Abstracts

Rationale: Up to a third of patients with mitochondrial disease develop severe epilepsy. This form of epilepsy is understudied and typically has a severe prognosis. We have developed a brain slice model of mitochondrial epilepsy using a combination of astrocytic aconitase inhibitor, fluorocitrate, and a cocktail of mitochondrial respiratory chain inhibitors; rotenone and potassium cyanide. We aim to characterize the role that astrocytes play in seizure generation in this novel model of mitochondrial epilepsy. Methods: Analytical biochemical techniques such as high performance liquid chromatography (HPLC) and gas chromatography mass spectrometry (GC-MS) were used to interrogate the astrocytic glutamate-glutamine cycle following induction of epileptic state in acutely prepared rat brain slices. Post-hoc immunohistochemistry was used to measure GABA expression in the neurons following epileptic induction. Response to external glutamine and GABA application was measured through electrophysiological response and validated by biochemical and immunohistochemical techniques. Finally, findings on relevant mechanism were validated in neuropathological study of patients with mitochondrial epilepsy. Results: Following induction of epileptic state, there was a pronounced reduction (p<0.05) in tissue glutamine concentration in the epileptic slice. GC-MS confirmed that there was reduced incorporation (p<0.0001) of [U-13C]-glucose into glutamine, especially the M+2 isotopologues. When applied with high concentration of L-glutamine (2.5mM) or GABA (2.5mM), there was significant (p<0.05) suppression of epileptic activity by 62.2% and 90.0% respectively. This rescue application of L-glutamine and GABA both significantly increased GABA expression (p<0.05) in the neurons without rescuing neuronal loss from epilepsy. Interestingly, application of L-glutamine did rescue the loss of tissue glutamine but the application of 2.5mM GABA did not. This indicates the loss of astrocytic glutamine synthetase activity in the epileptic tissue. To confirm this, we performed immunofluorescence detection of glutamine synthetase expression in reactive astrocytes in patients with mitochondrial epilepsy as compared with age-matched controls and patients with mitochondrial disease with no epilepsy phenotype. A significant reduction in glutamine synthetase optical densities (p<0.05) was observed within GFAP-positive astrocytes in 6 of 8 patients; out of which 5 had recorded history of epilepsy. Conclusions: Using our model of mitochondrial epilepsy, we found downregulation of the astrocytic glutamate-glutamine cycle in the epileptic tissue leading to a loss of tissue glutamine. This loss of tissue glutamine was postulated to be due to a loss of astrocytic glutamine synthetase activity which was demonstrated both in our model and post-mortem neuropathology of patients with mitochondrial epilepsy. Astrocytes seem to drive seizure generation in mitochondrial epilepsy through the loss of glutamine synthetase activity and could be a potential target for therapeutic development. Funding: EPSRC Industrial Case AwardNENS Exchange Grant