Deletions or mutations in survival of motor neuron 1 (SMN1) cause motor neuron loss and spinal muscular atrophy (SMA), a neuromuscular disorder, with the most severe type manifesting in utero. Whether SMA is a disease of defects in neurodevelopment and/or neuromaintenance remains unclear. We performed an analysis of Smn gene and protein expression during murine embryogenesis. Furthermore, we examined Smn(-/-);SMN2 mice, a model of very severe SMA, for developmental, morphological, and molecular abnormalities. We demonstrate that Smn transcript levels are regulated in a tissue- and developmental stage-specific manner and that the Smn protein expression pattern generally followed that of the Smn mRNA. Cell death and pathological foci were observed in E10.5 Smn-depleted embryos, and this increased in the telencephalon at E14.5. Furthermore, we show an altered morphology of cranial nerves as well as truncated lumbar spinal nerves in a subset of E10.5 Smn(-/-);SMN2 embryos. Finally, we compared the splicing of a subset of genes shown recently to be aberrantly spliced in phenotypic-stage SMA mice. Changes in alternative splicing of the Slc38a5 and Uspl1 genes were detectable in prephenotypic-stage embryos and neonates but became more pronounced with the severity of the phenotype. By comparison, Hif3a alternative splicing was affected only at the end stage of disease. This result suggests that alterations in mRNA splicing in SMA occur, in part, as a result of disease progression. Overall, we conclude that Smn depletion affects developmental processes, which ultimately may also contribute to SMA pathogenesis.

Huntington's disease (HD) is associated with a range of cellular consequences including selective neuronal death and decreased levels of neurogenesis. Ultimately, these altered processes are dependent upon proteins that interact with Huntingtin (Htt) such as the Huntingtin-interacting protein 1 (Hip1) which has a reduced binding preference to expanded Htt. These effects are similar to those observed with modified Notch signal transduction. As Hip1 plays a key role in endocytosis and intracellular transport, and activation of the Notch signal requires both, we investigated putative links between Hip1 and Notch signaling in flies. We have identified two forms of Hip1 that may be produced through the use of alternative first exons: a version of Hip1 with a lipid-binding ANTH domain and Hip1DeltaANTH lacking this domain. The directed expression of Hip1 decreases, while expression of Hip1DeltaANTH increases, the density of sensory microchaetae on the dorsal notum, a classical model of neurogenesis. A reduction in microchaetae density associated with Notch(Microchaetae Deficient (MCD)) (N(MCD) ) alleles is sensitive to both Hip1 and Hip1DeltaANTH levels, as are the bristle phenotypes generated by misexpression of deltex, a key mediator of Notch signaling. Genetic studies further demonstrate that the observed effects of Hip1 and of Hip1DeltaANTH are sensitive to achaete gene dosage while insensitive to the levels of E(Spl), suggesting a non-canonical Notch neurogenic signal through a deltex-dependent pathway. The novel role we describe for Hip1 in Notch-mediated neurogenesis provides a functional link between Notch signaling and proteins related to HD.

In vivo, intracellular recordings of mammalian brain stem motoneurones, followed by peroxidase staining and tridimensional reconstruction, suggest that the shape of the dendritic tree plays an important role in the processing of neural information. To test this hypothesis attempts were made to guide, in culture, the growth of neuritic branches of neurones dissociated from the hypoglossal nucleus of rat brain stem. This was performed using topographical and adhesive microstructures which were designed to control the shape of the neuritic tree. Guidance of the neuritic processes can be observed with small grooves engraved on quartz and plastic substrates, and simple shapes with few processes and bifurcations on each neurite could be obtained using adhesive microstructures. These procedures, which allow the shape of a neurone to be controlled, are very promising in the study, by means of classical electrophysiological methods as well as optical recordings, of the involvement of dendritic architecture in the processing of neural information.