Twenty-four-hour post-seizure inhibition during limbic kindling requires seizure generalization

Long-term seizure dynamics are determined by the nature of seizures and the mutual interactions between them

The seemingly random and unpredictable nature of seizures is a major debilitating factor for people with epilepsy. An increasing body of evidence demonstrates that the epileptic brain exhibits long-term fluctuations in seizure susceptibility, and seizure emergence seems to be a consequence of processes operating over multiple temporal scales. A deeper insight into the mechanisms responsible for long-term seizure fluctuations may provide important information for understanding the complex nature of seizure genesis. In this study, we explored the long-term dynamics of seizures in the tetanus toxin model of temporal lobe epilepsy. The results demonstrate the existence of long-term fluctuations in seizure probability, where seizures form clusters in time and are then followed by seizure-free periods. Within each cluster, seizure distribution is non-Poissonian, as demonstrated by the progressively increasing inter-seizure interval (ISI), which marks the approaching cluster termination. The lengthening of ISIs is paralleled by: increasing behavioral seizure severity, the occurrence of convulsive seizures, recruitment of extra-hippocampal structures and the spread of electrographic epileptiform activity outside of the limbic system. The results suggest that repeated non-convulsive seizures obey the 'seizures-beget-seizures' principle, leading to the occurrence of convulsive seizures, which decrease the probability of a subsequent seizure and, thus, increase the following ISI. The cumulative effect of repeated convulsive seizures leads to cluster termination, followed by a long inter-cluster period. We propose that seizures themselves are an endogenous factor that contributes to long-term fluctuations in seizure susceptibility and their mutual interaction determines the future evolution of disease activity.

Adeno-associated viral vector-mediated preprosomatostatin expression suppresses induced seizures in kindled rats

Somatostatin is expressed widely in the hippocampus and notably in hilar GABAergic neurons that are vulnerable to seizure neuropathology in chronic temporal lobe epilepsy. We previously demonstrated that sustained bilateral preprosomatostatin (preproSST) expression in the hippocampus prevents the development of generalized seizures in the amygdala kindling model of temporal lobe epilepsy. Here we tested whether sustained preproSST expression is anticonvulsant in rats already kindled to high-grade seizures. Rats were kindled until they exhibited 3 consecutive Racine Grade 5 seizures before adeno-associated virus serotype 5 (AAV5) vector driving either eGFP (AAV5-CBa-eGFP) or preproSST and eGFP (AAV5-CBa-preproSST-eGFP) expression was injected bilaterally into the hippocampal dentate gyrus and CA1 region. Retested 3 weeks later, rats that received control vector (AAV5-CBa-eGFP) continued to exhibit high-grade seizures whereas 6/13 rats that received preproSST vector (AAV5-CBa-preproSST-eGFP) were seizure-free. Of these rats, 5/6 remained seizure-free after repeated stimulation sessions and when the stimulation current was increased. These results suggest that vector-mediated expression of preproSST may be a viable therapeutic strategy for temporal lobe epilepsy.

Oxidative stress induced by epileptic seizure and its attenuation by melatonin

An epileptic seizure and postictal period in addition to well-known features are also characterized by massive consumption of energy. This is thought to lead to oxidative stress and increased generation of free radicals, which is reflected by increased levels of oxidative products. Our previous work described the neuroprotective effects of melatonin in preventing cognitive worsening after a single epileptic seizure. This work was aimed on direct measurement of free radicals in brain tissue using the EPR method 1, 15 and 60 minutes after seizure. The measurement was performed in adult male Wistar rats at the mentioned intervals after a single tonic-clonic seizure induced by flurothyl. In comparison to control animals there was a significant increase in hydroxyl and nitroxyl radicals 60 minutes after the seizure. The levels of hydroxyl radicals were significantly lower in animals that received melatonin 60 minutes before seizure induction compared to animals without preventive treatment. Therefore, melatonin affected the generation of the measured free radicals differently. An important finding was the delayed increase in free radicals after a single seizure in the later phases of recovery.

