Tohoku J. Exp. Med., 1990, 161, Suppl., 253-271

Intractability of Complex Partial with Secondary Generalization: Kindling Studies in Cats

MITSUMOTOSATO Department of Neuropsychiatry, Okayama University Medical School, Okayama 700

SATO, M. Intractability of Complex Partial Seizure with Secondary Generali-

zation: Kindling Studies in Cats. Tohoku J. Exp. Med., 1990, 161, Suppl., 253- 271•\Among many factors linked to an intractability of partial seizure secon-

darily generalized, changes in the brain resulting from repeated epileptic

are discussed mainly in light of our evidence obtained from kindling studies in cats. In addition, the literatures on biological mechanisms of the kindling effect are

reviewed briefly. The evidence demonstrated and reviewed here indicates that: 1) limbic structure is not susceptable to develop generalized convulsions initially,

2) repeated attacks of limbic seizures result in a profound reduction in seizure

threshold at the primary epileptogenic focus in the limbic structures, 3) once a limbic seizure developed to secondarily generalized convulsion, it seldom changes

into the original partial seizure, 4) kindled events in the limbic structures are more

profound and persistent than that in the cerebral cortex, and 5) repetition of focal cortical or limbic seizures may eventually produce spontaneous convulsive seizures

originating in the limbic structures. These findings strongly suggest that the limbic system, rather than the cerebral cortex, is more susceptible to a lasting

functional change resulted from seizure repetition, which can lead to an intractability of epilepsy with partial seizure. Lasting changes in the cell mem-

brane including long-lasting enhancement of inositol phospholipid hydrolysis of

the amygdala stimulated by excitatory amino acid appear important for develop- ment of trans-synaptic changes underlying the kindling-induced seizure susceptibil-

ity.•\intractability; limbic seizure; focal cortical seizure; secondary sei-

zure generalization; kindling mechanism

Epilepsy is one of the most common neurological diseases with an incidence estimated to be between 0.3 to 0.5%. Some epileptic patients suffer from intractability of epileptic seizures. Many factors that lead to a poor prognosis of epilepsy have been documented except for progressive excerbation of some under- lying neurological diseases. Rodin (1968) described 4 factors linked to poor prognosis : 1) partial or mixed seizure types, 2) presence of an abnormal neur- ological or mental state, 3) low IQ, and 4) increase in the number of seizures and duration of epilepsy before referral. Reynolds and others (1983) reported that

Present address and reprints requests to : Mitsumoto Sato, Department of Neurop- sychiatry,`Tohoku University School of Medicine, 1-1 Seiryo-machi, Aoba-ku 980, Sendai, Japan. 253 254 M. Sato there was less certainty about the influence of age of onset of epilepsy, of EEG abnormalities, and of genetic factors. The factor that the seizure type links to an intractability of epilepsy has been confirmed by a Japanese epilepsy study group in 1981 (Okuma and Kumashiro). According to the data on seizure control, the remission rates for partial seizures with secondary generalization and some of the combined seizure types such as elementary and complex partial seizures with generalized tonic-clonic seizures were significantly lower than the mean remission rate of seizures overall (Table 1). With regards to the number of seizures and duration of epilepsy as a poor prognostic factor, Gowers (1881) has stated : "when one attack has occurred, whether in apparent consequence of an immediate excitant or not, others usually follow without any immediate traceable cause. The effect of a convulsion on the nerve centres is such as to render the occurrence of another more easy, to intensify the predisposition that already exists. Thus every fit may be said to be, in part, the result of those which have preceded it, the cause of those which follow it "(Reynolds et al.1983). The following evidence that support this Gowers' hypothesis has been available thereafter. Firstly, there are reports of spontane- ous convulsions that appeared in non-epileptic patients with schizophrenia after repeated electroconvulsive therapy too many times. In fact, more than 200 such cases have been reported by at least 14 authors (Sato and Wada 1975). As an example, Naoi (1959) has reported 35 cases with spontaneous convulsive seizures out of 172 schizophrenic patients who had been treated with electroconvulsive therapy : 26 out of these 35 cases had received the therapy more than 50 times. Secondly, the occurrence of spontaneous convulsions as well as an augmented

TABLE.1 Outcome of seizure control

Remission means disappearance of seizure for more than 3 years. GTC, generalized tonic-clonic convulsion ; n, number of subjects (Okuma and Kumashiro 1981). Epilepsy and Kindling 255 epileptic response after repeated electroconvulsive seizures has been reported in animal experiments by several authors (Alonso-de Florida and Delgado 1958 ; Delgado 1959; Delgado and Sevillano 1961; Herberg and Watkins 1966). God- dard and others (1969) investigated such phenomenon systematically and used the term kindling effect to describe the phenomenon. The kindling effect seems to be an adequate animal model to study the Gowers' hypothesis.

