrenergic cell bodies in the locus coeruleus22. Finally, the fits were abolished by bilateral electrolytic lesion of the locus coeruleus23 Thus, the noradrenergic system at the brainstem level might play an essential role in the occurrence of seizures in qk mice. The data obtained from lesion of the locus coeruleus on the one hand, and by stimulation of this nucleus (present study) on the other, are in perfect agreement. Several low-threshold points were found at the diencephalic level, mainly in certain nuclei of the intralaminar and medial groups, in the nucleus reticularis thalami, and in the subthalamus (H field, zona incerta). Thalamocortical systems probably do not contribute much to this effect, since the cortical EEG does not show any epileptiform activity during the seizures. It is of interest that 2 main groups of ‘low-threshold structures’, namely the bulbo-ponto-mesencephalic RF and the reticular, intra-laminar and medial thalamic nuclei, have been referred to as ‘non-specific’ or ‘diffuse’ systems, both on the basis of the multiple sensory afferents they receive and of the global effects obtained after electrical stimulation4V’03’7Y27. These characteristics, and in particular the important role of the RF in the mechanisms of attention, may be related to the fact that, in qk mice, spontaneous or triggered seizures are more frequent during drowsiness or when the mice are quiet than when the animals are

restless. Except for a few points located in the globus pallidus, stimulations of the telencephalic structures were unable to elicit any convulsion. These results are in agreement with the absence of epileptogene-like activity on the cortical EEG and, moreover, they suggest that the telencephalic structures may not be necessary for the expression of tonic-clonic convulsions in qk mice. Concerning the auditory structures, we failed to obtain seizures by stimulating the inferior colliculus (IC), and only some HT stimuli were effective in triggering fits from the medial geniculate body (GM). Maxson and Cower? have suggested that the pathway from the IC to the RF is essential in the elaboration of all AGS. However, we can postulate that in qk mice the auditory relay nuclei do not play an important role in the elaboration of seizures. Interestingly, in the GEPR, which has been the best studied audiogenic model (see Faingold13 and Dailey et a1.9for recent reviews) the IC appears to have an important function in the initiation of seizures, since AGS can be elicited by electrical or chemical stimulation of this structurei5, and can be inhibited by local microinjections of GABA agonists or of an excitatory amino-acid antagonist3s11s14a15, Moreover, the IC of GEPR contains an increased number of GABAergic cells3’. On the contrary, the results of Caboche et al.’ obtained in the qk mouse, suggest that convulsions are not associated with a GABAergic dysfunction. Therefore, even though qk mice and GEPR exhibit AGS and have some features in common (no cortical epileptiform activity, implication of the RF, etc.), some important differences exist between them, suggesting that the mechanisms responsible for the seizures in these 2 models differ. It may be interesting to recall that the manipulation of qk mice also triggers convulsions and that spontaneous seizures are frequent in these animals. Therefore, it can be hypothesized that seizures are elaborated from the RF but are triggered either by an external stimulus (manipulation, sound) or by an unknown internal processus (spontaneous seizures). Thus, in the case of an acoustic stimulation, the stimulus would act primarily on the RF, but some auditory structures like the IC might be secondarily involved. This hy-

30

pothesis, which supposes a sharing between mechanisms responsible for the wild circling phase and for the convulsive phase during AGS, is supported by 3 observations: (1) a partial lesion of the IC, sparing the ventral aspect of this nucleus, blocks the wild circling activity in cats but not the convulsive phase of the AGS36; (2) in qk mice, the wild circling phase is often absent during the AGS; (3) when present, the wild circling phase is sometimes preceded by a short tonic convulsion. In the latter case, the short tonic convulsion would constitute the first phase of the seizure o~ginating in the RF, before the involvement of other structures in an abnormal way (due to hyperactivity of the RF) and responsible for the wild circling. The fact that, when animals are quiet, spontaneous and triggered seizures are more frequent than when they are restless, also argues in favor of the importance of the RF in the initiation of fits. Therefore, seizures would only be triggered when the RF is in a certain state of excitability corre-

REFERENCES 1 Bancaud, J., Talairach, J., Morel, P. and Bresson, M., La come d’Ammon et le noyau amygdalien: effets chimiques et tlectriques de leur stimulation chez I’homme, Rev. Neurol., 115 (1966) 329-352. 2 Bergmann, F., Costin, A. and Gutman, J., A low threshold convulsive area in the rabbit’s mesencephalon, Electroencephalogr. Ctin. Neurophysiot.,

15 (1963) 683-690.

