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

Obesity is a growing disorder in the United States, affecting over 60% of the population. We previously defined sympathetic nervous system (SNS) outflow from brain to white adipose tissue (WAT) using a viral transneuronal tract tracer. SNS innervation of WAT is the principle initiator of lipolysis, whereas decreases in sympathetic drive promote lipid accumulation. Which of the many origins of SNS outflow from brain to WAT results in SNS-mediated changes in lipid mobilization (increases in drive) or accumulation (decrease in drive) is unknown. Previous research indicates that sympathetic denervation blocks lipid mobilization; thus, rostral sites in the neuroaxis connected to WAT via the SNS may promote WAT lipid mobilization. The hypothalamic paraventricular nucleus (PVN) may play a role via its descending projections to the intermediolateral horn of the spinal cord. Therefore, the consequences of PVN lesions (PVNx) on WAT mobilization or accumulation were tested. PVNx resulted in increased lipid accumulation, indicated by increases in retroperitoneal (RWAT) , epididymal (EWAT) , and inguinal WAT (IWAT) pad masses, in fed hamsters, but PVNx did not block fasting (56 h)-induced lipid mobilization. Because adrenal medullary catecholamines, especially epinephrine, also play a minor role in lipid mobilization, we tested the contribution of catecholamine release on lipid mobilization through adrenal demedullation (ADMEDx), with and without PVNx, and found fastinginduced lipid mobilization was not blocked. There was, however, a suggestion that distal denervation of IWAT, with and without ADMEDx, partially blocked lipid mobilization. In addition, evidence suggests SNS also may be an important controller of fat cell proliferation. Surgical denervation of WAT triggers increases in fat cell number (FCN), but have not determined if this FCN increase is due to preadipocyte proliferation or differentiation of preadipocytes into mature fat cells. We also have not demonstrated what role sensory innervation may have in regulating white adipocyte proliferation. Therefore, the role of WAT sympathetic or sensory innervation on adipocyte proliferation was tested. The SNS but not sensory denervation triggered bona fide proliferation as indicated by bromodeoxyuridine plus AD3, a specific adipocyte membrane protein, colabeling. These and previous data suggest that the SNS plays a role in regulating adiposity.

No description available.

Tesis por compendio / La exposición crónica a sustancias de abuso lleva a cambios adaptativos en el cerebro que implican alteraciones en la expresión génica. Se ha propuesto que los factores de transcripción CREB y deltaFosB serían dianas moleculares para la regulación de la plasticidad, la cual lleva a la adicción.En este trabajo hemos estudiado los cambios en la activación de CREB, en PVN y NTS, y las quinasas que mediarían su activación durante la dependencia y síndrome de abstinencia a morfina, así como la respuesta del eje HHA durante dicho síndrome. También se investigó la posibilidad de que la activación de CREB y su coactivador transcripcional TORC1 dependan de la activación de receptores adrenérgicos. Además se evaluaron las posibles modificaciones en la expresión de FosB/deltaFosB en diferentes áreas cerebrales implicadas en la adicción, así como los cambios neuroendocrinos/neuroquímicos responsables de las alteraciones metabólicas observadas durante el tratamiento crónico con morfina. / Chronic exposure to opioids and other abused drugs results in adaptive changes in the brain involving alterations in gene expression. It is proposed that the transcription factors CREB and deltaFosB be molecular targets for the regulation of plasticity, which leads to addiction.In this work we studied changes in activation of the cAMP-response element binding protein (CREB) in PVN and NTS and the kinases that may mediate this activation during dependence and morphine withdrawal and the HPA axis response after naloxone-induced morphine withdrawal. We also investigated the possibility that the activation of CREB and the transcriptional coactivator of CREB, TORC1, arises from the activation of adrenergic receptors. We also evaluated the possible modifications in FosB/deltaFosB expression in several brain areas involved in addiction and neuroendocrine/neurochemical changes that are responsible for the metabolic alterations seen during chronic morphine treatment.

