The lithium/pilocarpine model of epilepsy mimics mesial temporal lobe epilepsy with hippocampal sclerosis (MTLE-HS) in humans. Systemic injection of pilocarpine in lithium chloride (LiCL) pretreated adult rats results in an acute episode of severe continuous seizure activity (status epilepticus, SE). SE causes a latent period, whereby the animal appears neurologically normal, with subsequent development of spontaneous recurrent seizures (SRSs). Neuropathological changes that occur during the latent period are believed to contribute to the epileptic condition. The present thesis characterized the development of neuronal death and behavioural alterations in rats after SE induced by the repeated low-dose pilocarpine procedure (RLDP), and investigated the causal relationship between these two processes. Our data demonstrated that the RLDP procedure for the induction of SE results in widespread neurodegeneration and behavioural alterations comparable to the pilocarpine and low-dose pilocarpine (LDP) procedures. However, the advantage to using this protocol was strain-dependent as it reduced mortality in Wistar, but not in Long Evans Hooded (LEH), rats. Stereological analysis of neurons (stained for the neuronal specific marker [NeuN]) at various times (1 hr to 3 months) following SE showed that different brain regions within the hippocampus, amygdala, thalamus and piriform cortex exhibited differential rates of neuronal loss, with the majority of SE-induced neuronal death present by 24 hours. SE resulted in decreased exploratory behavior as assessed in the open field test, increased aggression to handling, increased hyperreactivity as assessed in the touch-response test, and anxiolytic effects as measured in the elevated-plus maze. Furthermore, deficits in search strategies used, as well as impaired spatial learning and memory, contributed to poor Morris water maze (MWM) performance. Partial neuroprotection within the hippocampus (by tat-NR2B9c) had no effect on the number of rats developing SRSs or on behavioural alterations; this argues against a causal relationship between neurodegeneration within this region, genesis of SRSs, and behavioural morbidity.

The lithium/pilocarpine model of epilepsy mimics mesial temporal lobe epilepsy with hippocampal sclerosis (MTLE-HS) in humans. Systemic injection of pilocarpine in lithium chloride (LiCL) pretreated adult rats results in an acute episode of severe continuous seizure activity (status epilepticus, SE). SE causes a latent period, whereby the animal appears neurologically normal, with subsequent development of spontaneous recurrent seizures (SRSs). Neuropathological changes that occur during the latent period are believed to contribute to the epileptic condition. The present thesis characterized the development of neuronal death and behavioural alterations in rats after SE induced by the repeated low-dose pilocarpine procedure (RLDP), and investigated the causal relationship between these two processes. Our data demonstrated that the RLDP procedure for the induction of SE results in widespread neurodegeneration and behavioural alterations comparable to the pilocarpine and low-dose pilocarpine (LDP) procedures. However, the advantage to using this protocol was strain-dependent as it reduced mortality in Wistar, but not in Long Evans Hooded (LEH), rats. Stereological analysis of neurons (stained for the neuronal specific marker [NeuN]) at various times (1 hr to 3 months) following SE showed that different brain regions within the hippocampus, amygdala, thalamus and piriform cortex exhibited differential rates of neuronal loss, with the majority of SE-induced neuronal death present by 24 hours. SE resulted in decreased exploratory behavior as assessed in the open field test, increased aggression to handling, increased hyperreactivity as assessed in the touch-response test, and anxiolytic effects as measured in the elevated-plus maze. Furthermore, deficits in search strategies used, as well as impaired spatial learning and memory, contributed to poor Morris water maze (MWM) performance. Partial neuroprotection within the hippocampus (by tat-NR2B9c) had no effect on the number of rats developing SRSs or on behavioural alterations; this argues against a causal relationship between neurodegeneration within this region, genesis of SRSs, and behavioural morbidity.

