Mapping the Neural Dynamics of Locomotion across the Drosophila Brain

Abstract

SummaryWalking is a fundamental mode of locomotion, yet its neural correlates are unknown at brain-wide scale in any animal. We use volumetric two-photon imaging to map neural activity associated with walking across the entire brain of Drosophila. We detect locomotor signals in approximately 40% of the brain, identify a global signal associated with the transition from rest to walking, and define clustered neural signals selectively associated with changes in forward or angular velocity. These networks span functionally diverse brain regions, and include regions that have not been previously linked to locomotion. We also identify time-varying trajectories of neural activity that anticipate future movements, and that represent sequential engagement of clusters of neurons with different behavioral selectivity. These motor maps suggest a dynamical systems framework for constructing walking maneuvers reminiscent of models of forelimb reaching in primates and set a foundation for understanding how local circuits interact across large-scale networks.

Visual and motor signatures of locomotion dynamically shape a population code for feature detection in Drosophila

“…To extract glomerulus responses from our in vivo imaging volumes, we used techniques similar to other recent imaging alignment studies in the Drosophila brain ( Brezovec et al, 2022 ; Mann et al, 2017 ; Pacheco et al, 2021 ; Turner et al, 2021 ). First, we generated a ‘mean brain’ volume by iteratively aligning and averaging a collection of high-resolution, in vivo anatomical scans of the volume of interest ( Figure 2E , n=11 flies).…”

Descending neuron population dynamics during odor-evoked and spontaneous limb-dependent behaviors

“…This may be because grooming limb movements are more stereotyped and thus may only rely on controllers within the VNC (a notion that is supported by the ability of headless flies to perform spontaneous grooming [51]). Other studies have also shown brain-wide activity during walking but not during other behaviors [52][53][54]. This difference may arise because adaptive locomotion-to avoid obstacles [55], cross gaps [56], and court potential mates [57]-depends heavily on the brain's descending signals.…”

Visual and motor signatures of locomotion dynamically shape a population code for feature detection in Drosophila

“…To extract glomerulus responses from our in vivo imaging volumes, we used techniques similar to other recent imaging alignment studies in the Drosophila brain (Brezovec et al, 2022; Mann et al, 2017; Pacheco et al, 2021; Turner et al, 2021). First, we generated a “mean brain” volume by iteratively aligning and averaging a collection of high-resolution, in vivo anatomical scans of the volume of interest (Fig.…”

“…Natural locomotor behavior modulates the sensitivity of small object detectors Behavior-associated gain changes are widespread in visual systems across phyla (Maimon, 2011;Maimon et al, 2010;McAdams and Maunsell, 1999;McBride et al, 2019;Niell and Stryker, 2010). Recent work demonstrates that locomotor signals are prevalent throughout the Drosophila brain, including in the visual system (Aimon et al, 2019;Brezovec et al, 2022;Schaffer et al, 2021), but has been examined most extensively in circuits involved in elementary motion detection and widefield motion encoding. Behavioral activity has been shown to modulate response gain in widefield motion detecting lobula plate tangential cells (LPTCs) and some of their upstream circuitry (Chiappe et al, 2010;Kohn et al, 2021;Maimon et al, 2010;Strother et al, 2018;Suver et al, 2012), and LPTC membrane potential tightly tracks walking behavior, even in the absence of visual stimulation (Cruz et al, 2021;Fujiwara et al, 2017;Fujiwara et al, 2022).…”

“…For experiments with behavior tracking, we raised a patterned, air-suspended ball underneath the fly to monitor its fictive walking behavior, as in Brezovec et al, 2022. We monitored the fly and ball movement using IR illumination and a camera triggered by our imaging acquisition software at 50Hz frame rate.…”