Biomechanical origins of proprioceptor feature selectivity and topographic maps in the Drosophila leg

Abstract

Our ability to sense and move our bodies relies on proprioceptors, sensory neurons that detect mechanical forces within the body. Different subtypes of proprioceptors detect different kinematic features, such as joint position, movement, and vibration, but the mechanisms that underlie proprioceptor feature selectivity remain poorly understood. Using single-nucleus RNA sequencing (RNA-seq), we found that proprioceptor subtypes in the Drosophila leg lack differential expression of mechanosensitive ion channels. However, anatomical reconstruction of the proprioceptors and connected tendons revealed major biomechanical differences between subtypes. We built a model of the proprioceptors and tendons that identified a biomechanical mechanism for joint angle selectivity and predicted the existence of a topographic map of joint angle, which we confirmed using calcium imaging. Our findings suggest that biomechanical specialization is a key determinant of proprioceptor feature selectivity in Drosophila. More broadly, the discovery of proprioceptive maps reveals common organizational principles between proprioception and other topographically organized sensory systems.

Keywords: Drosophila; biomechanics; locomotion; mechanosensation; proprioception; somatosensation; topographic.

Figure 1.. Functional subtypes of FeCO proprioceptors are spatially clustered within three discrete compartments in the Drosophila leg.

(A) Top: peripheral anatomy of the femoral chordotonal organ (FeCO, labeled with GFP driven by iav-Gal4 (white); red is a phalloidin stain; blue is cuticle autofluorescence. Bottom: schematic of FeCO organization, based on X-ray reconstruction. (B) Dendrites of each pair of FeCO cells are surrounded by a scolopale cell that connect to a cap cell, which in turn connects to a tendon. (C) Split-Gal4 lines driving RFP label subtypes of FeCO cell bodies in specific locations in the femur. Bottom: a composite schematic showing the relative locations of cell bodies for each FeCO subtype. (D) Two-photon calcium imaging from axons of each FeCO subtype during controlled movements of the femur-tibia joint. Thin traces are from individual flies (n = 5–7) and the thick white line is the response average. (E) Reconstruction from an X-ray microscopy dataset of the fly femur reveals FeCO organization. Each image corresponds to a transverse section indicated in the schematic above. FeCO compartments and tendons are indicated by color shading. (F) The same image planes as C but visualized with confocal imaging (GFP driven by iav-Gal4; pseudo-colored to indicate compartment). Chordotonal cap cells are labeled with an antibody against phalloidin (white). See also Figure S1, and supplemental movies S1 and S2.

Figure 2.. Single-nucleus RNA-sequencing reveals overlapping expression of multiple mechanotransduction channels among FeCO subtypes.

(A) Femurs from 666 fly legs expressing GFP were dissected and nuclei were extracted and collected using FACS for droplet-based (10x) sequencing. (B) UMAP visualization of eight clusters with annotated cell types. See methods for clustering details. (C) Dot plot showing expression levels of cell-type and neuronal markers in each of the eight clusters. (D) UMAP of the three FeCO subtypes from B. (E) Left: Dot plot showing expression levels of cell-type candidate genes for FeCO club (AstaA-R1), claw (Dop2R), and hook (Ca-alpha1T). Right: Mechanotransduction channel expression for each of the three FeCO clusters. In dot plots, color intensity represents mean level of gene expression in a cluster relative to the level in other clusters, and size of dots represents the percent of nuclei in which gene expression was detected. See also Figure S2.

Figure. 3.. The arculum decomposes and transmits distinct mechanical signals to different subtypes of FeCO sensory neurons.

(A) FeCO neurons are mechanically coupled to the tibia via two sensory tendons that converge upon the arculum. (B) X-ray image sections showing how the lateral and medial tendons attach to the arculum. (C) The arculum and its tendons, segmented from a confocal image of the femur. (D) A 3D finite element model of the arculum, stimulated by low amplitude, periodic forces at its base (top). Yellow arrows represent the arculum movement during vibration; the arculum rotates and the attachment point for the medial tendon (right) moves mainly in the z direction, while the attachment point for the lateral tendon (bottom) moves in the x direction. (E) A schematic showing how linear motion is translated into rotation in a manner analogous to a slider-crank linkage. (F) Autofluorescence images of the arculum at different femur-tibia joint angles. As the tibia flexes, the arculum translates toward the femur/tibia joint. A white “x” marks the position of the arculum centroid that we tracked with in vivo imaging in G. (G) Measurements of arculum position during tibia flexion/extension with trans-cuticular two-photon imaging (setup schematized below). The plot above shows arculum centroid position during full tibia flexion (white lines; n = 6 flies). The thick colored line is the average trace (color indicates the tibia angle). See also Figure S3, and supplemental movies S3 and S4.

