It was originally observed, during kinetic studies of infection, that the intravenous injection of a small inoculum of dispersd Mycobacterium bovis (BCG) cells resulted in one of two distinct patterns of multiplication (growth vs non-growth) in the liver and spleen of mice of several inbred strains. Twenty-four inbred strains of mice which were so examined exhibited marked discontinuous variation in their level of early resistance to BCG infection, as estimated by the number of viable bacilli recovered from the spleen three weks after the intravenous injection of 2.5 x 10('4) BCG. The nature of this genetic variation was then investigated by a Mendelian segregation analysis: resistance of F(,1), F(,2) and backcross progeny bred from representative resistant and susceptible progenitors showed that the trait of resistance was controlled by a single, dominant, autosomal gene, designated Bcg. This gene exists in two allelic forms, Bcg('s) (susceptibility) and Bcg('r) (resistance). The Bcg locus was subsequently mapped to the Chromosome 1 after comparison of the respective strain distribution pattern (SDP) of the previously-mapped Idh-1 and Pep-3 alleles with that of Bcg alleles among mice of BXH and BXD recombinant inbred strains. Strong linkage between the Bcg gene and 2 other host resistance genes, namely Ity and Lsh, was detected by comparison of their respective SDP's. This tight linkage or identity was further confirmed by formal Bcg-Ity-Lsh linkage analysis performed on individual animals of the segregating populations and on inbred mouse strains congenic but for this portion of the first chromosome. / The phenotypic expression of the Bcg locus at the cellular level was investigated in vivo: the cell responsible for early resistance was found not to be a T-cell and most probably not a B cell. The genetically-controlled mechanism of defense proved to be resistant to 950 rads of X-irradiation and susceptible to the toxic effect of silica. The reticuloendothelial system of both Bcg('r) and Bcg('s) strains showed an equal ability to clear a bacterial inoculum from their blood and these animals were also comparable in the intensity of the inflammatory response to an intraperitoneal inoculum of BCG. Explanted pleural and peritoneal resident macrophages supported equally well, independently of their genotype, the intracellular growth of various strains of mycobacteria in vitro.

Gene targeting is a process of recombinational exchange between a foreign molecule of DNA and its homologous chromosomal counterpart whereby the cellular machinery involved in homologous recombination accomplishes sequence exchange. The widespread use of gene targeting as a basic research and/or therapeutic tool has been precluded by its inherent inefficiency. Although many studies have reported improved gene targeting efficacy via manipulation of homologous recombination-associated proteins, reports examining the underlying molecular mechanisms of gene targeting are scant. Improvement in our understanding of these molecular mechanisms will undoubtedly lead to advancements in the study of genetic disease, and perhaps to the development of viable human gene therapy. To this end, the following is a report of the development and preliminary assessment of a model system for the study of the molecular mechanisms of gene targeting. Specifically, this system entailed the investigation of recombinational exchange between isogenic, non-functional copies of HSV-tk contained in a targeting plasmid and a stably integrated target plasmid, in Ltk-cells. Results showed an absolute random integration frequency of our targeting construct in the range of 10-4---a value ten-fold lower than classically associated with random integration. Also, linearization of our targeting plasmid outside the region of homology to our stably integrated chromosomal target appeared to be associated with an undetectable level of targeting. Similarly, non-homology---in the form of three non-silent point mutations in our target sequence---appeared to impede gene targeting. These initial results suggest that the design of this model system needs to be revised to allow proper investigation of the molecular mechanisms underlying gene targeting.

