Fasting and ketogenic diet prevent seizures in epileptic children, magnesium-deficient rats and other animal models of seizure disorders. This effect has been attributed to increased levels of circulating ketone bodies. The purpose of this study was to determine the role of serum ketone bodies, measured as beta-hydroxybutyrate (BHB), in preventing audiogenically-induced seizures in weanling rats fed a magnesium-deficient diet for 17 days. The effect on seizure susceptibility was investigated by feeding a magnesium-deficient diet to weanling rats for 17 days. Fasting and ketogenic diet (dietary medium chain triglycerides, MCT) markedly decreased seizure incidence that was associated with increased serum BHB level. Also, rats fasted for 24h or fed 28 percent dietary MCT had decreased seizure incidence as compared with rats fed 3 percent dietary MCT or rats fasted for 6h. These effects were not caused by differences in caloric density or percentage of calories from fat in the diets. Gavaging 2 mmoles of BHB resulted in lower seizure incidence; as compared with rats gavaged with 0.5 mmoles BHB when measured 30 min after dosing. In contrast, gavaging 5.6 mmoles of glucose resulted in increased seizure incidence in 24-h-fasted rats. Gavaging 5.6 mmoles of glucose with 0.5 mmole of BHB simultaneously resulted in higher seizure incidence than gavaging with 2.0 mmole BHB and 1.4 mmole glucose simultaneuosly. In addition, gavaging 5.6 mmoles of glucose with 2 mmoles of BHB resulted in higher seizure incidence than gavaging 2 mmoles of BHB alone, which markedly reduced seizure incidence in fasted animals. Fasting, ketogenic diet (MCT) and gavaging BHB increase serum BHB and decrease serum glucose concentrations . Gavaging glucose reduced serum BHB and increased serum glucose concentration. There was an inverse relationship between serum BHB and glucose in all treatments of this study. Although some treatments affected serum minerals, these effects were not consistent among experiments. Therefore, fasting, ketogenic diet (MCT) and gavaging BHB or glucose does not affect serum minerals markedly or consistently; and modifications in serum minerals caused by these treatments do not account for this effect on seizure incidence and severity. Finally, increases in serum BHB and decreases in serum glucose were consistently associated with dose-dependent reductions in seizure susceptibility of rats fed a magnesium-deficient diet for 17 days.

Magnesium deficiency in rats is known to precipitate audiogenic seizures. An unknown mineral factor in a diet mixture was found to substantially reduce the seizure occurence in magnesium deficient vii animals. This was corrected when the same mineral mixture was remade. Subsequently the faulty mineral mixture was discarded. This research .. as aimed at determining the mineral factor responsible for the observed changes in seizure occurence, and also aimed at investigating the relationships of serum calcium, magnesium and phosphorus concentrations to seizure susceptibility. Treatments to change the serum concentrations of these minerals included dietary manipulation, subcutaneous injection of calcitonin, of 1 ,25-dihydroxycholecalciferol, of calcium and of phosphorus. Animals fed magnesium deficient, low phosphorus diet and magnesium deficient animals fasted over-night were found to have significantly lower audiogenic seizure susceptibility than all other magnesium deficient animals regardless of treatment. Reduced audiogenic seizure susceptibility was not produced by any of the injections. The reduction in seizure with the magnesium deficient, low phosphorus diet indicates that phosphorus is an important factor in the mechanism of audiogenic seizuring in magnesium deficient rats.

The effects of dietary calcium and phosphorus levels on audiogenic seizure susceptibility and brain neurotransmitter were investigated in magnesium deficient rats. For 17 days, male weanling rats were fed magnesium deficient diets which also contained deficient (-), adequate (=}, and excess (+) amounts of calcium and/or phosphorus. Reduction of seizure incidence was seen in low calcium and/or low phosphorus diets. High calcium, and high calcium in combination with high phosphorus increased the severity of seizures. High calcium and low phosphorus, and high phosphorus and low calcium diets prevented seizuring. Most magnesium deficient diets resulted in elevation of serum phosphorus, calcium, and sodium concentrations. Within diet treatments, animals that seized had higher serum mineral concentrations than animals that failed to seize. Magnesium deficiency increased serotonin in cerebral cortex, cerebellum, and medulla oblongata and pons, and 5-hydroxyindoleacetic acid concentrations in cerebral cortex. There were no significant diet effects in brain neurotransmitter concentrations in the midbrain. Calcium seemed to play an important role in increasing both audiogenic seizure susceptibility and brain neurotransmitters. However, brain neurotransmitter levels were not related to seizure susceptibility. Calcium increased serum potassium and sodium concentrations and it increased brain serotonin concentration overall. Increasing dietary phosphorus levels increased serum phosphorus decreased serum calcium concentrations, but did not affect brain neurotransmitters.