Kindling effect and its transference phenomenon The kindling effect is a process whereby repeated induction of local epileptic manifestation results in long-term changes in neural organization, and when repeated many times, eventually results in generalized convulsions (Goddard et al. 1969 ; Racine 1972) and spontaneous seizures (Goddard and Douglas 1976 ; Wada et al. 1974). The kindled events which include readiness for kindled generalized convulsions and produced interictal spike discharges persist for one year without further stimulation (Wada et al. 1974). Related to kindling is the phenomenon called transference, which is a significant saving of kindling stimuli required for re-kindling at a secondary brain site following a primary brain site kindling. Transfer of the primary site kin- dling to a secondary site was confirmed after removal of the primarily kindled brain site (Racine 1972). Accordingly, this phenomenon presumably based on a trans-synaptic change from the primarily kindled site to the secondary site. Moreover, the kindling transfer was found to occur after a 14 day interval between the end of the primary site kindling and the beginning of the secondary site re-kindling (McIntyre and Goddard 1973). These findings indicate that the kindling-induced functional change in the secondary brain site lasts at least for 2 weeks, and is independent of the change in the primary site. The necessary conditions for kindling are recurrent appearance of after-discharges (AD) with an adequate intertrial interval (Racine 1978). Repeated administration of a stimu- lus which intensity is insufficient to produce an AD can not preduce kindling, and repeated stimulations applied at less than 20 min interval often produce adapta- tion rather than kindling (Goddard and Douglas 1976). Thus, it is clear that repeated induction of localized AD results in a lasting increase in seizure suscepti- bility in the brain which presumably based upon a trans-synaptic change originat- ing in the primary epileptogenic focus.

General procedures A series of our data described here are restricted to those obtained from kindling in cats. The kindling stimulus was one sec trains of a 60-Hz sine wave at an intensity minimal to produce AD. In kindling session, the kindling stimu- lus was delivered once daily until generalized convulsions were produced on 5 consecutive days. After 7 days from the last convulsion, re-kindling from a secondary brain site was started to examine the transference phenomenon. Our 256 M. Sato previous classification of seizure stages in amygdaloid (Wada and Sato 1974), hippocampal (Sato 1975), septal (Sato 1976, 1980) and Sylvian gyrus (Sato et al. 1980b) kindling in cats was used to assess chronological changes in behavioral seizure manifestations during kindling and re-kindling sessions. Behavioral observation with simultaneous EEG recording was performed before and after each daily stimulation and during spontaneous seizures.

Which is more susceptible to produce generalized convulsions, the cerebral cortex or limbic structures ? A one-sec stimulation of one mA intensity was delivered to the sylvian gyrus and limbic structures to examine the first trial generalized convulsions. The first trial generalized convulsions were observed in 5 of 9 cats after Sylvian gyrus stimulation (Table 2). No epileptic response or partial seizure was produced in the other 4 cats. On the contrary, when the amygdala, hippocampus and septum were stimulated, no generalized convulsion was produced in any of the 15 cats examined (Ebara et al., unpublished). This result shows that the Sylvian gyrus is more susceptible to the first trial generalized convulsions than limbic structures when high current stimu- lation was applied. In other word, the limbic structures are less susceptible to produce generalized convulsion than the cerebral cortex, initially. The progres- sive development of partial limbic seizure to secondarily generalized convulsions after seizure repetition may depend upon increased functional connections (God- dard 1981) between the kindled limbic structure and the cerebral motor cortex. Tanaka (1977) reported the appearance and subsequent augmentation of cortical responses evoked by the stimulation of the amygdala during amygdaloid kindling. Nevertheless, the kindling phenomenon in the limbic system which was not susceptible initally but became susceptible to produce generalized convulsions later on, resembles the chronological pattern of the change in complex partial seizures eventually combined with generalized convulsive seizures. The process

TABLE2. Susceptibility to generalized convulsion by first trial stimulation

GTC, generalized tonic-clonic convulsion ; PS, partial seizure ; Nil, no epilep- tic response ; AD, afterdischarge ; n, number of animals in this and following tables. Epilepsy and Kindling 257 of such kindled seizure susceptibilty, probably due to an increased connections of the kindled primary focus with wide brain structures, appears to be a biological aspect of poor prognosis in epilepsy with complex partial seizure with secondary generalized convulsion. Which produces more profound and lasting brain changes by repeatedfits, a focal cortical seizure or a limbic seizure ? The following items were compared between limbic kindling and focal cortical kindling : 1) chronological patterns of seizure development during kin- dling, 2) kindling rate, i.e., the total number of daily stimulations required for the first kindled generalized convulsion, 3) all-or-none type of occurrence of the kindled generalized convulsions, that is, no transition of the generalized convul- sion into partial seizure by reducing the stimulus intensity after kindling, 4) AD threshold reduction in the kindled site after kindling procedure, and 5) persist- ence of the kindled events after ceasing stimulation. Patterns of behavioral seizure development during kindling are shown in Fig. 1. The frequent regression of the seizure stage from the final stage 5 seizure into lower stages of partial seizures was found in all 23 animals kindled from the

Fig. 1. Patterns of behavioral seizure development : Sylvian gyrus kindling in cats. Behavioral seizure stages were classified using our method (Sato et al. 1980a). Please note the frequent regression in seizure stages during kindling. A, superior sygmoid gyrus ; B, anterior sygmoid gyrus ; C, posterior sygmoid gyrus. 258 M. Sato

