Motor Cortex Embeds Muscle-like Commands in an Untangled Population Response

Area of Science:

• Neuroscience
• Motor Control
• Computational Neuroscience

Background:

• The primate motor cortex (MC) extensively projects to spinal circuits, suggesting its activity might primarily consist of muscle-like commands.
• Existing evidence from reaching tasks is ambiguous, leading to ongoing debate about the nature of motor cortex representations.

Purpose of the Study:

• To investigate the structure of population activity in the primate motor cortex using a novel behavioral paradigm.
• To compare time-evolving neural and muscle activity patterns to understand motor command encoding.

Main Methods:

• Utilized a novel behavioral paradigm enabling direct comparison of neural and muscle activity over time.
• Analyzed single motor cortex neuron properties and population-level activity structures.
• Employed network models to explore the functional implications of observed neural activity patterns.

Main Results:

• Individual motor cortex neurons exhibited muscle-like properties, but population activity structure differed significantly from muscle activity.
• Neural population activity was characterized by an avoidance of 'tangling'—a state where similar neural states diverge unpredictably.
• This low-tangling feature was consistent across different tasks and primate species, suggesting a fundamental principle.

Conclusions:

• Motor cortex population activity is not dominated by muscle-like commands, possessing a distinct structure.
• The avoidance of tangling in neural activity enhances noise robustness, crucial for reliable motor control.
• Muscle-like commands may be embedded within a larger, low-tangling structure within the motor cortex.