Type 1 Diabetes (T1D) is characterized by the autoimmune destruction of the insulin producing beta cells of the pancreas. Linkage with Type 1 Diabetes has been reported on chromosome 2q31-33. Within this 23 cM interval, T1D association with CTLA4 polymorphisms has been localized to a linkage disequilibrium (LD) block covering the D2S72- CTLA4-D2S116 interval. Within this interval, T1D is associated with a haplotype in the CTLA4 gene, containing several potentially functional polymorphisms. Due to the extent of LD in the region, genetics studies alone are not sufficient to identify the causative polymorphism(s). To this end, we undertook a systematic analysis of each of these polymorphisms and their functional relevance in the pathogenesis of Diabetes. / We investigated the role of the only coding polymorphism in the CTLA4 gene, a Threonine to Alanine substitution in the N-terminal signal sequence. The Ala17 (+49G) allele is associated with increased risk to T1D. Since the major function of the signal sequence is to direct trafficking of the pre-protein to the Endoplasmic Reticulum (ER), we hypothesized that a substitution could alter these events. The Ala17 is inefficiently N-glycosylated in the ER and its expression is lower than the Thr17 at the cell surface in COS1 cells. / Using the family-based transmission disequilibrium test, (TDT) we report an association of the -318C-allele at the proximal promoter polymorphism and T1D. Conditional TDT analysis dissected the contributions of the linked -318C>T and +6230G>A polymorphisms, as independent contributors to the genetic effect. We report transcriptional effects of the promoter alleles in vitro , and in vivo with the -318T allele inducing stronger transcriptional activity. This effect may be mediated by alternative transcription factor binding at the Single Nucleotide Polymorphism (SNP). / We confirmed association of the 3'+6230G>A SNP with T1D. In evaluating its contribution to the expression of soluble and full-length CTLA4 isoforms in an allele-dependent manner, we found no correlation between alleles at +6230 and levels of either isoform. The 3' SNP does not alter in cis the expression of the downstream ICOS gene. / We propose that susceptibility at the CTLA4 locus is mediated by a haplotype exerting both transcriptional and post-translational effects.

Interferon regulatory factors (IRFs) is a family of transcriptional regulators that are essential for cell differentiation and growth, oncogenesis and the regulation of immunity. IRF1 and IRF8 play a crucial role in immunity by regulating the differentiation of the cells of myeloid lineage, particularly the development of immune cells including macrophages, NK and dendritic cells. Both transcription factors are essential for the resistance to viral and bacterial infections. We optimized the chromatin immunoprecipitation (ChIP) technique to be used with IRF8 and IRF1 antibodies in a J774 line stimulated with IFNγ/CpG. Using ChIP coupled to an exon promoter 244K microarray (ChIP on Chip) we identified 201 and 303 binding sites of high confidence for IRF1 and IRF8 respectively. Previously known IRF1 and IRF8 transcription targets (OAS1b for IRF1 and IFNb for IRF8) as well as novel transcriptional control loci were identified. One of the major gene ontology (GO) functional categories in the two gene lists was "immune response". The overlap of IRF1 and IRF8 transcription target genes from ChIP on Chip revealed an overlap of 19 genes most of which belonged to the "immune regulation" GO pathway and included targets of high confidence: Gbp6, Mx2, Tnfsf13b, H2-T24, and Ifit1. A parallel microarray transcription profiling study on mouse bone marrow macrophages (BMDMs) from an F2 generation of Balb/C IRF8-wildtype and BXH2 IRF8-mutant cross was performed. The BMDMs possessing the wildtype and the mutant IRF8 alleles were infected with Legionella pneumophila or stimulated with IFNγ/CpG to identify genes regulated by IRF8 in the context of infection. 1171 and 852 differentially regulated genes were differentially regulated in the infection and stimulation experiments respectively. Genes that displayed interaction between IRF8 allele and the infection conditions were isolated to yield high confidence gene lists of 31 and 129 genes for the Legionella and IFNγ/CpG expe / Les facteurs régulateurs interférons (IRFs) sont une famille de régulateurs transcriptionnels essentiels à la différenciation et la croissance cellulaire, à l'oncogénèse et au bon maintien du système immunitaire. IRF1 et IRF8 jouent un role clé dans l'immunité de l'organisme par la régulation de la différenciation des lignées cellulaires myéloïdes, nottament le développement de cellules immunitaires comme les macrophages, les cellules tueuses naturelles (NK) et les cellules dendritiques. Ces deux facteurs de transcription sont indispensables pour la résistance aux infections virales et bactériennes. Nous avons optimisé une technique d'immunoprécipitation de la chromatine (ChIP) afin d'y utiliser des anticorps ciblant IRF1 et IRF8, et ce à partir d'une lignée cellulaire J774 suivant un stimulation par interféron gamma et par oligonucléotides CpG (IFNγ/CpG). En combinant cette technique ChIP à un microarray 244K comprenant exons et promoteurs (ChIP on Chip), nous avons identifiés respectivement 