Dans cette thèse nous étudions les algorithmes d'apprentissage automatique pour la reconnaissance visuelle. Un accent particulier est mis sur l'apprentissage automatique de représentations, c.à.d. l'apprentissage automatique d'extracteurs de caractéristiques; nous insistons également sur l'apprentissage conjoint de ces dernières avec le modèle de prédiction des problèmes traités, tels que la reconnaissance d'objets, la reconnaissance d'activités humaines, ou la segmentation d'objets. Dans ce contexte, nous proposons plusieurs contributions : Une première contribution concerne les modèles de type bags of words (BoW), où le dictionnaire est classiquement appris de manière non supervisée et de manière autonome. Nous proposons d'apprendre le dictionnaire de manière supervisée, c.à.d. en intégrant les étiquettes de classes issues de la base d'apprentissage. Pour cela, l'extraction de caractéristiques et la prédiction de la classe sont formulées en un seul modèle global de type réseau de neurones (end-to-end training). Deux algorithmes d'apprentissage différents sont proposés pour ce modèle : le premier est basé sur la retro-propagation du gradient de l'erreur, et le second procède par des mises à jour dans le diagramme de Voronoi calculé dans l'espace des caractéristiques. Une deuxième contribution concerne l'intégration d'informations géométriques dans l'apprentissage supervisé et non-supervisé. Elle se place dans le cadre d'applications nécessitant une segmentation d'un objet en un ensemble de régions avec des relations de voisinage définies a priori. Un exemple est la segmentation du corps humain en parties ou la segmentation d'objets spécifiques. Nous proposons une nouvelle approche intégrant les relations spatiales dans l'algorithme d'apprentissage du modèle de prédication. Contrairement aux méthodes existantes, les relations spatiales sont uniquement utilisées lors de la phase d'apprentissage. Les algorithmes de classification restent inchangés, ce qui permet d'obtenir une amélioration du taux de classification sans augmentation de la complexité de calcul lors de la phase de test. Nous proposons trois algorithmes différents intégrant ce principe dans trois modèles : - l'apprentissage du modèle de prédiction des forêts aléatoires, - l'apprentissage du modèle de prédiction des réseaux de neurones (et de la régression logistique), - l'apprentissage faiblement supervisé de caractéristiques visuelles à l'aide de réseaux de neurones convolutionnels. / In this thesis, we present our research on visual recognition and machine learning. Two types of visual recognition problems are investigated: action recognition and human body part segmentation problem. Our objective is to combine spatial information such as label configuration in feature space, or spatial layout of labels into an end-to-end framework to improve recognition performance. For human action recognition, we apply the bag-of-words model and reformulate it as a neural network for end-to-end learning. We propose two algorithms to make use of label configuration in feature space to optimize the codebook. One is based on classical error backpropagation. The codewords are adjusted by using gradient descent algorithm. The other is based on cluster reassignments, where the cluster labels are reassigned for all the feature vectors in a Voronoi diagram. As a result, the codebook is learned in a supervised way. We demonstrate the effectiveness of the proposed algorithms on the standard KTH human action dataset. For human body part segmentation, we treat the segmentation problem as classification problem, where a classifier acts on each pixel. Two machine learning frameworks are adopted: randomized decision forests and convolutional neural networks. We integrate a priori information on the spatial part layout in terms of pairs of labels or pairs of pixels into both frameworks in the training procedure to make the classifier more discriminative, but pixelwise classification is still performed in the testing stage. Three algorithms are proposed: (i) Spatial part layout is integrated into randomized decision forest training procedure; (ii) Spatial pre-training is proposed for the feature learning in the ConvNets; (iii) Spatial learning is proposed in the logistical regression (LR) or multilayer perceptron (MLP) for classification.

