Thèse numérisée par la Direction des bibliothèques de l'Université de Montréal.

Long-term depression (LTD) plays an important role in the refinement of neocortical maps during early postnatal development. Synapse formation and refinement in the cortex during development rely on synaptic plasticity, the cellular mechanisms of which are poorly understood. The aim of this thesis was to investigate timing- and pattern-dependent LTD at excitatory synapses in the mouse barrel cortex during development. This thesis first describes the developmental prole and N-methyl-D-aspartate (NMDA) receptor GluN2 subtype-dependence of timing-dependent plasticity at layer 4-to-layer 2/3 synapses. A developmental dissociation of timing-dependent plasticity was observed where timing-dependent LTD (t-LTD) was found during early development (postnatal day, P6-8) but disappeared after P25. In contrast, timing-dependent LTP (t-LTP) only appeared in the second postnatal week of development (P11-15) and persisted in the adult cortex. This bidirectional plasticity also showed a GluN2 subtype-dependent dissociation. Whereas t-LTP was dependent on GluN2A subunit-containing NMDA receptors, t-LTD was dependent on GluN2C/D subunit-containing NMDA receptors. This thesis also reports a novel anti-Hebbian form of NMDA receptor-dependent plasticity, in which presynaptic layer 4 neurons drive their presynaptic long-term self-depression without the involvement of postsynaptic action potentials or calcium. This mechanism suggests that, during development, presynaptic self-depression occurs when specific spike patterns (presynaptic burst-spike) in the presynaptic neuron are unsuccessful in driving postsynaptic activity. Finally, this thesis addresses how t-LTD induction rules differ in vertical intra-columnar layer 4-to-layer 2/3 and horizontal cross-columnar layer 2/3-to-layer 2/3 synapses in the barrel cortex. Distinct GluN2 subunit expression in vertical and horizontal synapses regulated the time-window of t-LTD induction. It is also shown that different excitatory intra- and cross-columnar synapses onto the same postsynaptic layer 2/3 neurons can have different molecular requirements for the induction of t-LTD, and that they also interact to induce heterosynaptic associative LTD. These findings may have important implications for understanding the cellular mechanisms of experience-dependent plasticity and its relevance to the computational principles of cortical circuit operation.

Thèse numérisée par la Division de la gestion de documents et des archives de l'Université de Montréal.

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal.

Les troubles schizophréniques (SCZ) ont une forte héritabilité, de l’ordre de 80%, mais, une très faible part du risque génétique a été identifiée. La plupart des études ont considéré l’implication de polymorphismes fréquents, chacun ayant un effet relativement faible individuellement, alors que les études de variants du nombre de copies (CNVs) ainsi que les études d’anomalies chromosomiques ont pointé l’implication possible de variants rares et de novo à une forte pénétrance. Dans une première partie, nous présentons une synthèse sur les facteurs génétiques dans la SCZ, puis une revue des arguments en faveur de l’implication d’anomalies du système glutamatergique dans la SCZ, domaine sur lequel s’est centré notre travail. Notre travail s’inscrit dans un projet plus vaste, Synapse to Disease (S2D) ayant pour objectif de séquencer 1000 gènes synaptiques dans des cohortes de patients atteints de schizophrénie ou de troubles du spectre autistique. Nous avons exploré en particulier le système glutamatergique et les récepteurs NMDA. Dans un premier article, nous montrons une association d’une mutation troncante de novo de la kinésine 17, impliquée dans le transport de la sous-unité GRIN2B des récepteurs NMDA. Dans un second article, nous explorons les mutations rares et de novo dans les sous-unités des récepteurs NMDA et montrons l’association de mutation de novo dans GRIN2A et GRIN2B avec des cas de SCZ et d’autisme. Nos résultats renforcent l’idée qu’une part des cas de schizophrénie pourrait être due à l’implication de mutations rare à effet majeur, hypothèse alternative mais non exclusive à l’hypothèse d’interactions entre variants génétiques fréquents à effet mineur. / Schizophrenic disorders (SCZ) have high heritability (around 80%), but only a small part has been characterized. Most studies have focussed on common polymorphisms, each having small individual effect, whereas copy number variant and chromosomal abnormalities studies have pointed to the possible involvement of rare and de novo mutations with high penetrance. In the first part of this manuscript, we will present a synthesis on genetic factors of SCZ and then a review of the arguments supporting an involvement of glutamatergic system abnormalities in SCZ, which is the focus of our research. Our work is part of a global project, Synapse to Disease (S2D), that aimed to sequence 1000 synaptic genes in cohort of patients affected with schizophrenia or autism spectrum disorders. We focussed in particular on the glutamatergic system and NMDA receptors. In a first publication we show an association between SCZ and a de novo truncating mutation of kinesin 17, wich has been implicated in the transport of the GRIN2B subunit of NMDA receptors. In a second publication we explore rare and de novo mutations in NMDA receptor subunits. We show an association between de novo mutations in GRIN2A and GRIN2B with cases of SCZ and autism. Our results strengthen the idea that a portion of schizophrenia cases could be related to rare mutations having a high penetrance, an alternative but not contradictory explanation to the hypothesis for an interaction between common variants having a small effect.