Functional, metabolic, and synaptic changes after seizures as potential targets for antiepileptic therapy

Little is known about how the brain limits seizure duration and terminates seizures. Depending on severity and duration, a single seizure is followed by various functional, metabolic, and synaptic changes that may form targets for novel therapeutic strategies. It is long known that most seizures are followed by a period of postictal refractoriness during which the threshold for induction of additional seizures is increased. The endogenous anticonvulsant mechanisms involved in this phenomenon may be relevant for both spontaneous seizure arrest and increase of seizure threshold after seizure arrest. Postictal refractoriness has been extensively studied in various seizure and epilepsy models, including electrically and chemically induced seizures, kindling, and genetic animal models of epilepsy. During kindling development, two antagonistic processes occur simultaneously, one responsible for kindling-like events and the other for terminating ictus and postictal refractoriness. Frequently occurring seizures may lead to an accumulation of postictal refractoriness that may last weeks. The mechanisms involved in seizure termination and postictal refractoriness include changes in ionic microenvironment, in pH, and in various endogenous neuromodulators such as adenosine and neuropeptides. In animal models, the anticonvulsant efficacy of several antiepileptic drugs (AEDs) is increased during postictal refractoriness, which is a logical consequence of the interaction between endogenous anticonvulsant processes and the mechanism of AEDs. As discussed in this review, enhanced understanding of these endogenous processes may lead to novel targets for AED development.

Reduction of seizure thresholds following electrical stimulation of sensorimotor cortex is dependent on stimulation intensity and is not related to synaptic potentiation

Epilepsy is characterized as a chronic brain state with a very low seizure threshold, and the occurrence of repeated seizure activity. Currently, there is no animal model of induced epilepsy that allows for the exploration of the brain mechanisms underlying a low seizure threshold without the elicitation of seizures. In this study, we employed repeated application of different intensities of electrical stimulation in an attempt to reduce afterdischarge (seizure) thresholds without eliciting seizures. We utilized an in vivo model of neocortical activation via stimulation of the corpus callosum of the adult rat. The intensities were chosen to be subthreshold (20, 30, 40, 50 microA), near threshold (150 microA), and suprathreshold (250, 500 microA) relative to the mean initial afterdischarge threshold (ADT). We also examined changes in the evoked field responses of the transcallosal pathway to the sensorimotor cortex as a measure of synaptic efficacy. Our results indicated that stimulation at 50 microA was effective at reducing the ADT, while minimizing the number of seizures elicited. Stimulation at 150 microA resulted in the concomitant reduction of ADT and repeated seizures typical of most electrical kindling studies. Finally, the 500 microA group showed repeated seizures, but no reduction of afterdischarge threshold. These stimulation intensities (50 microA, 150 microA, 500 microA and 0 microA-control) can be used to independently determine the brain mechanisms responsible for 1) the acquisition of a low afterdischarge threshold independent of the reorganizing effect of repeated seizures, and 2) the elicitation of repeated seizures independent of stimulation induced reduction of afterdischarge threshold.

Dissociation of mossy fiber sprouting and electrically-induced seizure sensitivity: rapid kindling versus adaptation

