research synopsis

How are flexible and robust animal behaviors orchestrated by the nervous system? Different forms of this general question have occupied neuroscientists for decades. Great strides have been made toward describing the elements of nervous system development, structure, and function. Our next challenge is to examine how behavior emerges from the interactions among genetic, cellular, cell-system, and organ-system levels of organization. My laboratory studies these interactions in a powerful model system, the fruit fly Drosophila melanogaster. Whereas research with Drosophila is most often focused within the molecular-genetic spectrum of modern biology, this animal also shows remarkable behavioral performance, making its living navigating vast distances through complex visual landscapes in search of the source of an attractive odor. A fly's sophisticated navigation capabilities emerge from the fusion of multiple sensory modalities and transformation of a robust motor code. By combining the rapidly expanding toolkit of fruit fly molecular genetics with state-of-the-art engineering techniques, we hope to reveal the functional mechanisms and structural circuits with which the fly brain coordinates the mechanics and dynamics of complex behavior. The results of this cross-disciplinary approach could have broad impact on our understanding of the general principles of sensory fusion and sensory-motor integration common among animal taxa, and also motivate specialized technical advances in bio-inspired robotic devices.

Advances in modern neuroengineering such as functional imaging, ensemble recording arrays, and optically activated ion channels are providing an ever clearer picture of high-level sensory processing and sensory-motor integration by the brain. The next challenge is to move beyond individual sense-by-sense analyses and explore how information from multiple sensory modalities is fused and then integrated with the biological mechanics of locomotion. To meet this challenge, we study one of the most sophisticated multi-sensory behaviors on the planet, fly flight. Neuroscientists and engineers alike marvel that the robustness and flexibility of fly behavior is somehow controlled by a "processor" the size of a poppy seed. It seems somewhat intuitive given the complexity of even the tiny fly brain that multi-sensory interactions represent something greater than the sum of their individual parts. A hypothesis central to our research is that cross-modal interactions are neither continuously independent, nor wholly synergistic, but rather show a combination of both depending upon stimulus conditions. For example, olfactory cues might enhance the sensitivity of visual processing in very dim or low contrast conditions, but have no influence under bright high contrast conditions. We will test for synergistic cross-modal integration according to formal mathematical criteria for linear superposition and multisensory enhancement.

We take advantage of powerful gaze-stabilizing visual reflexes displayed by flies and humans alike to map spatial, temporal, and contrast parameters onto motor control functions. These analyses are performed for individual flies suspended within electronic flight simulators and also for groups of flies walking within a "virtual treadmill" (see movie gallery). The unique aspect of our approach is that for each experiment we can precisely manipulate and measure sensory stimuli, yet the fly actively controls its own sensory-motor experience under closed-loop feedback conditions. For example, within the tethered flight simulator, when the fly steers left, the visual display moves right, as expected under natural conditions. Similarly, a magnetic tether system allows a fly to spin freely in the yaw plane beneath a radial array of odor plumes. Finally, we use achromatic motion or stationary color cues to repeatedly and reversibly "stampede" a population of flies into the center of a 20 centimeter tube. Risking overconfidence, we know of no other neurobiological research systems that afford similar quantitative measurements while at the same time providing the animals with vast output degrees-of-freedom. Using these systems, we quantify responses to variations in the spatial, temporal, and intensity characteristics of visual and olfactory stimuli, then evaluate functional roles of specific neural pathways using standard genetic techniques.

Ultimately, we are interested in identifying the neural circuits responsible for cross-modal processing. We will make use of state-of-the-art molecular-genetic techniques available in fruit flies to selectively identify and modulate the activity of specific classes of neurons. We will then characterize performance deficits with the quantitative behavioral analyses outlined above. Genetically introduced transcriptional activators fused to a gene-of-choice allow us to control the spatial and temporal expression of reporters like green fluorescent protein or effectors such as tetanus toxin or optically activated ion channels to stimulate and suppress neural activity. We have isolated several such lines localized within the early visual pathway. Flies for which these circuits are inactivated show specific deficits to motion or color vision, but not both. We shall expand these analyses to include circuits mediating visuo-olfactory fusion.

1) mapping the spatial, temporal, and contrast properties of optomotor responses during flight

2) using the p[Gal4] UAS-"gene X" transgene system for behavioral analysis of visual circuit function

3) investigating properties of contrast and texture based ("second-order") motion vision

4) an unbiased screen of mobilized p-element mutants, tested for walking optomotor performance

5) examining visual and olfactory sensory sensory fusion and sensory-motor integration during flight

7) mapping the spatial and temporal distribution of olfactory sensory activity onto behavior

8) exploring the cross-modal sensitization of optomotor behaviors by mechanosensory stimuli

9) examining the transition between sensory independent exploratory behavior and sensory dependent active search

10) fabrication of miniturized multi-unit recording electrodes for the fruit fly CNS