Fig. 2. Patterns of behavioral seizure development: Limbic kindling in cats. See text for explanation of the seizure stages during kindling. AM, Amyg- dala; H, hippocampus ; S, septum; NA, nucleus accumbens. anterior, mid- or posterior Sylvian gyrus (Sato 1980; Sato et al. 1980b). The same characteristic was reported in 6 cats kindled from the frontal cortex (Wake and Wada 1975). Patterns of behavioral seizure development in limbic kindling are shown in Fig. 2. In contrast to the cerebral cortical kindling, the behavioral seizure manifestations were augmented progressively along with marked prolonga- tion of AD duration, and the final stage 6 kindled generalized convulsions rarely regress into partial seizures, once the limbic seizure developed to generalized convulsions (Wada and Sato 1974 ; Sato 1975). The pattern of seizure develop- ment of a mesolimbic structure kindling, i.e., the nucleus accumbens kindling, had the same characteristic as focal cerebral cortical kindling (Sato 1976) : a regression of the kindled generalized convulsion into partial seizures. Thus, the limbic seizure, rather than the cerebral cortical or mesolimbic seizure, can result in more stable and profound seizure susceptibility when the seizure developed secondarily generalized convulsions after a long-term seizure repetition.

AD duration during kindling and re-kindling, AD threshold reduction and the “all -or-none phenomenon”

Changes in both the AD threshold and AD duration in a representative cat are shown in Figs. 3 and 4. Epilepsy and Kindling 259

Fig. 3. Afterdischarge (AD) threshold, AD during kindling, reduction in AD threshold and all-or--none phenomenon in the amygdaloid kindled cat. PS, partial seizure; GCS, generalized convulsive seizure; L-AM, kindled left amygadala ; R-AM, kindled right amygdala.

Fig. 4. The secondary site kindling. The secondary site right amygdala (R-AM) was stimulated 7 days after the left amygdala (L-AM) kindling.

On the first day, a 200 ƒÊA stimulus was applied to the left amygdala to decide the AD threshold in this cat. No AD was produced. Then, the stimulus inten- sity was increased in 50 ,u A steps at 10 min intervals. At 300 ƒÊA, an AD was triggered in the stimulated amygdala as shown in the second trace from the top of

Fig. 3. (This stimulus intensity was arbitrarily regarded as the threshold inten- sity to produce AD threshold), and was maintained during the kindling session 260 M. Sato thereafter. On the 6th day, the AD duration was prolonged, and the AD propagated to the opposite amygdala with appearance of partial seizure manifesta- tions such as facial twitching, wretching-like movements and head nodding. On the 18th day, the animal developed secondarily generalized convulsions with AD changes such as prolonged duration, increased amplitude and appearance of self-sustained discharges in various brain sites (Wada and Sato 1974) which spike frequency was independent of the AD in the stimulated amygdala. After trigger- ing kindled convulsions on 5 successive days, the stimulus intensity was reduced into 100 and 150 ƒÊA, but behavioral or electrographic epileptic response was produced. When the stimulus intensity was increased up to 200 ƒÊA, the kindled generalized convulsion was produced again. Thus, the reduction in AD threshold from 300 ƒÊA to 200 p A after kindling, and the all-or-none type of appearance in the kindled convulsions by reduction in stimulus intensity was confirmed.

One week after the left amygdala kindling (primary site), the right amygdala was stimulated in the same manner as the primary site (Fig. 4). Although 100 and 150 ƒÊA of stimulus intensity was insufficient to produce AD, stimulation at

200 ƒÊA produced not only AD but also generalized convulsions in the first trial stimulation. Thus, kindling transfer in cats is much clearer than in rats.

A range of reduction in AD threshold after kindling varied from the brain site kindled (Table 3).

A marked reduction in the threshold (34.1 to 48.9% of the initial AD thresh- old) was found after kindling from limbic structures such as the amygdala, hippocampus and septal area as well as from the nucleus accumbens (Sato 1982),

TABLE3. Kindling-induced reduction of afterdischarge threshold

AD, afterdischarge. Epilepsy and Kindling 261

TABLE 4. Kindling rate and all-or-none phenomenon

while there was no or less reduction in AD threshold after cerebral cortical kindling. These different AD threshold reduction observed after the limbic and cerebral cortical kindling may suggest again that repetition of limbic seizure results in more profound seizure susceptibility than cerebral cortical seizure . The kindling rate and the all-or-none phenomenon in kindled convulsion of various brain sites are summarized in Table 4. The numcer of daily stimulations required for kindling ranged from 23 to 40. The amygdala and Sylvian gyrus appear to be more susceptible to kindling than the other brain structures. The most obvious difference between limbic kindling and cerebral cortical kindling was found in the all-or-none type of response of kindled generalized convulsions. The kindled convulsions from the amygdala , hippocampus and septum had the all-or-none type of response in all the cats examined. In contrast, such pattern of response was not found in most of the cats kindled from the Sylvian gyrus or frontal cortex. As for the persistence of the kindled events in amygdaloid kindled cats, we have reported previously (Wada et al. 1974) that established interictal spike discharges during kindling and the acquired pattern of response of secondarily generalized convulsions to kindling stimulus remained unchanged for 1 year without further stimulation. Contrary to the limbic kindling , less susceptibility to develop generalized convulsion was found 100 days after the Sylvian gyrus kindling (Sato et al. 1980b). These findings suggest that : 1) limbic structures are not susceptible to generalized convulsions initially, 2) repeated attacks of limbic seizures result in a profound reduction in seizure threshold at the primary epileptogenic focus, 3) 262M. Sato once a limbic seizure developed kindled generalized convulsions, it rarely changes into partial seizures, and 4) kindled events in limbic structures are more persist- ent than that in cerebral cortex.

