Fruit flies produce sophisticated adaptive behavior using relatively few neurons. The fly olfactory system is an excellent system for the study of learned behavior due to its genetic manipulability, measurable behavior, and well-studied neural circuitry. Moreover, the olfactory system is mapped up to the 3rd-order neurons: olfactory sensory neurons project to antennal lobe projection neurons, which in turn innervate two structures called the mushroom body (MB) and the lateral horn (LH). Each of these structures is believed to have a specific function; the MB is required for learned behaviors, whereas the LH is thought to support innate behaviors. However, about 30% of outputs from MB project to the LH, suggesting that in addition to its role in innate behavior, the LH may play a key role in learned behavior. Recent data shows an LH-projecting Mushroom Body output neuron (MBON), V2a, is needed for learned avoidance. This suggests that V2a plays an important role in driving learned avoidance in LH. The anatomical convergence of direct sensory inputs to LH and indirect sensory inputs to LH via MB may be a conserved motif in brain pathways that perform learning or adaptive functions. Indeed, the confluence of more and less processed versions of the same information is seen in many circuits of the mammalian brain. The experiments proposed here test the hypothesis that LH is crucial for learned behavior. The hypothesis predicts that sensory inputs to LH provide raw representations of odors, and MBONs provide representations of those odors that have been modified by learning. It is posited that neurons in LH compare these two information streams to select a behavioral response. To test this hypothesis, light and temperature-dependent tools will be used to manipulate the activity of certain sets of LH neurons, in order to determine roles these neurons play in learned behavior. Then, anatomical tracing will be used to identify LH neurons that are downstream targets of the V2a MBONs, marking an important step towards understanding the neural mechanisms for learned avoidance. These experiments will reveal the roles of identified LH neurons in learned behavior, and reveal the anatomical structure, at the single- neuron level, of the neural circuits beyond MB that support learned behavior. This work provides a tractable means to shed light on the circuits used for learning in Drosophila, and will inform future studies of learning in both insects and mammals.
The ability to adapt to new information and associate sensory stimuli with reward or danger is critical for an animal to flourish in an ever-changing world. Reinforcement learning is crucial for healthy cognitive function, and impaired reinforcement learning is implicated in disorders such as addiction, schizophrenia, and depression, yet fundamental questions remain about the neural circuits that mediate such learning. This proposal harnesses a powerful model system in order to shed light on the neural circuit mechanisms of associative learning and thereby advance our understanding of the elements that are required for normal learning and how perturbation of these mechanisms lead to disease.