It has been shown that massed stimulation (MS) of the amygdala or hippocampus does not result in seizure progression but in the 'phenomenon of adaptation', whereas alternate day rapid kindling (ADRK) produces reliable kindling (Lothman, E.W., Williamson, J.M., 1994. Brain Res. 649, 71-84). The goal of the present experiment was to determine if the two different effects are due to differences in mossy fiber sprouting and/or different seizure and postictal spike propagation patterns. Nine rats underwent MS (66-70 stimulations separated by 5-min interstimulus interval), six were exposed to ADRK (12 stimulations/day, every 30 min, with 4 stimulus days, each separated by 1 stimulus-free day), five rats served as control. All rats had electrodes implanted bilaterally in dorsal and ventral hippocampi (VH) and 14 of them had additional electrodes in the piriform cortex. Animals were stimulated in the left VH at afterdischarge threshold. There was no potentiation in seizure response 4-7 weeks after MS. In contrast, ADRK produced not only kindling but also ongoing epileptogenesis resulting 4-7 weeks later in spontaneous seizures and development of a prolonged convulsive state in response to the initially subconvulsive stimulus. Epileptiform activity during MS was mostly restricted to VH, whereas during ADRK it spread widely among studied structures including piriform cortex. Afterdischarges during MS were elicited frequently but seizures did not progress beyond stage 2-3. During ADRK, afterdischarges were evoked less frequently but seizures reached stage 4-7 by the end of the 3rd and 4th stimulus days. The fully kindled state was not reached at this time, but epileptogenic changes continued to progress. Seven weeks after the initial stimulation, both groups demonstrated mossy fiber sprouting of similar intensity in VH. We suggest, (1) frequent but predominantly local hippocampal afterdischarges induce mossy fiber sprouting, but this is not sufficient to produce significant enhancement in seizure susceptibility, and (2) the involvement of extra-hippocampal structures, possibly piriform cortex, and formation of an aberrant hippocampal-para-hippocampal circuit is required to result in a condition of progressive epileptogenesis.

Delayed development of spontaneous seizures and prolonged convulsive state in rats after massed stimulation of the anterior piriform cortex

We studied the short- and long-term epileptogenic effects of massed stimulation (MS) of the piriform cortex. Sprague-Dawley rats with electrodes implanted bilaterally in the anterior piriform cortex and the dorsal and ventral hippocampi underwent MS: electrical stimulation of the left piriform cortex every 5 min for 6 h (afterdischarge threshold, 60 Hz, 1 ms, 1 s). Animals were retested (5 stimulations) 3-4 times later at different time points to check for the kindled state. Our data showed that MS resulted in delayed development of severe epilepsy. The interval between MS and the first appearance of convulsive response (2 weeks) was characterized by deep refractoriness to seizure (silent period). Unexpectedly, dramatic seizure activity occurred 4-7 weeks after MS. This was manifested by (1) generalized tonic-clonic convulsions with multiple failings, which were elicited repeatedly during retest; (2) frequent progression of elicited generalized convulsions into a prolonged (> 8 min) postictal convulsive state expressed mainly by continuous partial seizures and even new bouts of generalized seizures, and (3) development of mild spontaneous seizures. We found that epileptiform activity predominated in the ventral hippocampus. Mossy fiber sprouting was also most pronounced in this area. We propose that the MS resulted in formation of pathological circuits which involve both piriform cortex and ventral hippocampus and lead to severe epilepsy.

Differential sensitivity of various temporal lobe structures in the rat to kindling and status epilepticus induction

Using focal brain stimulation (kindling), discrete seizures can be triggered from many neuroanatomic sites with varying degrees of facility. From several of these sites, protracted seizures or status epilepticus (SE) also can be triggered. To date, no comparison has been made between different brain sites in their sensitivity both to kindling and to SE development. In this report, we have compared the kindling profiles of three amygdala nuclei, namely the basal (BL), central (CE), and medial (ME) nuclei, to the adjacent piriform (PIR) and perirhinal (PRH) cortices. In addition, three weeks following kindling, the susceptibility of each kindled site to status epilepticus (SE) was assessed by exposing the site to 60 min of electrical stimulation. We observed that (a) during the course of daily kindling, the afterdischarge threshold dropped progressively and significantly in all structures, (b) the rate of kindling in the PRH and PIR cortices and the CE amygdala was significantly faster than either the BL or ME amygdala, (c) when discrete convulsions were triggered, the latency to forelimb clonus in the PRH cortex and CE amygdala was significantly shorter than the other three structures, and (d) despite being slower to kindle than most other sites, stimulation of the BL nucleus most readily triggered SE. The kindling data suggest that discharges triggered from the PRH and CE more readily access motor systems supporting limbic convulsions than discharges triggered from the BL, ME nuclei or the PIR cortex. On the other hand, the SE data indicate that the mechanisms and circuits associated with the development of discrete kindled seizures are not identical to those associated with the induction of limbic SE.

