La crête neurale crâniale (CNC) est une structure transitoire et pluripotente de l’embryon des Vertébrés qui génère la totalité du squelette de la face et de la voûte crânienne et fournit les méninges et une vascularisation fonctionnelle au cerveau antérieur. Précocement, la CNC contrôle également la croissance du cerveau. Pour identifier les mécanismes par lesquels la CNC exerce son rôle trophique sur le cerveau, nous nous sommes intéressés à l’expression du gène Smad1, qui transduit divers voies de signalisation, et est massivement exprimé par les cellules de la CNC juste avant leur migration. L’inactivation de Smad1 par l'interférence ARN dans les CCN conduit à une microcéphalie sévère et une holoprosencéphalie partielle, qui résulte de la perte de l’expression de Fgf8 et Foxg1. Les expériences de sauvetage montrent que les cellules de la CNC régulent positivement Foxg1 indépendamment de Fgf8. De plus, nous montrons que la perte de fonction de Foxg1 dans le télencéphale affecte le développement du thalamus et du toit optique en dérégulant l’expression de Otx2 et de Foxa2 à leur niveau. Nous avons identifié les molécules médiatrices produites par les cellules de la CNC nécessaire au contrôle de l’expression de Foxg1. Nous montrons que les antagonsites de Bmp and Wnt, Noggin, Gremlin et Dkk1 sont indispensable pour initier la spécification du télencéphale. De plus, la régionalisation moléculaire des territoires télencephalique et di/mésencéphalique, requiert l’activité conjointe de Sfrp1 et Sfrp2, d’une part, et de Cerberus, d’autre part. L’ensemble des données acquises au cours de ces travaux documente les mécanismes moléculaires par lesquels la CNC participe de façon essentielle à la régionalisation moléculaire du cerveau des Vertébrés. / The cranial NC (CNC) is a transient structure of the vertebrate embryo, which is essential for brain ontogenesis and provides the developing brain with a skeletal and meningeal protection and functional vasculature. Early in development, CNC cells also control morphogenetic activities of brain organizers and stimulate the growth of prosencephalic alar plate. To understand how CNC conveys its trophic effect on the telencephalon, we have silenced the gene encoding for Smad1, a transcription factor expressed in the CNC cells, which transduces diverse morphogenetic pathways. Smad1 silencing results in microcephaly and partial holoprosencephaly, which early coincide with the loss of Fgf8 and Foxg1 in the telencephalon. Rescue experiments show that CNC cells can positively regulate Foxg1 expression independently of Fgf8 activity in the prosencephalic organizer. Furthermore, the depletion of Foxg1 activity in the telencephalon alters Otx2 and Foxa2 expression in the thalamus and tectum. We have identified the mediators produced by the CNC to control Foxg1 activity and showed that Bmp and Wnt antagonists, Noggin, Gremlin and Dkk1 initiate the specification of the telencephalon. Additionally, the molecular patterning of the telencephalic and di/mesencephalic compartments requires the activity of Sfrp1 and Sfrp2, and Cerberus, respectively. Altogether, we show that CNC cells controls brain patterning by regulating Foxg1 expression through a network of morphogen modulators controlled by Smad1 activity. / A crista neural cranial (CNC) é uma estrutura transiente em embriões de vertebrado, que possui um papel crucial no desenvolvimento da cabeça. A CNC é uma importante fonte de derivados mesenquimais. Recentes descobertas mostraram que as células da CNC possuem uma atividade trófica no desenvolvimento do tubo neural anterior, estimulando e organizando o desenvolvimento prosencefálico em oposição à sinalização Bmp presente nos tecidos adjacentes. Com o objetivo de entender como as células da CNC controlam a atividade de morfógenos durante o desenvolvimento do cérebro. Nós focamos nossos estudos no fator de transcrição Smad1, expresso pelas células da CNC, que controla a transcrição de Noggin. Noggin é um antagonista de Bmp que por sua vez controla a atividade de sua via de sinalização. Além disso, Smad1 interage com outras vias de sinalização com Fgf8 e Wnt. Para testar o papel de Smad1 nas células da CNC, nós eletroporamos o RNA dupla fita de Smad1 (dsSmad1) nas células da CNC em embriões de galinha no estágio de 4 somitos com a finalidade de bloquear sua tradução. Estes espécimes foram analisados em estágio mais avançados do desenvolvimento embrionário. A perda de função de Smad1 compromete o desenvolvimento das vesículas cefálicas, nos estágios de 26 somitos, E4, E6 e E8. Em cortes histológicos em E8, observou-se o aumento do volume ventral do cérebro destes embriões. Com o objetivo de entender como Smad1 controla o desenvolvimento das vesículas cefálicas, embriões no estágio de 26ss foram analisado por hibridização in situ. Nós observamos em embriões dsSmad1 a diminuição da expressão de Fgf8 na borda neural anterior e a completa ausência de expressão de Foxg1 no neuroectoderma prosencéfalico. A falta de Smad1, também gera a diminuição da expressão de Otx2 nos limites ventrais e laterais do telencéfalo, diencéfalo e mesencéfalo. Em contrapartida, nestes embriões observa-se o aumento da zona de expressão de Foxa2 na porção ventral do diencéfalo e mesencéfalo. O bloqueio de Smad1 também acarreta no aumento dos níveis de Dkk1, que é um importante inibidor da via de sinalização Wnt. Com o intuito de entender o mecanismo sobre o controle de Smad1, nós aumentamos os níveis de transcritos nas células da CNC de Dkk1. Como resultado deste aumento, observamos as mesmas modificações nos níveis dos transcritos de Fgf8, Foxg1, Otx2 e Foxa2. Interessantemente os efeitos do excesso de Dkk1 podem ser revertidos com a co-eletroporação do Smad1 constitutivamente fosforilada. Nós também analisamos a expressão de Foxg1 e Otx2 em embriões privados de Cubilin nas células da CNC. Estes embriões apresentam o mesmo padrão de expressão encontrados nos embriões dsSmad1. Interessantemente os nocautes para Cubilin apresentam diminuição da fosforilação de Smad1. Nossos resultados mostram que a presença de Smad1 nas células da CNC é extremamente importante para padronização e desenvolvimento do cérebro. Smad1 nas células da CNC funciona como um regulador da via de Bmp, através do controle transcricional de Noggin impedindo que o excesso de Bmp chegue até o tubo neural. Sendo assim, Smad1 controla o excesso de Bmp permitindo a indução e o desenvolvimento da região anterior por Fgf8.