The glutamate system dysfunctions present in schizophrenia raise new questions about possible glutamatergic actions of the atypical antipsychotic clozapine. While clozapine has been shown to partially substitute for the discriminative stimulus of the glutamate agonist N-methyl D-aspartate (NMDA) in rats, NMDA discrimination has not previously been established in mice. The present study was designed to explore the possible role of NMDA activity in clozapine’s discriminative stimulus. Two groups of C57BL/6 mice were trained to discriminate either 2.5 mg/kg CLZ from vehicle or 30 mg/kg NMDA from vehicle in a standard two-lever drug discrimination task. NMDA drug discrimination was successfully established in C57BL/6 mice. While NMDA did not substitute for clozapine, clozapine partially substituted for NMDA at the 0.625 mg/kg dose, demonstrating an asymmetrical relationship between clozapine’s and NMDA’s discriminative stimuli. Dose combination tests further investigated this relationship. It was found that 0.625 mg/kg CLZ + 30 mg/kg NMDA produced partial substitution (61.82% DLR), while 0.625 mg/kg CLZ + 56 mg/kg NDMA full substitution (92.82% DLR) in CLZ-trained mice. In addition, combination testing with 10 mg/kg NMDA + 2.5 mg/kg CLZ and 10 mg/kg NMDA + 5.0 mg/kg CLZ produced full substitution in NMDA-trained mice ((80.04% DLR and 100% DLR, respectively). Finally, it was found that the α1- adrenoreceptor antagonist prazosin fully substituted for both CLZ (3.0 mg/kg = 92.20% DLR) and NMDA (1.0 mg/kg = 98.77% DLR and 3.0 mg/kg = 99.62% DLR). These findings suggest that interactions between clozapine’s and NMDA’s discriminative stimuli may involve antagonism of α1- adrenoreceptors, but further research of other mechanisms including serotonergic, histaminergic, and cholinergic receptor activity or metabolic interactions is needed. Finally, these initial finding suggest that that drugs active at glutamatergic receptors may have potential as therapeutic drugs for treatment of schizophrenia.