L’épilepsie du lobe temporal (ELT) est une pathologie neurologique sévère dont le fort taux de pharmacorésistance nécessite de nouveaux traitements. Le modèle lithium-pilocarpine modélise les caractéristiques et le développement de l’ELT. L’administration du carisbamate au début de l’épileptogenèse empêche l’apparition de l’ELT dans une sous-population de rats et la remplace par une épilepsie de type absence (ETA). L’évaluation cognitive effectuée durant la phase chronique a permis de distinguer les deux sous-populations : le groupe épilepsie de type absence ne développe pas le déficit cognitif sévère observé dans le modèle lithium pilocarpine. La spectroscopie du 13C n’a pas révélé de différence métabolique majeure entre les deux sous populations traitées, qu’elles développent une ELT ou une ETA. Ce travail souligne que l’administration précoce de carisbamate peut transformer l’ELT en une épilepsie moins sévère et fortement améliorer les comorbidités cognitives qui accompagnent l’ELT. / Temporal lobe epilepsy (TLE) is a severe neurological disease with a high refractory rate, which requires new treatments. The lithium-pilocarpine model allows reproducing human TLE features and development. Carisbamate administration at epileptogenesis onset prevents TLE incidence in a rats’ subpopulation, which is substituted by absence-like epilepsy (ALE). Behavioral and cognitive assessment performed during chronic period allowed differentiating the two subpopulations: ALE group did not develop the severe cognitive impairment shown in the lithium-pilocarpine model. 13C spectroscopy did not show major metabolism difference between the two treated subpopulations, whatever they develop TLE or ALE. This work demonstrates that early carisbamate administration can induce a shift from TLE in a less severe epilepsy form, and can strikingly improve TLE-related cognitive comorbidities.

Introdução: A epilepsia do lobo temporal mesial (ELTM) é a forma mais comum de epilepsia na população adulta, tendo a esclerose hipocampal como principal substrato neuropatológico. Uma elevada proporção dos pacientes com ELTM apresentam história familiar positiva para a epilepsia, sugerindo a participação de fatores genéticos nesta síndrome. O citoesqueleto neuronal é essencial para os processos fisiológicos das células nervosas e crises epilépticas parecem afetar esta estrutura. O citoesqueleto neuronal possui importantes componentes regulatórios, dentre eles as proteínas associadas aos microtúbulos do tipo 2 (MAP2) e tau. Objetivos: Avaliar a densidade neuronal (células imunomarcadas com NeuN), neobrotamento das fibras musgosas (por histoquímica de neo-Timm) e a expressão imunohistoquímica das proteínas MAP2 e tau no hipocampo de casos cirúrgicos de ELTM (subtipos familiar e esporádico) e controles. As mesmas análises foram conduzidas no modelo animal de lítiopilocarpina. Métodos: Casos ELTM (n=38) foram divididos em ELTM familiar (n=20) e ETLM esporádica (n=18). Hipocampos controles (n=10) foram provenientes de necrópsias de sujeitos sem histórico de problemas neurológicos. No modelo animal de lítio-pilocarpina, animais submetidos ao status epilepticus (SE) foram sacrificados nos seguintes tempos pós-SE: 1, 7, 15 e 60 dias e classificados como grupos SE1, SE7, SE15 e SE60. Animais controle foram injetados com salina. Resultados: O grupo ELTM apresentou menor densidade neuronal que o controle nas regiões da camada granular, hilo, CA4, CA3, CA1 e prosubículo. O grupo ELTM mostrou maior valor de cinza da histoquímica de neo-Timm na camada granular, molecular interna e molecular externa e maior espessura do neobrotamento axonal na camada molecular interna em relação ao grupo controle. O grupo ELTM esporádica teve maior valor de cinza da histoquímica de neo-Timm e tendência à maior espessura do neobrotamento que o grupo ELTM familiar na camada molecular interna. O grupo ELTM apresentou menor expressão de MAP2 que o grupo controle no hilo, CA4, CA3, CA1 e prosubículo e maior expressão que o controle na camada granular, CA2 e parasubículo. O grupo ELTM esporádica teve maior expressão de MAP2 que o grupo ELTM familiar em CA1 e córtex entorrinal. O grupo ELTM apresentou maior expressão de tau que o grupo controle nas regiões da camada granular, hilo, CA3, CA2 e córtex entorrinal. O grupo ELTM esporádica demonstrou menor imunorreatividade para tau que o grupo ELTM familiar no córtex entorrinal. No modelo de lítio-pilocarpina, os grupos submetidos ao SE apresentaram menor densidade neuronal e maior neobrotamento que o grupo salina. Quanto à expressão de MAP2, os grupos SE60 e SE15 apresentaram maiores valores de expressão que os demais grupos em todas as regiões analisadas. A expressão de tau em CA3 não foi diferente entre os grupos animais analisados. Conclusão: Nossos resultados existem perda neuronal, reorganização sináptica das fibras musgosas