Odors dispersed by the wind form turbulent plumes that contain only stochastic information about source location. A fly navigating towards an attractive odor must therefore combine olfactory cues with information about wind direction and self-motion to correctly locate its source. I propose to use olfactory navigation in the fruit-fly Drosophila as a model system for studying general questions about how neurons represent noisy real-world stimuli, how animals adapt their behavior to changing environmental conditions, and how neural circuits integrate information from multiple senses to guide behavior. My hope is that the genetic tools available in Drosophila will ultimately allow us to answer these questions at a mechanistic biophysical level. In the first part of my post-doc I examined how dynamic stimuli, including plumes, are encoded by olfactory receptor neurons (ORNs) in the Drosophila olfactory periphery (Nagel and Wilson, 2011). Previous studies had observed that ORNs show odor- and cell-dependent dynamics that would seem to make them poorly suited for encoding the rapid fluctuations seen in natural plumes. I found that I could explain these dynamics in terms of two biophysical processes, odor transduction and spiking. Odor transduction gives rise to the odor- and cell-dependence of ORN dynamics, while spiking increases both the complexity of responses, and their speed. This work drew on my graduate training quantifying the response properties of auditory neurons in the songbird (Nagel and Doupe, 2006, 2008). However, it also relied on genetic techniques that I learned during my post-doc. For the second part of my post-doc, I propose to extend this type of analysis to second order olfactory neurons. Specifically I propose to ask how second order neurons encode dynamic plume stimuli, and what circuit and synaptic mechanisms contribute to their responses. This project forms Aim #1 of this proposal. In working on this project I will learn new techniques, such as intracellular recording from central fly neurons. I will also learn to manipulate different parts of a neural circuit and to analyze the results of these experiments critically. Together with the first part of my post-doc, this study will form a template for how to link neural representations of sensory stimuli to biophysical mechanisms. As an independent investigator, I plan to expand my focus to look at the algorithms flies use to localize odor sources (Aim #2) and the central circuits involved in this behavior (Aim #3). This will allow me to differentiate my research program from that of my post-doctoral advisor, Rachel Wilson, and to begin to address larger questions about multi-sensory integration and behavioral choice.
In Specific Aim #2, I propose three novel methodologies for studying olfactory navigation behavior. These approaches will allow me to quantify how flies integrate cues from multiple modalities to decide when to turn, stop, and advance.
In Specific Aim #3, I propose to study how a candidate brain area, the central complex, contributes to these behaviors. Using intracellular recordings, I will ask whether neurons in this area carry the sort of spatial or directional information necessary for navigation. Using genetic lesions I will ask whether mutations of this area disrupt sensory integration or behavioral choice in predictable ways. Together these experiments will allow me to identify the main computations that the fly nervous system must perform in order to successfully localize an attractive odor and to test whether a particular brain area is likely to play an important role in these computations. Most importantly, these experiments will provide a basis for asking mechanistic questions about how sensory input is integrated to guide on-going behavior. Answering these questions is my long-term research goal.
Malaria-bearing mosquitoes and pests that destroy human crops both find their hosts largely by smell. I propose to investigate the behavioral algorithms and neural mechanisms that allow insects to find food by smell. These experiments will allow us to better control harmful insect populations and will provide general insights into how nervous systems guide behavior.