The general problem that I will address is how neurons integrate their inputs and compute their outputs. Central to integration and computation is the morphology of neurons, and in particular that of their highly branched dendritic arbors, which receive synaptic input from other neurons or sensory input from the outside world. The connections between axonal and dendritic processes define the nervous system's structure, which is viewed as a prerequisite for understanding neural function; connectomics, the global study of neuronal connectivity, has emerged as a major goal of neuroscience. In this Pioneer proposal, I want to take an orthogonal approach to neuronal morphology. My hypothesis is that the cell biology of the neuron-the transport and turnover of materials-places very strong constraints on both building and maintaining dendrites. Furthermore, I propose that these constraints are so strong that they actually compromise the functioning of neurons: I hypothesize, for example, that the changes in diameters of dendritic processes across branch junctions are dictated by transport constraints and that they actually degrade signal propagation. If this is true, then morphology is a compromise between cell biology and neuronal function, and determining the nature of the tradeoff is likely to provide key insight into connectivity. The morphological rules that I will uncover will provide powerful a prioris for determining connectivit maps, and may help to solve a major problem in connectomics: how well does the connectivity map need to be in order to understand the function? To test this hypothesis, one needs a system in which morphology can be measured precisely (and in the most general sense, which includes protein localization), where it can be manipulated in a controlled way, and where morphology can be correlated with function. The Class IV dendritic arborization mechanoreceptor of Drosophila larvae meets these requirements, and will be the
! The relationship between the morphology of neurons and the way that they process information is poorly understood. We will investigate the following questions: how do cell-biological limitations constrain neuronal morphology and how does this impact neuronal function? The transformative nature of this approach is that the principles uncovered here are likely to have a large impact on our understanding of neuronal connectivity and architecture.