Une cascade de facteurs de transcription composée de SIM1, ARNT2, OTP, BRN2 et SIM2 est requise pour la différenciation des cinq types cellulaires qui peuplent le noyau paraventriculaire (PVN) de l’hypothalamus, un régulateur critique de plusieurs processus physiologiques essentiels à la survie. De plus, l’haploinsuffisance de Sim1 est aussi une cause d’hyperphagie isolée chez la souris et chez l’homme. Nous désirons disséquer le programme développemental du PVN, via une approche intégrative, afin d’identifier de nouveaux gènes qui ont le potentiel de réguler l’homéostasie chez l’individu adulte. Premièrement, nous avons utilisé une approche incluant l’analyse du transcriptome du PVN à différents stades du développement de la souris pour identifier de tels gènes. Nous avons comparé les transcriptomes de l’hypothalamus antérieur chez des embryons de souris Sim1+/+ et Sim1-/- à E12.5 issus de la même portée. De cette manière, nous avons identifié 56 gènes agissant en aval de Sim1 dont 5 facteurs de transcription - Irx3, Sax1, Rxrg, Ror et Neurod6. Nous avons également proposé un modèle de développement à deux couches de l’hypothalamus antérieur. Selon ce modèle, les gènes qui occupent un domaine médial dans la zone du manteau caractérisent des cellules qui peupleront le PVN alors que les gènes qui ont une expression latérale identifient des cellules qui donneront plus tard naissance aux structures ventrolatérales de l’hypothalamus. Nous avons aussi démontré que Sim1 est impliqué à la fois dans la différenciation, la migration et la prolifération des neurones qui peuplent le PVN tout comme Otp. Nous avons également isolé par microdissection au laser le PVN et l’hypothalamus médiobasal chez des souris de type sauvage à E14.5 pour en comparer les transcriptomes. Ceci nous a permis d’identifier 34 facteurs de transcription spécifiques au PVN et 76 facteurs spécifiques à l’hypothalamus médiobasal. Ces gènes représentent des régulateurs potentiels du développement hypothalamique. Deuxièmement, nous avons identifié 3 blocs de séquences au sein de la région 5’ d’Otp qui sont conservés chez l’homme, la souris et le poisson. Nous avons construit un transgène qui est composé d’un fragment de 7 kb contenant ces blocs de séquences et d’un gène rapporteur. L’analyse de 4 lignées de souris a montré que ce transgène est uniquement exprimé dans le PVN en développement. Nous avons généré un deuxième transgène dans lequel le fragment de 7 kb est inséré en amont de l’ADNc de Brn2 ou Sim1 et de Gfp. Nous avons obtenu quatre lignées de souris dans lesquels le profil d’expression de Brn2 et de Gfp reproduit celui d’Otp. Nous étudierons le développement du PVN et la prise alimentaire chez ces souris. En parallèle, nous croisons ces lignées avec les souris déficientes en Sim1 pour déterminer si l’expression de Brn2 permet le développement des cellules du PVN en absence de Sim1. En résumé, nous avons généré le premier transgène qui est exprimé spécifiquement dans le PVN. Ce transgène constitue un outil critique pour la dissection du programme développemental de l’hypothalamus. Troisièmement, nous avons caractérisé le développement de l’hypothalamus antérieur chez l’embryon de poulet qui représente un modèle intéressant pour réaliser des études de perte et de gain de fonction au cours du développement de cette structure. Il faut souligner que le modèle de développement à deux couches de l’hypothalamus antérieur semble être conservé chez l’embryon de poulet où il est aussi possible de classer les gènes selon leur profil d’expression médio-latéral et le devenir des régions qu’ils définissent. Finalement, nous croyons que cette approche intégrative nous permettra d’identifier et de caractériser des régulateurs du développement