Le co-transporteur KCC2 spécifique au potassium et chlore a pour rôle principal de réduire la concentration intracellulaire de chlore, entraînant l’hyperpolarisation des courants GABAergic l’autorisant ainsi à devenir inhibiteur dans le cerveau mature. De plus, il est aussi impliqué dans le développement des synapses excitatrices, nommées aussi les épines dendritiques. Le but de notre projet est d’étudier l’effet des modifications concernant l'expression et la fonction de KCC2 dans le cortex du cerveau en développement dans un contexte de convulsions précoces. Les convulsions fébriles affectent environ 5% des enfants, et ce dès la première année de vie. Les enfants atteints de convulsions fébriles prolongées et atypiques sont plus susceptibles à développer l’épilepsie. De plus, la présence d’une malformation cérébrale prédispose au développement de convulsions fébriles atypiques, et d’épilepsie du lobe temporal. Ceci suggère que ces pathologies néonatales peuvent altérer le développement des circuits neuronaux irréversiblement. Cependant, les mécanismes qui sous-tendent ces effets ne sont pas encore compris. Nous avons pour but de comprendre l'impact des altérations de KCC2 sur la survenue des convulsions et dans la formation des épines dendritiques. Nous avons étudié KCC2 dans un modèle animal de convulsions précédemment validé, qui combine une lésion corticale à P1 (premier jour de vie postnatale), suivie d'une convulsion induite par hyperthermie à P10 (nommés rats LHS). À la suite de ces insultes, 86% des rats mâles LHS développent l’épilepsie à l’âge adulte, au même titre que des troubles d’apprentissage. À P20, ces animaux presentent une augmentation de l'expression de KCC2 associée à une hyperpolarisation du potentiel de réversion de GABA. De plus, nous avons observé des réductions dans la taille des épines dendritiques et l'amplitude des courants post-synaptiques excitateurs miniatures, ainsi qu’un déficit de mémoire spatial, et ce avant le développement des convulsions spontanées. Dans le but de rétablir les déficits observés chez les rats LHS, nous avons alors réalisé un knock-down de KCC2 par shARN spécifique par électroporation in utero. Nos résultats ont montré une diminution de la susceptibilité aux convulsions due à la lésion corticale, ainsi qu'une restauration de la taille des épines. Ainsi, l’augmentation de KCC2 à la suite d'une convulsion précoce, augmente la susceptibilité aux convulsions modifiant la morphologie des épines dendritiques, probable facteur contribuant à l’atrophie de l’hippocampe et l’occurrence des déficits cognitifs. Le deuxième objectif a été d'inspecter l’effet de la surexpression précoce de KCC2 dans le développement des épines dendritiques de l’hippocampe. Nous avons ainsi surexprimé KCC2 aussi bien in vitro dans des cultures organotypiques d’hippocampe, qu' in vivo par électroporation in utero. À l'inverse des résultats publiés dans le cortex, nous avons observé une diminution de la densité d’épines dendritiques et une augmentation de la taille des épines. Afin de confirmer la spécificité du rôle de KCC2 face à la région néocorticale étudiée, nous avons surexprimé KCC2 dans le cortex par électroporation in utero. Cette manipulation a eu pour conséquences d’augmenter la densité et la longueur des épines synaptiques de l’arbre dendritique des cellules glutamatergiques. En conséquent, ces résultats ont démontré pour la première fois, que les modifications de l’expression de KCC2 sont spécifiques à la région affectée. Ceci souligne les obstacles auxquels nous faisons face dans le développement de thérapie adéquat pour l’épilepsie ayant pour but de moduler l’expression de KCC2 de façon spécifique. / The potassium-chloride cotransporter KCC2 decreases intracellular Cl- levels and renders GABA responses inhibitory. In addition, it has also been shown to modulate excitatory synapse development. In this project, we investigated how alterations of KCC2 expression levels affect these two key processes in cortical structures of a normal and/or epileptic developing brain. First, we demonstrate that KCC2 expression is altered by early-life febrile status epilepticus. Febrile seizures affect about 5% of children during the first year of life. Atypical febrile seizures, particularly febrile status epilepticus, correlate with a higher risk of developing cognitive deficits and temporal lobe epilepsy as adults, suggesting that they may permanently change the developmental trajectory of neuronal circuits. In fact, the presence of a cerebral malformation predisposes to the development of atypical febrile seizures and temporal lobe epilepsy. The mechanisms underlying these effects are not clear. Here, we investigated the functional impact of this alteration on subsequent synapse formation and seizure susceptibility. We analyzed KCC2 expression and spine density in the hippocampus of a well-established rodent model of atypical febrile seizures, combining a cortical freeze lesion at post-natal day 1 (P1) and hyperthermia-induced seizure at P10 (LHS rats). 86% of these LHS males develop epilepsy and learning and memory deficits in adulthood. At P20, we found a precocious increase in KCC2 protein levels, accompanied by a negative shift of the reversal potential of GABA (EGABA) by gramicidin-perforated patch. In parallel, we observed a reduction in dendritic spine size by DiI labelling and a reduction of miniature excitatory postsynaptic current (mEPSC) amplitude in CA1 pyramidal neurons, as well as impaired spatial memory. To investigate whether the premature expression of KCC2 played a role in these alterations in the LHS model, and on seizure susceptibility, we reduced KCC2 expression in CA1 pyramidal neurons by in utero electroporation of shRNA using a triple-probe electrode. This approach lead to reduced febrile seizure susceptibility, and rescued spine size shrinkage in LHS rats. Our results show that an increase of KCC2 levels induced by early-life insults affect seizure susceptibility and spine development and may be a contributing factor to the occurrence of hippocampal atrophy and associated cognitive deficits in LHS rats. Second, we investigated whether KCC2 premature overexpression plays a role in spine alterations in the hippocampus. We overexpressed KCC2 in hippocampal organotypic cultures by biolistic transfection and in vivo by in utero electroporation. In contrast to what was previously published, we observed that both manipulations lead to a decrease in spine density in the hippocampus, as well as an increase in spine head size in vivo. In fact, it has been previously shown that overexpressing KCC2 leads to an increase of spine density in the cortex in vivo. To prove that this discrepancy is due to brain regional differences, we overexpressed KCC2 in the cortex by in utero electroporation, and similarly found an increase in spine density and length. Altogether, our results demonstrate for the first time, that alterations of KCC2 expression are brain circuit-specific. These findings highlights the obstacles we will face to find adequate pharmacological treatment to specifically modulate KCC2 in a region-specific and time-sensitive manner in epilepsy.