Seres humanos com danos nos corpos mamilares exibem prejuízos no funcionamento da memória episódica. O presente estudo investigou a possível participação do núcleo mamilar medial (MM), estrutura densamente inervada por fibras imunorreativas ao hormônio concentrador de melanina (MCH) potencialmente envolvida no comportamento alimentar, em processos de evocação de memórias sobre locais em que alimento foi previamente encontrado, em ratos. Investigou-se ainda seu envolvimento na memória espacial motivada aversivamente, por meio do labirinto aquático de Morris, e em processos de ansiedade, por meio do labirinto em cruz elevado. Descreve-se uma tarefa comportamental envolvendo busca alimentar orientada por memória espacial em ratos, funcionalmente análoga ao teste de aprendizagem de pares associados, utilizado para avaliar memória episódica em seres humanos. Ratos treinados a escavar numa arena para obter alimento foram expostos ao pareamento de um determinado sabor, e.g., amendoim, com um local da arena e, numa segunda tentativa, ao pareamento de um outro sabor, e.g., avelã, com um local diferente da mesma arena. Antes de uma terceira tentativa, ainda na caixa de espera, os animais receberam uma amostra de amendoim ou avelã. Então, inseridos na arena para uma terceira tentativa, deveriam procurar e escavar por alimento enterrado no local sinalizado pela amostra. As associações sabor-local, a seqüência das associações, os locais do pareamento e a amostra oferecida antes da terceira tentativa, variaram quase aleatoriamente a cada dia. O treinamento dos animais para a aquisição da tarefa foi prolongado. O número de visitas e o tempo de permanência no local da arena sinalizado pela amostra foram significativamente maiores em relação aos parâmetros correspondentes ao local pareado com alimento mas não sinalizado pela amostra, quando o intervalo de tempo entre a 2ª e 3ª tentativas foi 60 ou 120 minutos; Testes de Avaliação sem oferta de alimento na arena na 3ª tentativa, permitiram excluir a hipótese de que os animais desempenharam essa tarefa com base em estímulos olfativos, levando à conclusão que o fizeram com base na memória das associações sabor-local. Ratos com danos seletivos do MM, induzidos pela aplicação tópica, micro-iontoforética, de N-Metil-D-Aspartato (NMDA), testados na tarefa de pares associados sabor-local, não exibiram prejuízo de desempenho em relação aos animais controle. Análises adicionais revelaram que houve uma redução da busca por alimento em ambos os grupos (lesado e controle) nos testes pós-operatórios, em relação ao seu próprio desempenho nos testes pré-operatórios; porém, essa redução foi mais acentuada em relação ao local sinalizado pela amostra, particularmente nos animais do grupo lesado. Estes resultados, embora não conclusivos, podem representar indícios em favor da idéia de que o MM estaria envolvido numa circuitaria relacionada à memória espacial voltada à busca alimentar. Mais estudos seriam necessários para avaliar essa possibilidade. Além disso, esses animais não exibiram qualquer prejuízo no teste de memória de referência espacial no labirinto aquático de Morris e no paradigma do teste e re-teste no labirinto em cruz elevado, sugerindo que esse tipo de lesão não leva a deficiência no funcionamento da memória espacial motivada por estimulação aversiva, na memória aversiva por espaços abertos ou nos mecanismos responsáveis pela ansiedade. / Human beings with damage to the mammillary bodies exhibit episodic memory disruption. This study investigated the potential participation of the medial mammillary nucleus (MM), a structure that receives a high density of melaninconcentrating hormone immunoreactive fibers (MCH) potentially involved in food intake, in processes of retrieval of memory about places where food was previously found, in rats. In addition, we investigated the possible involvement of this brain structure in both aversively motivated spatial memory, using the Morris´ water maze, and in anxiety, using the elevated plus maze. A behavioral task involving search for food guided by spatial memory, in rats, functionally analogous to the paired-associate learning task used for studying episodic memory in humans, was described. Rats trained to dig for food in an arena were exposed both to the pairing of a taste, peanut, with a specific place in the arena and, in a second trial, to the pairing of another taste, hazelnut, with another place in the arena. Immediately before a third trial, while still in the waiting box, the animals received a sample of either peanut or hazelnut. Then, they were inserted into the arena for a third trial and could search and dig for food buried in the place previously associated with the sample. The taste-place associations, sequence of associations, pairing places and sample offered before the third trial, were all varied quasi-randomly every day. The animals took extensive training to learn this task. It was found that the number of visits and the time spent in the place indicated by the sample were significantly greater as compared to the corresponding parameters of the place previously associated with a taste but not indicated by the sample, when the time interval between the second and the third trials was both 60 and 120 min; probe tests with no food in the arena in the third trial allowed to exclude the hypothesis that the rats were performing the task relying on olfactory cues, leading to the conclusion that the animals performed the task relying on the memory for the associations tastefood. Rats with selective damage to the MM, induced by topic, microiontophoretic, injections of N-Methyl-D-Aspartate (NMDA), tested in the taste-place pairedassociate learning task, did not exhibit any disruption of performance relative to the controls. Additional analyses revealed that both MM and control subjects exhibited a reduction of food search during the post-operatory tests as compared to their own performance in pre-operatory tests; however, this reduction was stronger towards the place indicated by the sample, particularly for the animals subjected to MM damage. Even though not conclusive, these results may suggest that the MM might be involved in search for food guided by spatial memory. More studies would be required to evaluate this possibility. Further, these rats did not exhibit any disruption of performance in the reference memory test in the water maze and in the test/re-test elevated plus maze, suggesting that this type of damage does not disrupt aversively motivated spatial memory, open-space aversive memory and anxiety.