e alterações na expressão de proteínas MAP2 e tau no hipocampo de pacientes com ELTM. Dentre os casos ELTM, encontramos diferenças no neobrotamento de fibras musgosas e na expressão de MAP2 e tau entre os grupos ELTM familiar e esporádica, indicando que estes grupos se apresentam de forma distinta em relação à epilepsia. Encontramos resultados II semelhantes em relação à densidade neuronal e o neobrotamento de fibras musgosas entre pacientes com ELTM e no modelo experimental. Porém, em relação à expressão das proteínas MAP2 e tau, encontramos algumas diferenças entre humanos e animais com ELTM. Nossos resultados sugerem que apesar do modelo de epilepsia induzida por lítio-pilocarpina ser relevante para o estudo da ELTM humana e mimetizar importantes aspectos neuropatológicos, a correspondência quanto à expressão de algumas proteínas estruturais não é completa. Estudos adicionais de expressão de proteínas do citoesqueleto em outros modelos animais de ELTM serão de grande importância para o melhor entendimento do processo epileptogênco. / Introduction: Mesial temporal lobe epilepsy (MTLE) is the most common form of epilepsy in adults and has hippocampal sclerosis as the main pathological substrate. A high proportion of patients with MTLE have positive familial history for epilepsy, suggesting the involvement of genetic factors in this syndrome. Seizures may affect the neuronal cytoskeleton, an structure that is essential in the physiological processes of nerve cells. Components of the neuronal cytoskeleton include microtubule-associated protein type 2 (MAP2) and tau. Objectives: Our aim was to evaluate neuronal density, mossy fiber sprouting and immunohistochemical expression of MAP2 and tau in the hippocampus of surgical cases of MTLE (familial and sporadic subtypes) and controls. The same analysis were conducted in the MTLE lithiumpilocarpine animal model.Methods: MTLE cases (n = 38) were divided into familial MTLE (n = 20) and sporadic MTLE (n = 18). Control hippocampi (n = 10) were obtained from autopsies of subjects without history of epilepsy. In the lithium-pilocarpine animal model, male Wistar rats were submitted to status epilepticus (SE) and were killed at the following post-SE days: 1, 7, 15 and 60, and were further classified SE1, SE7, SE15 and SE60. Control animals were injected with saline. Results: MTLE showed decreased neuronal density than controls in the granular layer, hilus, CA4, CA3, CA1 and prosubiculum. MTLE group showed increased neo-Timm gray value in the granular layer, inner and outer molecular layer, as well as increased mossy fiber length of mossy fiber in the inner molecular layer when compared to controls. Sporadic MTLE specimens exhibited increased inner molecular layer gray value than familial MTLE. MTLE hihppocampi showed decreased MAP2 expression in the hilus, CA4, CA3, CA1 and prosubiculum. In the granular layer, CA2 and parasubiculum, MAP2 expression was higher in MTLE specimens than in controls. CA1 and entorhinal cortex from sporadic MTLE hippocampi showed increased MAP2 expression than familial MTLE. Tau expression was increased in the granular layer, hilus, CA3, CA2 and entorhinal cortex of MTLE specimens. Sporadic MTLE exhibited lower tau immunoreactivity in the entorhinal cortex than familial MTLE. In the lithiumpilocarpine animal model, rats submitted to SE presented lower values of neuronal density and mossy fiber sprouting than controls. SE15 and SE60 showed increased MAP2 expression in all hippocampal subfields. Tau expression in CA3 was not different among the groups. Conclusion: Our results indicate that besides neuronal loss and axonal sprouting, MTLE specimens also exhibit dendritic abnormalities related to MAP2 expression. Increased tau immunoreactivity in epileptogenic hippocampi indicates possible abnormal expression related to mossy fiber sprouting in chronic MTLE. We found differential sprouting, MAP2 and tau expression between sporadic and familial MTLE. It is possible that different genetic background might result in somehow distinct neuropathological substrates between the two MTLE subtypes, although their clinical manifestation is quite similar. While neuronal loss and axonal sprouting profiles in human MTLE and chronic phase of the lithium-pilocarpine animal model are comparable, we could not find corresponding results regarding MAP2 and tau expression. IV Several studies have shown that chemo-convulsant as pilocarpine and kainic acid result in widespread brain epileptic discharges, which are different from the more focal hippocampal discharges seen in human MTLE. Despite its limitations, the lithium-pilocarpine model stands as an important and widely used animal model of epilepsy. Besides that, other animal models in which ictal discharges and lesions are more limited to the hippocampal formation might better mimic what we see in human MTLE.