du PVN qui pourront potentiellement être associés à des pathologies chez l’adulte telles que l’obésité ou l’hypertension. / A cascade of transcription factors composed of SIM1, ARNT2, OTP, BRN2 and SIM2 is required for the differentiation of the five major cell types populating the paraventricular nucleus (PVN) of the hypothalamus, a critical integrator of several homeostatic processes that are required for the survival of vertebrates. Haploinsufficency of Sim1 also causes isolated hyperphagia in mice and humans. The goal of our study is to dissect the developmental program of the PVN using an integrative approach in order to identify new genes that could potentially be implicated in the regulation of homeostasis in adults. First, we used a comparative approach to analyse the PVN transcriptome at different developmental stages in mice embryos in order to identity new genes implicated in PVN development. We compared gene expression in the anterior hypothalamus of E12.5 Sim1-/- and Sim1+/+ littermate embryos using a microarray approach. We identified 56 genes acting downstream of Sim1 including 5 transcription factors - Irx3, Sax1, Rxrg, Ror and Neurod6. We proposed a model for the development of the anterior hypothalamus. In this model, the genes expressed in the medial domain of the mantle layer characterise cells that will form the PVN and genes expressed in the lateral domain identify cells that will give rise to ventrolateral areas of the hypothalamus. We also showed that Sim1, like Otp, is implicated in the differentiation, migration and proliferation of the neurons populating the PVN. Furthermore, we have isolated by laser captured microdissection the PVN and ARC nucleus in wild type mice at E14.5 and compared their transcriptomes. This technique allowed us to identity 34 transcription factors specific to the PVN and 76 factors specific to the ARC. These genes represent potential regulators of hypothalamic development. Second, we identified 3 blocks of sequence in the 5’ region of Otp that are conserved between human, mouse and fish. We constructed a transgene which included a 7 Kb fragment encompassing these sequences followed by a reporter gene. The analysis of 4 mice strains showed that this transgene is specifically expressed in the prospective PVN. We have generated a second transgene in which the 7 Kb fragment is located upstream of the cDNA encoding Brn2 or Sim1 and Gfp. We obtained 4 mice strains in which the Brn2 and Gfp expression pattern is similar to the Otp expression pattern. These mice will be used to study PVN development and food intake. Also, to determine if Brn2 expression only – without Sim1 gene expression - allows the development of PVN cells, we are presently crossing these mice with Sim1 deficient mice. In conclusion, we have generated the first transgene that is specifically expressed in the PVN. This transgene constitutes a critical tool for dissecting the developmental program of the hypothalamus. Third, we have characterised the development of the anterior hypothalamus of chick embryos which represent an interesting model for loss and gain of function experiments during the development of this brain region. Interestingly, our proposed model for the development of the anterior hypothalamus seems to be conserved in chick embryos. As a matter of fact, it is possible to classify genes according to their medio-lateral expression patterns and the outcome of the regions that they are defining. Finally, we believe that this integrative approach will allow us to identify and characterize factors implicated in PVN development. From a clinical point of view, these factors could potentially be associated with pathologies such as obesity or arterial hypertension.