Le syndrome de Coffin Lowry est une forme syndromique rare de déficience intellectuelle liée au chromosome X. Ce syndrome est dû à des mutations du gène Rsk2 codant la protéine kinase RSK2 dans la voie de signalisation des MAPK/ERK. La caractérisation phénotypique du modèle murin Rsk2-KO a principalement mis en évidence un retard d’acquisition ainsi qu’un déficit de mémoire spatiale à long terme, suggérant une altération des fonctions hippocampiques. Nous avons montré que les souris Rsk2-KO présentent également des déficits dans une forme d’apprentissage et de mémoire mettant en jeu la fonction de séparation de patterns dépendante du gyrus denté. Plusieurs études montrent que la genèse de nouveaux neurones dans le gyrus denté chez l’adulte constitue une forme de plasticité jouant un rôle important dans l’apprentissage et la mémoire dépendante de l’hippocampe, en particulier dans les tâches spatiales et de séparation de patterns. En raison des déficits observés chez les souris Rsk2-KO, nous nous sommes intéressés à la neurogenèse adulte chez ce modèle murin. Aucune différence de prolifération, de survie ou de maturation n’a été observée dans le gyrus denté des souris Rsk2-KO à l’état basal, ni après une tâche de séparation de patterns. Cependant, nous avons observé un déficit de survie des nouvelles cellules chez les souris Rsk2-KO après apprentissage dans la piscine de Morris. La littérature montre que l’enrichissement environnemental a des effets bénéfiques sur les performances cognitives des rongeurs et est notamment capable d’augmenter la neurogenèse adulte hippocampique. Nous avons donc analysé les effets de l’enrichissement sur les performances comportementales et la neurogenèse adulte des souris Rsk2-KO. Nos résultats montrent qu’un protocole d’enrichissement environnemental de 3 heures par jours durant 24 jours est capable de compenser ou d’améliorer les performances des souris Rsk2-KO dans les tâches de mémoire spatiale et de séparation de patterns et aussi d’augmenter la neurogenèse hippocampique adulte. / The Coffin-Lowry Syndrome is a rare syndromic form of X-linked intellectual disability. This syndrome is caused by mutations of the Rsk2 gene that encodes a protein kinase, RSK2, in the MAPK/ERK signaling pathway. Characterization of the behavioural phenotype of Rsk2-KO mice mainly showed that they display delayed acquisition and long-term deficits in a spatial reference memory task, suggesting an alteration in hippocampal function. Here, we show that Rsk2-KO mice are also deficient in a learning and memory task that involves dentate gyrus-dependent pattern separation function. Several studies showed the formation of new neurons in the adult dentate gyrus by neurogenesis is a form of plasticity that plays a significant role in hippocampal-dependent learning and memory, in particular for spatial learning and memory and pattern separation. As these functions are altered in Rsk2-KO mice, we studied hippocampal adult neurogenesis in these mice. No difference in proliferation, survival and maturation of newborn neurons was found in the dentate gyrus of the mutant mice in basal conditions, nor after a pattern separation task. However, we found a deficit in the survival of newborn cells in Rsk2-KO mice submitted to spatial learning and memory in the Morris water maze task. According to several studies, environmental enrichment in rodents has beneficial effects on cognitive performance and is associated with an enhancement of adult hippocampal neurogenesis. Thus, we assessed the potential effect of environmental enrichment on spatial learning and memory performance and adult hippocampal neurogenesis in Rsk2-KO mice. Our results show that an environmental enrichment protocol of 3h per day during 24 days can rescue or ameliorate spatial learning and memory performance and pattern separation function in Rsk2-KO mice and increase adult hippocampal neurogenesis.

