Nested mechanosensory feedback actively damps visually guided head movements in Drosophila
- Department of Mechanical Engineering, Pennsylvania State University, United States
Figures
Figure 1
Parallel and nested sensory fusion in biological systems.
(A) Control model of parallel sensory fusion. Multiple sensory systems, S1 and S2, measure an external reference state with respect to the system’s motion . The information measured by S1 and S2…
Figure 2
Control model of visual and nested mechanosensory feedback during gaze stabilization in fly flight.
(A) The control framework used to model and analyze the gaze stabilization system in body-free flies. Flies respond to an external visual perturbation by attempting to minimize the sensory visual …
Figure 3 with 6 supplements
Sensory feedback generated from body movements alters the magnitude, timing, and performance of head responses.
(A) The body response (red) of body-free flies to single-sine visual perturbations (grey) with varying frequency. The x-axis is normalized to show four oscillations at each frequency. Note that the …
Figure 3—figure supplement 1
Sum-of-sines body and head responses.
(A) Top: the body response (red) of body-free flies to a sum-of-sines visual perturbation (black). Bottom: the head response of body-free (blue) and body-fixed (violet) flies to the same visual …
Figure 3—figure supplement 2
Saturation-corrected head response and LSSA sensitivity analysis.
(A) A saturation-correction routine applied to the mean head response of body-fixed flies at the 0.7 Hz perturbation frequency. We removed the saturated data points and fit a sine wave to the …
Figure 3—video 1
Comparison of a body-free fly (magnetic tether, left) and body-fixed fly (rigid tether, right) head response to a 1 Hz sine wave visual perturbation.
The body-stabilized head view is shown below each raw video. Note that head movements are larger in the body-fixed fly.
Figure 3—video 2
Same as Figure 3—video 1 but for a 2.1 Hz visual perturbation.
Figure 3—video 3
Same as Figure 3—video 1 but for a 5.3 Hz visual perturbation.
Figure 3—video 4
Same as Figure 3—video 1 but for a sum-of-sines visual perturbation.
Figure 4 with 3 supplements
Visual and mechanosensory feedback mediate head control.
The legend indicates whether the data corresponds to the head or body, , the sources of sensory feedback, and the relevant experiment (or prediction). (A) The closed-loop transform from the visual …
Figure 4—figure supplement 1
Sum-of-sines transforms.
A Same as Figure 4 and Figure 4—figure supplement 3, but for sum-of-sines visual perturbations with different normalized velocities. (B) Same as Figure 5C but for the sum-of-sines perturbations. …
Figure 4—figure supplement 2
Control diagrams and replay experiment for manipulations of sensory feedback.
(A) Control diagram for the prediction of the head response with head and body visual feedback (corresponding to Equation 7). (B) Control diagram for the prediction of the head response with head …
Figure 4—figure supplement 3
Overlaid transforms.
(A) Same as Figure 4, but with all transforms plotted on the same axes and the coherence of the experimentally measured responses. Shaded regions: ±1 STD. Body-free: flies, Body-fixed: flies, …
Figure 5
Nested mechanosensory feedback damps head movements.
(A) The control diagram of the sensory error to head transform in body-free flies. Note that this transform includes nested mechanosensory feedback from body motion. (B) The control diagram …
Figure 6 with 7 supplements
Head damping is present during self-generated but not externally generated body motion.
(A) The predicted transforms from body mechanosensory information to the head response for self-generated body motion (pink) and externally generated body motion (blue). Shaded regions: ±1 STD. (B…
Figure 6—figure supplement 1
Passive head movements.
(A) Flies were anesthetized with triethylamine (commercially available as FlyNap, Carolina Biological Supply) and rigidly tethered to the shaft of a stepper motor (Nema 17). The body motion of …
Figure 6—video 1
A fly was mounted on a motor and rotated about the vertical (yaw) axis such that the body angle matched that of a magnetically tethered fly in response to a 1 Hz sine wave visual perturbation (see Figure 3A and Figure 3—video 1).
Body visual feedback was removed by mounting the visual scene to the motor. The head response (right) of the same fly in the body-stabilized coordinate frame. Note that the head response was …
Figure 6—video 2
Same as Figure 6—video 1 but for a 2.1 Hz perturbation.
Figure 6—video 3
Same as Figure 6—video 1 but for a 5.3 Hz perturbation.
Figure 6—video 4
Same as Figure 6—video 1 but for an anesthetized fly.
Figure 6—video 5
Same as Figure 6—video 2 but for an anesthetized fly.
Figure 6—video 6
Same as Figure 6—video 3 but for an anesthetized fly.
Figure 7 with 2 supplements
Head saccades are actively damped by mechanosensory feedback.
(A) Example body (red) and head (blue) trajectories for a body-free fly in the magnetic tether presented with a static visual stimulus. Rapid flight turns called saccades are highlighted. Note that …
Figure 7—figure supplement 1
Head saccades in darkness.
A Same as Figure 7D, but for a dark visual environment. Shaded regions: ±1 STD. Body-free: flies, saccades. Body-fixed: flies, saccades.
Figure 7—video 1
A body-free fly (magnetic tether) presented with a static visual background performing simultaneous body and head saccades.
The left video shows the raw images and the right video shows the images in the body-stabilized coordinate frame. See Cellini et al., 2021 for videos of saccades in body-fixed flies (Cellini et al., …
Author response image 1
Left: the idealized smooth sine wave at the highest frequency designed for our flight simulator (black) vs the actual displayed signal (red).
Note that our flight simulator display has an angular resolution of 3.75°. Right: same as the left, but for the Fast-Fourier Transform (FFT) magnitude of the two signals.