Introdução: A epilepsia do lobo temporal mesial (ELTM) é a forma mais comum de epilepsia na população adulta, tendo a esclerose hipocampal como principal substrato neuropatológico. Uma elevada proporção dos pacientes com ELTM apresentam história familiar positiva para a epilepsia, sugerindo a participação de fatores genéticos nesta síndrome. O citoesqueleto neuronal é essencial para os processos fisiológicos das células nervosas e crises epilépticas parecem afetar esta estrutura. O citoesqueleto neuronal possui importantes componentes regulatórios, dentre eles as proteínas associadas aos microtúbulos do tipo 2 (MAP2) e tau. Objetivos: Avaliar a densidade neuronal (células imunomarcadas com NeuN), neobrotamento das fibras musgosas (por histoquímica de neo-Timm) e a expressão imunohistoquímica das proteínas MAP2 e tau no hipocampo de casos cirúrgicos de ELTM (subtipos familiar e esporádico) e controles. As mesmas análises foram conduzidas no modelo animal de lítiopilocarpina. Métodos: Casos ELTM (n=38) foram divididos em ELTM familiar (n=20) e ETLM esporádica (n=18). Hipocampos controles (n=10) foram provenientes de necrópsias de sujeitos sem histórico de problemas neurológicos. No modelo animal de lítio-pilocarpina, animais submetidos ao status epilepticus (SE) foram sacrificados nos seguintes tempos pós-SE: 1, 7, 15 e 60 dias e classificados como grupos SE1, SE7, SE15 e SE60. Animais controle foram injetados com salina. Resultados: O grupo ELTM apresentou menor densidade neuronal que o controle nas regiões da camada granular, hilo, CA4, CA3, CA1 e prosubículo. O grupo ELTM mostrou maior valor de cinza da histoquímica de neo-Timm na camada granular, molecular interna e molecular externa e maior espessura do neobrotamento axonal na camada molecular interna em relação ao grupo controle. O grupo ELTM esporádica teve maior valor de cinza da histoquímica de neo-Timm e tendência à maior espessura do neobrotamento que o grupo ELTM familiar na camada molecular interna. O grupo ELTM apresentou menor expressão de MAP2 que o grupo controle no hilo, CA4, CA3, CA1 e prosubículo e maior expressão que o controle na camada granular, CA2 e parasubículo. O grupo ELTM esporádica teve maior expressão de MAP2 que o grupo ELTM familiar em CA1 e córtex entorrinal. O grupo ELTM apresentou maior expressão de tau que o grupo controle nas regiões da camada granular, hilo, CA3, CA2 e córtex entorrinal. O grupo ELTM esporádica demonstrou menor imunorreatividade para tau que o grupo ELTM familiar no córtex entorrinal. No modelo de lítio-pilocarpina, os grupos submetidos ao SE apresentaram menor densidade neuronal e maior neobrotamento que o grupo salina. Quanto à expressão de MAP2, os grupos SE60 e SE15 apresentaram maiores valores de expressão que os demais grupos em todas as regiões analisadas. A expressão de tau em CA3 não foi diferente entre os grupos animais analisados. Conclusão: Nossos resultados existem perda neuronal, reorganização sináptica das fibras musgosas e alterações na expressão de proteínas MAP2 e tau no hipocampo de pacientes com ELTM. Dentre os casos ELTM, encontramos diferenças no neobrotamento de fibras musgosas e na expressão de MAP2 e tau entre os grupos ELTM familiar e esporádica, indicando que estes grupos se apresentam de forma distinta em relação à epilepsia. Encontramos resultados II semelhantes em relação à densidade neuronal e o neobrotamento de fibras musgosas entre pacientes com ELTM e no modelo experimental. Porém, em relação à expressão das proteínas MAP2 e tau, encontramos algumas diferenças entre humanos e animais com ELTM. Nossos resultados sugerem que apesar do modelo de epilepsia induzida por lítio-pilocarpina ser relevante para o estudo da ELTM humana e mimetizar importantes aspectos neuropatológicos, a correspondência quanto à expressão de algumas proteínas estruturais não é completa. Estudos adicionais de expressão de proteínas do citoesqueleto em outros