Do recurrent attacks of partial seizures result in a lasting functional changein the limbic structures ? This question was examined mainly using the transference phenomenon in cats. The results of kindling transference from various brain sites to the ipsilater- al hippocampus are shown in Table 5. To produce primary hippocampal kindling, an average of 40.5 stimuli were necessary whereas only 1.8, 1.5 and 1,4 stimuli were needed for hippocampal

TABLE5. Transfer to the hippocampus (Sato 1982). Please note obvious saving of kindling stimuli required for re-kindling in the hippocampus after kindling other limbic structures.

TABLE6. Trannsfer to the amygdala Epilepsy and Kindling 263 re-kindling after amygdaloid, septal and nucleus accumbens kindling , respective- ly. When the primary kindling site was the amygdala , 24.1 kindling stimuli were needed to obtain the first generalized convulsion. A significant saving of amyg- daloid stimuli in re-kindling sessions was also found after the hippocampal , septal, and nucleus accumbens kindling (Table 6). In the case of Sylvian gyrus kin- dling, such saving of kindling stimuli was found only in the cats whose initial AD propagated to the amygdala (Sato 1976; Sato et al. 1980b). This finding suggests that repeated bombardment of AD to the amygdala may be a necessary condition to produce the kindling transfer besed upon a trans-synaptic change . As shown in Table 7, re-kindling from the contralateral same brain structure was facilitated significantly after amygdaloid, hippocampal and nucleus acum- bens kindling, while no transfer was found from one side of the Sylvian gyrus to the contralateral side. These results of the transfer study suggest that repeated attacks of partial seizures originating in a limbic structure may result in a lasting change in other limbic structures bilaterally, and that repeated focal seizures originating in a cortical area can also produce a secondary functional alteration of the limbic structures (Goddard et al. 1969; Sato 1980). Moreover, we have reported previously (Wada and Sato 1974; Sato 1976) that interictal spike discharges were established in other limbic structures independent of those in the originally kindled site during the kindling session. A representative case of the amygdaloid kindled cats is shown in Fig . 5. At the beginning of the session, IIDs were localized in the stimulated amygdala , which may propagate to other brain sites. Later , IIDs independent of those in the kindled amygdala were observed. Another representative case of the Sylvian gyrus kindling is shown in Fig. 6. Like amygdaloid kindling, the IIDs indepen- dent of IIDs at the kindled cortical site were observed during the Sylvian gyrus kindling. These findings again indicate that a functional change in the limbic system can be resulted from the recurrent epileptic attacks. Thus, repeated

TABLE7. Transfer to the same structure on the contralateral side

PS, primarily kindled site ; SS, secondarily kindled site. 264 M. Sato

Fig. 5. Appearance of independent interictal discharges (IDD) during amygdaloid kindling. L-AM, left amygdala; R-AM, right amygdala ; H, hippocampus; MRF, midbrain reticular formation ; FC, frontal cortex; GP, globus pallidus, A, IDD localized in the kindled amygdala (L-AM); B, propagated IID; C, IIDs which appeared independently of the IIDs in the stimulated site.

attacks of partial seizures originating in a limbic structure or related cortical area may produce a lasting functional change in the whole limbic system. This view may be consistent with our previous finding that augmented behavioral and autonomic responses to a challenge dose of the catecholamine protagonist, meth- , were observed after amygdaloid kindling (Sato 1983).

Recurrent spontaneous seizures Recurrent spontaneous convulsions may be observed in the amygdaloid kindled cat (Wada et al. 1974). Those behavioral seizure patterns were almost identical to or a mirror image of that of the originally kindled convulsions, suggesting that the seizure may depend upon a kindling-induced trans-synaptic trace in the brain. In fact, the onset of electrographic seizure discharges of the spontaneous seizures occured in the contralateral amygdala and globus pallidus (Wada et al. 1974). In addition, we could record recently the EEG during a spontaneous seizure in cats kindled from the left Sylvian gyrus (Fig. 7). The onset of active seizure discharges was found in the contralateral limbic structure instead of the originally kindled Sylvian gyrus or motor cortex. These data indicate that a secondary epileptogenic change can be established after repeated limbic and focal cortical seizures, which is sufficient to produce the seizure spontaneously.

Kindling mechanism Considering the data presented so far, it is presumed that an increase in Epilepsy and Kindling 265 functional connections between the limbic system and the motor cortex after limbic kindling may participate to the secondary seizure generalization of the limbic structure origin, and that a persistent change in limbic function resulting from repeated fits of a partial seizure is a major factor related to intractability of complex partial seizure with secondary generalization. Since the increase in functional connections is presumed to be a basic change concerning with increase in seizure susceptibility or intractability of partial seizures, it is necessary to determine what mechanisms underlies the increased functional connections produced by kindling. Although the mechanism remains unknown, evidence concerning the mechanism is available, which may be summarized as follows :

Morphological change No morphological change sufficient to explain the kindling was found at the kindled brain site or its first link synaptic structure (Racine et al. 1976 ; Goddard 1981). However, Langmeier et al. (1980) have reported a significant increase in the number of synaptic vesicles in rats kindled from the sensorimotor cortex, suggesting increased transmitter release. Recently, Shin et al. (1985) reported their quantitative radiohistochemical study indicating a transient increase in fascia dentata of the hippocampal formation after amygdaloid kindling in rats. This result is consistent with enhanced GABAergic inhibition of dentate granule cells after kindling reported by Tuff et al. (1983a, b).

