Figure 5. Motor neurons are recruited in a specific order across different motor regimes.

(A) Membrane potential, EMG, and probe movement, for two example epochs of spontaneous leg movement during a whole-cell recording of the fast motor neuron. Highlighted events are plotted in D. The EMG was recorded in the proximal femur to pick up both fast and intermediate spikes (scale is 10 pA). (B) Same as A, but for an intermediate motor neuron. The EMG for the intermediate neuron was recorded in the proximal part of the femur, the largest events reflect activity in the fast motor neuron (scale is 100 pA). (C) Same as A-B, but for a slow motor neuron. The EMG was recorded in the proximal femur, near the terminal bristle, and large events are likely from the intermediate neuron (scale is 100 pA). (D) Example force probe trajectories from movement examples. The grayscale in D matches that of the trajectories in A-C. Each vertex is a single frame; vertices with dots indicate the occurrence of a spike in the whole-cell recording. (E) 2D histograms of probe force and velocity for video frames within 25 ms following a spike across recordings in cells of each type (fast – five neurons, 4666 frames; intermediate – four neurons, 11,833 frames; slow – nine neurons, 65,473 frames). Crosses indicate the centroids of the histograms for individual neurons. Dots outlined in white indicate centroid of the 2D histograms of all frames, within and outside the 25 ms window. Color scale shows the log number of frames in each bin, normalized to the total number of frames. (F) Schematic illustrating firing rate predictions of the recruitment hierarchy. As force increases, left to right, first the slow neuron begins firing, then the intermediate neuron (magenta line). Once the fast neuron spikes, both intermediate and slow neurons are already firing (blue dotted line). (G) Left) Spike rate of intermediate neuron in 30 ms preceding fast EMG spike (n = 4 pairs); center) slow neuron spike rate preceding fast EMG spike (n = 2 pairs); right) and slow neuron spike rate preceding intermediate EMG spike (n = 3 pairs).

Figure 5—figure supplement 1. Testing the recruitment hierarchy of motor neurons, in paired recordings and as a function of force probe position and velocity.

(A) The effective spike rate of fast, intermediate and slow motor neurons, over the course of all trials with spontaneous leg movements. Instantaneous spike rates could be much higher. (B) From the 2D spike-triggered phase histograms and Bayes’ rule, we calculated a likelihood function in order to compare spiking regimes in different neurons. If the force probe has a particular position, p, and velocity, q, the likelihood that a motor neuron fired a spike in the preceding 25 ms (right column, linear color scale from 0 to 1) is given by the number of frames with p,q that follow a spike (middle column, shown in Figure 5) divided by the full distribution of p,q (left column). The centroids of the full histograms (left column) are shown in Figure 5. (C) Comparison of activity in recordings of pairs of motor neurons: the number of EMG spikes preceded by a spike in the whole-cell recording, i.e. in a neuron lower down the recruitment hierarchy. The left hand axis indicates the probability of a spike in each millisecond bin for 30 ms before the EMG spike. The right hand axis indicates the cumulative probability of observing a spike in the period before the EMG spike. We found that a small number of intermediate neurons were not preceded by slow neuron spikes (N = 110/3082 spikes). (D) An example trial in which the normal recruitment hierarchy was violated, which only occurred when the leg was unloaded (tibia angle, bottom). During rapid shaking of the leg (highlighted region), the intermediate neuron could spike before the slow motor neuron and have a higher instantaneous spike rate.