Seres humanos com danos nos corpos mamilares exibem prejuízos no funcionamento da memória episódica. O presente estudo investigou a possível participação do núcleo mamilar medial (MM), estrutura densamente inervada por fibras imunorreativas ao hormônio concentrador de melanina (MCH) potencialmente envolvida no comportamento alimentar, em processos de evocação de memórias sobre locais em que alimento foi previamente encontrado, em ratos. Investigou-se ainda seu envolvimento na memória espacial motivada aversivamente, por meio do labirinto aquático de Morris, e em processos de ansiedade, por meio do labirinto em cruz elevado. Descreve-se uma tarefa comportamental envolvendo busca alimentar orientada por memória espacial em ratos, funcionalmente análoga ao teste de aprendizagem de pares associados, utilizado para avaliar memória episódica em seres humanos. Ratos treinados a escavar numa arena para obter alimento foram expostos ao pareamento de um determinado sabor, e.g., amendoim, com um local da arena e, numa segunda tentativa, ao pareamento de um outro sabor, e.g., avelã, com um local diferente da mesma arena. Antes de uma terceira tentativa, ainda na caixa de espera, os animais receberam uma amostra de amendoim ou avelã. Então, inseridos na arena para uma terceira tentativa, deveriam procurar e escavar por alimento enterrado no local sinalizado pela amostra. As associações sabor-local, a seqüência das associações, os locais do pareamento e a amostra oferecida antes da terceira tentativa, variaram quase aleatoriamente a cada dia. O treinamento dos animais para a aquisição da tarefa foi prolongado. O número de visitas e o tempo de permanência no local da arena sinalizado pela amostra foram significativamente maiores em relação aos parâmetros correspondentes ao local pareado com alimento mas não sinalizado pela amostra, quando o intervalo de tempo entre a 2ª e 3ª tentativas foi 60 ou 120 minutos; Testes de Avaliação sem oferta de alimento na arena na 3ª tentativa, permitiram excluir a hipótese de que os animais desempenharam essa tarefa com base em estímulos olfativos, levando à conclusão que o fizeram com base na memória das associações sabor-local. Ratos com danos seletivos do MM, induzidos pela aplicação tópica, micro-iontoforética, de N-Metil-D-Aspartato (NMDA), testados na tarefa de pares associados sabor-local, não exibiram prejuízo de desempenho em relação aos animais controle. Análises adicionais revelaram que houve uma redução da busca por alimento em ambos os grupos (lesado e controle) nos testes pós-operatórios, em relação ao seu próprio desempenho nos testes pré-operatórios; porém, essa redução foi mais acentuada em relação ao local sinalizado pela amostra, particularmente nos animais do grupo lesado. Estes resultados, embora não conclusivos, podem representar indícios em favor da idéia de que o MM estaria envolvido numa circuitaria relacionada à memória espacial voltada à busca alimentar. Mais estudos seriam necessários para avaliar essa possibilidade. Além disso, esses animais não exibiram qualquer prejuízo no teste de memória de referência espacial no labirinto aquático de Morris e no paradigma do teste e re-teste no labirinto em cruz elevado, sugerindo que esse tipo de lesão não leva a deficiência no funcionamento da memória espacial motivada por estimulação aversiva, na memória aversiva por espaços abertos ou nos mecanismos responsáveis pela ansiedade. / Human beings with damage to the mammillary bodies exhibit episodic memory disruption. This study investigated the potential participation of the medial mammillary nucleus (MM), a structure that receives a high density of melaninconcentrating hormone immunoreactive fibers (MCH) potentially involved in food intake, in processes of retrieval of memory about places where food was previously found, in rats. In addition, we investigated the possible involvement of this brain structure in both aversively motivated spatial memory, using the Morris´ water maze, and in anxiety, using the elevated plus maze. A behavioral task involving search for food guided by spatial memory, in rats, functionally analogous to the paired-associate learning task used for studying episodic memory in humans, was described. Rats trained to dig for food in an arena were exposed both to the pairing of a taste, peanut, with a specific place in the arena and, in a second trial, to the pairing of another taste, hazelnut, with another place in the arena. Immediately before a third trial, while still in the waiting box, the animals received a sample of either peanut or hazelnut. Then, they were inserted into the arena for a third trial