🧠 neuroscience

Motor unit mechanisms of speed control in mouse locomotion

By combining behavioral analysis with single motor unit recordings in mice, this study reveals that the nervous system controls locomotor speed primarily through the probabilistic recruitment of diverse motor unit subsets across different muscles, rather than solely through changes in firing rates, thereby linking individual unit activity to specific limb kinematic adjustments.

Original authors: Thomas, K., Gibbs, R., Marques, H., Carey, M. R., Sober, S. J.

Published 2026-02-21

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Thomas, K., Gibbs, R., Marques, H., Carey, M. R., Sober, S. J.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The Big Picture: The Orchestra of Your Muscles

Imagine your body is a massive orchestra, and your muscles are the instruments. For a mouse to walk, run, or change speed, this orchestra needs to play in perfect harmony.

For a long time, scientists listened to this orchestra using a "bulk microphone." This microphone could hear the total volume of the music (the whole muscle working together), but it couldn't tell you which specific violinist or drummer was playing, or how hard they were hitting their drum.

This paper is like putting a tiny, high-tech microphone on every single musician (every single motor unit) in the mouse's arm. A "motor unit" is just one nerve cell and the bundle of muscle fibers it controls. By listening to them individually, the researchers discovered that the mouse's nervous system has a much more clever and flexible way of controlling speed than we previously thought.

The Main Characters: The Two "Arms" of the Muscle

The researchers focused on the mouse's triceps (the muscle on the back of the upper arm). In mice, this muscle has two main "heads" or sections:

The Long Head: Think of this as the Stabilizer. It connects to the shoulder and the elbow. Its job is to keep the arm steady and balanced, especially when the foot first hits the ground.
The Lateral Head: Think of this as the Propeller. It only connects to the elbow. Its job is to push the body forward and extend the arm to get ready for the next step.

The Big Discovery: It's Not Just "Turn Up the Volume"

Old Idea: To walk faster, the brain tells the muscle to "turn up the volume" by making the existing musicians play louder (firing faster).

New Discovery: The mouse brain uses a mix of two strategies, but recruiting new musicians is the most important one.

The "Roll Call" Strategy (Recruitment):
Imagine a band leader who doesn't just ask the current players to play louder. Instead, they call out names from the back of the room.
- At a slow walk, only a few musicians are playing.
- As the mouse speeds up, the leader calls out more names.
- The Twist: These musicians don't play on every single step. They play probabilistically. It's like a game of chance. On some steps, a specific musician is "off duty," and on others, they jump in. As the mouse runs faster, the odds of them showing up increase.
The "Tempo" Strategy (Firing Rate):
Once a musician is on stage, they can also play their notes faster. The researchers found that while the musicians did play faster when the mouse ran faster, this wasn't the main driver. The main driver was getting more people on stage.

Why This Matters: The "Specialized Roles"

The study found that the two "heads" of the muscle have very different personalities:

The Long Head (Stabilizer): These musicians are a bit more unpredictable. Some of them only show up for a few steps out of ten. They seem to be the "emergency responders" or the "specialists" who step in when the mouse needs extra stability or is dealing with a tricky step.
The Lateral Head (Propeller): These musicians are the reliable workhorses. They show up almost every single step, especially right before the mouse lifts its foot to swing it forward.

The "Speed vs. Movement" Connection

The researchers also noticed something fascinating: Who shows up changes how the leg moves.

When the "Stabilizer" (Long Head) musicians showed up, the mouse's elbow didn't extend as far. It was keeping the arm tight and stable.
When the "Propeller" (Lateral Head) musicians showed up, the elbow extended further, helping push the mouse forward.

This means the mouse isn't just running in a straight line with a robotic motion. By deciding which specific motor units to recruit on which specific step, the mouse can subtly adjust its stride, balance, and push-off power in real-time.

The Takeaway

Think of walking like driving a car.

Old View: To go faster, you just press the gas pedal harder (increase firing rate).
New View: To go faster, you also shift into a higher gear (recruit more motor units) and maybe engage the turbo (increase firing rate a little bit).

But the coolest part is that the mouse's brain is like a smart conductor. It doesn't just tell the whole orchestra to play louder. It selectively calls up specific sections of the orchestra at specific times to fine-tune the movement. This allows the mouse to be incredibly agile, adjusting its stride instantly without needing to think about it.

In short: The mouse controls its speed not just by making its muscles work harder, but by constantly shuffling the deck of which specific muscle fibers get to play, creating a flexible and adaptable way to move.

1. Problem Statement

While it is well-established that the nervous system controls locomotor speed by modulating muscle force, the specific mechanisms at the single motor unit (MU) level remain poorly understood. Previous studies relying on bulk electromyography (EMG) have shown that muscle activity increases with speed, but bulk signals obscure the heterogeneity of individual MU firing patterns. Key unknowns include:

How individual MUs are recruited across strides (is it deterministic or probabilistic?).
How MU firing rates and recruitment probabilities scale with walking speed.
Whether closely apposed muscles (e.g., the long and lateral heads of the triceps brachii) exhibit distinct coordination strategies despite performing synergistic functions.
How MU activity correlates with specific limb kinematics (e.g., elbow extension) independent of speed changes.

