known that enhancing NMDA receptor-mediated excitatory actions (e.g. by lowering extracellular Mg) produces epileptiform activity in experimental models of ‘kindled’ epilepsy (Chapman, 1998, 2000). It has been postulated that NMDA receptors may change after neuronal damage (Rice and DeLorenzo, 1998). New receptors are formed that have either less sensitivity to ambient Mg or more sensitivity to ambient glycine; increased excitability could occur within local circuits where the circuitry itself is not altered (or may occur in addition to circuit alterations) (Meldrum et al., 1999). As it is known that the NMDA receptor is subject to modulation by a variety of endogenous agents, including glycine (as a co-agonist with glutamate), polyamines, steroids, neuropeptides (Vezzani et al., 2000b), pH, the redox state of the receptor, and NO, there are many chronic alterations in NMDA receptors that could underlie long-term changes in excitability and, thereby, epilepsy. Presently, there are no data to support changes in any of these regulatory factors in chronic epilepsy, but it is distinctly possible that alterations in one or more of these will be shown to be responsible for one or another form of inherited epilepsy.
Kindling is the most extensively studied animal model of epileptogenesis, and this has demonstrated the unique importance of NMDA receptors in the creation of seizure activity (Bengzon et al., 1999; Meldrum et al., 1999). In kindling, repeated electrical stimuli in the limbic system lead to a progressive increase of seizure susceptibility. When the animal responds to stimuli with generalized convulsions, it has developed a permanent epileptic condition. Activation of NMDA receptors and levels of NMDA receptor function are critical in kindling epilepsy (Bengzon et al., 1999). Selective NMDA-receptor antagonists retard kindling development and can also, at higher doses, have an anticonvulsant effect (Bengzon et al., 1999; Trist, 2000).
Metabotropic receptors
On account of these receptors’ responsibility for regulating glutamatergic and GABA-ergic neurotransmission, it is not surprising that mGluRs strongly influence the induction, propagation and termination of epileptic activity in the central nervous system (CNS) (Doherty and Dingledine, 2002). Pharmacological studies with mGluR group specific agonists and antagonists provide a relatively clear picture for Group I, with agonists being convulsant and antagonists being anticonvulsant (Meldrum et al., 1999; Doherty and Dingledine, 2002; Sayin and Rutecki, 2003). The picture is more complicated for the Group II and III receptors but anticonvulsant effects have been described for agonists of both these groups (Meldrum et al., 1999).
Glutamate transporters
In addition to receptor abnormalities, glutamate transporters, responsible for the removal of glutamate from the extracellular fluid, have been implicated in epilepsy (Meldrum et al., 1999). In situ hybridization studies have shown that the mRNA responsible for the rat glial glutamate transporter (GLT) is reduced in several brain regions in epilepsy-prone rats (Meldrum et al., 1999). GLT ‘knockout’ mice have been bred to provide homozygous mice, in which the GLT protein is not detected. In such mutant mice, glutamate uptake in cortical synaptosomes is 5.8% compared with the wild-type (Meldrum et al., 1999). The mutant mice show spontaneous seizures, with wild running and tonic extension, which is frequently fatal. In chronic seizure models (kindled seizures, spontaneous seizures and genetically epilepsy-prone rats), there are numerous reports of increases in extracellular glutamate during seizures (Meldrum et al., 1999). This strongly suggests that in these chronic models there are sustained functional alterations in mechanisms relating to the synaptic release of glutamate or its transport. GLT-1 astrocytic expression was reduced in four Shetland sheepdogs with IE (Morita et al., 2005). In these dogs it was suggested that decreased expression of the transporter might be related to development of status epilepticus.
There is not as yet any genetically determined epilepsy syndrome occurring spontaneously in man or mouse that can be ascribed to a primary gene defect involving a glutamate receptor or transporter.
Targets for treatment
In animal models of epilepsy, antagonists acting at NMDA receptors, AMPA receptors or at Group I metabotropic receptors have potent anticonvulsant actions (Meldrum and Chapman, 1999; Rogawski and Donevan, 1999; Chapman, 2000; Moldrich et al., 2003).
NMDA receptor antagonists have been successful in stopping the maintenance phase of self-sustaining status epilepticus (SE) in rats, which suggests that these compounds may have a promising role in the treatment of unrelenting seizure activity such as SE (Mazarati and Wasterlain, 1999). Studies with selective AMPA receptor antagonists have indicated that AMPA receptors are potentially promising anticonvulsant drug targets, but at present this is uncertain (Rogawski and Donevan, 1999).
In genetic mouse models, mGlu1/5 antagonists and mGlu2/3 agonists are effective against absence seizures. Thus, antagonists at Group I mGlu receptors and agonists at Groups II and III mGlu receptors are potential anti-epileptic agents, but their clinical usefulness will depend on their acute and chronic side-effects (Moldrich et al., 2003). Potential also exists for combining mGlu receptor ligands with other glutamatergic and non-glutamatergic agents to produce an enhanced anticonvulsant effect (Moldrich et al., 2003).
The Veterinary Perspective
Idiopathic epilepsy (see Chapter 6) is the most common cause of seizures in dogs (Podell and Hadjiconstantinou, 1997). Low levels of GABA and high levels of glutamate have been detected in the cerebrospinal fluid of epileptic dogs independent of time relation to recent seizure activity (Podell and Hadjiconstantinou, 1997). The glutamate elevations are not related whether the seizures were focal or generalized in character (Podell and Hadjiconstantinou, 1997). These findings may indicate the brains of epileptic dogs are under a state of chronic over-excitation. Although a separate study found that lower CSF GABA concentration was associated with a reduced response to phenobarbital therapy in dogs, there was no association between CSF glutamate and response to this therapy (Podell and Hadjiconstantinou, 1999). However, a negative association was found between CSF glutamate:GABA ratio and response to phenobarbital therapy (Podell and Hadjiconstantinou, 1999). Therefore glutamate-mediated mechanisms may be useful targets for anticonvulsant therapy in dogs. Intracerebral microdialysis was used to demonstrate elevation of extracellular levels of glutamate in four Shetland sheep-dogs with IE, suggesting an important role in the occurrence of seizure activity (Morita et al., 2005).
Gabapentin (see Chapter 17), a relatively new human anticonvulsant, has been evaluated in dogs refractory to phenobarbitone and potassium bromide with an approximate 50% success rate.
Gabapentin has been shown to modestly decrease glutamate levels in the brain (Errante and Petroff, 2003). Another new anticonvulsant, topiramate (see Chapter 19), produces its antiepileptic effect by several mechanisms, one of which is inhibition of kainite-mediated glutamate receptors (Angehagen et al., 2003a). This drug has also been demonstrated to protect neurons from excitotoxic levels of glutamate, potentially preventing brain damage during seizure activity (Angehagen et al., 2003b).
Catecholamines
Abnormalities of CNS catecholamines have been reported in several genetic models of epilepsy. In the spontaneous epileptic rat, dopamine was decreased in the caudate nucleus whereas noradrenaline was increased in the midbrain and brainstem (Hara et al., 1993). Decreased levels of dopamine have been found in epileptic foci of epilepsy patients (Mori et al., 1987). In animal models of absence epilepsy, seizures are exacerbated by dopamine antagonists while alleviated by