and could search and dig for food buried in the place previously associated with the sample. The taste-place associations, sequence of associations, pairing places and sample offered before the third trial, were all varied quasi-randomly every day. The animals took extensive training to learn this task. It was found that the number of visits and the time spent in the place indicated by the sample were significantly greater as compared to the corresponding parameters of the place previously associated with a taste but not indicated by the sample, when the time interval between the second and the third trials was both 60 and 120 min; probe tests with no food in the arena in the third trial allowed to exclude the hypothesis that the rats were performing the task relying on olfactory cues, leading to the conclusion that the animals performed the task relying on the memory for the associations tastefood. Rats with selective damage to the MM, induced by topic, microiontophoretic, injections of N-Methyl-D-Aspartate (NMDA), tested in the taste-place pairedassociate learning task, did not exhibit any disruption of performance relative to the controls. Additional analyses revealed that both MM and control subjects exhibited a reduction of food search during the post-operatory tests as compared to their own performance in pre-operatory tests; however, this reduction was stronger towards the place indicated by the sample, particularly for the animals subjected to MM damage. Even though not conclusive, these results may suggest that the MM might be involved in search for food guided by spatial memory. More studies would be required to evaluate this possibility. Further, these rats did not exhibit any disruption of performance in the reference memory test in the water maze and in the test/re-test elevated plus maze, suggesting that this type of damage does not disrupt aversively motivated spatial memory, open-space aversive memory and anxiety.

No description available.

Le but de cette thèse est premièrement d’évaluer l’effet du vieillissement sur les fonctions psychomotrices des souches de souris sélectionnées génétiquement en fonction de leur tension artérielle (TA); deuxièmement, de localiser les déterminants génétiques des phénotypes psychophysiologiques à partir de souches recombinantes congéniques (RCS). Ces travaux ont mené à la publication de 4 articles. Le premier article décrit l’évaluation des fonctions psychomotrices des souches avec une tension artérielle élevée (HBP), basse (LBP) et normale (NBP). La performance aux épreuves d’exploration, d’habiletés motrices et d’apprentissage spatial, a été mesurée sur deux cohortes âgées respectivement de 12 mois et de trois mois. Indépendamment de l’âge, les HBPs sont hyperactives dans l’open-field (OF), mais pas dans le test d’exploration de trous. Inversement, les LBP explorent moins d’espaces que les NBP et, à trois mois seulement, sont hypoactives dans l’OF. Par ailleurs, les HBPs et les LBP présentent des déficits précoces de coordination motrice et des fonctions visuo-motrices. Le second article concerne l’évaluation longitudinale de la coordination motrice, de l’anxiété et de l’apprentissage spatial des souches HBP, LBP et NBP, à l’âge de deux mois et de 12 mois. Le vieillissement accentue l’hyperactivité des HBPs dans l’OF. Par contre, l’hypoactivité des souris LBP est détectable seulement à l’âge de deux mois. Indépendamment de l’âge, les souris HBP et LBP montrent une perception réduite du danger dans l’épreuve d’anxiété et des dysfonctions visuo-motrices au labyrinthe aquatique. Enfin, des déficits précoces de coordination motrice se manifestent seulement chez les HBPs. Il reste à déterminer si les déficits observés sont liés à des déterminants génétiques indépendants ou secondaires aux altérations de la tension artérielle. Le troisième article présente la comparaison entre les souches consanguines A/J et C57Bl/6J (B6) aux épreuves de l’OF, de la planche à trous, du labyrinthe aquatique et du cintre (coordination motrice). Les B6 explore d’avantage l’OF et la planche à trous. Les B6 sont moins rapides sur le cintre, mais supérieurs aux A/J dans le labyrinthe aquatique, avec une plate-forme invisible ou visible. Ces résultats démontrent l’implication de déterminants génétiques. Cette thèse se termine par un quatrième article sur la localisation des déterminants génétiques de la susceptibilité au stress dans les RCS, dérivées de A/J et B6, et présentant un agencement spécifique de 12.5% du génome. La réactivité émotionnelle est évaluée dans l’OF et le plus-maze; la réponse de stress est mesurée par radio télémétrie de