Monosynaptic neurophysiology Monosynaptic neurophysiological studies have been done mainly using the perforant path originating in the entorhinal cortex and terminating on the dendrites of the granule cells of the fascia dentata. The main results obtained up to date are : a long-term potentiation (LTP) of population excitatory post- synaptic potentials of the perforant path (Douglas and Goddard 1975), a long- lasting increase in the amplitude of the synaptically evoked potentials and development of the burst response (Racine et al. 1981, 1983), a transient increase in seizure inhibition including trans-commissural inhibition (Goddard and Maru 1986) and GABA-mediated recurrent inhibition in the dentate gyrus (Liver and Miller 1985, Miller and Baimbridge 1983 ; Tuff et al. 1983a, b), and a transient reduction of cellular excitability after amygdaloid kindling (Douglas and Goddar- d 1975). Because of the following reasons that 1) LTP can be produced by tetanic stimulation without accompanying AD which is ineffective to produce kindling and 2) LTP occurs after a few kindling stimuli and remains unchanged during the kindling sessions thereafter, LTP can not fully account for all aspects of the kindled events of the brain. In fact, Giacchino et al. (1984) reported that kindling from the entorhinal cortex developed without accompanying LTP of the perforant path. Recently, Goddard and Maru (1986) reported that the kindling- induced changes in the granule cells are not the basis of the kindled state. 266 M. Sato

Nakatsu (1985) also reported no significant effect of bilateral ventral hippocampal lesions on amygdaloid kindling in cats. Synaptic and cellular mechanisms for reported enhancement and development of the burst response during kindling (Racine et al. 1981) should be further studied in the amygdala in where long- lasting increase in ibotenate-stimulated phosphatidylinositol hydrolysis was found in rats after kindling (Akiyama et al. 1987).

Neuropharmacological study The reported evidence indicates that classical neurotransmitters may partici- pate differently in the seizure development (kindling) process and the process to produce the fully kindled seizures. Brain catecholamines, especially nora- drenalin, inhibit the kindling process (Sato et al. 1980a ; Mohr and Corcoran 1981), while noradrenalin exerts no significant inhibitory effect on the kindled seizures (Westerberg et al. 1984). On the other hand, GABA agonists have an action on kindled seizures (Myslobodsky et al. 1979; Le Gal La Salle 1980; Kalichman 1982; Kalichman et al, 1982), while these agonists have no effect on the kindling process (Joy et al. 1984). Microinjections of GABA agonists into both the substantia nigra (McNamara et al. 1984) and substantia innominata (Okamoto and Wada 1984; Morita 1985) inhibit the initiation and seizure generalization of the fully kindled amygdaloid seizures in rats- respective- ly. The role of serotonin and acetylcholine in these two processes remains unclear due to inconsistent data (Kalichman 1982). Recently, the anticonvulsant action of amino acid antagonists (Croucher et al. 1982 ; Meldrum 1984) against the kindled hippocampal and amygdaloid seizures was reported. 2-Amino-4-phosphonobutyric acid, a selective blocker of the receptor, can suppress the kindled amygdaloid seizures (Peterson et al. 1983) and 2-amino-7-phosphonobutyric acid, a selective blocker of th N-methyl-D- aspartate receptor, can suppress the kindled hippocampal seizures (Peterson et al. 1984). More recently, we have demonstrated that TRH and an analog, DN-1417, have a potent inhibitory action on both the kindling process and the occurrence of kindled convulsions (Sato et al. 1984, 1985). When applied intra- ventricularly, TRH and this analog exhibit a potent antiepileptic properties without any accompanying behavioral toxic signs. The antiepileptic activity is dose-dependent, suggesting that TRH may be an endogenous antiepileptic sub- stance of the brain (Sato et al. 1986).

Neurochemical data Kindling-induced changes in the regional brain concentration of monoamines, the metabolites and receptor bindings as well as GABA, , , opiates and nucleotides have been reviewed (Kalichman 1982). The reported evidence is too inconsistent to lead to a certain conclusion. However, recent evidence that indicate changes in synaptic cell membrane and neuropeptides is Epilepsy and Kindling 267 available : a marked enhancement in extracellular calcium changes induced by electrical stimulation or by iontophoresis of excitatory amino acids (Wadman et al. 1985), changes in the activity of a major Ca2+/calmodulin dependent kinase system in hippocampal plasma membrane (Goldenring et al. 1986), significant (33%) and long-lasting depletion of calcium binding protein (Miller and Bairn- bridge 1983), increased phosphorylation of a 45K membrane protein (Patel et al. 1984), significant enhancement of -stimulated inositol phospholipid hydrolysis in the hippocampus (Iadorola et al. 1986) and the amygdala (Yamada et al. 1986 ; Akiyama et al. 1987), significant (about 2-fold) and lasting increase in striatal TRH receptors (Sato et al. 1986), and lasting increase in release of somatostatin (Higuchi et al. 1984). The decrease in striatal postsynaptic TRH receptors may represent a change in the brain TRH mechanism, possibly a seizure inhibitory system as well as the BABAergic mechanism, while the enhancement of ibotenic acid-stimulated inositol phospholipid hydrolysis in the amygdala may represent a changes the brain excitatory amino acid mechanism that is possibly a seizure facilitating system. Since an antagonistic action between excitatory amino acids and TRH has been reported (Renaud et al. 1979; Manaker et al. 1985), further study on this interaction is required for understanding a neuro- chemical mechanism of the kindling-induced seizure susceptibility.

