The nBmp2 protein was first identified in a DNA affinity chromatography/mass spectrometry screen designed to detect proteins that interact with a cartilage-specific enhancer element (called D/E) from the type XI collagen gene Col11a2. The transcription factor SOX9, a protein from the Sox (SRY-related HMG box) family, binds to and activates gene expression from this enhancer. nBmp2 has no transcriptional activity of its own on this enhancer, but when co-transfected with SOX9 it increases SOX9's activation of D/E nearly 2-fold. SOX9 also activates cartilage-specific enhancer elements from the Col2a1, Col27a1, and Col9a1 genes. The purpose of this project was to determine 1) whether nBmp2 similarly effects SOX9-dependent expression from these enhancers, and 2) whether it does so by binding (either directly or indirectly) to the Col2a1, Col27a1, and Col9a1 enhancers. The work described in this thesis has shown that nBmp2 increases luciferase levels produced from three enhancer/reporter plasmids, but it does so without binding directly to the enhancers. This work has opened up a new area of exploration into the function of the novel protein nBmp2 to examine its potential effects on a variety of different gene regulatory processes.

SOX9 est un facteur de transcription à domaine HMG. Il est impliqué dans de multiples processus biologiques au cours du développement et de la vie adulte. En particulier, SOX9 joue un rôle important dans l'homéostasie de l'épithélium intestinal. Nous avons montré que SOX9, cible positive de la voie de signalisation oncogénique Wnt/(beta)-caténine, réprime l'expression de PKC(alpha). Cette répression implique un nouveau mécanisme d'action qui ne nécessite ni la fixation du domaine HMG à l'ADN ni le domaine de transactivation de SOX9. Nous avons également identifié MiniSOX9, un nouveau variant d'épissage de SOX9, résultant de la rétention du second intron. MiniSOX9 est fortement exprimé dans les tumeurs coliques. Il agit en tant que dominant négatif de SOX9, inhibiteur de l'expression du suppresseur de tumeurs PKC(alpha) et activateur de la voie de signalisation oncogénique Wnt/(beta)-caténine. Nos données suggèrent ainsi un rôle primordial de MiniSOX9 dans la tumorigenèse intestinale. Enfin, notre étude protéomique des partenaires de SOX9 et de MiniSOX9 permet d'ouvrir de nouvelles perspectives quant aux rôles de ces deux protéines dans l'homéostasie et la tumorigenèse intestinale / SOX9 is an HMG transcription factor involved in numerous biological processes during development and adult life. It plays an important role especially in the intestinal epithelium homeostasis. In the present study, we demonstrate that SOX9, a positive target of the oncogenic signaling pathway Wnt/(beta)-catenin, represses PKC(alpha) expression. This repression involves a new mechanism of action requiring neither HMG domain binding to DNA nor the transactivation domain of SOX9. We also report the discovery of MiniSOX9, a new SOX9 splice variant, resulting from the second intron retention. MiniSOX9 is highly expressed in colon tumors. It acts as a SOX9 dominant negative, as a repressor of the expression of the tumor suppressor PKC(alpha), and as an activator of the oncogenic signaling pathway Wnt/(beta)-catenin. Our data suggest a crucial role of MiniSOX9 in intestinal tumorigenesis. Finally, a proteomic analysis allowed us to identify potential new SOX9 and MiniSOX9 partners which will be useful to decipher the roles of these two proteins in intestinal homeostasis and tumorigenesis

SOX9 est un facteur de transcription, appartenant à la famille des protéines à domaine HMG, et connu pour réguler la transcription grâce à la liaison de ce domaine à l'ADN. Au laboratoire, il a été montré que SOX9 possédait des propriétés anti-oncogéniques, cependant, de façon paradoxale, SOX9 est surexprimé dans les tumeurs colorectales. Nous avons mis en évidence l'existence d'un nouveau variant d'épissage de SOX9, MiniSOX9, qui possède un effet dominant négatif vis-à-vis de l'activité transcriptionnelle de SOX9. MiniSOX9 est fortement exprimé dans les tumeurs en comparaison avec le tissu sain adjacent à la tumeur. Nous avons émis l'hypothèse que MiniSOX9 pourrait donc avoir, dans les tumeurs, un effet antagoniste à SOX9 et s'opposer à ses propriétés anti-oncogéniques. Grâce à la mise en place de modèles cellulaires tumoraux de surexpression de SOX9 et MiniSOX9, inductibles à la doxycycline, nous avons mis en évidence que SOX9 réduit la prolifération, la migration et l'invasion cellulaire. De manière surprenante, MiniSOX9 ne possède aucun effet sur la prolifération cellulaire, suggérant que les effets de SOX9 pourraient être dus à son activité transcriptionnelle. En revanche, SOX9 ainsi que MiniSOX9 réduisent la capacité clonale des cellules et leur capacité à former des tumorosphères. Dans ce cas, il serait probable que SOX9 et MiniSOX9 modulent l'activité de protéines partenaires / SOX9 is an HMG transcription factor known to regulate transcription by binding of its HMG domain to DNA. We previously demonstrated that SOX9 has anti-oncogenic-properties but SOX9 is overexpressed in colon tumors when compared to adjacent healthy tissu. This overexpression appears paradoxical, unless its anti-oncogenic activity cannot be exert. In this study, we report the discovery of MiniSOX9, a new SOX9 splice variant, which is highly expressed in colon tumors. MiniSOX9 acts as a SOX9 dominant negative isoform. Our hypothesis was that MiniSOX9 antagonizes the SOX9 anti-oncogenic activity in tumors.Using tumors cells lines inducible for SOX9 and MiniSOX9 overexpression, we showed that SOX9 reduces cell proliferation, migration and invasion. Surprisingly, MiniSOX9 has no effect on cell proliferation, suggesting that SOX9 effects could be du to his transcriptionnal activity. However, SOX9 and MiniSOX9 decrease cell clonal ability and tumorosphere formation. In this case, it is likely that SOX9 and MiniSOX9 modulate the activity of proteins partners.