la température interne pendant le stress d’immobilisation (SI) sous diète régulière et riche en sel; l’excrétion des électrolytes urinaires est dosée après 24 heures de diète salée. Les loci les plus significatifs sont situés dans les régions suivantes: de l’émotionalité dans l’OF (Emo1) sur le chr. 1 (LOD=4.6) correspondant à la région homologue impliquée dans la cohorte d’hypertension familiale du Saguenay; de la dopa décarboxylase (ddc) sur le chr. 11 pour l’émergence du plus-maze (LOD=4.7); de la protéine liant l’endotoxine (lbp) sur le chr. 2 pour l’hypothermie initiale en réponse au SI (LOD=4); et de HSP90 sur le chr. 12 pour l’excrétion de Ca++ (LOD=4.6). Des banques de données sont ensuite interrogées pour recenser les polymorphismes des régions régulatrices ou codantes des gènes candidats chez les souches ancestrales A/J et B6, dont les séquences sont disponibles pour le génome entier. Des utilitaires web permettent de dévoiler les changements dans la structure secondaire de l’ARNm, l’interférence avec des microARN ou avec d’autres motifs de liaison. Plusieurs SNPs fonctionnels ont été identifiés pour le QTL du chr. 1, particulièrement dans les éléments de régulation; ceux-ci impliquant des gènes reliés avec les réponses inflammatoire/immunitaire ou avec le système cardiovasculaire. La quantification par la PCR confirme une régulation à la baisse d’atp1a2 dans le cœur et le cerveau des souches susceptibles à l’anxiété. Ces résultats confirment l’intrication des altérations de la susceptibilité au stress et de la régulation de la TA. / Our studies in this thesis, which led to 4 publications, are divided in two parts. The first part describes the neuropsychological effects of aging in strains of mice genetically selected for high (HBP), low (LBP) or normal blood pressure (NBP). The second part focuses on the genetic determinants of these neuropsychological phenotypes in recombinant congenic strains (RCS) of mice. The first manuscript compares HBP or LBP mice to normotensive controls in tests of exploration, motor coordination, and spatial learning at two age levels: 3 and 12 months. At either age, HBPs were hyperactive in an open field (OF) but not in terms of hole-poking responses. On the contrary, LBPs were hypoactive in the OF and in the hole-board, with the effect on the former measure being limited to the younger cohort. In either cohort, HBP and LBP mice were deficient in subtle aspects of motor coordination, and visuomotor function. These strains may serve as experimental models for the evaluation of beneficial early antihypertensive or antihypotensive treatments on brain function. The second study uses a longitudinal design to compare either HBP or LBP mice to normotensive controls at 2 and 12 months of age for motor coordination, anxiety, and spatial learning. Hyperactivity of HBPs in the OF increased with aging; whereas LBP mice were hypoactive only at 2 months of age. At both age, HBP and LBP mice displayed reduced levels of anxiety in the elevated plus-maze (EPM), abnormal coordination and visuomotor guidance. It remains to determine if these strain-, age-, and test-specific abnormalities are genetically related or secondary to uncontrolled hypertension or hypotension. The following article compares the C57BL/6J (B6) to the A/J inbred mouse strain in exploration of the OF and the hole-board, in the coat-hanger coordination test, and in spatial learning of a water maze. B6 mice displayed a higher number of segment crossings in the open-field and of hole-poking responses than A/J mice. By contrast to their hypo activity, A/J strain were faster in the coat-hanger motor test, but deficient in the submerged but also in the visible platform version of the water maze. These results indicate the considerable potential of genetic models derived from B6 and A/J mice for discerning the determinants of several behavioural phenotypes. In the last manuscript, the genomic loci bearing stress-related phenotypes were dissected by genome wide analysis of linkage in the recombinant congenic strains (RCS), resulting from a cross of B6 and A/J progenitors, each strain bearing 12.5% of specific parts of one progenitor on the background of the other. Adult male mice from 14 A/J and 22 B6 background lines were evaluated for emotional reactivity in the OF and the EPM. Core temperature was monitored by radio-telemetry during immobilization (IM), under standard and salt-enriched diets. In addition, urinary electrolytes were measured. The highest LOD scores strengthen the evidence for a previously reported locus for emotionality in the open-field on Chr 1 (LOD=4.6), in the