CONCLUSION The kindling studies in cats reviewed suggest that: 1) recurrent fits of a partial seizure result in a progressive augmentation of seizure manifestations, which eventually lead to secondarily generalized convulsions, 2) once a partial seizure developed to generalized convulsions, the long-lasting susceptibility to generalized convulsions is apparently more profound and persistent in the limbic seizure than cerebral cortical seizure, 3) both reduction of the seizure threshold and lasting functional changes due to seizure repetition can occur much greater in the limbic system than cerebral cortex, 4) these changes in limbic system func- tion may be a major factor of the poor prognosis of partial epilepsy, 5) progres- sive augmentation of seizure susceptibility resulting from seizure repetition may be due to a change in cell membranes of the brain, upon which widely distributed trans-synaptic changes originating in the primary epileptogenic focus depend, and 6) biochemical difference may exsist in the two process-the process of progressive increase in seizure susceptibility in kindling fashion and the process to trigger the established seizure. The data presented here point to the importance of intensive treatment during the early stage of a partial seizure in order to prevent persistent changes in limbic system function, which may result in poor seizure control in clinically. New prophylactic and anticonvulsant agents for intractable epilepsy could be found when using a relevent animal model of epilepsy such as kindling. 268 M. Sato

References 1) Akiyama, K., Yamada, N. & Sato, M. (1987) Increase in ibotenate-stimulated phos- phatidylinositol hydrolysis in slices of the amygdala/pyriform cortex and hippocampus of rat by amygdaloid kindling. Exp. Neurol., 98, 499-508 2) Alonso-de Florida, F. & Delgado, J.M.R. (1958) Lasting behavioral and EEG changes in cats induced by prolonged stimulation of amygdala. Am J. Physiol., 193, 223-229. 3) Croucher, M.J., Collins, J.F. & Meldrum, B.S. (1982) Anticonvulsant action of excitatory amino acid antagonists. Science, 215, 899-901. 4) Delgado, J.M.R. (1959) Prolonged stimulation of brain in awake monkeys. J. Neurophysiol., 22, 465-475. 5) Delgado, J.M.R. & Sevillano, M. (1961) Evolution of repeated hippocampal seizure in the cat. Electroencephalogr. Clin. Neurophysiol., 13, 722-733. 6) Douglas, R.M. & Goddard, G.V. (1975) Long-term potentiation of the perforant path-granule cell synapse in the rat hippocampus. Brain Res., 86, 205-215. 7) Giacchino, J.L., Somjen, G.G., Frush, D.P. & McNamara, J.O. (1984) Lateral entor- hinal cortical kindling can be established without potentiation of the entorhinal- granule cell synapse. Exp. Neurol., 86, 483-492. 8) Goddard, G.V. (1981) The continuing search for mechanism. In: Kindling 2, edited by J.A. Wada, Raven Press, New York, pp. 1-14. 9) Goddard, G.V. & Douglas, R.M. (1976) Does the engram of kindling model the engram of normal long term memory ? In: Kindling, edited by J.A. Wada, Raven Press, New York, pp. 1-18. 10) Goddard, G.V. & Maru, E. (1986) Forces for and against the kindled state as revealed by EEG and field potential analysis in the hippocampal dentate area of perforant path kindled rats. In: Kindling 3, edited by J.A. Wada, Raven Press, New York, pp. 1-16. 11) Goddard, G.V., McIntyre, D.C. & Leech, C.K. (1969) A permanent change in brain function resulting from daily electrical stimuation. Exp. Neurol., 25, 295-330. 12) Goldenring, J.R., Wasterlain, C.G., Beate-Oestreicher, A., de Graan, P.N.E., Farber, D.B., Glaser, G. & de Lorenzo, R.J. (1986) Kindling induces a long-lasting chage in the activity of a hippocampal membrane calmodulin-dependent protein kinase system. Brain Res., 377, 47-53. 13) Gowers, W.R. (1881) Epilepsy and Other Chronic Convulsive Diseasses, Churchil, London (referred from Reynolds et al. 1983) 14) Herberg, L.J. & Watkins, P.J. (1966) Epileptiform seizures induced by hypoth- alamic stimulation in the rat: Resistance to fits following fits. Nature, 209, 515-516. 15) Higuchi, T., Kokubu, T., Silkand, S., Wada, J.A. & Friesen, H.G. (1984) A study of somatostatin receptors in amygdaloid-kindled rat brain. J. Neurochem., 43, 1271 - 1276. 16) Iadorola, M.J., Nicoletti, F., Naranjo, J.R. Putnam, F. & Costa, E. (1986) Kindling enhances the stimulation of inositol phospholipid hydrolysis elicited by ibotenic acid in rat hippocampal slices. Brain Res., 374, 174-178. 17) Joy, R.M., Albertson, T.E. & Stark, L.G. (1984) An analysis of the action of probagide, a specific GABA receptor agonist, on kindling and kindled seizures. Exp. Neurol., 83, 144-154. 18) Kalichman, M.W. (1982) Neurochemical correlates of the kindling model of epilepsy. Neurosci. Biobehav. Rev., 6, 165-181. 19) Kalichman, M.W., Burnham, W.M. & Livingston, K.E. (1982) Pharmacological investigation of gamma-aminobutyric acid (GABA) and fully-developed generalized seizures in the amygdala-kindled rat. Neuropharmacology, 21, 127-131. Epilepsy and Kindling 269