Kindled epileptic seizures, postictal refractoriness, status epilepticus, and electrical self-stimulation

A single stimulus applied once daily to the limbic system commonly leads to convulsive seizures yet seizures are relatively infrequent during intracranial self-stimulation (ICSS), a procedure that involves many hundreds of similar stimuli. The present study examined the possible role of electrode site, interstimulus interval, afterdischarge and reinforcement thresholds and postictal refractoriness in accounting for this paradox. Electrode location was an overriding factor: seizures were never seen with hypothalamic implants posterior to the level of the ventromedial nucleus but were elicited by the majority of more rostral reward sites. Frequent repeated stimulation by ICSS did not in itself prevent subsequent kindling or reverse the effects of earlier kindling; on the contrary, seizures induced by ICSS showed a progressive increase in severity similar to the progression produced by conventional kindling. Individual convulsive seizures, as in previous studies, conferred transient protection against further seizures whether from ICSS or from kindling. More prolonged protection occassionally developed after repeated convulsive seizures: protection was accompanied by continuous EEG slow-waves corresponding in presentation to clinical petit mal status. Prolonged resistance to seizures has also been reported after tonic-clonic status epilepticus causing temporal lobe damage. The relative infrequency of seizures during ICSS ordinarily appears to depend on the siting of the electrodes, on distinct short- and long-term postictal refractory states, and on the rat learning to restrict stimulus input to subseizural levels.

Formalin induced FOS-like immunoreactive neurons in the trigeminal spinal caudal subnucleus project to contralateral parabrachial nucleus in the rat

By combining the retrograde-labeling method of injecting Fluoro-Gold (FG) into the parabrachial nucleus (PB) and the immunocytochemical staining of the FOS-like immunoreactive neurons (FLNs) in the trigeminal spinal caudal subnucleus (TSCS) induced by s.c. formalin injection into the perioral region in the rat, it was demonstrated that there are FLNs, FG-labeled neurons and neurons containing both FOS-like immunoreactivity and FG fluorescence in the TSCS. The three kinds of labeled neurons are distributed mainly in laminae I, II and V of the TSCS and there are also some neurons containing both FOS-like immunoreactivity and FG distributed in the adjacent ventrolateral reticular formation. The retrograde-labeling of FG- and double-labeled neurons showed contralateral predominance. In addition, we found that there are retrogradely labeled neurons in bilateral nuclei of the solitary tract with a contralateral predominance. The results suggest that FOS-like immunoreactivity might serve as an indicator for the nociceptive response after formalin injection into the trigeminal region and that the PB might be an important relay station for the further processing of the nociceptive information relayed from the trigeminal afferents. As the PB is known as a relay structure for visceral sensory pathway, it is proposed that there might be viscero-somatic convergence in this nucleus.

Rapid kindling with recurrent hippocampal seizures

Kindling is a widely used model of epilepsy. While intriguing hypotheses have recently emerged about how kindling occurs, the mechanisms behind kindling remain to be elucidated. In order to test whether certain anatomical changes that have been detected in the brains of animals that have completed kindling are necessary for the expression of kindled seizures, means to achieve kindling over a time course too rapid for the anatomical changes to take place were sought. Stimulus trains of various durations (2 and 10 s) and frequencies (20, 50, and 100 Hz) were given every 30 min, 12 times a day for 4 consecutive days to rats through bipolar electrodes stereotactically positioned in the ventral hippocampus. Responses were monitored with conventional kindling behavioral seizure scores and afterdischarge durations. The frequencies studied were chosen to survey the range that has been previously used to determine the optimal frequency for eliciting maximal dentate activation. Maximal dentate activation is a paroxysmal process that has been postulated to play both a role in regulating epileptiform activity in the hippocampus and adjacent regions that are coupled in a functional hippocampal-parahippocampal loop and a role in kindling. All types of trains resulted in rapid kindling in which kindled motor seizures emerged after several stimuli and then were consistently elicited with each stimulus; there was also retention of the kindled state after periods of 18 h of withholding the stimuli. Thus, the overall response profile of the rapid kindling demonstrated in this study was phenomenologically similar to the profile of traditional kindling. Yet rapid kindling developed more quickly than did mossy fiber sprouting, determined in prior work, thereby excluding the latter as a necessary factor in rapid kindling. Stimulus frequency significantly influenced the rate of rapid kindling. Trains of 20 Hz, the optimal frequency for eliciting maximal dentate activation, yielded the fastest kindling. This finding supports the proposed role of maximal dentate activation and the hippocampal-parahippocampal loop in kindling. Longer (10 s) trains consistently triggered shorter afterdischarge durations. We postulate that this may reflect a tighter linkage between seizure-terminating processes with the 10 s trains than with 2 s trains.

