Alzheimer’s disease (AD) is the most prevalent form of dementia. There is considerable evidence that AD is caused by accumulating amyloid beta peptides in the brain, as a result of amyloid precursor protein (APP) cleavage by secretase enzymes. The presenilin proteins are central to the gamma-secretase cleavage of the intramembrane domain of APP. Aberrant splicing and point mutations in the human presenilin genes, PSEN1 and PSEN2, have been linked to familial forms of AD, through aberrant APP cleavage resulting in irregular amyloid beta formation. Paper 1 gives a review of the literature on AD research and how animal models are used to elucidate mechanisms of AD pathogenesis. The zebrafish model is used in this thesis to investigate genes with potential relevance to AD initiation and pathogenesis. Paper 2 demonstrates that lowlevel aberrant splicing of exon 8 in psen1 transcripts in zebrafish embryos produces potent dominant negative effects that increased psen1 transcription, cause a dramatic hydrocephalus phenotype, decreased pigmentation and other developmental defects. Similar effects are also observed after low-level interference with splicing of exon 8 in psen2 transcripts. In paper 3, a microarray analysis was performed to analyse global gene expression changes to illuminate the molecular aetiology of the phenotypic effects described in paper 2. Of the 100 genes that showed greatest dysregulation after psen1 or psen2 manipulation, 12 genes were common to both treatments. Five of these have known function and showed increased expression. Cyclin G1 (ccng1) was of particular interest as the human CCNG1 protein shows increased immunoreactivity in the cytoplasm of neurons in human AD brains. Phylogenetic and conserved synteny analysis confirmed the orthology of zebrafish ccng1 with human CCNG1. Expression of zebrafish ccng1 in developing embryos at 24 hours post fertilization (hpf) was observed in the eye, tectum and somites. Decreased Ccng1 expression does not lead to any developmental defects and also cannot rescue the hydrocephalus or pigmentation phenotypes of embryos with aberrant splicing of psen1 exon 8. An analysis of zebrafish ccng1 function in paper 4 (thesis chapter in the form of a manuscript) indicates that truncation of Ccng1 appears to cause developmental defects in the brain, notochord and somites, however, it does not decrease the level of normal ccng1 transcript. The CCNG1 paralogue, Cyclin G2, (CCNG2), is also expressed in zebrafiish (ccng2). Decreasing the expression of Ccng2 results in similar effects on embryo development as truncating Ccng1. Therefore, the truncated forms of Ccng1 potentially interfere with Ccng2 function in a dominant negative manner. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1342482 / Thesis (Ph.D.) -- University of Adelaide, School of Molecular and Biomedical Science, 2008

Alzheimer’s disease (AD) is the most prevalent form of dementia. There is considerable evidence that AD is caused by accumulating amyloid beta peptides in the brain, as a result of amyloid precursor protein (APP) cleavage by secretase enzymes. The presenilin proteins are central to the gamma-secretase cleavage of the intramembrane domain of APP. Aberrant splicing and point mutations in the human presenilin genes, PSEN1 and PSEN2, have been linked to familial forms of AD, through aberrant APP cleavage resulting in irregular amyloid beta formation. Paper 1 gives a review of the literature on AD research and how animal models are used to elucidate mechanisms of AD pathogenesis. The zebrafish model is used in this thesis to investigate genes with potential relevance to AD initiation and pathogenesis. Paper 2 demonstrates that lowlevel aberrant splicing of exon 8 in psen1 transcripts in zebrafish embryos produces potent dominant negative effects that increased psen1 transcription, cause a dramatic hydrocephalus phenotype, decreased pigmentation and other developmental defects. Similar effects are also observed after low-level interference with splicing of exon 8 in psen2 transcripts. In paper 3, a microarray analysis was performed to analyse global gene expression changes to illuminate the molecular aetiology of the phenotypic effects described in paper 2. Of the 100 genes that showed greatest dysregulation after psen1 or psen2 manipulation, 12 genes were common to both treatments. Five of these have known function and showed increased expression. Cyclin