Ddc region encoding dopa decarboxylase, on Chr 11 in the EPM (LOD=4.7), near Lbp (lipopolysaccharide binding protein), on Chr 2 for initial hypothermia during IM (LOD=4), as well as in the region of Hspca, encoding heat shock protein 1 alpha (48.0 cM) on Chr 12 for Ca++ excretion after a 24 hr-salt load (LOD=4.6). RCS stress QTL overlapped with several candidate loci for cardiovascular disease. In silico evidence of functional polymorphisms by comparative sequence analysis of progenitor strains assisted to ascertain this convergence, then further tested using quantitative PCR for releant genes mRNA. The anxious BcA70 strain showed down regulation of the Atp1a2 gene expression in the heart (P < 0.001) and brain (P < 0.05) compared to its parental B6 strain, compatible with the enhanced emotionality described in knock out animals for this gene, also involved in the salt-sensitive component of hypertension. Functional polymorphisms in regulatory elements of candidate genes of the cardiovascular / inflammatory / immune systems support the hypothesis of genetically-altered environmental susceptibility in cardiovascular disease development.

Cette recherche comparative – terrains français et polonais – a été menée à Lyon et à Cracovie. L’objet du travail concerne le tourisme social destiné aux enfants, et ce dans le cadre des classes de découvertes en France et écoles vertes en Pologne, ces voyages scolaires constituant à la fois un temps de vie collective et un temps d’éducation. Le cadre scolaire de la recherche concerne l’école primaire. Le travail s’inscrit dans la vaste problématique de l’acquisition de savoirs géographiques. Le sujet de la recherche mobilise le champ du tourisme dans la mesure où les pratiques qui en relèvent constituent des moyens et modalités d’appropriation de l’espace que les individus mettent en œuvre dans l’élaboration de leurs représentations spatiales. A travers ses expériences, l’individu construit un modèle intériorisé de son environnement. Cette recherche questionne les modalités de l’apprentissage spatial chez les jeunes enfants, questionnement dont l’objectif essentiel est de valider l’hypothèse selon laquelle l’expérience du voyage scolaire est susceptible de modifier leurs représentations spatiales. Pour vérifier si tous les enfants ont des chances égales pour se construire leur propre capital spatial, la recherche examine aussi l’influence sur leurs représentations de la mobilité spatiale antérieure. Afin d’approcher les représentations spatiales des écoliers, sont mobilisés à la fois les données discursives (questionnaires) et graphiques (cartes mentales). L’enquête, menée auprès de 192 élèves de Lyon et de Cracovie, explore donc finalement l’interface entre deux champs scientifiques, ceux de la géographie et ceux de la psychologie, en s’appuyant sur le postulat de représentations spatiales intrinsèques au paradigme qui conçoit l’espace comme un construit. Le travail s’inscrit dans l’approche d’une géographie mettant l’espace vécu au centre de ses préoccupations. / This comparative research - in France and Poland - was conducted in Lyon and Cracow. The purpose of the work is social tourism for children, more specifically the “discovery classes” in France and the “green schools” in Poland, the school trips constituting both a life experience and a time of collective education. The object of the research is primary schools.The work is part of a broader problem of the acquisition of geographical knowledge. The subject of the research mobilizes the field of tourism since the practices that stem from it are ways and means of appropriation of space that individuals implement in the construction of representations of space. Through experience, the individual constructs an interior model of his or her environment. This research project studies the modalities of spatial learning. The objective is to understand if the experience of school trips modifies children’s spatial representations. To ensure that all children have an equal opportunity to build their own spatial capital, research also examines the influence on their representations of previous spatial mobility. In order to examine the children’s spatial representations, both discursive (questionnaires) and graphics data (mental maps) are mobilized. The research, carried out with 192 pupils in Lyon and Cracow, is related to two disciplines, geography and psychology. It is supported by the postulate of spatial representations within the framework of the paradigm of spatial production. The work is part of an approach to geography, in which the “espace vécu” is the central preoccupation.