20) Langmeier, M., Fisher, J. & Mares, J. (1980) Numbers of synaptic vesicles in the rat somatosensory cortex after repetitive electrical stimulation prolonging self-sustained after-discharges. Epilepsia, 21, 255-260. 21) Le Gal La Salle, G. (1980) Inhibition of kindling-induced generalized seizures by aminooxyacetic acid. Can. J . Physiol. Pharmacol., 58, 7-11. 22) Liver, M.W. & Miller J.J. (1985) Alterations of inhibitory processes in the dentate gyrus following kindling-induced epilepsy. Exp. Brain Res., 57, 443-447. 23) Manaker, S., Caine, S.B. & Winokur, A. (1985) Ibotenic acid decreases the thyrotropin-releasing hormone receptor binding in the rat amygdala. Brain Res., 347, 144-148. 24) McIntyre, D.C. & Goddard, G.V. (1973) Transfer, interference and spontaneous recovery of convulsions kindled from the rat amygdala. Electroencephalogr. Clin. Neurophysiol., 35, 533-543. 25) McNamara, J.O., Rigsbee, L.C., Galloway, M.T. & Shin, C. (1984) Evidence that activation of substance nigra is necessary for limbic seizures. In : Neurotransmitters, Seizures, and Limbic Seizures, edited by R.G. Farriello, P.L. Morselli, A.G. Lloyd, L.F. Quesney & J. Engel, Jr., Raven Press, New York, pp. 49-56. 26) Meldrum, B. (1984) Amino acid neurotransmitters and new approaches to anticonvulsant drug action. Epilepsia, Suppl. 2, 140-149. 27) Miller, J.J. & Baimbridge, K.G. (1983) Biochemical and immunohistochemical correlates of kindling-induced epilepsy: Role of calcium binding protein. Brain Res., 278, 322-326. 28) Mohr, E. & Corcoran, M.E. (1981) Depletion of noradrenalin and amygdaloid kin- dling. Exp. Neurol., 72, 507-511. 29) Morita, K. (1985) GABAergic system in substantia innominata inhibits the secon- dary generalization of amygdaloid kinded seizure: Studies using the cerebral microin- jection of GABAergic drugs. Okayama Igaku Zasshi, 97, 973-991. (Japanese) 30) Myslobodsky, M.S., Ackermann, R.F. & Engel, J., Jr. (1979) Effects of gamma- acetylenic GABA and gamma-vinyl GABA on metrazol-activated, and kindled sei- zures. Pharmacol. Biochem. Behay., 11, 265-271. 31) Nakatsu, T. (1985) The effects of bilateral ventral hippocampal lesions on amyg- daloid kindling in cats. Okayama Igaku Zasshi, 97, 855-870. (Japanese) 32) Naoi, T. (1959) Electroencephalographic study on the electric convulsive treatment ― the study of latent brain dysfunction by pentazol-activation. Psychiatr. Neurol. Jpn., 61, 871-894. 33) Okamoto, M. & Wada, J.A. (1984) Reversible suppression of amygdloid kindled convulsion following unilateral gabacullin injection into the substantia innominata. Brain Res., 305, 389-392. 34) Okuma, T. & Kumashiro, H. (1981) Natural history and prognosis of epilepsy: Reports of a multi-institutional study in Japan. Epilepsia, 22, 35-53. 35) Patal, J., Marangos, P.J., Contel, N., Gardner, G. & Post, R.M. (1984) Increased phospholylation of a membrane protein consequent to amygdalid kindling. J. Neurochem., 43, 169-173. 36) Peterson, D.W., Collins, J.F. & Bradford, H.F. (1983) The kindled amygdala model of epilepsy: Anticonvulsant action of amino acid antagonists. Brain Res., 275, 169- 172. 37) Peterson, D.W., Collins., J.F. & Bradford, H.F. (1984) Anticonvulsant action of amino acid antagonists against kindled hippocampal seizures. Brain Res., 331, 176- 180. 38) Racine, R.J. (1972) Modification of seizure activity by electrical stimulation: II. Motor seizure. Electroencephalogr. Clin. Neurophysiol., 32, 281-284 39) Racine, R.J. (1978) Kindling: The first decada. Neurosurgery, 3,234-251. 40) Racine, R.J., Tuff, L. & Zaide, J. (1976) Kindling, unit discharge patterns and 270 M. Sato