3 Browning, R.A., Lanker, M.L. and Faingold, C.L., Injec-

tions of noradrenergic and GABAergic agonists into the inferior colliculus: effects on audiogenic seizures in genetically epilepsy-prone rats, Epilepsy Res., 4 (1989) 119-125. 4 Buser, P. and Bignall, K.E., Non-primary sensory projections on the cat neocortex, int. Rev. Neurobiol., 10 (1967) 111-165. 5 Caboche, J., Mitrovic, N., Le Saux, F., Besson, M.J., Sauter, A. and Maurin, Y., Postnatal evolution of the gammaaminobutyric acid/benzodiazepine receptor complex in a model of inherited epilepsy: the quaking mouse, J. Neurothem., 52 (1989) 419-427. 6 Chauvel, P., Louvel, J., Kurcewicz, I. and DeBono,

M., Epileptic seizures of the quaking mouse: electroclinicai correlations. In: N. Baumann (Ed.), Neurological Mutates Affecfing Mye~~~~on, Elsevier, Amsterdam, 1980, pp. 513-516. 7 Chermat, R., Doare, L., LaChapelle, F. and Simon, P., Effects of drugs affecting the noradrenergic system on convulsions in the quaking mouse, Naunyn-Schmiedeberg’s Arch. Pharmacol., 318 (1981) 94-99.

sponding to the quiet state of the animal. In conclusion, we can stress that, as in several genetic models of generalized convulsions, the RF seems to be strongly implicated in the genesis of seizures in the qk mouse. Nevertheless, other structures like the locus coeruleus and the thalamic medial and reticular nuclei might also play a major role. Interestingly enough, in addition to the triggered fits, qk mice exhibit frequent spontaneous seizures. More studies, particularly involving cellular recordings, will be necessary for further understanding of this model. ACKNOWLEDGEMENTS The authors are grateful to M. Rogard for the breeding of mice, J. Prevost and A. Sansonetti for helpful technical support, and G. Chaillou for the histology. We would also like to thank Dr. Micelli for helpful participation in the elaboration of the manuscript.

8 Chermat, R., LaChapelle, F., Baumann, N. and Simon, P., Anticonvulsant effect of yohimbine in quaking mice: antagonism by clonidine and prazosin, Life Sci., 25 (1979) 1471-1476. 9 Dailey, J. W., Reigel, C.E., Mishra, P.K. and Jobe, C.J., Neurobiology of seizure predisposition in the genetically epilepsy-prone rat, Epilepsy Res., 3 (1989) 3-17. 10 Dempsey, E.W. and Morison, R.S., The production of rhythmically recurrent cortical potentials after localized thalamic stimulation, jr.Physiof., 135 (1942) 293-300. 11 Duplisse, B.R., Picchioni, A.L., Chin; L. and Consroe, P.F., Relationship of the inferior cofliculus and gammaaminobutyric acid (GABA) to audiogenic seizure in the rat, Fed. Proc., 33 (1974) 468. 12 Faingold, CL., The role of the brain stem in generalized epileptic seizures, Metab. Brain Dis., 2 (1987) 81-112. 13 Faingold, CL., The genetically epilepsy-prone rat, Gen. Pharmac.,

19 (1988) 331-338.

14 Faingold,

CL., Millan, M.H., Boersma, CA. and Meldrum, B.S., Excitant amino acids and audiogenic seizures in the genetically epilepsy-prone rat. I. Afferent seizure initiation pathway, Exp. Neural., 99 (1988) 678-686. 15 Huxtable, R. and Laird, H., The prolonged anticon~sant action of taurine on geneticatfy determined seizure-susceptibility, Can. J. Neural. Sci., 5 (1978) 215-221, 16 Kreindler, A., Zuckermann, E., Steriade, M. and Chimion, D., Electroclinical features of convulsions induced by stimulation of brain stem, J. Neurophysiol., 21 (1958) 430-436. 17 Krupp, P. and Monnier, M., The unspecific intralaminary