2. Methodology

The study employed a high-resolution, multi-modal approach combining behavioral kinematics with single-unit neurophysiology in freely moving mice.

Subjects & Surgery: Six mice were implanted with Myomatrix arrays (32-electrode, high-density flexible electrodes) targeting the long head and lateral head of the triceps brachii. The arrays were subcutaneously implanted with threads sutured directly into the muscle bellies.
Behavioral Paradigm: Mice walked on a transparent treadmill at varying speeds (10 cm/s to 27.5 cm/s). Trials were organized in pseudo-random blocks to minimize fatigue effects.
Data Acquisition:
- Kinematics: High-speed video (330 fps) captured via DeepLabCut to track anatomical landmarks (shoulder, elbow, wrist, paw). Stride cycles were defined by the minimum elbow angle (occurring ~50ms before footstrike).
- Neurophysiology: Bipolar EMG signals were recorded at 30 kHz (300–7500 Hz bandpass).
Signal Processing:
- Spike Sorting: Motor units were isolated using Kilosort 2.5, with specific algorithmic adjustments for MU activity (e.g., removing spatial decay assumptions).
- Validation: Units were verified by ensuring <2% inter-spike interval (ISI) violations of a 1ms refractory period and checking waveform stability across trials and stride phases.
- Statistical Modeling: A joint probabilistic model (combining recruitment probability $p$ and Poisson-distributed firing rate $\lambda$ ) was used to distinguish between non-recruitment and recruitment with zero firing.
Analysis: Researchers analyzed recruitment probability, peak firing rates, burst timing, and correlations with elbow angle ( $\Delta\theta$ ) and velocity ( $\Delta\omega$ ) across four speed quartiles.

3. Key Contributions

Probabilistic Recruitment: Demonstrated that individual MUs are not recruited in every stride but rather probabilistically (firing in a subset of strides), challenging the view that bulk EMG consistency implies uniform MU activation.
Muscle-Specific Coordination: Revealed systematic differences in firing patterns between the long head (biarticular: elbow/shoulder) and lateral head (monoarticular: elbow) of the triceps, despite their synergistic role in elbow extension.
Speed Control Mechanism: Quantified that while both recruitment and firing rate increase with speed, recruitment probability is the dominant mechanism for modulating force output in mice, with rate changes being proportionally smaller.
Kinematic Correlation: Established that MU recruitment correlates with stride-by-stride variations in limb kinematics (elbow extension) even when controlling for walking speed.

4. Key Results

A. Probabilistic Recruitment

Individual MUs fired in only a fraction of strides. For example, some long head units fired in <50% of strides, while lateral head units were more consistently active but still showed probabilistic gaps.
Correlated Recruitment: The probability of a unit firing increased significantly when other units in the same muscle were active, suggesting shared synaptic inputs (common drive) rather than independent firing.

B. Muscle-Specific Firing Patterns

Timing: Long head units activated ~100 ms earlier than lateral head units (often before footstrike), whereas lateral head units peaked just prior to liftoff.
Spike Patterns: Long head units showed a larger magnitude ratio between the first and second ISI ( $ISI_1/ISI_2$ ) compared to lateral head units, indicating distinct synaptic input patterns or intrinsic properties.
Functional Implication: The early activation of the long head suggests a role in limb stabilization (biarticular function), while the late activation of the lateral head suggests a role in propulsion (monoarticular power generation).

C. Speed Modulation

Recruitment vs. Rate: As walking speed increased from the slowest to the fastest quartile:
- Recruitment Probability: Increased significantly for 31/33 units (mean increase ~79%).
- Peak Firing Rate: Increased significantly for 22/33 units, but the magnitude of change was smaller (mean increase ~11%).
Conclusion: The nervous system primarily controls speed in mice by recruiting more units rather than drastically increasing the firing rate of already active units.

D. Kinematic Correlations

Lateral Head: Recruitment was associated with greater elbow extension and higher angular velocity.
Long Head: Recruitment was associated with smaller elbow extensions.
These correlations persisted even when analyzing strides within the same speed quartile, suggesting MU recruitment directly influences limb mechanics beyond just speed scaling.

5. Significance

Paradigm Shift: The study moves beyond bulk EMG to reveal that "reliable" muscle activity is actually an emergent property of highly variable, probabilistic single-unit recruitment.
Species Specificity: The findings highlight unique features of mouse motor control, such as extremely high firing rates (50–175 Hz) driven by short afterhyperpolarization (AHP) durations and persistent inward currents (PICs), which differ significantly from larger mammals (cats, humans).
Biomechanical Insight: The distinct timing and recruitment strategies of the triceps heads provide a neural basis for the complex biomechanical roles of biarticular (stabilization) vs. monoarticular (propulsion) muscles.
Future Directions: The results suggest that future models of locomotion must account for probabilistic recruitment and muscle-specific synaptic inputs. Furthermore, the study sets the stage for causal perturbation experiments to definitively link specific MU firing patterns to kinematic outcomes.