Effects of long-stimulus intervals and scopolamine administration on hippocampal kindling

Using the low-frequency kindling procedure, we studied the effects of periodic 2-week stimulus-free intervals and chronic scopolamine administration on hippocampal kindling seizure development. In Experiment 1, rats were divided into two groups, interval group and no-interval group. In the interval group a 2-week stimulus interval was set after every five consecutive daily stimulations until the 21st stimulation. The number of stimulating pulses required for the triggering of epileptic afterdischarge, pulse-number threshold (PNT), was used as an indicator of the seizure threshold. PNT, afterdischarge duration (ADD) and behavioral seizure stage (BSS) of each induced seizure in the initial stage of kindling, kindling rate, seizure parameters at the completion of kindling were recorded and compared to the values of the no-interval, control group. Our result showed that PNT at the 6th stimulation, the first stimulation after the first 2-week stimulus interval increased significantly compared to control. Other seizure parameters did not differ significantly between the two groups. In Experiment 2 scopolamine hydrobromide, 0.5 and 1.0 mg/kg i.p., was administered 1 h before each electrical stimulation until each rat showed the stage-3 seizure. PNT, ADD and BSS in the initial stage of kindling, kindling rate for the stage-3 and -5 seizures, seizure parameters at the first stage-3 and -5 were recorded and compared to the values of saline-treated, control group. Although scopolamine 1.0 mg/kg increased PNT at the 5th stimulation compared to control, no other significant changes of the seizure parameters were found by scopolamine.(ABSTRACT TRUNCATED AT 250 WORDS)

Two types of neuroplasticities in the kindling phenomenon: effects of chronic MK-801 and methamphetamine

Using the low-frequency kindling technique, we studied the effects of chronic MK-801 and chronic methamphetamine (MAP) administration on hippocampal kindling seizure development. In experiment 1, MK-801 (0.05, 0.1 mg/kg i.p.) was administered 2 h before each electrical stimulation until kindling developed into stage-3 seizure. In experiment 2, we started daily electrical stimulations two weeks after the last injection of chronic MAP administration (6 mg/kg/day, 14 days). The number of stimulating pulses required for the triggering of epileptic afterdischarge (pulse-number threshold, PNT) was used as an indicator of the seizure threshold. PNT, afterdischarge duration (ADD) and behavioral seizure stage (BSS) of each induced seizure in the initial stage of kindling; the kindling rates for stage 3 and stage 5 seizures; seizure parameters at the completion of kindling of the drug-treated groups were recorded and compared to the values of each saline-treated control group. Our result showed that MK-801 administration prior to each electrical stimulation selectively and significantly increased PNT in the initial stage of kindling without affecting other seizure parameters. Chronic pretreatment of MAP caused a selective and significant decrease of PNT of the first two stimulations in the kindling process. Taken together with the previous studies, these results suggest that long-term potentiation plays an important role in the development of the excitability of seizure focus but not of the induced seizure's propagation in the hippocampal kindling phenomenon. Clinically MK-801 seems to be a more efficacious drug in preventing the induction of seizures than in suppressing the acquired seizures.