neural plasticity. In: Kindling, edited by J.A. Wada, Raven Press, New York, pp. 19-39. 41) Racine, R.J., Kairiss, E. & Smith, G. (1981) Kindling mechanism: The evolution of the burst response versus enhancement. In: Kindling 2, edited by J.A. Wada, Raven Press, New York, pp. 15-29. 42) Racine, R.J., Milgram, N.W. & Hafner, S. (1983) Long-term potentiation phenomena in the rat limbic forebrain. Brain Res., 260, 217-231. 43) Renaud, L.P., Blume, H.W., Pittman, Q.J., Lamour, Y. & Tan, A.T. (1979) Thyrotropin-releasing hormone selectively depresses glutamate excitation of cerebral cortical neurons. Science, 205, 1275-1277. 44) Reynolds, E.H., Elwes, R.D.C. & Shorvon, S.D. (1983) Why does epilepsy become intractable? Lancet, 2, 952-954. 45) Rodin, E. A. (1968) The Prognosis of Patients with Epilepsy, Springfield, Thomas, Ilinois. 46) Sato, M. (1975) Hippocampal seizure and secondary epileptogenesis in the kindled cat preparation. Folia Psychiat. Neurol. Jpn., 29, 239-250. 47) Sato, M. (1976) A study on psychomotor epilepsy with kindled cat preparations. Folia Psychiat. Neurol. Jpn., 27, 257-273. 48) Sato, M (1980) Kindling, transference phenomenon between temporal cortex and limbic structure in cats. In: Adv. Physiol. Sci., Vol 17, Brain and Behavior, edited by G. Adams, I. Meszaeros & E.I. Banyai, Pergamon Press, Budapest, pp. 509-516. 49) Sato, M. (1982) Mesolimbic system and amygdaloid kindling. In: Kyoto Symposia, EEG Suppl., No. 36, edited by P.A. Buser, W.A. Cobb & T. Okuma, Elsevier, Biomedical Press, Amsterdam, pp. 249-256. 50) Sato, M. (1983) Long-lasting hypersensitivity to amphetamine following amyg- daloid kindling in cats: The relationship between limbic epilepsy and the psychotic state. Biol. Psychiatry., 18, 525-536. 51) Sato, M. & Wada, J.A. (1975) Review on the kindling preparation: A new exper- imental model of epilepsy. Brain Nerve, 27, 257-273. (Japanese) 52) Sato, M., Tomoda, T., Hisaka, N. & Otsuki, S. (1980a) Inhibition of amygdaloid kindling by chronic pretreatment with cocaine or methamphetamine. Epelepsia, 21, 497-504. 53) Sato, M., Tomoda, T., Yamashita, M. Kakashima, T. & Wake, A. (1980b) Seizure development of temporal cortical seizure. Psychiat. Neurol. jpn., 82, 378-391. (Japanese) 54) Sato, M., Morimoto, K. & Wada, J.A. (1984) Antiepileptic effects of thyotropin- releasing hormone and its new derivative, DN-1417, examined in feline amygdaloid kindling preparation. Epilepsia, 25, 537-544. 55) Sato, M., Kajita, S., Okamoto, M., Morita, K., Ogawa, T., Otsuki, S., Ogawa, N., Nagai, Y. & Narumi, S. (1985) Anticonvulsant action of a thyrotropin-releasing hormone analog (DN-1417) and changes in brain monoamines. Psychiatr, Neurol. jpn., 87, 176-185. 56) Sato, M., Kajita, S., Ogawa, N. & Okamoto, M. (1986) Amygdaloid kindking and thyrotropin-releasing hormone. In: Kindling 3, edited by J.A. Wada, Raven Press, New York, pp. 319-332. 57) Shin, C., Pedersen, H.B. & McNamara, J.O. (1985) Gamma-aminobutyric acid and receptors in the kindling model of epilepsy: A quantitative radiohistochemical study. J. Neurosci., 4, 2696-2701. 58) Tanaka, T. (1977) Modification of amygdaloid-cortical evoked potentials by kin- dling and pentetrazol-induced generalized convulsions in cats. Electroencephalogr. Clin. Neurophysiol., 43, 675-678. 60) Tuff, L.P., Racine, R.J. & Adamec, R. (1983a) The effects of kindling on GABA- mediated inhibition in the dentate gyrus of the rat. I. Paired-pulse depression. Epilepsy and Kindling 271

Brain Res., 277, 79-90. 61) Tuff, L.P., Racine, R.J. & Mishra, R.K. (1983b) The effects of kindling on GABA- mediated inhibition in the dentate gyms of the rat. II Receptor binding. Brain Res., 277, 91-98. 62) Wada, J.A. & Sato, M. (1974) Generalized convulsive seizure induced by daily electrical stimulation of the amygdala in cats: Correlative electrographic and behavioral features. Neurology, 24, 565-574. 63) Wada, J.A., Sato, M. & Corcoran, M.E. (1974) Persistent seizure susceptibility and recurrent spontaneous seizures in kindled cats. Epilepsia, 15, 465-478. 64) Wadman, W.J., Heinemann, U., Kennerth, A. & Neuhaus, S. (1985) Hippocampal slices of kindled rats reveal calsium involvement in epileptogenesis. Exp. Neurol., 57, 404-407. 65) Wake, A. & Wada, J.A. (1975) Frontal cortical kindling in cats. Can. J. Neurol. Sci., 2, 493-499. 66) Westerberg, V., Lewis, J. & Corcoran, M.E. (1984) Depletion of noradrenalin fails to affect kindled seizures. Exp. Neurol., 84, 237-240. 67) Yamada, N., Akiyama, K., Ujike, H., Sato, M. & Otsuki, S. (1986) Changes in ibotenic acid-stimulated phosphatidylinositol turnover in amygdala and hippocampal slices from amygdaloid kindled rats. Bull. Jpn. Neurochem. Soc., 24, 133-135. (Japanese)