The cerebellum has historically been implicated solely in motor coordination. However, we now know it is a major associative center for sensory input. For example, the vestibulocerebellum (VbC) receives a large projection from retinal recipient nuclei in the Accessory Optic System and pretectum; these nuclei analyze visual optic flow information during self-motion and ultimately function in gaze stabilization. In birds, these nuclei project to the cerebellum directly, as mossy fibres, and indirectly as climbing fibres via the inferior olive. Therefore, there are parallel pathways carrying visual information to the cerebellum; these pathways show a remarkable modular organization. In fact, the pattern of cerebellar afferent and efferent connections, the physiological response properties of Purkinje cells, and a variety of molecular markers, all provide abundant evidence that the vertebrate cerebellum is organized into parasagittal zones. This modular organization exists across several different vertebrate species, suggesting that this basic principle of organization has been evolutionarily conserved and is necessary for fundamental information processing in the cerebellum. Although recent advances have increased our understanding of cerebellar organization, the relationship between these parasagittal patterns and the functional behavior of the cerebellum remains unknown. This dissertation seeks to tie together various features of the zonal organization of the cerebellum, using molecular, electrophysiological and neuroanatomical techniques, and to relate this organization to visual motion processing in the VbC. To investigate the parasagittal biochemical organization we use the molecular marker, zebrin (aldolase C), which is expressed by a subset of Purkinje cells, creating a striped appearance in the cerebellar cortex. We provide the first evidence of the zebrin parasagittal expression pattern in an avian species and then further elucidated the relationship between zebrin stripes and the visual afferent climbing fibre input, mossy fibre input, and the physiological zones in the VbC. There is a substantial, and growing, body of evidence that now suggests a more fundamental cerebellar architecture is built around arrays of parasagittal modules. By understanding the basic principles underlying this organization, we gain insights that may lead to a better understanding of the principles of modular organization in the central nervous system in general. / Neuroscience