modelos animais de ELTM serão de grande importância para o melhor entendimento do processo epileptogênco. / Introduction: Mesial temporal lobe epilepsy (MTLE) is the most common form of epilepsy in adults and has hippocampal sclerosis as the main pathological substrate. A high proportion of patients with MTLE have positive familial history for epilepsy, suggesting the involvement of genetic factors in this syndrome. Seizures may affect the neuronal cytoskeleton, an structure that is essential in the physiological processes of nerve cells. Components of the neuronal cytoskeleton include microtubule-associated protein type 2 (MAP2) and tau. Objectives: Our aim was to evaluate neuronal density, mossy fiber sprouting and immunohistochemical expression of MAP2 and tau in the hippocampus of surgical cases of MTLE (familial and sporadic subtypes) and controls. The same analysis were conducted in the MTLE lithiumpilocarpine animal model.Methods: MTLE cases (n = 38) were divided into familial MTLE (n = 20) and sporadic MTLE (n = 18). Control hippocampi (n = 10) were obtained from autopsies of subjects without history of epilepsy. In the lithium-pilocarpine animal model, male Wistar rats were submitted to status epilepticus (SE) and were killed at the following post-SE days: 1, 7, 15 and 60, and were further classified SE1, SE7, SE15 and SE60. Control animals were injected with saline. Results: MTLE showed decreased neuronal density than controls in the granular layer, hilus, CA4, CA3, CA1 and prosubiculum. MTLE group showed increased neo-Timm gray value in the granular layer, inner and outer molecular layer, as well as increased mossy fiber length of mossy fiber in the inner molecular layer when compared to controls. Sporadic MTLE specimens exhibited increased inner molecular layer gray value than familial MTLE. MTLE hihppocampi showed decreased MAP2 expression in the hilus, CA4, CA3, CA1 and prosubiculum. In the granular layer, CA2 and parasubiculum, MAP2 expression was higher in MTLE specimens than in controls. CA1 and entorhinal cortex from sporadic MTLE hippocampi showed increased MAP2 expression than familial MTLE. Tau expression was increased in the granular layer, hilus, CA3, CA2 and entorhinal cortex of MTLE specimens. Sporadic MTLE exhibited lower tau immunoreactivity in the entorhinal cortex than familial MTLE. In the lithiumpilocarpine animal model, rats submitted to SE presented lower values of neuronal density and mossy fiber sprouting than controls. SE15 and SE60 showed increased MAP2 expression in all hippocampal subfields. Tau expression in CA3 was not different among the groups. Conclusion: Our results indicate that besides neuronal loss and axonal sprouting, MTLE specimens also exhibit dendritic abnormalities related to MAP2 expression. Increased tau immunoreactivity in epileptogenic hippocampi indicates possible abnormal expression related to mossy fiber sprouting in chronic MTLE. We found differential sprouting, MAP2 and tau expression between sporadic and familial MTLE. It is possible that different genetic background might result in somehow distinct neuropathological substrates between the two MTLE subtypes, although their clinical manifestation is quite similar. While neuronal loss and axonal sprouting profiles in human MTLE and chronic phase of the lithium-pilocarpine animal model are comparable, we could not find corresponding results regarding MAP2 and tau expression. IV Several studies have shown that chemo-convulsant as pilocarpine and kainic acid result in widespread brain epileptic discharges, which are different from the more focal hippocampal discharges seen in human MTLE. Despite its limitations, the lithium-pilocarpine model stands as an important and widely used animal model of epilepsy. Besides that, other animal models in which ictal discharges and lesions are more limited to the hippocampal formation might better mimic what we see in human MTLE.

