Straw Group Research

The neural basis of locomotory visual guidance in Drosophila

The brain of a fly is capable of steering the animal through a complex environment at high relative speeds, avoiding stationary obstacles and moving predators. Because it is relatively easy to study how flies do this at several levels, from the behavioral to the cellular, fly vision has long been recognized as an ideal system to address a fundamental question in neuroscience -- how does the distributed activity of neurons orchestrate animal-environment interactions to result in successful coordinated behavior? We work on this basic question using techniques including automated realtime fly tracking, virtual reality displays, molecular genetic tools, and neuroanatomy.

A powerful toolkit to study neural basis of visual behavior

Figure 1 (Click to view legend)

To enable our experiments, we use and develop vir tual reality systems. In one such apparatus, freely flying flies are tracked in realtime by a multi-camera computer vision system. This unique technology allows rapid testing of physically unmanipulated and unrestrained flies to repeated presentations of arbitrary visual stimuli projected on the arena walls and floor. This device is used to perform experiments that allow the fly to exhibit a large portion of its natural behavioral repertoire while simultaneously measuring and limiting the effect of behavioral variability. Thousands of digitized 3D flight trajectories are collected in virtual environments with the capability of experimentally triggered events such as object disappearance.

Such behavioral experiments are combined with targeted genetic manipulation of the nervous system and analyzed to reveal the magnitude and reliability of effects. The activity of individually identified neurons is perturbed via transgenic expression of exogenous ion channels or altered synaptic machinery, and using statistical tests from the field of machine learning, the effects of these genetic manipulations on flight control are quantified. Ultimately, these techniques to control stimulus conditions and measure behavioral responses in detail allow us to show the contribution of individual neurons to behavior.

Mapping the visual circuits of the fly brain

Figure 2 (Click to view legend)

Neurons in the fly visual system, especially the lobula plate tangential cells, have been used for decades as a model to study cellular neurophysiology including the basis of visual motion detection and spike timing precision, the behavioral role of most of these cells remains unclear. To link our knowledge of cellular properties with natural behavior, we are performing genetic ablations on small numbers of neurons and measuring behavioral deficits in response to precisely defined visual stimuli. To attain greater cellular specificity than allowed by the standard Drosophila GAL 4-UA S system, we are making use of recent intersectional techniques such as split-GAL 4 and alternative binary expression systems such as LexA. Flies with such manipulations are tested behaviorally in our virtual reality free flight arena, and the affected neurons are identified using immunostaining and confocal microscopy. The aim of this work is to establish a causative link between identified cells and their function in visual guidance.

Figure 3 (Click to view legend)
Figure 4 (Click to view legend)

Further reading:

Straw, A.D., Lee, S., and Dickinson, M.H. (2010) Visual control of altitude in flying Drosophila. Current Biology
Straw, A.D., Branson, K., Neumann, T.R., and Dickinson, M.H. (2011) Multi-camera Realtime 3D Tracking of Multiple Flying Animals. Journal of the Royal Society Interface.