Apparent disappearance of postseizure inhibitions and intensity of seizures during the development of rapid kindling in rabbits

The evolution of seizures and postseizure inhibitions in the course of 'rapid kindling' and after the termination of stimulation were studied in rabbits with chronically implanted electrodes (neocortex, dorsal hippocampus, amygdala, caudate nucleus). The amygdala (n = 4) or hippocampus (n = 7) was electrically stimulated every 5 min. Generalized convulsions and wide-spread electrographic epileptic changes together with a striking shortening of postictal refractory periods were produced by this procedure within 2-6 h. In most cases, these epileptogenic effects continued their progression after the termination of stimulation for more than 2-4 weeks. The degree of reduction of postseizure inhibition durations was significantly greater than the degree of increase of generalized motor seizure durations. These may be mediated by mechanisms which facilitate the onset of seizure but do not significantly influence seizure expression.

Two types of neuroplasticities in the kindling phenomenon

We stimulated the dorsal hippocampus of the rat with 2 Hz electrical stimulation to induce kindling seizures. As we reported previously using cats, pulse-number threshold (PNT), the number of stimulating-pulses required for the triggering of epileptic afterdischarge, decreased profoundly in the initial stage of the kindling process and the behavioral seizure stage (BSS) developed in the later stage. After the completion of kindling, a 4 week interval elevated PNT significantly compared to the value at the completion of kindling, whereas BSS showed no regression. These results suggest that there are two types of neuroplasticities which are independent of each other in the kindling phenomenon, one is the early-short type which is involved in the susceptibility of epileptic focus in hippocampus and the other is the late-long type which is involved in the full propagation of kindled seizures to the extra-limbic area.

The effect of interstimulation interval on the assessment of anticonvulsant drug potency in fully kindled rats

Multiple stimulation regimes with interstimulation intervals of 0.25-2 h were investigated in fully kindled rats as methods of testing the time course and potency of anticonvulsant drugs after a single administration. Protocols with conventional interstimulation intervals of 1-3 days were used for comparison. Prior to the drug experiments, the different stimulation regimes were examined in drug-naive, kindled rats in order to determine the extent of postseizure inhibition occurring with such protocols. All stimulations were carried out with a suprathreshold current of 500 microA. Using protocols with 3-9 stimulations within 8 h, seizure severity was relatively stable, but motor seizure duration was reduced in most experiments. Both decreases and increases were observed with respect to afterdischarge duration (ADD). The increases in ADD were primarily due to the appearance of 'secondary' afterdischarges with small amplitude, which were associated with immobility, intermittent facial clonus and head nodding. After a series of stimulations at short intervals, reduced seizure severity was observed after this series for at least 1 week, so that an interval of at least 2 weeks had to be interposed between multistimulation experiments in the same group of rats. When the effects of carbamazepine, 15 mg/kg, were determined with 4 different stimulation regimes, it was found that the anticonvulsant potency of the drug was higher in experiments with short interstimulation intervals compared to conventional protocols with interstimulation intervals of 1-3 days, indicating synergistic effects between the drug and postictal inhibition. Indication for such synergism was also found when the animals were only stimulated once daily during the drug experiments. With higher doses of carbamazepine or phenobarbital, 30 mg/kg, the difference between the stimulation protocols was less marked. Furthermore, the time course of anticonvulsant action determined with different stimulation regimes was similar. The data indicate that multistimulation regimes can be used in kindled rats to determine the time course following single administration of anticonvulsant drugs, but such protocols may lead to an overestimation of anticonvulsant potency.

A biphasic change of afterdischarge threshold during the kindling process

Cats were stimulated in the ventral hippocampus with low-frequency (about 3 Hz) square wave pulses. All subjects were kindled until generalized convulsion occurred. During the kindling process, the number of stimulating pulses required for the provocation of afterdischarge (AD), which was used as an indicator of AD threshold, decreased suddenly in the initial stage and increased gradually in the late stage. We consider this phenomenon to be important in deepening the understanding of seizure generation.