201 et 303 sites d'ancrages pour chacun de IRF1 et IRF8. Certaines cibles transcriptionnelles déjà connues pour IRF1 (OAS1b) et IRF8 (IFNb) ainsi que de nouveaux loci de control transcriptionel ont été identifiés. Parmi les catégories par fonction relevées par Ontologie de gène (GO), une des catégories majeures des deux listes de gènes générées fut pour les gènes reliés à la "réponse immunitaire''. La comparaison entre les gènes cibles de transcription pour IRF1 et IRF8 par "ChIP on Chip'' a révélé 19 gènes représentés dans les deux cas. La majeure partie de ces gènes sont reliés à la "réponse immunitaire'' par GO et ils incluent les cibles suivantes avec détection très significative: Gbp6, Mx2, Tnfsf13b, H2-T24, et Ifit1. En parallèle, des macrophages dérivés de la moëlle osseuse de souris (BMDMs) furent utilisés pour l'étude du profile transcriptionel par microarray d'une génération F2 générée avec des so

5-10% of hereditary breast cancer cases are caused by germline mutations in well-defined, dominantly acting susceptibility genes such as BRCA1 and BRCA2. However, more than 50% of the genetic predisposition to hereditary breast cancer remains unexplained. In the following thesis, we present a multifaceted approach aimed at further elucidating hereditary breast cancer associated with BRCA1 and BRCA2 interacting genes; specifically, by analyzing the potential contribution from of two previously unscreened BRCA1-associating genes, RAP80 and Abraxas, assessing the presence and risk associated with CHEK2 susceptibility alleles in the previously uninvestigated French Canadian population and by investigating molecular and cellular mechanisms underlying the increased risk associated with PALB2 susceptibility alleles. / A combination of genotyping 96 BRCA1/2 negative, high risk breast cancer patients and segregation analysis was utilized in the determination of whether or not RAP80 and Abraxas are breast cancer susceptibility genes. The contribution of CHEK2 associated breast cancer amongst the French Canadian population was determined through the genotyping 25 BRCA1/2 negative, high risk breast cancer and a cohort of 25 controls. Finally, the biological significance of four PALB2 susceptibility alleles was investigated through the use of the cellular cytotoxicity assay WST-1, telomere specific Q-FISH, centromere specific FISH and spectral karyotyping. / The results presented herein suggest that both RAP80 and Abraxas are not high to moderately penetrant breast cancer susceptibility genes. Further, our results suggest that alleles other than the CHEK2 1100delC are unlikely to significantly contribute to the hereditary breast cancer risk in the French Canadian population. Lastly, the results obtained throughout our analysis of PALB2 heterozygous cell lines may be suggestive of a possible chromosomal instability phenotype predisposing carriers to additional tumourgenic mechanisms. / 5-10% des cas de cancer héréditaire du sein sont causés par des mutations germinales dans des gènes des susceptibilité bien caractérisés et à l'effet dominant tel les gènes BRCA1 et BRCA2. Cependant, plus de 50% de la prédisposition génétique au cancer du sein héréditaire demeure inexpliquée. Dans cette thèse, nous présentons une approche à trois volets ayant pour but d'étudier les cas de cancer héréditaire du sein associés avec des gènes interagissant avec BRCA1 et BRCA2. D'abord, nous analysons la contribution potentielle de deux gènes peu caractérisés qui sont partenaires de BRCA1 : RAP80 et Abraxas. Nous étudions ensuite le risque associé avec la présence d'allèles nouveaux ou connus du gène CHEK2 jamais encore caractérisés dans la population canadienne française. Enfin, nous examinons les mécanismes cellulaires et moléculaires responsables de l'augmentation du risque de cancer du sein conférée par des allèles à risque du gène PALB2. / Nous avons utilisé une combinaison de génotypage chez 96 patients souffrant du cancer du sein mais étant non porteurs de mutations dans BRCA1/2 et d'analyse de ségrégation des mutations et des phénotypes dans leurs familles afin de déterminer si RAP80 et Abraxas sont ou non des gènes de prédisposition au cancer héréditaire du sein. La contribution au risque de cancer du sein du gène CHEK2 fût déterminée grâce au génotypage de 25 cas à haut risque, non porteurs de mutations chez BRCA1/2, et de 25 contrôles sans cancer. Finalement, nous avons étudié 4 allèles nonsense du gène PALB2 à l'aide du test de toxicité cellulaire WST-1 ainsi qu'en utilisant l'analyse Q-FISH spécifique aux télomères, l'analyse FISH spécifique aux centromères et finalement par caryotype spectral (SKY). / Les résultats présentés dans cet ouvrage suggèrent que RAP80 et Abraxas ne sont pas des gènes de susceptibilité au cancer du sein à pénétrance moyenne ou élevée. De plus, il est peu probable que des allèles du gène CHEK2 autres que l'allèle connu 1100delC contribuent de façon significative au risque de cancer du sein héréditaire dans la population canadienne française. Par contre, les résultats de notre analyse du gène PALB2 dans les lignées cellulaires hétérozygotes pour un allèle nonsense suggèrent la possibilité que la présence de ces allèles crée de l'instabilité chromosomique chez les porteurs de mutations qui puissent prédisposer à la progression tumorale.