No description available.

Data mining has become a commonly used method for the analysis of organisational data, for purposes of summarizing data in useful ways and identifying non-trivial patterns and relationships in the data. Given the large volumes of data that are collected by business, government, non-government and scientific research organizations, a major challenge for data mining researchers and practitioners is how to select relevant data for analysis in sufficient quantities, in order to meet the objectives of a data mining task. This thesis addresses the problem of dataset selection for predictive data mining. Dataset selection was studied in the context of aggregate modeling for classification. The central argument of this thesis is that, for predictive data mining, it is possible to systematically select many dataset samples and employ different approaches (different from current practice) to feature selection, training dataset selection, and model construction. When a large amount of information in a large dataset is utilised in the modeling process, the resulting models will have a high level of predictive performance and should be more reliable. Aggregate classification models, also known as ensemble classifiers, have been shown to provide a high level of predictive accuracy on small datasets. Such models are known to achieve a reduction in the bias and variance components of the prediction error of a model. The research for this thesis was aimed at the design of aggregate models and the selection of training datasets from large amounts of available data. The objectives for the model design and dataset selection were to reduce the bias and variance components of the prediction error for the aggregate models. Design science research was adopted as the paradigm for the research. Large datasets obtained from the UCI KDD Archive were used in the experiments. Two classification algorithms: See5 for classification tree modeling and K-Nearest Neighbour, were used in the experiments. The two methods of aggregate modeling that were studied are One-Vs-All (OVA) and positive-Vs-negative (pVn) modeling. While OVA is an existing method that has been used for small datasets, pVn is a new method of aggregate modeling, proposed in this thesis. Methods for feature selection from large datasets, and methods for training dataset selection from large datasets, for OVA and pVn aggregate modeling, were studied. The experiments of feature selection revealed that the use of many samples, robust measures of correlation, and validation procedures result in the reliable selection of relevant features for classification. A new algorithm for feature subset search, based on the decision rule-based approach to heuristic search, was designed and the performance of this algorithm was compared to two existing algorithms for feature subset search. The experimental results revealed that the new algorithm makes better decisions for feature subset search. The information provided by a confusion matrix was used as a basis for the design of OVA and pVn base models which aren combined into one aggregate model. A new construct called a confusion graph was used in conjunction with new algorithms for the design of pVn base models. A new algorithm for combining base model predictions and resolving conflicting predictions was designed and implemented. Experiments to study the performance of the OVA and pVn aggregate models revealed the aggregate models provide a high level of predictive accuracy compared to single models. Finally, theoretical models to depict the relationships between the factors that influence feature selection and training dataset selection for aggregate models are proposed, based on the experimental results. / Thesis (PhD)--University of Pretoria, 2010. / Computer Science / unrestricted

No description available.

No description available.

Les facteurs de transcription SIM1, OTP, POU3F2 et ARNT2 interagissent ensemble en orchestrant le développement complexe de l’hypothalamus, une région du cerveau contenant plusieurs petites populations circonscrites de neurones, dont le noyau paraventriculaire (PVN). Ce noyau est un important centre intégrateur et l’haploinsuffisance du facteur de transcription Sim1, essentiel au développement du PVN, mène à l’hyperphagie autant chez la souris que l’homme. Différentes souris ont été générées par génie génétique afin de nous aider à trouver d’autres gènes essentiels qui participent à cette cascade transcriptionnelle. Premièrement, la partie antérieure de l’hypothalamus a été récoltée chez des embryons (E12.5) de souris qui surexprimaient le gène Pou3f2 sous le promoteur de Otp7. L’analyse transcriptionnelle de cette partie a été comparée avec les embryons (E12.5) wt de la même portée, de sorte qu’il a été possible de constater que différents gènes ont été régulés à la hausse et d’autres à la baisse. Les gènes en question ont été choisis selon leur pertinence au développement de la région d’intérêt, l’hypothalamus, et de trois autres critères : un accroissement de plus de 1.5, une expression minimale dans l’embryon d’au moins 1000 lectures et le rang le plus haut. Cette méthode discriminatoire a permis d’identifier les gènes les plus affectés dont Lgi2, Fezf2, Sema3c, Six6, Sox14, Lmo4, Nwd2 et Nkx2.1. Les gènes régulés à la baisse étaient Six6, Sox14 et Nkx2.1, tandis que tous les autres étaient à la hausse. Afin de confirmer les résultats obtenus, une validation par hybridation in situ a été utilisée sur des tranches d’hypothalamus d’embryons E12.5. Nous avons pu confirmer la surexpression des marqueurs Lgi2, Fezf2, Sema3c, Lmo4 et de Pou3f2 dans le domaine du PVN. La diminution de l’expression des marqueurs Sox14 et Nkx2.1 a pu être détectée dans el domaine basal de l’hypothalamus, ce qui suggère que l’effet d’une surexpression de Pou3f2 dans le domaine du PVN ait un effet cellulaire non autonome. La diminution de l’expression de Six6 dans le domaine basal n’a pas pu être confirmer de façon reproductible. Deuxièmement, il semblerait qu’il y ait une redondance de rôle entre les facteurs de transcription Otp et Sim1, tous les deux agissant en amont de Pou3f2. Afin de comparer leur impact dans le programme transcriptionnel, la technologie CrisPR-cas9 a été utilisée pour faire un KO du 2e exon d’Otp. Nous avons pu confirmer notre modèle de mutation en le comparant aux autres mutants de la littérature par une réduction de l’expression d’OT, d’OTP, d’AVP et de TRH. Troisièmement, les souris hétérozygotes pour Otp et Sim1 seront croisées, de sorte d’obtenir quatre génotypes : wt, Otp+/-, Sim1+/tlz+ et Otp+/-Sim1+/tlz+. Puisqu’une redondance des rôles de Sim1 et Otp est soupçonnée, le phénotype du double mutant devrait présenter une obésité par hyperphagie plus importante que celle des souris hétérozygotes pour le gène Sim1 ou Otp. Les souris sont pesées à partir de la 5e semaine de vie jusqu’à l’âge de 6 mois à une fréquence d’une fois par semaine. L’apport calorique est également mesuré une fois par semaine sur période de 24h. La double mutation (Otp+/-Sim1+/tlz+) chez les souris mâles causait un phénotype d’obésité plus important que la singularité des mutations, mais ce n’était pas le cas chez les souris femelles. Les souris portant la mutation d’Otp étaient tout de même plus obèses que les souris sauvages pour les deux sexes. Plus de souris seront nécessaire pour déterminer si un apport calorique sans changement au niveau des dépenses énergétiques est la cause de ce gain pondéral. / The transcription factors SIM1, OTP and POU3F2 interact together to orchestrate the complex development of the hypothalamus, a region of the brain containing several small, circumscribed populations of neurons, including the paraventricular nucleus (PVN). This nucleus is an important integrating center, and the haploinsufficiency of the transcription factor Sim1, essential for the development of PVN, leads to overeating in both mice and humans. Different mice have been genetically engineered to help us find other essential genes that participate in this transcriptional cascade. Firstly, the anterior part of the hypothalamus was collected from mice embryos (E12.5) which overexpressed the Pou3f2 gene under the OTP7 promotor. The transcriptional analysis was compared to wt embryos from the same litter to see the different upregulated and downregulated genes. These genes were chosen according to their relevance to the development of the anterior hypothalamus. Three criteria were used to discriminate genes from one another: 1) an increase of more than 1.5, 2) a minimal expression in the embryo (E12.5) of at least 1000 reads and 3) the highest rank. This discriminatory method allowed us to identify the genes Lgi2, Fezf2, Sema3c, Six6, Sox14, Lmo4, Nwd2, Nkx2.1. To have a visuospatial idea of these affected genes, the validation of these results was done by in situ hybridization on E12.5 embryo hypothalamus. We have been able to see an overexpression in the PVN domain for the Lgi2, Fezf2, Sema3c, Lmo4 and Pou3f2 markers. The reduction of expression for Sox14 and Nkx2.1 markers were visible in the basal domain of the hypothalamus, which suggest a non cell autonomous effect of Pou3f2 being overexpressed. The reduction of Six6 couldn’t be consistently visible with repetition. Secondly, a redundant role of OTP and SIM1 seems to occur in the development of the hypothalamus. We created a KO line of the Otp gene by deleting the second exon with CrispR-cas9 and characterized it. We then compared it to the Sim1+/tlz+ line that we already generated in the lab. We were able to confirm our mutation model by seeing a reduction in the expression of crucial markers such as OT, OTP, AVP and TRH. Thirdly, we crossed Otp+/- with Sim1+/tlz+ mice to obtain four different genotypes: wt, Otp+/-, Sim1+/tlz+ and Otp+/- Sim1+/tlz+. Since a redundant aspect has been observed for SIM1 and OTP transcription factors, we were wondering if the obesity phenotype would be worsened by carrying both mutation or not. These mice were weighted every week from 5-week-old up to 6 months old. Food intake has also been measured since the obesity has been reported to be caused by hyperphagia in Sim1 mutant mice. The male mice carrying the double mutation (Otp+/-Sim1+/tlz+) showed a more important weight gain than only Sim1+/tlz+ or Otp+/- mutants, but it was not the case for the female mice. The mice carrying the Otp mutation still got a more important weight gain than the wt mice (females and males). More mice would be necessary to determine if this weight gain is caused by hyperphagia only or if unbalance energy cost is part of the cause.