Myomatrix arrays for high-definition muscle recording

Abstract

Neurons coordinate their activity to produce an astonishing variety of motor behaviors. Our present understanding of motor control has grown rapidly thanks to new methods for recording and analyzing populations of many individual neurons over time. In contrast, current methods for recording the nervous system's actual motor output - the activation of muscle fibers by motor neurons - typically cannot detect the individual electrical events produced by muscle fibers during natural behaviors and scale poorly across species and muscle groups. Here we present a novel class of electrode devices ("Myomatrix arrays") that record muscle activity at unprecedented resolution across muscles and behaviors. High-density, flexible electrode arrays allow for stable recordings from the muscle fibers activated by a single motor neuron, called a "motor unit", during natural behaviors in many species, including mice, rats, primates, songbirds, frogs, and insects. This technology therefore allows the nervous system's motor output to be monitored in unprecedented detail during complex behaviors across species and muscle morphologies. We anticipate that this technology will allow rapid advances in understanding the neural control of behavior and in identifying pathologies of the motor system.

Figure 1:. Myomatrix arrays record muscle activity at motor unit resolution.

(a) The nervous system controls behavior via motor units, each consisting of a single motor neuron and the muscle fibers it innervates. Each motor neuron’s spiking evokes motor unit action potentials in the corresponding muscle fibers. Myomatrix arrays (right) bearing 32 electrode contacts on a flexible substrate (Supplemental Fig. 1) can be targeted to one or more muscles and yield high-resolution recordings of motor activity during free behavior. Motor neurons, muscle fibers, and electrode arrays are not shown to scale. (b,c) Example recordings from the right triceps muscle of a freely behaving mouse. (b) Top, bipolar Myomatrix recording from the mouse triceps during locomotion. Blue dots indicate the spike times of one motor unit isolated from the data using a spike sorting method based on principal components analysis (Supplemental Fig. 2a–d). Bottom, example data (from , used with permission) from traditional fine-wire EMG recording of triceps activity during locomotion. Applying the PCA-based spike sorting method to the fine-wire data did not isolate any individual motor units. (c) Unipolar Myomatrix recording during quiet stance. Colored boxes illustrate motor unit action potentials from four identified units. Spike waveforms from some units, including those highlighted with gray and orange boxes, appear on multiple electrode channels, requiring the use of a multi-channel spike sorting algorithm (Kilosort 2.5, see Supplemental Fig. 2e–h). (d) Spiking pattern (tick marks) of six individual motor units recorded simultaneously during locomotion on a treadmill. The three bursts of motor unit action potentials correspond to triceps activity during three stride cycles. Motor unit 4 (cyan) is the same motor unit represented by cyan dots in (b). The other motor units in this recording, including the smaller amplitude units at top in (b), were isolated using Kilosort but could not be isolated with the PCA-based method applied to data from only the single recording channel shown (b).

Figure 2:. Myomatrix recordings across muscles and species.

(a) Example recording from mouse triceps during a head-fixed pellet reaching task. Arrows at top indicate the approximate time that the animal’s paw leaves a rest position and first contacts the target. Bottom, colored boxes highlight motor unit action potentials identified using Kilosort . Different box colors on the same voltage trace indicate distinct motor units. (b) Recordings from the mouse superficial masseter muscle were obtained in anesthetized, head-fixed mice when passive mandible displacement evoked reflexive muscle contractions. Top trace shows the lateral component of jaw displacement, with arrows indicating the direction and approximate time of displacement onset. (c) In a recording from rat triceps during head-free locomotion, the arrowhead indicates the time that the mouse’s paw touched the treadmill surface, marking the beginning of the stance phase. (d) Recording from the rat flexor digitorum profundus muscle during a pellet reaching task, arrow indicates the time of grasp initiation. (e) Myomatrix recording of motor unit activity in the mouse bulbospongiosus muscle evoked by optical stimulation of spinal motor neurons, producing motor unit spikes at latencies between 10–15 msec, consistent with results obtained from traditional fine-wire electrodes in mice . (f-j) Recordings from the cat soleus (f) during sensory nerve stimulation, songbird vocal (ventral syringeal) muscle (g) and expiratory muscle (h) during quiet respiration, hawkmoth larva dorsal internal medial (DIM) muscle (i) during fictive locomotion, and bull frog semimembranosus (SM) muscle (j) in response to cutaneous (foot) stimulation. Spike times from individual motor units are indicated by colored tick marks under each voltage trace in f–j. Recordings shown in panels (a, c, g, h, i, and j) were collected using bipolar amplification, data in panels (b, d, e, and f) were collected using unipolar recording. See Methods for details of each experimental preparation.

Figure 3:. Motor unit recordings during active movement in primates.

(a) An injectable version of the Myomatrix array (Supplemental Fig. 1g) was inserted percutaneously (Supplemental Fig. 1i) into the right biceps of a rhesus macaque performing a cued reaching task. Green and red dots: reach start and endpoints, respectively; grey regions: start and target zones. (b) Recording from five of 32 unipolar channels showing spikes from three individual motor units isolated from the multichannel recording using Kilosort (Supplemental Fig. 2). (c) At trial onset (dotted line), a sudden force perturbation extends the elbow, signaling the animal to reach to the target. (d) Spike times (tick marks) from 13 simultaneously recorded motor units. (e) Example voltage data from a Myomatrix array (top) and traditional fine-wire EMG (middle, bottom) collected from the same biceps muscle in the same animal performing the same task, but in a separate recording session. Gray traces (bottom) show smoothed EMG data from the fine-wire electrodes in all trials, orange trace shows trial-averaged smoothed fine-wire EMG, dark gray trace represents the fine-wire trial shown at middle. (f) Spike times of four motor units (of the 13 shown in d) recorded simultaneously over 144 trials.