31

18

19 20

21

modulating system of the thalamus, Int. Rev. Neurobiol., 9 (1966) 45-94. Krushinski, L.V., Etude physiologique des differents types de crises convulsives de l’epilepsie audiogtne du rat, Coil. ht. CNRS, 112 (1963) 71-92. Lehmann, A., Atlas Stt!r&otaxique du Cerveau de la Souris, C.N.R.S.,Paris, 1974. Marescaux, C., Vergnes, M., Kiesmann, M., Depaulis, A., Micheletti, G. and Warter, J.M., Kindling of audiogenic seizures in Wistar rats: an EEG study, Exp. Neurol., 97 (1987) 160-168. Maurin, Y., Arbilla, S., Dedek, J., Lee, C.R., Baumann, N. and Langer, S.Z., Noradrenergic neurotransmission in the brain of a convulsive mutant mouse. Differences between the cerebral cortex and the brain stem, Naunyn-

mation and activation of E.E.G., Electroencephalogr. Clin. l(l949) 455-473. 28 Naritoku, D.K., Mecoxxi, L.B. and Faingold, CL., Effects of repeated audiogenic seizures (AGS) on seizure severity and EEG in two substrains of the genetically epilepsyprone rats (GEPRS), Sot. Neurosci. Abstr., 14 (1988) 252. 29 Nellhaus, G., Experimental epilepsy in rabbits: observations in a strain susceptible to audiogenic seizures. In: R.G. Busnel (Ed.), Psychopharmacologic, Neuropharmacologie et Biochimie de la Crise Audiogene, C.N.R.S., Paris, 1963, pp. 131-150. 30 Roberts, R.C., Ribak, C.E. and Oertel, W.H., Increased numbers of GABAergic neurons occur in the inferior colliculus of an audiogenic model of genetic epilepsy, Brain

Schmiedeberg’s

31 Sidman, R.L., Angevine,

Arch. Pharmacol.,

320 (1982) 26-33.

22 Maurin, Y., Berger, B., Le Saux, F., Gay, M. and Bau-

mann, N., Increased number of locus coeruleus noradrenergic neurons in the convulsive mutant quaking mouse, Neurosci. Len., 57 (1985) 313-318. 23 Maurin, Y., Enz, A., Le Saux, F. and Besson, M.J., Super-

numerary locus coeruleus neurons as a determinant of inherited epilepsy in the convulsive mutant mouse quaking, Brain Res., 366 (1986) 379-384. 24 Maurin, Y., Le Saux, F., Graillot, C. and Baumann,

N., Altered postnatal ontogeny of al- and a2-adrenoceptor binding sites in the brain of a convulsive mutant mouse (quaking), Dev. Brain Res., 22 (1985) 229-235. 25 Maxson, S.C. and Cowen, J.S., Electroencephalographic correlates of the audiogenic seizure response of inbred mice, Physiol. Behav., 16 (1976) 623-629. 26 Mitrovic, N., Besson, M.J. and Maurin, Y., Anticonvulsant effects of antagonists of the N-methyl-o-aspartate receptor complex in a genetic model of epilepsy: the quaking mouse, Eur. J. Pharmacol., 176 (1990) 357-361. 27 Moruzzi, G. and Magoun, H.W., Brain stem reticular for-

Neurophysiol.,

Res., 361(1985)

324-338.

J.B. and Pierce, E.T., Atlas of Harvard University Press, Cambridge, MA, 1971. 32 Sidman, R.L., Dickie, M.M. and Appel, S.H., Mutant mice (Quaking and Jimpy) with deficient myelination in the central nervous system, Science, 144 (1964) 309-311. 33 Tacke, U., Tuomisto, L. and Danner, R., Cortical spike wave discharges during audiogenic convulsions in rats, the Mouse

Brain and Spinal Cord,

Exp. Neural., 85 (1984) 233-238. 34 Valatx, J.L. and Jouvet, M., Souris quaking: etude du som-

meil et des crises comitiales, C.R. Sot. Biol., 165 (1971) 2131-2136. 35 Wada, J. and Ikeda, H., Phenomenological

and pharmacological identity of audiogenic-sensitive state in animals treated with methionine sulfoximide and in genetically audiogenic-seizure-susceptible animals, Recent Adv. Biol. Psychiatry, 9 (1969) 131-153.

36 Wada, J.A., Tevao, A., White, B. and Jung, E., Inferior

colliculus lesion and audiogenic seizure susceptibility, Exp. Neural., 28 (1970) 326-332.