国立情報学研究所で電子化したコンテンツを使用している。

El cerebelo es el órgano encargado de integrar las vías sensitivas y las motoras del sistema nervioso controlando, por ejemplo, los movimientos involuntarios del cuerpo. Su desarrollo comienza en la etapa embrionaria, en la que tiene lugar la formación del órgano y de sus principales capas. Conforme avanza el desarrollo, tiene lugar un crecimiento del órgano debido a una elevada proliferación de los precursores de las células granulares (PNGCs), localizados en su capa más externa. Su desarrollo finaliza en la etapa postnatal, en la que el cerebelo aumenta aun más su tamaño debido al mantenimiento de esta elevada actividad mitogénica. Alcanza su madurez cuando los PNGCs se diferencian y migran hacia la capa más interna. El morfógeno Sonic Hedgehog (Shh) es el responsable de la proliferación de los PNGCs. Esta proteína activa al complejo de receptores constituído por Patched (Ptch) y Smoothened (Smo), que desencadenaran una cascada de señalización dando lugar a la división celular. La actividad de Shh puede ser regulada negativamente por la proteína quinasa A dependiente de AMP cíclico (PKA). En su estado basal, la subunidad catalítica de PKA (PKA-C) se encuentra inhibida por su asociación con la subunidad reguladora (R-PKA). En la presente tesis se demuestra que en presencia de Shh la subunidad PKA-C inactiva se localiza en la base del cilio de PNGCs, estructura celular necesaria para que la señalización de Shh tenga lugar y donde se acumulan varias moléculas de la vía como Smo y Ptch. Cuando los PNGCs son deprivados de Shh tiene lugar una activación y una parcial dispersión de PKA-C de la base del cilio. Mostramos que la subunidad reguladora RII (PKA-RII) de PKA es la responsable del anclaje de la holoenzima a través de su unión a las proteínas de anclaje de PKA (AKAPs). La disrupción de la interacción entre PKA y AKAPs inhibe la actividad de Shh bloqueando la proliferación de los PNGCs. Estos resultados, por lo tanto, demuestran que el pool de PKA localizado en la base del cilio juega un papel esencial en la integración de la transducción de la señal de Shh. En la segunda parte de la presente tesis estudiamos la activación de proteínas G heterotriméricas mediante Smo. Aunque en su secuencia Smo sugiere que pertenece a la familia de los receptores acoplados a proteína G (GPCR), únicamente en Drosophila melanogaster se demostró que la proteína Gαi es necesaria para la actividad de Hedgehog (Hh), mientras que las evidencias de la implicación de las proteínas G como efectoras de la vía de Shh de los vertebrados no son muy claras. En este trabajo demostramos que la inducción de la proliferación de PNGCs se potencia por la expresión de formas activas de proteínas de la clase Gαi/o (Gαi1, Gαi2, Gαi3 y Gαo) pero no por las de la clase Gα12 de proteínas G. Además, observamos que los miembros de la familia de Gαi poseen un patrón de expresión específico en el cerebelo en desarrollo: únicamente Gαi2 y Gαi3 se encuentran altamente expresados en la parte más externa de la capa granular externa, donde tiene lugar la proliferación de los PNGCs. En concordancia con este resultado observamos que la proliferación de los PNGCs se reduce de manera significativa mediante la eliminación de Gαi2 y Gαi3 pero no cuando silenciamos a otros miembros de la familia de Gαi/o. Finalmente, los resultados de la presente tesis demuestran que las subunidades Gαi2 y Gαi3 se localizan en el cilio primario cuando son sobre-expresados en cultivos de PCGCs. Por lo tanto, los efectos proliferativos de Shh en estas células están mediados por la activación combinada de las proteínas Gαi2 y Gαi3 de proteínas G. / During cerebellum development, clonal expansion of cerebellar granular neuronal precursors (CGNPs) takes place in response to Sonic Hedgehog (Shh) signalling pathway. Shh transduces its signals through the Patched (Ptc) and Smoothened (Smo) receptor complex, delivering activated forms of Gli (Ci) transcription factors to the nucleus. PKA is a known negative regulator of the CGNPs mitogenic activity by inducing Gli2/3 phosphorylation and their subsequent degradation by the proteasome generating potent transcriptional repressors of the pathway. In mammals, Shh pathway requires and takes place at a specific cellular structure: the primary cilum, where many of the proteins of the pathway localize. In this thesis we demonstrated that PKA-C subunit accumulates at the base of the primary cilium through PKA-RII subunit that joins to an unknown AKAP. In the absence of Shh, PKA activates and spreads from the primary cilium base and phosphorylates its target proteins. PKA needs to be accumulated at this specific location, since the inhibition of PKA localization to the cilium base reduces CGNPs mitosis. In the second part of the present thesis we demonstrate that any active member of the Gai/o but not of the Ga12/13 family of heterotrimeric G proteins can induce CGNPs proliferation at low but non active doses of Shh. We also show that Gai2 and Gai3 are the only expressed proteins of this family in the external granular layer of the developing cerebellum (EGL), where proliferation takes place. These proteins are the specific members responsible for Shh activity as their ablation by RNAi has a synergic effect in reducing CGNPs mitogenic activity. Moreover, these subunits can exclusively accumulate at the cilium shaft. All together these results indicate that Gai2 and Gai3 are the two specific members that mediate Shh pathway in this cells.

Duchenne muscular dystrophy (DMD) and congenital muscular dystrophy 1A (MDC1A) are the two most common forms of muscular dystrophy in humans, caused by mutations in dystrophin and laminin α2 genes respectively. Both are severe forms of the disease that lead to premature death due and are both now known to have a significant effect on the central nervous system. This project investigated the role of both proteins involved in each of these diseases in cerebellar Purkinje cells of two murine models of disease: the mdx mouse a dystrophin-deficient model of DMD and the dy2J a laminin α2-deficient murine model of MDC1A. In the case of dystrophin further studies were undertaken in order to determine if increasing age had any effects on cerebellar function. It was found that there is no difference in electrophysiological characteristics (RMP, IR, eEPSP) of the cells when compared to appropriate control groups, nor was there any difference when young and aged dystrophin-deficient mdx groups were compared. Evoked IPSP characteristics were examined in young mdx cerebellar Purkinje cells and again no difference was found when compared to wildtype. There was a significant difference in response to the GABAA antagonist bicuculline, with wildtype increasing eEPSP amplitude by almost double that found in mdx. There was no difference in short term plasticity as measured by paired pulse facilitation in any of these groups. There was no difference in paired pulse depression at the inhibitory interneuron- Purkinje cell synapse of young wildtype and mdx cerebellar Purkinje cells. There a significant blunting of long term depression (LTD, (a form of long term synaptic plasticity) between young wildtype and mdx. When young wildtype animals were compared to aged wildtype animals LTD was found to be similar, when young mdx was compared to aged mdx, there was a recovery of LTD seen in the aged population. There was also significant differences in LTD found when littermate controls were compared to dy2J (laminin α2 mutants). A third of the phenotypic animals (dy2J) potentiated. Finally when rebound potentiation (a GABA-ergic form of long term synaptic plasticity in the cerebellum) was compared in young wildtype and mdx mice, mdx mice displayed depression, rather than the expected potentiation in contrast to potentiation (or no change) as seen in all wildtype cells.

