Activity-dependent adaptation in single neurons in the Drosophila mushroom body

How do neural circuits maintain stable function in the face of developmental changes and natural variability? Many neurons use homeostatic plasticity to adjust their activity to maintain a ‘set point’ level of activity. I investigate Kenyon cells, a population of ~2000 neurons that store olfactory associative memories in the Drosophila mushroom body. Our computational models suggest individual Kenyon cells should equalize the average level of activity across the population to avoid a few neurons responding to most odours while a minority are silent or respond very little. This would increase overlap between representations and impair learned odour discrimination. In this thesis I tested whether single Kenyon cells have activity-dependent compensation mechanisms to prevent such a situation. I tested this by artificially activating single Kenyon cells using the heat-activated cation channel TrpA. I find that in flies heated for 24-96 hours, TrpA-expressing γ Kenyon cells have fewer claws (dendritic input sites) than controls. However, the other two Kenyon cell subtypes, α’/β’ and α/β showed no such adaptation, suggesting different compensatory mechanisms exist between subtypes. Furthermore, this adaptation is not a result of early development flexibility but does disappear with significant age. I propose that this morphological change compensates for increased activity by reducing the number of excitatory inputs the cell receives and thereby decreasing Kenyon cell activity. This is supported by preliminary single cell live imaging experiments using the calcium indicator GCaMP, which showed that the activity of γ KCs hyperactivated for 24 hours is reduced. Together, these results reveal previously unidentified compensatory, potentially homeostatic, mechanisms in single Kenyon cells.