Dynamic top-down biasing implements rapid adaptive changes to individual movements.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Behavioral Neuroscience

Background:

• Complex behaviors rely on coordinated neural activity across brain regions.
• The function of neural coordination, especially rapid co-fluctuations, in supporting behavior is not well understood.
• Top-down influences are hypothesized to guide motor learning and adaptation.

Purpose of the Study:

• To investigate if rapid neural co-fluctuations reflect moment-by-moment task-relevant influences between brain areas.
• To test this hypothesis in the context of birdsong error-corrective adaptation.
• To determine if the lateral magnocellular nucleus of the anterior nidopallium (LMAN) dynamically influences the robust nucleus of the arcopallium (RA) during motor learning.

Main Methods:

• Paired recordings of LMAN and RA activity in singing birds.
• Analysis of neural co-fluctuations to identify top-down influence signatures.
• Perturbation of LMAN activity during specific premotor temporal windows.
• Assessment of effects on birdsong pitch modification during learning.

Main Results:

• A signature of top-down LMAN influence on RA was identified as an LMAN-leading co-fluctuation.
• This co-fluctuation strengthened during learning within a specific premotor temporal window.
• Perturbing LMAN activity in this window rapidly occluded pitch modifications, indicating a localized motor-biasing signal.
• The dynamic top-down influence varied on the timescale of individual movements and was context-dependent.

Conclusions:

• Inter-area neural co-fluctuations can signify dynamic top-down influences crucial for behavior.
• LMAN exerts a temporally localized and context-flexible motor-biasing signal on RA during birdsong adaptation.
• This study reveals a mechanism for how neural coordination supports adaptive motor learning.