No matter their living environment, the long-term survival of cells requires proliferation. This implies that they have to replicate their genetic material and to divide into two cells. But going through the mitotic division cycle is a major cellular commitment that needs to be balanced between the capacity to do it and the necessity to survive (adapt) against immediate threats. Outside of a few keynote species our understanding of the connections between cell cycle progression and environmental sensing is far from clear. In this thesis, I have explored the control of the cell cycle in Candida albicans, a fungal pathogen that is also growing in importance as a model organism compared with Saccharomyces cerevisiae. To do so, I have used the fungal mating pheromone-triggered pathway which, upon stimulation, is known to influence mitotic cycle progression. Although the mating pathway has been extensively studied in S. cerevisiae, our understanding of this pathway in C. albicans is still at its early stage. In this thesis, I will 1) describe the functional characterization of a “barrier” activity in C. albicans that functions to restrict the impact of pheromone stimulation, 2) characterize the Candida homolog of the mating associate d cyclin-dependent kinase inhibitor Far1p, 3) analyze the mitotic-dependent transcriptional regulation profile and 4) identify a novel protein connecting environmental stress adaptation and virulence to C. albicans cell cycle progression. These findings highlight that one of the biggest molecular biology challenges still remains the identification of each gene's function. Although bioinformatical comparisons are tremendously helpful is this quest, ultimately experimental characterization is required to reveal the subtleties beneath adaptive evolution that have occurred among the species being compared. / Qu'importe là ou elles vivent, chaque cellule se doit de proliférer pour survivre. Ceci implique que les cellules doivent obligatoirement répliquer leur matériel génétique et se diviser en deux cellules-filles. Toutefois, s'engager dans le cycle cellulaire représente une implication considérable et chaque cellule doit prendre en considération à la fois sa propre capacité à compléter le cycle et la nécessité de le faire dans des circonstances qui ne nuiront pas à sa survie. Malheureusement, en dehors de quelques rares espèces, notre compréhension des liens unissant le cycle mitotique et l'adaptabilité à l'environnement est relativement superficielle. Dans cette thèse, je propose d'explorer ces liens chez Candida albicans, une levure pathogène dont l'importance en tant qu'organisme modèle grandit. Pour ce faire, j'ai utilisé la voie de signalisation de l'accouplement (mating) connue chez la levure Saccharomyces cerevisiae pour influencer la course du cycle cellulaire. Au fil de ce manuscrit, je 1) démontrerai qu'une activation enzymatique contribuant à restreindre la stimulation par la phéromone de mating existe chez C. albicans (activité connue sous le nom de « barrière »), 2) proposerai la caractérisation fonctionnelle de la protéine inhibitrice du cyclin-dépendent kinase (Far1p), 3) analyserai la régulation transcriptionnelle du cycle cellulaire chez Candida et finalement 4) révélerai l'existence d'une protéine périodiquement exprimée liant l'adaptation aux stress environnementaux et la virulence chez C. albicans. Ensemble, ces résultats mettent en évidence que l'un des plus gros défis de la biologie moléculaire reste, encore et toujours, la détermination fonctionnelle de chaque gène. Bien que la comparaison bioinformatique inter-espèce soit une aide fantastique dans cette tâche, les analyses expérimentales s'avèrent indispensables et bien souvent mettent en évidence des subtilités fonctionnelles qui, ensemble,

