A circuit for context-dependent behavioral bias in Drosophila melanogaster


April 11, 2018 - 1:00pm
Northwest Building Room 243
About the Speaker
Kyobi Skutt-Kakaria
Speaker Title: 
Graduate Student
Speaker Affiliation: 
de Bivort Lab

Variability across populations of animals is often considered to be a result of experimental noise. Nevertheless, it is clear that individuals display considerable idiosyncrasies in behavior, beyond that expected from sampling error. Individual Drosophila locomoting in a simple Y-shaped maze display individual biases in their preference to turn in one direction or the other, as they repeatedly pass through the center point of the maze. The distribution of individual locomotor biases for a population of flies is unimodal, centered around no bias (choosing randomly), and is greatly over-dispersed in comparison to a null distribution driven purely by sampling error. These biases are generally stable, as flies maintain similar biases in identical conditions for at least 28 days. Surprisingly however, I have recently found that individual biases are context dependent. When flies are exposed to whole field illumination with white light, their locomotor turning bias can dramatically shift from their normal tendency to new biases that are different, but not entirely independent of their original bias and are rapidly reversed by turning the illumination off. The change in bias requires vision as blind flies do not display this modulation. The changes occur rapidly, consistent with a neural dynamics model. I have found that the deep brain circuits in the fly, namely the central complex, is required for robust state switching. Furthermore, I have identified output neurons of the central complex that display morphological correlates of light-dependent modulation of locomotor biases.