Glioblastoma (GBM) is the most common intrinsic primary brain tumour. It is uniformly fatal, with median survival approximately 14 months. These tumours comprise a mixture of neural stem cell-like cells and more differentiated astrocytic cells. The former are thought to be responsible for tumour development and recurrence, and display self-renewal and differentiation capacity in vitro. Glioma stem cells (GSCs) are defined operationally by their capacity to initiate tumours on orthotopic transplant into immunocompromised mice. The Pollard lab has identified the neural developmental transcription factor Forkhead Box G1 (FOXG1) as the most consistently overexpressed gene in GBM-derived neural stem (GNS) cells compared to their genetically normal neural stem (NS) cell counterparts. Here we explore the function and critical downstream effectors of FOXG1 in NS and GNS cells. We find that, although FOXG1 is not essential for sustaining proliferation of NS or GNS cells (in vitro), high FOXG1 restricts astrocyte differentiation in response to BMP and can drive dedifferentiation of postmitotic astrocytes. We identify a potential cooperation with SOX2. ChIP-Seq and RNA-Seq were used to define transcriptional targets. FOXG1 directly controls critical cell cycle regulators FOXO3 and FOXO6 (two forkhead family proteins), as well as the proto-oncogene MYCN and key regulators of both DNA and chromatin methylation, including TET3 and CHD3. Pharmacological inhibitors of MYC block FOXG1-driven de-differentiation, whereas Vitamin C and 5-azacytidine – agents that disrupt DNA and chromatin methylation – can facilitate de-differentiation. CRISPR/Cas genome editing was used to genetically ablate the cell cycle inhibitor FOXO3, or remove the FOXG1-bound cis-regulatory region. These data suggest direct transcriptional repression of FOXO3 by FOXG1 may drive cells into cycle. We conclude that high levels of FOXG1 in GBM limit astrocyte differentiation commitment by direct transcriptional control of core cell cycle regulators and DNA/histone methylation.