Background: Angiotensinogen (AGT) is a vasopressor agent and is part of the Renin-Angiotensin System (RAS) which plays an essential role in blood pressure regulation. The M235T polymorphism in AGT has been inconsistently associated with both hypertension and cardiovascular disease. We examined the role of M235T in hypertension and myocardial infarction (MI) and investigated whether this polymorphism interacts with other risk factors. Methods: We genotyped a subset of 7,232 individuals from the INTERHEART population, a global case-control study and sought replication in a subset of 12,206 individuals from the DREAM/EpiDREAM trial. Results: The T variant was significantly associated with MI in the two largest INTERHEART populations, Europeans (P<0.036) and South Asians (P<0.0005). Interestingly, we observed a more significant result in females (P< 0.0002) compared to males (P< 0.01). M235T was also associated with hypertension in a weighted meta-analysis adjusted for age and sex in both INTERHEART (P<0.01) and DREAM/EpiDREAM (P<0.001). Testing for an association with MI after adjusting for hypertension did not greatly alter the odds ratios obtained prior to adjustment, suggesting that only part of the effect on MI can be attributed to hypertension. We further hypothesized that M235T may have a different effect in certain sub-groups. We tested for the association with MI within low risk and high risk subgroups separately and observed an interaction with MI risk factors across multiple ethnicities. The Breslow-Day test for homogeneity of effect size between the low and high risk groups in the combined analysis yielded P<0.002. This trend was also observed in the DREAM/EpiDREAM subset. Conclusion: These results confirm the association of the T variant of M235T with both MI and hypertension. In addition, they demonstrate that M235T interacts with known risk factors and is involved in the pathogenesis of MI independently of hypertension. / Contexte: L'angiotensinogène (AGT) est un agent vasopresseur qui fait partie du système rénine-angiotensine (SRA) et qui joue un rôle essentiel dans la régulation de la pression artérielle. Le rôle du polymorphisme M235T, situé dans l'AGT, dans l'hypertension et la maladie cardiovasculaire est inconsistent. Nous avons voulu examiner le rôle de ce polymorphisme dans l'hypertension et dans l'infarctus du myocarde (IM) et étudier son interaction avec d'autres facteurs de risque. Méthodes: Nous avons génotypé un sous-ensemble de 7,232 individus provenant du projet INTERHEART, une étude de cas témoins internationale. Nous avons voulu confirmer nos résultats en analysant un sous-ensemble de 13,669 individus provenant de l'étude DREAM/EpiDREAM. Résultats: Le variant T est associé de façon significative avec l'IM dans les deux plus importantes populations d'INTERHEART, soient les populations européenne (P<0.04) et de l'Asie du Sud (P<0.0004). Il est intéressant de noter que nous avons observé un effet plus important chez les femmes (P< 0.0002) que chez les hommes (P< 0.01). M235T a aussi été associé avec l'hypertension dans une méta-analyse pondérée ajustée pour le sexe et l'âge (P<0.01). L'étude DREAM/EpiDREAM nous a permis de confirmer cette association dans le cas de l'hypertension. Le variant T a été associé avec l'hypertension auto-déclarée (OR=1.08, P=0.018), et avec la tension artérielle systolique (P=0.05) and diastolique (P=0.001). Un test d'association avec l'IM après ajustement pour l'hypertensions n'a pas considérablement modifié le rapport de cotes, suggérant que seule une partie de l'effet sur l'IM peut être attribuable à l'hypertension. Nous avons formulé l'hypothèse que M235T puisse avoir un effet différent dans certains groupes. Nous avons testé l'association avec l'IM parmi les groupes à faible et à haut risque séparément et avons observé une interaction avec les facteurs de risque de l'IM à travers toutes les ethnies testées. Le test de Breslow-Day pour l'homogénéité de la taille d'effet entre les groupes à risque faible et à risque élevé dans l'analyse combinée a produit une valeur de P<0.002. Cette tendance a aussi été observée dans les échantillons de DREAM/EpiDREAM. Ajuster pour l'hypertension ou exclure les individus hypertendus n'a pas modifié la taille d'effet. Conclusion: Ces résultats confirment l'association du variant T à la fois avec l'IM et l'hypertension. De plus, ils démontrent l'interaction entre le polymorphisme M235T et des facteurs de risque connus ainsi que son implication dans la pathogénie de l'IM, indépendamment de celle de l'hypertension.

Bril is a 134 amino acid-long protein compartmentalized exclusively on the osteoblast cell surface, with two transmembrane domains. Its osteoblast-specific expression coincides with the onset of bone matrix formation and mineralization and has been shown to have an essential role in bone mineralization in vitro, although the mechanism involved is still unknown. The purpose of this study was to characterize the regulation pattern of the Bril gene by identifying the cis-elements on the promoter and the transcription factors that modulate its activation. The promoter for the mouse, rat, and human Bril gene were independently cloned into a luciferase (Luc) reporter plasmid. Transient transfection studies were then performed to compare the activity of the Luc reporter in permissive (UMR106, MC3T3-E1) and non-expressing cell lines (HEK293). Deletion mutant analyses revealed that the proximal 300bp promoter was sufficient to confer highest activation in the three species studied. In silico analysis, EMSA and DNAseI footprinting studies revealed that this region is rich in GC-boxes and contains many potential regulatory elements known to be involved in osteoblastic differentiation (e.g. Runx2, TCF1/LEF, Osx/Sp7/Sp1) that are fully functional. Co-transfection experiments in HEK293 and MC3T3-E1 cells showed that the Bril promoter is most strongly trans-activated by forced expression of Sp1 and the long variant of Sp3, moderately by members of the Wnt pathway (i.e. β-catenin and TCF/LEF), yet is not considerably affected by bone-specific factors. Additionally, shRNA-mediated knockdown of Sp1 (but not Sp3) in UMR106 cells resulted in complete abrogation of Bril protein. In vitro interference of the GC-rich boxes via GC-bisintercalating agents and CpG methylation resulted in a dramatic drop in Bril promoter activity; which highlights the importance of these GC-sequences in promoting Bril regulation. In a search for molecules that might regulate Bril expression, the parathyroid hormone (PTH) was found to be a potent negative regulator. Endogenous Bril expression in UMR106 was dramatically downregulated by PTH in a time- and dose-dependent manner, which coincided with suppressed osteoblast mineralization. These results provide the first mechanistic evidence for regulation of the Bril gene, which could help situate its function in the bone mineralization process in vitro. / Bril est une protéine de 134 acides aminés, localisée exclusivement sur la membrane des ostéoblastes avec deux domaines transmembranaires. Son expression exclusive dans les ostéoblastes coïncide avec le début de la formation osseuse et la minéralisation de la matrice. Ainsi, un rôle essentiel a été démontré dans la minéralisation osseuse in vitro, bien que le mécanisme en cause n'est pas encore connu. L'objectif de cette étude était de caractériser le mode de régulation du gène Bril en identifiant les éléments cis sur le promoteur et les facteurs de transcription qui modulent son activation. Le promoteur du gène Bril de la souris, du rat et de l'homme ont été indépendamment clonés dans un plasmide rapporteur de la luciférase (Luc). Des transfections transitoires ont ensuite été effectuées pour comparer l'activité Luc dans des cellules qui expriment (UMR106, MC3T3-E1) ou non (HEK293) le gène Bril de manière constitutive. L'analyse des mutants de délétion a révélé que le promoteur proximal 300 pb est suffisant pour conférer la majorité d'activation dans les trois espèces étudiées. Ainsi, les analyses in silico ont révélé que cette région est riche en séquences de GC et contient des éléments de régulation putatifs connus pour être impliqués dans la différenciation ostéoblastique (par exemple Runx2, TCF1/LEF, Osterix ou Sp1). Des analyses de retard sur gel (EMSA) et de cartographie à la DNAseI ont indiqué que beaucoup d'entre eux sont des sites fonctionnels. Les co-transféctions dans des cellules HEK293 et MC3T3-E1 ont montré que le promoteur Bril est le plus fortement trans-activé par l'expression forcée de Sp1 et la variante longue du Sp3, modérément par des membres de la voie Wnt (β-caténine et TCF /LEF), mais ne sont pas encore fortement influencés par des facteurs spécifiques à l'os. En plus, l'invalidation dans les cellules UMR106 de Sp1 (et non Sp3) par les shRNAs mène à l'abrogation complète de la protéine Bril. In vitro, l'interférence des séquences GC (via les drogues MMA et WP631 et la méthylation) a entraîné une forte baisse de l'activité promoteur de Bril; ce qui souligne l'importance de ces GC-séquences dans la promotion et la réglementation du Bril et propose la méthylation de l'ADN comme un mécanisme à considérer dans la spécificité d'expression cellulaire de Bril. La recherche de molécules qui pourraient réguler l'expression de Bril a identifié l'hormone parathyroïde (PTH) comme un puissant régulateur négatif. L'expression endogène de Bril dans les UMR106 a été considérablement diminuée par la PTH dans une manière dose-dépendante, qui coïncide avec une reduction de la minéralisation des ostéoblastes. Ces résultats fournissent les premières preuves d'un mécanisme pour la régulation du gène Bril, qui pourrait aider à démontrer sa fonction dans le processus de minéralisation osseuse in vitro.

