Distinct sensorimotor encoding in tuft dendrites and somata associated with action, correction, and learning.

Area of Science:

• Neuroscience
• Motor Control
• Dendritic Computation

Background:

• Frontal cortex is crucial for action control and motor learning.
• Layer 5 (L5) neurons' apical tuft dendrites in the frontal cortex may support flexible computation and learning through regenerative events.

Purpose of the Study:

• To investigate sensorimotor encoding differences between apical tuft dendrites and somata of L5 extratelencephalic (ET) neurons during motor skill learning.
• To understand how these neuronal compartments contribute to action control and learning plasticity.

Main Methods:

• Longitudinal two-photon calcium imaging in frontal cortex L5 ET neurons during a cued dexterous action learning task.
• Dissociating error signals from corrective action signals by analyzing movement errors during learning.

Main Results:

• Somatic activity encoded both sensory cues and action, while tuft activity primarily encoded sensory cues.
• Distinct tuft activity selectively associated with corrective actions was observed during learning.
• Learning induced divergent plasticity in response gain and selectivity between tuft dendrites and somata.

Conclusions:

• Apical tuft dendrites and somata of frontal cortical L5 neurons exhibit distinct sensorimotor encoding, sensitivity to corrective actions, and functional plasticity.
• These findings provide a basis for exploring dendritic computation's role in motor skill learning.