La métabolomique a pour objet l’identification et la quantification de métabolites dans un échantillon biologique. Cette discipline s’inscrit dans une approche du vivant connue sous le terme de « biologie des systèmes ». La spectroscopie par résonance magnétique nucléaire haute résolution à l’angle magique (RMN HRMAS) est une méthode de choix pour l’obtention de ce type de profilage métabolique. L’épilepsie du lobe temporal (ELT) est une épilepsie focale fréquente associée le plus souvent à des pertes neuronales sélectives, une gliose réactionnelle et une plasticité cellulaire spécifique. Bien que restant débattue, une origine neurométabolique reste un axe de recherche majeur. A ce jour une caractérisation métabolomique des tissus épileptogènes par RMN HRMAS reste à effectuer. Notre travail a consisté dans un premier temps à caractériser, chez le rat, les effets des méthodes de prélèvement et de fixation sur le métabolome cérébral dans le cadre des acquisitions RMN HRMAS. Dans un second temps, nous avons travaillé sur le modèle animal lithium-pilocarpine d’ELT. Nous avons pu décrire le métabolome issu des données RMN 1H HRMAS de différentes structures cérébrales impliquéesdans l’épileptogénèse. Des analyses multivariées de type PLS-DA ont pu mettre en évidence des profils métaboliques pathologiques au sein du cortex entorhinal et de l’hippocampe. A l’aide de substrats marqués au carbone 13 ([1-13C]glucose et de [1,2-13C]acétate) nous avons étudié les voies métaboliques neuronales et gliales. Nos résultats suggèrent l’absence d’anomalies métaboliques au sein des astrocytes. Enfin dans un dernier temps, nous avons effectué des analyses RMN 1H HRMAS sur près de 200 échantillons cérébraux de patients atteints d’ELT. Une analyse multivariée a permis de distinguer les profils métaboliques des hippocampes sclérosés et non sclérosés. En revanche la construction de modèles sur la base d’hypothèses clinico métaboliques (durée de la maladie, fréquence de crises, antécédents de convulsions fébriles) n’a pas permis d’identifier de profils métaboliques spécifiques. L’ensemble de ces données suggère l’existence de profils métabolomiques distincts en fonction des caractéristiques neuropathologiques des patients atteints d’ELT. Notre travail confirme la nécessité d’une approche intégrée de type « biologie des systèmes » pour l’étude de l’ELT aussi bien chez l’homme que dans des modèles animaux. / Metabolomics relates to the identification and quantification of metabolites in biological samples. This discipline is part of an approach known under the term of "systems biology". High Resolution Nuclear Magnetic Resonance Spectroscopy at Magic Angle Spinning (HRMAS NMR) is a method for obtaining metabolic profiling in such sample. Temporal Lobe Epilepsy (TLE ) is a common focal epilepsy often associated with selective neuronal loss, reactive gliosis and specific cellular plasticity. A neurometabolic origin of this epilepsy is a major area of research. To date no characterization of human cerebral biopsy from TLE patients has been conducted using HRMAS NMR. In the present work we aimed first at characterizing, in rats, the effects of sampling methods and fixation on brain metabolome under HRMAS NMR acquisitions. In a second step, we studied the lithium-pilocarpine model of TLE. In this model, we could describe the metabolome from HRMAS 1H NMR data of different brain structures involved in epileptogenesis. Multivariate analysis could highlight pathological metabolic profiles in the entorhinal cortex and hippocampus. Using substrates labeled with carbon 13 ( [1 -13C ]-glucose and [1,2-13C ]-acetate) we studied neuronal and glial metabolic pathways. Our results suggest the absence of metabolic abnormalities in astrocytes metabolism as previously reported. Finally, we conducted HRMAS 1H NMR analysis in nearly 200 brain samples from TLEpatients. Multivariate analysis was able to distinguish metabolic profiles between sclerotic and non sclerotic hippocampi. However mutlivariate models based on clinico- metabolic assumptions (disease duration, frequency of seizures, history of febrile seizures ) did not identify specific metabolic profile. All these data suggest the existence of distinct metabolomic profile based on neuropathological features of patients with TLE. Our work confirm the need of an integrated approach such as " systems biology" for the study of TLE in humans as long as in animal models.