G1 (ccng1) was of particular interest as the human CCNG1 protein shows increased immunoreactivity in the cytoplasm of neurons in human AD brains. Phylogenetic and conserved synteny analysis confirmed the orthology of zebrafish ccng1 with human CCNG1. Expression of zebrafish ccng1 in developing embryos at 24 hours post fertilization (hpf) was observed in the eye, tectum and somites. Decreased Ccng1 expression does not lead to any developmental defects and also cannot rescue the hydrocephalus or pigmentation phenotypes of embryos with aberrant splicing of psen1 exon 8. An analysis of zebrafish ccng1 function in paper 4 (thesis chapter in the form of a manuscript) indicates that truncation of Ccng1 appears to cause developmental defects in the brain, notochord and somites, however, it does not decrease the level of normal ccng1 transcript. The CCNG1 paralogue, Cyclin G2, (CCNG2), is also expressed in zebrafiish (ccng2). Decreasing the expression of Ccng2 results in similar effects on embryo development as truncating Ccng1. Therefore, the truncated forms of Ccng1 potentially interfere with Ccng2 function in a dominant negative manner. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1342482 / Thesis (Ph.D.) -- University of Adelaide, School of Molecular and Biomedical Science, 2008

BMI1 est une protéine appartenant à la famille des polycombs impliquée dans la régulation épigénétique de la transcription. Il a été montré que cette protéine est essentielle à la régulation de la prolifération, de la sénescence et du métabolisme ainsi qu’à l’auto-Renouvellement des cellules souches hématopoïétiques et cancéreuses. Ce répresseur transcriptionnel au fort potentiel oncogénique est retrouvé surexprimé dans de nombreux types de cancer ; dans le cas de la Leucémie Myéloïde Chronique (LMC) le niveau d’expression de BMI1 augmente avec l’aggravation de la pathologie. Cependant, les voies de signalisation impliquées dans sa surexpression et le rôle qu’il joue au sein de cette maladie demeurent méconnus. En réprimant l’expression de BMI1 par ARN interférence nous avons pu mettre en évidence que ce polycomb était essentiel à la prolifération cellulaire ainsi qu’au potentiel clonogénique des cellules de LMC. Nous avons également démontré pour la première fois que BMI1 soutenait la croissance tumorale à travers la répression d’un processus autophagique délétère pour la cellule cancéreuse. Une approche transcriptomique nous a permis d’identifier la cible transcriptionnelle impliquée dans ce processus, la Cycline G2. Nous avons, pour finir, trouvé une molécule, via une approche bioinformatique, capable de réinduire l’expression de la Cycline G2 dans les cellules de LMC, l’alexidine dihydrochloride. Cette molécule induit une forte autophagie dans les cellules cancéreuses ainsi que de l’apoptose. Elle s’est également montrée capable de resensibiliser à l’imatinib (un inhibiteur de BCR-ABL) une lignée pourtant résistante. / The polycomb protein Bmi1 is a major epigenetic regulator. It has been shown that this protein is essential for the regulation of cell proliferation, senescence and metabolism but also self-Renewal of hematopoïetic and cancer stem cells. This transcriptional repressor, with a strong oncogenic potential, is overexpressed in many types of cancer. In case of Chronic Myeloid Leukemia (CML) the expression level of BMI1 is associated with worsening prognosis. However, the signaling pathways involved in its overexpression and its role in this disease remains unclear. By using RNAi to repress BMI1 expression we highlighted that this polycomb was essential for proliferation and clonogenicity of CML cells. We also demonstrated, for the first time, that BMI1 supported tumor growth through repression of deleterious cancer cell autophagy. A transcriptomic approach allowed us to identify a transcriptional target involved in this process: the Cyclin G2. Through a bioinformatic approach, we finally found a molecule capable of expression re-Induction of Cyclin G2 in CML cells : alexidine dihydrochloride. This molecule induced a high level of autophagy as well as apopotosis in cancer cells. It had also been able to re-Sensitize to imatinib a resistant cell line. In conclusion, our results revealed a new role for the polycomb BMI1 in supporting the CML pathology. Moreover, our work allowed the identification of two new approaches for therapeutically targeting this oncogene functions.