Neural Ensemble Codes in Drosophila Olfaction

点击翻译按钮获取中文摘要

Novel experimental technologies were developed to monitor and manipulate, through light, genetically circumscribed populations of neurons in vivo. These technologies will be harnessed to study information processing in the olfactory system of the fruit fly, Drosophila melanogaster. Cell-type specific promoters and enhancers will target genetically encoded optical sensors and effectors to neurons in four synaptically coupled processing layers: 1) To olfactory receptor neurons, which are located at the sensory surface and project axons to the antennal lobe, the analog of the vertebrate olfactory bulb; 2) to local inhibitory interneurons of the antennal lobe; 3) to projection neurons that relay olfactory information from the antennal lobe to the mushroom body, the analog of the vertebrate olfactory cortex; and 4) to Kenyon cells, the intrinsic neurons of the mushroom body. Odor-evoked neural activity patterns will be visualized by optical microscopy at each of these genetically highlighted processing stages, and the propagation of activity from processing stage to processing stage will be traced. With this innovative strategy at their core, three specific aims directed at the fundamental operational principles of a multilayered neural network will be addressed: 1) The codes employed to represent odor information at each of the four olfactory processing stages, and the mapping rules that relate these codes, will be characterized. 2) The computational origin, physical form, and semantic content of neural symbols that express abstractions, classifications, and behavioral connotations will be elucidated in experiments that combine optical imaging with behavioral assays. The focus of these experiments will be on the neuronal substrate for attaching hedonic valence (i.e., attractiveness or repulsiveness) to odors, and on its modification by experience. 3) Mutations with defined neural or behavioral phenotypes, or artificial neural signals that will be inserted into genetically designated target neurons by selective photostimulation, will be used to perturb neuronal physiology; the impact of these perturbations at the systems level will be analyzed by optical imaging. This opens an opportunity to understand collective network properties in cellular or even molecular terms, and to transcend thereby some of the boundaries that currently divide neuroscience into mo ecu ar, ce u ar, and systems branches.