Indiana University-Purdue University Indianapolis (IUPUI) / Despite warnings about the dangers of drinking during pregnancy, little progress has been made in reducing alcohol drinking among women of childbearing age. Even after the recognition of pregnancy, 15% of women continue to drink, 3% of which admit to binge drinking. Because we cannot stop women from drinking during pregnancy, and many children with fetal alcohol spectrum disorders (FASD) are adopted, there is a significant need to develop postnatal interventions that can improve the long-term outcome of children adversely affected by prenatal alcohol exposure. This thesis aims to evaluate one promising new treatment in the rehabilitation or rescue of specific learning deficits long after the damage has occurred. The treatment evaluated herein (40µg D-NAP + 40µg D-SAL) has long been used in the prevention of the detrimental effects of long-term and binge-like alcohol exposures in rodent models of fetal alcohol syndrome and FASD. Until recently this peptide treatment had only been shown to be effective in preventing some of the consequences of alcohol exposure when administered concurrently with the prenatal alcohol exposure. A recent report by Incerti and colleagues (2010c), however, reported that these peptides could completely reverse a profound spatial learning deficit induced by one episode of a heavy binge-like alcohol exposure (5.9g.kg in a single intraperitoneal injection) on gestational day 8 (G8) in C57BL/6 mice. In that report, the peptide treatment was administered starting in late adolescence, beginning three days prior to and throughout water maze training, and the profound deficits in their alcohol-placebo group were completely eliminated in the alcohol-peptide group. There are currently no FDA-approved treatments for FASD. An effective treatment for the cognitive and behavioral dysfunctions suffered by the 1% of people born today could potentially improve the lives of millions of children and adults. The first aim of this thesis was to determine whether the peptide treatment could reverse the significant spatial learning deficits we have demonstrated in adult C57BL/6 mice given high-dose binge-like alcohol exposure (2.5 g/kg in each of two intraperitoneal injections separated by two hours) on postnatal day (P)7. When administered three days prior to and throughout water maze testing (P67-76), the peptide treatment had no effect on spatial learning. The second aim sought to determine whether the same peptide treatment could reverse water maze spatial learning deficits in G8 binge-like exposure models, as reported by Incerti et al. (2010c). For this analysis, the first study used a different binge-like alcohol exposure model that is more commonly used than that employed by the Incerti et al. (2010c) study, namely administration of 2.8g/kg in each of two intraperitoneal injections separated by four hours (Sulik et al., 1981). This model has been shown to produce high peak blood alcohol concentrations and neuroanatomical aberrations in the hippocampal formation and septal regions (Parnell et al., 2009), which have been implicated in learning and memory. Surprisingly, this G8 binge-like alcohol exposure failed to produce a spatial learning deficit, undermining the usefulness of this model in evaluating the peptide effects. In direct contrast to the outcomes of Incerti et al. (2010c), the G8 Webster alcohol exposure was also unable to produce any deficits in acquisition of spatial learning in the Morris water maze. Surprisingly, neither of the heavy binge-like alcohol exposures on G8 were able to produce spatial learning deficits in the Morris water maze. The binge-like alcohol exposure on P7 did yield the expected spatial learning deficit, but the peptide treatment was unsuccessful in recovering water maze learning. These findings fail to support oral administration of 40µg D-NAP and 40 µg D-SAL as a potential therapy for postnatal alcohol-induced spatial learning deficits in adult mice.