Despite an extensive research on the molecular basis of epilepsy, the essential players in the epileptogenic process leading to epilepsy are not known. Gene expression analysis is one strategy to enhance our understanding of the genes contributing to the functional neuronal changes underlying epileptogenesis. In the present study, we used the novel MPSS (massively parallel signature sequencing) method for analysis of gene expression in the rat kindling model of temporal lobe epilepsy. Kindling by repeated electrical stimulation of the amygdala resulted in the differential expression of 264 genes in the hippocampus compared to sham controls. The most strongly induced gene was Homer 1A , an immediate early gene involved in the modulation of glutamate receptor function. The overexpression of Homer 1A in the hippocampus of kindled rats was confirmed by RT-PCR. In order to evaluate the functional implications of Homer 1A overexpression for kindling, we used transgenic mice that permanently overexpress Homer 1A. Immunohistochemical characterization of these mice showed a marked Homer 1A overexpression in glutamatergic neurons of the hippocampus. Kindling of Homer 1A overexpressing mice resulted in a retardation of seizure generalization compared to wild-type controls. The data demonstrate that kindling-induced epileptogenesis leads to a striking overexpression of Homer 1A in the hippocampus, which may represent an intrinsic antiepileptogenic and anticonvulsant mechanism in the course of epileptogenesis that counteracts progression of the disease.

PURPOSELamotrigine (LTG) is an anticonvulsant that is currently in use for the treatment of various seizure disorders and that shows promise in the treatment of affective illness. LTG is also effective in the suppression of amygdala-kindled seizures. Because many drugs show a differential efficacy profile as a function of the phase of kindling evolution, we evaluated LTG for its potential antiepileptogenic effects on the development of amygdala-kindled seizures.METHODSIn two separate studies, LTG (5 or 15 mg/kg versus vehicle) was administered before each daily amygdala stimulation (biphasic square wave pulses, 100 pulse pairs per second for a total of 0.5 second, 1-millisecond pulse width) at an intensity of 50 microA over the AD threshold. Seizure development was assessed, as well as the effect of this pretreatment on subsequent efficacy of LTG on completed kindled seizures.RESULTSLTG at 5 mg/kg failed to block seizure development. At 15 mg/kg, LTG paradoxically enhanced seizure development and produced running fits in four of the nine animals tested. Animals previously treated with either dose of LTG during kindling development showed a diminished response to the anticonvulsant effects of LTG on fully kindled seizures compared with the vehicle-treated controls.CONCLUSIONSAlthough LTG possesses potent anticonvulsant effects on completed amygdala-kindled seizures, it is either without effect (5 mg/kg) or facilitates (15 mg/kg) the initial phase of kindling development. In addition, exposure to LTG during kindled seizure development leads to a reduced subsequent response to the drug in fully kindled animals. These observations parallel those with carbamazepine and suggest that different stages of kindling (epileptogenesis versus fully manifest seizures) may have different underlying neural mechanisms that require distinct pharmacotherapies.

Using an amygdala-kindled seizure paradigm, we evaluated the acute and chronic anticonvulsant effects of lamotrigine (LTG). Lamotrigine produced dose-dependent inhibitory effects on seizure stage, afterdischarge (AD), and seizure duration. Lamotrigine (15 mg/kg) also increased the afterdischarge and seizure thresholds. Following repeated LTG administration and stimulation at 48-h intervals, tolerance developed to LTG's (15 mg/kg) anticonvulsant effects, and cross-tolerance was observed to the anticonvulsant effects of carbamazepine (CBZ, 15 mg/kg). In a separate group of kindled rats, CBZ (15 mg/kg) was repeatedly administered to induce tolerance. This led to a partial cross-tolerance to LTG, manifesting as an increased rate of tolerance development to LTG, and seizures following the first injection in some animals, which were not observed in CBZ-nontolerant controls. When these rats were made fully tolerant to LTG and then exposed to higher doses of LTG (30 and 50 mg/kg), no anticonvulsant effects were observed. In contrast, higher doses of CBZ (30 mg/kg) did restore efficacy in CBZ-tolerant animals. Cross-tolerance from LTG to valproate and diazepam was not observed, although cross-tolerance from CBZ to valproate has been reported previously. These data suggest that LTG has both shared and distinct anticonvulsant mechanisms from those of CBZ on amygdala-kindled seizures. The implications of these results for clinical therapeutics remain to be evaluated.

A modified cortical ramp stimulation (CRS) model has been developed allowing repeated determinations of seizure threshold at short time intervals in individual rats without inducing postictal threshold increases. Anticonvulsant potency of the standard antiepileptic drugs carbamazepine, phenytoin, phenobarbital, valproate, diazepam and ethosuximide in the CRS model was compared with respective drug potencies in two more traditional seizure models with transcorneal stimulus application, i.e., the minimal electroshock seizure threshold (minEST) and the maximal electroshock seizure threshold (maxEST). In the CRS model, two different types of threshold were determined, the threshold for localized seizures (TLS) and the threshold for generalized seizures (TGS). When screw electrodes were implanted over the primary motor cortex, TLS was characterized by unilateral forelimb clonus, tonic abduction of contralateral forelimb, and head adversion. When ramp-shaped stimulation was continued above the TLS current, bilateral clonic forelimb seizures with loss of posture developed, which was defined as TGS. In contrast to TLS, TGS could not be repeatedly determined at short time intervals because of postictal threshold increase. TLS was dose-dependently increased by carbamazepine, phenobarbital, valproate and diazepam, although phenytoin showed a truncated dose-response, and ethosuximide was ineffective. In comparison to TLS, drug-induced increases in TGS were more marked. All drugs dose-dependently increased minEST and, except ethosuximide, maxEST. For comparison of drug potencies, doses increasing seizure thresholds by 20 or 50% were calculated from dose-response curves. Respective comparisons showed marked differences in drug potencies between models, indicating that the CRS method presents a model of another, more pharmacoresistant seizure type than seizure types induced in traditional models, such as transcorneal electroshock. Based on the location of electrodes in the frontal neocortex, the characteristic seizure pattern, and the low pharmacological sensitivity of the seizures to standard antiepileptics, the modified CRS model most likely represents a new model of localization-related seizures occurring in frontal lobe epilepsy and may thus be used in the search for novel drugs with higher efficacy against this difficult-to-treat type of epilepsy.