Manipulating Neural Dynamics to Tune Motion Computation in Drosophila

Date of Award

Spring 2022

Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Clark, Damon


Virtually every neuron integrates inputs from excitatory and inhibitory presynaptic cells toproduce its output response. Despite this, it remains poorly understood how the dynamics of these inputs influence the neuron’s output. This is in part due to the difficulty of manipulating neural response dynamics with high specificity. Previous studies have used temperature and pharmacology to alter circuit dynamics (Arenz et al., 2017; Long and Fee, 2008; Suver et al., 2012; Tang et al., 2010), but these methods affect entire circuits, making it difficult to investigate how the dynamics of single excitatory and inhibitory input neurons drive computation. In this study, we use the powerful genetic tools in Drosophila to manipulate the dynamics of individual excitatory and inhibitory visual neuron types, in order to examine how these neural dynamics tune downstream computations. Circuits that detect visual motion offer a compelling framework for understanding howexcitatory and inhibitory input dynamics contribute to the tuning of downstream neurons. This is because in motion detection, responses to visual signals are highly interpretable in their selectivity for the direction and speed of motion. Specifically, Drosophila’s motion detection circuits are anatomically and functionally well-characterized, making this proposal experimentally feasible. Therefore, to study how upstream input dynamics tune downstream computations, we altered the expression of specific membrane ion channels in five cell types in the fly motion detection circuit. By mis-expressing channels in these upstream neurons, we sped up or slowed down their responses and in so doing, identified channels required for their native dynamics. Then, to test models of motion computation, we asked how those manipulations of neural dynamics influence the tuning of downstream motion signals in the direction-selective neuron T4. To do this, we manipulated membrane ion channel expression in individual neuron types upstream of T4 neurons while measuring the responses of T4 to different speeds of visual motion. This experimental protocol allowed us to record how the different biophysical manipulations changed T4’s sensitivity to motion of different speeds. As a result, we measured changes in T4’s tuning to motion velocity and uncovered an amacrine cell’s role in regulating this tuning. We also showed that perturbation of membrane channel expression in interneurons upstream of motion detectors similarly altered the fly’s behavioral response to motion. Last, we developed a data-driven circuit model that is strongly constrained by the anatomical and functional connectivity of cells in the fly’s visual circuit, as well as by our measurements of dynamics. We compared these models to our experimental data, and found that parallel, redundant excitatory and inhibitory inputs are required to explain our experimental data. We also found that the full filtering properties of the inputs—rather than just their response kinetics—are necessary to reproduce our experimental observations. Together, these results reveal how the dynamics of excitatory and inhibitory inputs jointly tune a canonical circuit computation.

This document is currently not available here.