Skip to contentSkip to filters

We use cookies for analytics.

Descending control of turning during walking

Neuroscience ePoster by Helen Yang, Quinn Vanderbeck, Rachel Wilson, Laia Serratosa Capdevila, Anna Li, Jasper Phelps and 4 more

Part of COSYNE 2023 posters, Harvard Medical School; Department of Neurobiology; Harvard Medical School; EPFL

Saturday

COSYNE 2023
Montreal, Canada
Related content

Abstract

To understand how the brain controls behavior, we need to understand the neural code for movement. Descending neurons (DNs) represent a bottleneck for motor control signals, and so provide an opportunity to study neural coding in a compact form. In Drosophila, almost all DNs are uniquely identifiable in the connectome of the central nervous system and targetable with specific genetic drivers, allowing us to study their connectivity and physiology in parallel. Previous studies have concluded that these DNs generate high-level motor commands, such as “stop” or “turn”, leaving the detailed implementation to the circuitry of the ventral nerve cord. Here, we challenge this conclusion. First, we use two-photon imaging to show that there are many DNs whose activity correlates with turning during walking. To understand their potential specializations, we focused on two of these cells, DNa02 and DNg13. We found that they have distinct synaptic inputs in the brain connectome and also distinct synaptic outputs in the ventral nerve cord connectome; this implies they can be recruited independently and that they drive different features of movement. We found that directly activating DNa02 shortens the stride on the inside of a turn, while activating DNg13 lengthens the stride on the outside of a turn. Using electrophysiological recordings in walking flies, we found that DNa02 is mainly active during fast turns that arrest forward movement, whereas DNg13 is more broadly active, spiking during fast and slow turns regardless of forward movement. Moreover, DNa02 is recruited abruptly, whereas DNg13 is recruited gradually. Together, these results clearly show that DNs specify limb movements, rather than merely issuing high-level commands, and that different DNs are specialized to independently control different limb movement features. Our work illustrates how DNs can be recruited to execute a complex motor output with flexible kinematics and vigor.

Topics

Details

Session
Poster Session II