No description available.

La composition chimique des aliments est perçue par les bourgeons du goût et transmise au cerveau postérieur par des nerfs viscérosensoriels particuliers, les nerfs du goût. L’intégrité de ces nerfs est essentielle au maintien des bourgeons du goût chez les animaux adultes. Cependant, leur rôle dans l’ontogénie des bourgeons, chez l’embryon et aux premiers stades postnataux, est controversé and reste non résolu. Dans cette étude, j’ai établi de façon définitive que la formation embryonnaire des bourgeons du goût dépend des nerfs gustatifs chez la souris, unifiant ainsi les mécanismes de maintien/régénération et d’ontogénie de ces organes. En parallèle, j’ai réexaminé la possibilité (jusque-là exclue par d’autres auteurs) d’un rôle du facteur de transcription Foxg1 dans la formation des ganglions sensoriels épibranchiaux. J’ai découvert que Foxg1 est un déterminant des neurones gustatifs dans le ganglion géniculé. Ce nouveau rôle, de pair avec ceux décrits précédemment dans l’épithélium olfactif, la placode otique et la rétine, révèle une cohérence physiologique remarquable des fonctions de Foxg1 (en dehors de son rôle bien établi dans le cortex) en tant que facteur de transcription maître des neurones impliqués dans les « sens spéciaux » : vision, ouïe, odorat et goût. / Taste information is received by taste buds and transmitted to the hindbrain by special visceral sensory nerves, the taste nerves. The integrity of taste nerves is essential for the maintenance of taste buds in adult animals. However, a role for taste nerves in the ontogeny of taste buds, in the embryo and at early postnatal stages, has been controversial and is still unresolved. In this study, I establish in a definitive manner that embryonic taste bud formation is nerve-dependent in mouse, thus unifying mechanistically the maintenance/regeneration and ontogeny of these organs. Parallel to this work, I re-examined the possibility (previously excluded by other authors) of a role for the transcription factor Foxg1 in epibranchial ganglion formation. I find that Foxg1 is essential for the differentiation of gustatory neurons in the geniculate ganglion. This novel role, together with previously described ones in the olfactory epithelium, otic placode and retina, unveils a striking physiological coherence of the functions of Foxg1 (outside its well established one in the cortex), as a master transcription factor for neurons involved in “special senses”: vision, hearing, smell and taste.

Over the past five years the role of epigenetic modifiers in brain development has become increasingly evident. In this regard, Snf2l, a homolog of the chromatin remodeling protein ISWI, was shown to have enriched expression in the brain and be important for neuronal differentiation. Mice lacking functional Snf2l have hypercellularity of the cerebral cortex due to increased cell cycle re-entry. In this thesis I demonstrate the effects of Snf2l-ablation on cortical progenitor cells including increased proliferation and cell cycle deregulation, the consequence of which is a delay in neuronal migration and altered numbers of mature cortical neurons. This phenotype arises from increased expression of Foxg1, a winged-helix repressor expressed in the forebrain and anterior optic vesicle. Moreover, genetically reducing its overexpression rescues the Snf2l-ablated phenotype. Snf2l is bound directly to a promoter region of Foxg1 suggesting that it acts as a repressive regulator in vivo and is an important factor in forebrain differentiation.

Over the past five years the role of epigenetic modifiers in brain development has become increasingly evident. In this regard, Snf2l, a homolog of the chromatin remodeling protein ISWI, was shown to have enriched expression in the brain and be important for neuronal differentiation. Mice lacking functional Snf2l have hypercellularity of the cerebral cortex due to increased cell cycle re-entry. In this thesis I demonstrate the effects of Snf2l-ablation on cortical progenitor cells including increased proliferation and cell cycle deregulation, the consequence of which is a delay in neuronal migration and altered numbers of mature cortical neurons. This phenotype arises from increased expression of Foxg1, a winged-helix repressor expressed in the forebrain and anterior optic vesicle. Moreover, genetically reducing its overexpression rescues the Snf2l-ablated phenotype. Snf2l is bound directly to a promoter region of Foxg1 suggesting that it acts as a repressive regulator in vivo and is an important factor in forebrain differentiation.