The development of healthy gametes is paramount to the health of progeny and to the survival of a species. Epigenetic information contained within gametic DNA in the form of DNA methylation is essential for germ cell and embryo development. DNA methylation is a genome-wide phenomenon involved in the control of gene expression and chromosome structure and stability. During germ line development, patterns of DNA methylation are established in a sex- and sequence-specific manner. The primary goal of the work presented in this thesis is to gain an understanding of the nature of the genome-wide pattern of DNA methylation in germ cells and to study its progression during germ cell development. The complexity of male germ cell development has been well studied in mice and thus makes an excellent system in which to study germ cell DNA methylation. Firstly, genome-wide patterns of DNA methylation in adult male germ cells were determined using a variety of techniques. Results from these studies demonstrate that the DNA methylation pattern in male germ cells is highly distinct from that of somatic cells. The reorganization of the germ cell pattern is associated with chromosomal features such as the chromosomal banding pattern and regional GC content. Secondly, by examining purified populations of male germ cells, we have determined that patterns of DNA methylation are being acquired during spermatogenesis. De novo methylation and demethylation events occur in a sequence-specific manner prior to the meiotic phase of germ cell development. Finally, the stability of these patterns was studied by perturbing DNA methyltransferase activity. The study of germ cells lacking a functional Dnmt3L gene demonstrates that the abnormalities displayed in these cells are associated with a failure to acquire normal levels of DNA methylation. In addition, the treatment of mice with the hypomethylating agent, 5-aza-2'-deoxycytidine, results in adverse effects on sperm function and is associated with sequence-specific hypomethylation. Collectively, these studies have uncovered several novel aspects of DNA methylation in male germ cells and contribute to our understanding of the roles) for epigenetic phenomena in the development and maintenance of healthy gametes.

While studying the unique Nramp1 (Slcl1a1)-independent susceptibility of BXH-2 mice to Mycobacterium bovis (BCG) infection, we noted that these mice develop splenomegaly associated with important infiltration of Mac-1+/GR-1+ (macrophage antigen-1 +/granulocyte differentiation antigen 1+) granulocyte precursors in spleen, lymph nodes and bone marrow resembling a myeloproliferative syndrome. This syndrome was subsequently found to be independent of the BCG infection. Segregation analyses revealed that the myeloproliferative syndrome was controlled by a single recessive locus (Myls) mapping to a 9 megabases (Mb) region of chromosome 8. Myls contained one of the two replication-defective proviruses (Emv2) integrated into BXH-2 genome, and thought to play a role in generating the replication competent B-tropic ecotropic munne leukemia virus (MuLV) known to cause clonal tumours in BXH-2 mice. By narrowing down the Myls interval to 2 Mb, we could exclude Emv2 as a potential candidate for Myls. However, several other candidates remained, among which the interferon consensus sequence binding protein 1 (Icsbp1) gene. We showed that BXH-2 mice carry a mutation (915 C to T) resulting in an arginine-to-cysteine substitution at position 294 within the transactivation domain of the Icsbp1 protein. We found that Icsbp1C294 behaves as a hypomorphic allele which appears to have a threshold effect on maturation of the myeloid lineage, as well as pleiotropic consequences on resistance to different types of infectious agents including Mycobacterium bovis (BCG). Our results suggest a two-step mutagenic model in which first, inactivation of Icsbp1 predisposes to myeloproliferation and immunodeficiency and favours the production of retroviruses. Then, subsequent insertional mutagenesis of oncogenes or tumour suppressors by the retrovirus results in clonal expansion of leukemic cells characteristic of BXH-2 mice.