Die HIV-Infektion des Gehirns induzierte eine Vielzahl neurologischer und neuropsychiatrischer Veränderungen, die gemeinsam als HIV-assoziierte Demenz bezeichnet werden. Insbesondere eine Beteiligung der glutamate-vermittelten Toxizität wird im Rahmen der HIV-Demenz diskutiert. Die vorliegende Arbeit verwendete das wichtigste Tiermodell der HIV-Infektion, mit simianen Immundefizienzviren (SIV) infizierte Rhesusaffen, um die komplexen Wechselwirkungen zwischen Immunaktivierung und adaptiven Veränderungen der glutamatergen Synapse und ihrer Umgebung zu untersuchen. Mittels einer Microarrayanalyse wurden allgemein Unterscheide im Genexpressionsprofil der Basalganglien von Rhesusaffen während dem Fortschreiten der Infektion erfasst. Die Ergebnisse zeigten Unterschiede in der transkriptionalen Regulation zwischen Chinesischen und Indischen Makakkensubspezies, sowie ein einzigartig differenziertes Genexpressionsmuster als Antwort auf die SIV-Infektion selbst. Um adaptive Veränderungen in den funktionalen Elementen der glutamatergen Synapse weiter zu charakterisieren, wurde die Expression der N-methyl-D-aspartat-Rezeptoruntereinheiten gemessen, wobei ein progressiver Verlust der Untereinheitenexpression im Putamen sowie ein differenziellen Expressionsmuster in Abhängigkeit sowohl von der Makakkensubspezies sowie von der Hirnregion im Nucleus accumbens SIV-infizierter Rhesusaffen festgestellt wurde. Außerdem konnte die vorliegende Arbeite zum ersten Mal eine Störung der exzitatorischen Aminosäuretransporter, dem wichtigsten Glutamatwiederaufnahmesystem, im Putamen, aber nicht im Nucleus accumbens, während der SIV-Infektion nachweisen. Zuvor durchgeführte Arbeiten wiesen darauf hin, dass neurochemischen Dysfunktionen unter Umständen das Ergebnis einer indirekten Toxizität vermittelt durch aktivierte Mikroglia und der daraus resultierenden Ausschüttung schädlicher Faktoren sein könnten. Entsprechend zeigten unsere Daten einen Anstieg in der MHC-II- und TNF-alpha-Expression bereits während der asymptomatischen Phase, und eine noch deutlichere Heraufregulierung in AIDS-Tieren. Zusammenfassend weist die vorliegen Arbeit auf komplexe und kombinierte Mechanismen von einem Anstieg in der Glutamatkonzentration mit einer Dysregulation der NMDA-Rezeptorfunktion sowie der Glutamatwiederaufnahmesysteme verursacht durch eine Immunaktivierung hin. Außerdem konnte in Versuchen zur pharmakologischen Beeinflussung der glutamatergen Synapse mittels dopaminerger Substanzen, NMDAR-Antagonisten und Antioxidantien eine Verbesse-rung der neurochemischen Funktion durch den nichtkompetitiven Antagonisten Memantin und die Antioxidantien Melatonin sowie N-Acetylcystein gezeigt werden. Hingegen trugen die Ergebnisse mit dem MAO-B-Inhibitor Selegilin weiter zu Bedenken zur Sicherheit und Effizienz dopaminerger Substanzen in der Behandlung von HIV-Patienten bei. Außerdem zeigte die vorliegende Arbeit, dass Memantin spezifisch die mRNA- und Proteinexpression des Neurotrophins BNDF heraufregulierte und weist somit auf eine neuartige pharmakologische Wirkung Antidementivums hin. In einer ergänzenden Studie wurden Schritte zur Entwicklung und Verbesserung adeno-assoziierten viraler und foamyviraler Vektoren, die shRNAs für die effiziente Ausschaltung der Expression des murinen und humanen Dopamintransporters, durchgeführt. Es wurden verschieden virale Vektorplasmide kloniert, Effektormoleküle mit einer hohen Knockdown-Aktivität identifiziert, und infektiöse rekombinante Viren hergestellt. Zudem konnte die Expression des humanen Dopamintransporters auf Lymphozyten bestätigt werden. Diese Ergebnisse werden die Entwicklung transgener Tiere und Zelllinien erleichtern und tragen so zur Analyse der natürlichen dopaminergen Neurotransmission in der psychiatrischen Forschung bei, wobei die durch endogene Anpassungen der dopaminerge Systeme in klassischen Knockout-Systemen hervorgerufenen Schwierigkeiten umgangen werden können. / HIV infection of the brain induces a wide range of neurological and neuropsychiatric symptoms colletively defined as HIV-associated dementia. In particular, glutamate-mediated toxicity is discussed to be involved in neurodysfunction during HIV dementia. The present thesis used the most relevant animal model for HIV infection, the simian immunodeficiency virus (SIV)-infected rhesus macaques, to explore the complex interaction between immuneactivation and adaptive alterations of the glutamatergic synapse and its environment. A microarray analyses was performed to asses general differences in gene expression pofiling in the basal ganglia of rhesus macaques during diseases progression. The results showed considerable differences in transcriptional regulation between Chinese and Indian macaque subspecies, as well as a uniques differential gene expression pattern in response to SIV infection itself. To further characterize adaptative changes in functional elements of the glutamatergic synapse we measured expression of N-methyl-D-aspartate receptor subunits and found a progressive loss of subunit expression in putamen as well as a differential expression pattern with respect to macaque subspecies and brain region in nucleus accumbens of SIV-infected monkey. In addition, the current thesis demonstrated for the first time a disruption of excitatory amino acid transporters, the most important glutamate clearing system, in putamen but not in nucleus accumbens during