Over the past five years the role of epigenetic modifiers in brain development has become increasingly evident. In this regard, Snf2l, a homolog of the chromatin remodeling protein ISWI, was shown to have enriched expression in the brain and be important for neuronal differentiation. Mice lacking functional Snf2l have hypercellularity of the cerebral cortex due to increased cell cycle re-entry. In this thesis I demonstrate the effects of Snf2l-ablation on cortical progenitor cells including increased proliferation and cell cycle deregulation, the consequence of which is a delay in neuronal migration and altered numbers of mature cortical neurons. This phenotype arises from increased expression of Foxg1, a winged-helix repressor expressed in the forebrain and anterior optic vesicle. Moreover, genetically reducing its overexpression rescues the Snf2l-ablated phenotype. Snf2l is bound directly to a promoter region of Foxg1 suggesting that it acts as a repressive regulator in vivo and is an important factor in forebrain differentiation.

La crête neurale crâniale (CNC) est une structure transitoire et pluripotente de l'embryon des Vertébrés qui génère la totalité du squelette de la face et de la voûte crânienne et fournit les méninges et une vascularisation fonctionnelle au cerveau antérieur. Précocement, la CNC contrôle également la croissance du cerveau. Pour identifier les mécanismes par lesquels la CNC exerce son rôle trophique sur le cerveau, nous nous sommes intéressés à l'expression du gène Smad1, qui transduit divers voies de signalisation, et est massivement exprimé par les cellules de la CNC juste avant leur migration. L'inactivation de Smad1 par l'interférence ARN dans les CCN conduit à une microcéphalie sévère et une holoprosencéphalie partielle, qui résulte de la perte de l'expression de Fgf8 et Foxg1. Les expériences de sauvetage montrent que les cellules de la CNC régulent positivement Foxg1 indépendamment de Fgf8. De plus, nous montrons que la perte de fonction de Foxg1 dans le télencéphale affecte le développement du thalamus et du toit optique en dérégulant l'expression de Otx2 et de Foxa2 à leur niveau. Nous avons identifié les molécules médiatrices produites par les cellules de la CNC nécessaire au contrôle de l'expression de Foxg1. Nous montrons que les antagonsites de Bmp and Wnt, Noggin, Gremlin et Dkk1 sont indispensable pour initier la spécification du télencéphale. De plus, la régionalisation moléculaire des territoires télencephalique et di/mésencéphalique, requiert l'activité conjointe de Sfrp1 et Sfrp2, d'une part, et de Cerberus, d'autre part. L'ensemble des données acquises au cours de ces travaux documente les mécanismes moléculaires par lesquels la CNC participe de façon essentielle à la régionalisation moléculaire du cerveau des Vertébrés.

京都大学 / 0048 / 新制・課程博士 / 博士(生命科学) / 甲第22133号 / 生博第420号 / 新制||生||55(附属図書館) / 京都大学大学院生命科学研究科高次生命科学専攻 / (主査)教授 渡邊 直樹, 教授 千坂 修, 教授 原田 浩 / 学位規則第4条第1項該当 / Doctor of Philosophy in Life Sciences / Kyoto University / DFAM

The main functional cells of the inner ear are neurons and sensory cells that are formed from a common embryonic epithelial neurosensory domain. Discovering genes important for specification and differentiation of sensory cells and neurons in the inner ear is a crucial basis for understanding the pathophysiology of hearing loss. Some of these factors are necessary not only for the inner ear but also for the development of other neurosensory systems such as the visual and olfactory system. The aim of this work was to reveal functions of transcription factor SOX2 in inner ear development by using mouse models with different conditional deletions of Sox2 gene. Sox2 gene was deleted by cre-loxP recombination. In Isl1-cre, Sox2 CKO mutant, reduced number of hair cells differentiated only in some inner ear organs (utricle, saccule and cochlear base) and not in others (cristae and cochlear apex). Early forming inner ear neurons in the vestibular ganglion and neurons innervating the cochlear base developed in these mutants but died by apoptosis due to the lack of neurotrophic support from sensory cells. Late forming neurons in the cochlear apex never formed. In Foxg1-cre, Sox2 CKO mutant, only rudimental ear with no sensory cells was formed. The initial formation of vestibular ganglion with peripheral and...