Thesis (M. Phil.)--Hong Kong University of Science and Technology, 2004. / Includes bibliographical references (leaves 84-94). Also available in electronic version. Access restricted to campus users.

Thesis (Ph. D.)--University of Hong Kong, 1999. / Includes bibliographical references (leaves 181-203).

Thesis (Ph. D.)--University of California, San Diego, 1999. / Vita. Includes bibliographical references.

Thesis (Ph.D.)--University of Hong Kong, 1999. / Includes bibliographical references (leaves 181-203) Also available in print.

Recent studies have uncovered de novo mutations of the gene encoding the chromatin remodelling protein Snf2l in patients with schizophrenia, Rett-like syndrome and intellectual disability. Snf2l and its closely related protein, Snf2h, play a critical role in embryonic and post-natal brain development. Murine models lacking functional Snf2h or Snf2l point to complementary activities of these remodelers; Snf2h cKO mice present with a significantly reduced cerebellum, while Snf2l Ex6DEL (exon 6 deleted) cerebella are larger than their wild-type counterparts. Granule neuron progenitors (GNPs) isolated from Ex6DEL cerebella display delayed cell cycle exit and hindered terminal differentiation compared to wild-type controls. Moreover, loss of Snf2l activity results in widespread transcriptome shifts which underlie the Ex6DEL GNP differentiation phenotype. In particular, key transcription factors are differentially expressed without Snf2l remodelling activity. We confirm that ERK pathway activation is misregulated in Ex6DEL GNPs, possibly in response to elevated fibroblast growth factor 8 (Fgf8) expression in these cultures. We find that Snf2l activity maintains the chromatin landscape throughout GNP differentiation, as Ex6DEL cultures have a global increase in chromatin accessibility. We suggest that Snf2l-mediated chromatin condensation is responsible for proper regulation of gene expression programs in GNP differentiation.

Spinocerebellar ataxia type 1 (SCA1) is an adult onset, autosomal dominant neurodegenerative disease caused by a CAG repeat expansion in ataxin-1, which encodes the ataxin-1 protein. SCA1 is one of nine polyQ-expansion gain-of-function diseases which includes Huntington's disease, spinal-bulbar muscular atrophy, dentatorubral-pallidoluysian atrophy and other ataxias. Clinical symptoms of SCA1 include ataxia, dysarthria, ophthalmoparesis, muscle wasting, and extrapyramidal and bulbar dysfunction. Cerebellar Purkinje cells (PCs), neurons in the inferior olive and nuclei of the brainstem are affected. No disease-modifying therapy exists for SCA1. The goals of my thesis were to assess the safety and efficacy of AAV-delivered artificial miRNAs targeting ataxin-1 to alleviate neuropathological and behavioral phenotypes in the knock-in and transgenic SCA1 mouse models. In the knock-in SCA1 mouse model AAVs expressing an artificial miRNA (miSCA1) targeting sequences conserved in mouse and human ataxin-1 were injected directly to the deep cerebellar nuclei. This achieved long term silencing of ataxin-1 mRNA and significantly improved rotarod performance, gait deficiencies, and neuropathology of the cerebellum. In the transgenic SCA1 mouse model the same method of delivery was executed with an artificial microRNA (miR) (miS1) designed to optimize potency, efficacy and safety to suppress Atxn1 expression. Additionally the therapeutic potential of continuous overexpression of ataxin-1-like was examined. Delivery of either ataxin-1-like or miS1 viral vectors to SCA1 mouse cerebellum resulted in widespread cerebellar Purkinje cell transduction. There was significant improvement to rotarod performance, gait deficiencies, coordination and balance, as well as the neuropathology of cerebellar Purkinje cells. In summary, these data indicate the utility of these approaches as possible therapies for SCA1 patients.