SIV-infection. Previous studies suggested that neurochemical dysfunction may be the result of indirect toxicity mediated by activated microglia and subsequent release of ditrimental factors. Accordingly, our data demonstrated an increase in MHC-II and TNF-alpha expression during the asymptomatic stage of infection which was further upregulated in AIDS animals. Taken together the present thesis points out to complex and combined mechanisms of an increase glutamate concentration with dysregulation of NMDA receptor function and glutamate clearing system caused by immuneactivation. In addition, experiments concerning pharmacological manipulation of the glutamatergic synapse by dopaminergic drugs, N-methyl-D-aspartate antagonists or antioxidant treatment revelead an improvement of neurochemical function by the uncompetitive NMDA antagonist memantine and by the antioxidants melatonin and N-acetylcysteine. In contrast, administration of the MAO-B inhibitor selegiline contributed to concerns about the saftey and efficiency of dopaminergic substances in treatment of HIV-patients. In addition, the current thesis demonstrated that memantine specifically upregulates mRNA and protein expression of the neurotrophic factor brain-derived neurotrophic factor, and therefore, refers to a novel pharmacological action of the antidementivum. In a complementary study we presented steps in the devolpment and improvement of adeno-associated viral and foamyviral vectors expressing short hairpin RNAs for an efficient knockdown of the murine and human dopamine transporter. We cloned different viral vector plasmids, identified effector molecues with high knockdown activity, and prepared infective recombinant viruses. In addition, we confirmed the expression of human dopamine transporter by lymphocytes. The results will facilitate the generation of transgenic animals and cell lines, and therefor contribute to the analysis of natural dopaminergic neurotransmission in psychiatric research overcoming the pitfalls of endogeneous adaptatations observerd in classical knockout technology.

Consolidated memories can become reactivated in order to permit the integration of new information into the memory trace. Blockade of the resultant process, reconsolidation, with NMDA receptor antagonists or protein synthesis inhibition can lead to a decrease in subsequent memory expression. This may offer a potential tool for the treatment of psychiatric disorders characterised by maladaptive memories, including drug addiction and post-traumatic disorder. Given the importance of instrumental associations in supporting drug addiction experiments in Chapters 3 & 4 aimed to disrupt reconsolidation of these memories. Treatment with an NMDA receptor antagonist prior to retrieval sessions of various durations was not able to consistently prevent reconsolidation of these associations. Drug addiction is characterised by memories that have been formed not over days or weeks, but months or years. Experiments in Chapters 5 & 6 therefore investigated how the extent of training affects the propensity of an appetitive pavlovian memory to reconsolidate. Experiments in Chapter 5 were not able disrupt reconsolidation of these memories after a relatively short period of training. In Chapter 6 attempts to disrupt reconsolidation of a cocaine-seeking memory having undergone extensive training (>1 month, designed to promote the formation of drug-seeking habits) were also unsuccessful. However, when animals were trained in a similar fashion to respond for a food reinforcer treatment with a NMDA receptor antagonist prior to a reactivation session resulted in a decrease in food-seeking behaviour the following day. However, this deficit was only found in the first test session; drug treatment had no effect on responding following reminder of the memory. If data from preclinical studies are to inform future psychiatric treatments the findings from these works must be robust and replicable. Experiments in previous chapters encountered several issues in this regard, namely the repeated inability to prevent reconsolidation with NMDA receptor antagonism. Given that reconsolidation of auditory fear memories is well characterised a final series of experiments in Chapter 7 used this procedure to explore the possible reasons for the fleeting or absent effects of disrupted memory reconsolidation in previous chapters. Despite the use of similar methods as published reports showing decreases in memory expression as a result of blockade of reconsolidation it was not possible to disrupt this process with NMDA receptor antagonism or protein synthesis inhibition. Results suggested that the failure to observe reactivation-dependent amnesia was due to the amnestic agent used not being able to prevent reconsolidation, should it be taking place, and a failure of the given retrieval trial to result in memory reactivation. On numerous occasions throughout this thesis it was not possible to disrupt memory reconsolidation. One difficulty in interpreting null data of this nature is that it is often unclear whether the results are due to insufficient retrieval conditions to result in memory reconsolidation, or an inability of the pharmacological agent to disrupt this process. The final experiments of this thesis raised the possibility both of these issues may have contributed in tandem towards this inability to prevent memory reconsolidation.

