| An important problem in neuroscience is to understand how the brain encodes information. A hypothesis is that the timing of action potentials, reflecting changes in synchronization among neuronal ensembles, can be meaningful to downstream neurons detecting coincident input. Several properties, such as active conductances, feed-forward inhibition and oscillatory input, can determine whether a neuron acts as a coincidence detector. The insect olfactory system, sharing many design similarities with other systems while having a reduced complexity, provides an excellent model in which to study the functional interactions of all these coding features.; This dissertation focuses on the decoding of olfactory information by the mushroom body, the second relay of the insect olfactory system. Kenyon cells, its intrinsic neurons, were found to have very specific and brief odor responses, typically consisting of only one or two reliable spikes, phase-locked to input oscillations. This leads to a dramatic sparsening of the olfactory representation in the mushroom body. Several circuit and intrinsic properties were found to take part in this transformation. Feed-forward inhibition contributes to odor specificity and sparseness: blocking inhibitory input to the Kenyon cells broadened their odor tuning and abolished their phase-locking, supporting the idea that feed-forward inhibition limits the temporal window in which Kenyon cells integrate their inputs. Voltage-dependent conductances contribute to a supra-linear summation of coincident post-synaptic potentials and a reduction of their half-widths, indicating that Kenyon cell intrinsic properties further contribute to coincidence detection. These results describe a mechanism for decoding timing information: not all spikes will be equally relevant, their specific relevance depending on whether they are correlated with other input spikes. In this scheme, oscillations serve as a framework on which Kenyon cells act as coincidence detectors and sparsen the representation. Abolishing the input oscillations disrupted Kenyon cell odor responses, decreasing their specificity and the sparseness of the olfactory representation.; These general features can provide useful insights into coding in more complex neural systems. By illustrating how different neural mechanisms interact to code information and bring about a drastic transformation of sensory representations, this work increases our understanding of how nervous systems can process information. |