Medulloblastoma (MB) is the most common malignant pediatric brain tumor. Overall survival is about 70% and in cases where current treatment fails, the disease recurs and most often is fatal. At the molecular level, MB can be divided into four defined subgroups: WNT, SHH, Group 3 and Group 4. Amplification of MYC family genes is common in MB and correlates with poor prognosis and tumor relapse. In this thesis we showed how MYCN initiates brain tumors when transduced in neural stem cells (NSCs). Prior to transduction, NSCs were isolated from different brain regions and at various time points. While overexpression of wild-type MYCN did not generate any tumors, orthotopic transplantation of MYCNT58A-expressing forebrain, brain stem and cerebellar NSCs induced diffuse malignant glioma, PNET-like tumors and MB, respectively. Interestingly, MYCNT58A-expressing cerebellar NSCs induced SHH-dependent MB from embryonic cells but SHH-independent MB from postnatal cells. We further showed that cerebellar NSCs transduced with both MYCNT58A and transcription factor SOX9 developed tumors faster and promoted distant migration into the forebrain. The function and regulation of SOX9 in MB cells is poorly understood. We identified SOX9 protein as target of FBW7 ubiquitin ligase and demonstrated the effects of SOX9 on MB cells migration, metastasis and drug resistance. We further blocked PI3K pathway to destabilize SOX9 which sensitized cells to cytostatic treatment. We used a (TetOFF) transgenic mouse model of MYCN-induced MB (GTML) and crossed it with a (TetON) transgene which allowed us to specifically target rare SOX9-positive cells in the tumor. In this system, MB develops spontaneously and SOX9-negative tumor cells can be killed off by doxycycline. The few remaining SOX9-positive cancer cells were able to promote distant MB recurrences. Such a pattern of relapse was recently shown for Group 3 and 4 human MB where about 90% of the recurrences were distant. In summary, this thesis demonstrates that MYCN can generate various types of brain tumors depending on the timing and location of its expression. It further defines the existence of a rare population of SOX9-expressing MB cells that are involved in causing distant MB recurrences. Finally, it describes how SOX9 is stabilized in MB cells and increases MB migration and therapy resistance.