Dor neuropática (DN) é definida como “dor iniciada ou causada por lesão primária ou disfunção em sistema nervoso”, porém sua prevalência depende do tipo de trauma e da disfunção relacionada. Apesar desta condição dolorosa ser considerada altamente prevalente e debilitante, os tratamentos disponíveis são relacionados a efeitos adversos dificultando a adesão. Devido a isso, buscam-se alternativas não farmacológicas para o tratamento deste tipo de dor, entre elas, as técnicas de neuromodulação periférica, como acupuntura (AC) e eletroacupuntura (EA). Estas técnicas podem ser combinadas com intervenções farmacológicas e não farmacológicas e têm apresentado resultados promissores no tratamento da dor neuropática. No entanto, seus mecanismos de ação não estão totalmente elucidados, desta forma a utilização de modelos animais é de grande valia para o estudo destes mecanismos no tratamento da dor neuropática e da patofisiologia deste tipo de dor crônica. É importante salientar que a aplicação de AC e EA em animais acordados é complexa, visto que gera desconforto e pode alterar a analgesia induzida pelo tratamento. Em muitos estudos a anestesia com isoflurano é utilizada durante a aplicação dos tratamentos, porém sua utilização pode gerar um viés no estudo, considerando a possível interferência do fármaco nos resultados comportamentais e neuroquímicos. Outro importante foco de estudo consiste em comparar as duas técnicas, AC e EA, buscando determinar qual destas é a mais eficaz no tratamento da dor neuropática. Considerando o exposto acima, os objetivos desta tese foram: 1) avaliar os parâmetros comportamentais e neuroquímicos dos efeitos da utilização de anestesia na aplicação de AC e EA em ratos submetidos ao modelo de DN; 2) comparar os efeitos da AC e EA em modelo animal de DN por meio de parâmentros comportamentais, neuroquímicos e histológicos. Considerando os resultados obtidos nesta tese, concluímos que o isoflurano aumenta a analgesia promovida por AC e EA, provavelmente diminuindo o efeito do estresse gerado pela aplicação dos tratamentos em animais acordados, resultado que é corroborado pela diminuição do nível de S100β periférico (marcador de morte neuronal central); Por outro lado, o isoflurano diminuiu os níveis de fator de crescimento neuronal (NGF) no nervo periférico lesado, indicando diminuição do processo de regeneração neural, enquanto a EA aumentou. Ao mesmo tempo, o isoflurano alterou os efeitos dos tratamentos nos comportamentos exploratórios e nos níveis de N-metil D-aspartato em tronco encefáfio e medula espinhal. A AC apresentou-se mais eficaz no tratamento da DN em comparação à EA, porém nenhum dos tratamentos foi capaz de alterar os danos causados pela indução da DN no músculo gastrocnemio esquerdo dos animais demonstrado na histologia. Todavia, este resultado não alterou a analgesia gerada pelos tratamentos. / Neuropathic pain (NP) is defined as "pain initiated or caused by primary injury or dysfunction in the nervous system," but its prevalence depends on the type of trauma and related dysfunction. Although this painful condition is considered to be highly prevalent and debilitating, the available treatments are related to adverse effects, making adherence difficult. Because of this, non-pharmacological alternatives for the treatment of this type of pain are sought, among them, the techniques of peripheral neuromodulation, such as acupuncture (AC) and electroacupuncture (EA). These techniques can be combined with pharmacological and non-pharmacological interventions and have shown promising results in the treatment of neuropathic pain. However, its mechanisms of action are not fully elucidated, so the use of animal models is of great value for the study of these mechanisms in the treatment of neuropathic pain and the pathophysiology of this type of chronic pain. It is important to emphasize that the application of AC and EA in awake animals is complex, since it generates discomfort and can alter the analgesia induced by the treatment. In many studies, anesthesia with isoflurane is used during the application of the treatments, but its use may generate a bias in the study, considering the possible interference of the drug in the behavioral and neurochemical results. Another important focus of the study is to compare the two techniques, AC and EA, seeking to determine which is the most effective in the treatment of neuropathic pain. Considering the above, the objectives of this thesis were: 1) to evaluate the behavioral and neurochemical parameters of the effects of the use of anesthesia in the application of AC and EA in rats submitted to the DN model; 2) to compare the effects of AC and EA on animal model of DN by means of behavioral, neurochemical and morphological parameters. Considering the results obtained in this thesis, we conclude that isoflurane increases the analgesia promoted by AC and EA, probably decreasing the effect of the stress generated by the application of the treatments in agreed animals, a result that is corroborated by the decrease in the level of peripheral S100β (biomarker of central neuronal injury); On the other hand, isoflurane decreased the levels of neural grown factor (NGF) in the injured peripheral nerve, indicating a decrease in the neural regeneration process, while the EA increased. At the same time, isoflurane altered the effects of treatments on exploratory behaviors and N-metil-D-aspartato (NMDA) levels in the brainstem and spinal cord. AC was more effective in the treatment of DN compared to EA, but none of the treatments was able to alter the damage caused by DN induction in the left gastrocnemius muscle of the animals showed in histology. However, this result did not alter the analgesia generated by the treatments.

Stress has been shown to influence learning and memory in humans and rats (Diamond et al, 1996; Diamond et al, 1999; Krugers et al, 1997; Kirschbaum et al, 1996; Lupien et al, 1997). The hippocampus and is an area of the brain involved in memory function in humans and rats (Kirschbaum et al, 1996; Lupien et al, 1997) and is highly susceptible to stress (Diamond et al, 1990). Research has indicated that a number of stressors such as exposure to a predator (Diamond et al, 1999) can lead to stress effects. Recently efforts have been made to counteract the effects of stress on brain function and related behavioral performance. The antidepressant drug tianeptine has been used in this setting. Little is known about tianeptine's role in blocking stress effects on behavior and memory performance with regard to interactions with stress hormones, such as corticosterone. Here a set of experiments delineates the role of corticosterone and its link to stress effects on memory as well as an investigation into the actions of tianeptine and ADX in the blockade of stress effects on memory. First, I examined the effects of tianeptine on multi-day RAWM working memory training and a novel one-day learning and memory training task. Second, the effects of propranolol, an anti-anxiety medication, were tested with regard to the alleviation of stress effects on memory, allowing for a comparison between two anti-anxiety drugs, tianeptine and propranolol. Third, adrenalectomy (ADX) and the resultant depletion of adrenal hormones were examined in connection with learning and memory in the one-day learning task. Fourth, the effects and interactions of tianeptine and ADX were examined to see if tianeptine can exert its effects in the absence of adrenal hormones. Tianeptine blocked stress-induced memory errors in two different tasks and under ADX conditions. All effects were independent of corticosterone levels. In contrast, propranolol was ineffective in blocking stress-induced memory changes. The current data may prove useful in the development of antidepressant drugs and further the study of the mechanisms by which stress affects memory.