Figure. 4.. Position-tuned claw neurons exhibit a spatial gradient of cell movement and mechanical strain.

(A) We imaged RFP labeled claw cell nuclei and GFP labeled cap cells (each colored circle indicates the tracked cell) in the femur using transcuticular two-photon microscopy, while swinging the tibia from extension to flexion with a magnetic control system. (B) Example traces showing the position of claw cells (brown circles) and cap cells (grey triangles) during full tibia flexion. The color of each trace indicates tibia angle. (C) The position of cap cells (left) and claw cells (right) along the distal-proximal axis of the femur as the tibia moved from full extension to full flexion (n = 6 flies). We mean-centered the cell position within each fly by subtracting the average claw cell position when tibia is at 90° for each fly. (D) A finite element model of claw and cap cells. Model claw cells connect to the cap cells via their dendrites. The cap cells in turn connects to the medial tendon via tendon fibrils that fan out from the end of the medial tendon. Only 1/2 of the cells in the model are shown for display purposes. (E) Movement of claw and cap cells during tibia flexion (left) and extension (right) vs the cell’s position along the distal to proximal axis of the femur (n = 6 flies). We mean-centered the cell displacement within each fly by subtracting the average claw displacement for each fly. (F) Same as C, but for the movement of cells in the model. (G) A map of strain in the dendrites of model claw cells at different tibia angles. (H) The strain (1^st principal invariant strain) in different claw dendrites plotted against the tibia angle. The dendrites stretched and strain increased as tibia flexed. The strain was always larger in proximal cells and the difference increased as the tibia was flexed. The color of each line represents the cell position along the distal-proximal axis of the femur when tibia is at 90°. See also Figures S4 and S5, and supplemental movie S5.

Figure. 5.. The claw neuron array contains goniotopic map of tibia joint angle.

(A) Example images of GCaMP7f fluorescence (left column) and normalized activity (right column) of flexion-selective claw neurons during slow tibia flexion (6°/s) recorded with two photon microscopy. Each colored circle indicates a tracked cell. Claw cell bodies move distally and increase their calcium activity as the tibia flexes. (B) Tibia angle (top) and normalized calcium activity (bottom) during the example recording shown in A, indicating the tibia angle where each cell reached 50% of its peak activity. The color scheme is the same as in A. More distal cells (darker shades of copper) reached 50% of their peak activity at more acute angles. (C) A linear relationship between each claw cell’s position along the proximal-distal axis of the femur and the tibia angle when the cell reached 50% of its maximum activity. Within each fly, we subtracted the mean of the tibia angle at which cells reached their 50% maximum activity (see Methods for details). (D) Schematic illustrating how the goniotopic map of tibia angle is represented by the activity of the flexion selective claw cells. (E) Illustration of how the strain gradient predicted by the finite element model combined with a uniform threshold for the flexion and extension selective claw neurons lead to a goniotopic map of tibia angle represented by the activities in the array of claw neurons. Strain plot is the same as the one shown in Fig. 4H. Dotted lines represent hypothetical thresholds for the activation of the flexion and extension selective claw neurons. See also Figure S6 and supplemental movie S6.

Figure. 6.. The dendrites of club neurons contain a tonotopic map of tibia vibration frequency.

(A) Schematic of calcium imaging from the club neuron dendrites during tibia vibration using transcuticular two-photon microscopy (left). (B) Examples of calcium activity (ΔF/F, normalized for visualization) in the club dendrites during vibration of the tibia. White circles represent the weighted center of the calcium activity. Green dotted lines show ROI of club dendrites. (C) Average calcium activity (ΔF/F) in the club dendrites of individual flies (n= 17 flies) during the vibration of the tibia at different frequencies. Circles and thin lines are from individual flies and the thick black line represents the average across flies. (D) A scatter plot showing the weighted center of calcium activity for each vibration frequency (200 to 1600 Hz; n = 17 flies). Larger X’s with green outlines indicate averages. The gray line shows the best-fit for the response centers and is also indicated in A to illustrate the primary spatial axis of the tonotopic map. See also Figure S6.