Les malformations obstructives congénitales de l’uretère sont parmi les défauts les plus fréquents à la naissance chez l’Homme mais leur étiologie reste peu comprise au niveau moléculaire.L’uretère assure la fonction essentielle d’acheminer l’urine du rein à la vessie. Mon projet porte sur l’analyse d’un modèle murin d’obstruction fonctionnelle de l’uretère pour lequel le gène Teashirt 3(Tshz3) a été inactivé. J’ai montré que le phénotype d’hydro uretère (dilatation de l’uretère) de ce mutant résulte d’un défaut de différenciation des progéniteurs des cellules musculaires lisses pendant le développement et par voie de conséquence de l’absence de muscle lisse (ML). Chez le mutantTshz3, les cellules mésenchymateuses de l’uretère n’expriment pas le gène myocardin (Myocd) ni ses gènes cibles codant pour les protéines marqueurs du ML. Le facteur de transcription TSHZ3 est donc nécessaire à l’expression de Myocd et à la mise en place du ML. Afin de comprendre le contrôle moléculaire de la différenciation du ML, un crible double hybride a permis d’identifier parmi les partenaires de TSHZ3, la protéine à domaine HMG : SOX9. Au cours du développement de l’uretère,SOX9 s’exprime de manière similaire à TSHZ3 dans le mésenchyme jusqu’à l’apparition des marqueurs du ML, étape qui voit la perte d’expression de Sox9. Pendant l’étape précoce d’initiation dela différenciation TSHZ3, SOX9 et MYOCD sont co-exprimés. Un modèle cellulaire a permis de mieux comprendre les mécanismes mis en jeu pendant cette étape. Dans ce modèle, TSHZ3 et SOX9répriment l’activité promyogénique de MYOCD en l’empêchant de former un complexe activateur avec SRF. Dans l’uretère, la présence de TSHZ3 et SOX9 simultanément avec MYOCD suggère que ces facteurs participent au contrôle temporel de l’initiation du programme myogénique à un moment précis du développement de l’uretère. L’ensemble de ces résultats démontre que Tshz3 remplit des fonctions successives au cours du développement de l’uretère / Congenital obstructives disorders of the ureter are amongst the most common human birthdefects but the molecular networks underlying these defects remain unknown. Ureter plays a pivotalrole in propelling urine from kidney down to the bladder. My project aims at understanding a mouse model of functional obstructive ureter which possesses a null allele for the gene Teashirt3 (Tshz3). Ishowed that the hydroureter phenotype (dilated ureter) in the mutant embryos results of a failure ofureteric smooth muscle (SM) formation occurring since early embryonic steps of SM differentiation.Indeed, mutant ureteric mesenchymal cells do neither express the myocardin (Myocd) gene nor theMyocd targeted genes, encoding for contractile proteins of the SM. Thus, the transcription factorTSHZ3 is necessary for the expression of Myocd and subsequent SM differentiation. To understandthe molecular networks underlying the differentiation of SM, we performed a yeast two-hybrid screenthat allowed us to identify the HMG domain protein SOX9 as a TSHZ3 partner. During ureter development, SOX9 is expressed in an overlapping expression pattern with TSHZ3 until expression ofSM markers genes, a step whereby Sox9 is downregulated. An early initiation step of SM differentiation corresponds to the simultaneous expression of MYOCD with TSHZ3 and SOX9. A cellular model allowed to mimic this stage. TSHZ3 and SOX9 act as repressing factors of the myogenic activity of MYOCD by disrupting the MYOCD/SRF complex activator. These results suggest that TSHZ3 and SOX9 influence MYOCD activity and participate to the temporal control ofthe on set of ureteric myogenesis in the developing ureter. Together, these results demonstrate thatTSHZ3 fulfills successive functions during ureter development

Heart valve structures open and close during the cardiac cycle to provide unidirectional blood flow through the heart, critical for efficient cardiovascular function. Valve dysfunction results in either incomplete opening or incomplete closure of the valve. Both types of valve dysfunction decrease efficiency of blood flow, increasing the load on the myocardium and leading to secondary heart disease such as pathological hypertrophy and heart failure. There are currently no effective treatments to prevent or slow the progression of valve disease, and there are no pharmacological treatments for advanced valve disease. Although most valve disease is associated with aging, increasing evidence suggests that valve disease often has origins in development. Congenital valvuloseptal defects affect many newborns, ranging from life-threatening malformations requiring immediate repair to more subtle, often undiagnosed defects that increase susceptibility to valve disease later in life. Therefore, an improved understanding of the mechanisms of heart valve formation and maintenance of adult valves may serve as an important step in improving valve disease treatment options. In this work, the mechanisms of normal valve development and the role of Sox9 in developing and mature valves are further studied. The temporal and spatial expression of extracellular matrix genes and proteins are examined throughout normal murine valve development. Sox9 function in the processes of valve development and valve maintenance is examined using mouse models of conditional Sox9 loss-of-function. Heart valve phenotypes in mice with reduced Sox9 function are examined throughout development and in adult mice with resultant calcific valve disease. The possible causative mechanisms of calcific valve disease in mice with reduced Sox9 function are further investigated by identification of novel possible targets of Sox9 transcriptional regulation. Together these studies improve our understanding of heart valve development, characterize a model of heart valve calcification with genetic etiology, and identify and characterize novel targets of Sox9.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第18510号 / 医博第3930号 / 新制||医||1006(附属図書館) / 31396 / 京都大学大学院医学研究科医学専攻 / (主査)教授 妻木 範行, 教授 開 祐司, 教授 戸口田 淳也 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(医科学) / 甲第19231号 / 医科博第63号 / 新制||医||科5(附属図書館) / 32230 / 京都大学大学院医学研究科医科学専攻 / (主査)教授 長船 健二, 教授 柳田 素子, 教授 斎藤 通紀 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第20793号 / 医博第4293号 / 新制||医||1025(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 妻木 範行, 教授 鈴木 茂彦, 教授 瀬原 淳子 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第21012号 / 医博第4358号 / 新制||医||1028(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 戸口田 淳也, 教授 鈴木 茂彦, 教授 開 祐司 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM