Coupling between motor cortex and striatum increases during sleep over long-term skill learning

Stefan Lemke, Dhakshin Ramanathan, David Darevksy, Daniel Egert, Joshua D. Berke, Karunesh GangulyeLife 2021 · DOI

TL;DR

This study reveals that sleep, specifically NREM sleep, plays a crucial role in strengthening the neural connections between the motor cortex and the striatum during long-term skill learning. Blocking NMDA receptors in the striatum during offline periods (like sleep) impairs skill acquisition, suggesting that sleep-dependent plasticity in this pathway is essential for learning.

What Problem Does This Paper Address?

Learning new motor skills, like playing a musical instrument or mastering a sport, involves refining the coordination between different brain regions. The motor cortex (responsible for planning and executing movements) and the striatum (involved in habit formation and skill learning) are key players. While we know that the connections between these areas change with learning, it's been unclear when this crucial plasticity happens: during active training or during rest periods like sleep. This paper investigates whether "offline" periods, particularly sleep, are critical for strengthening corticostriatal connections and facilitating skill learning.

Methods

The researchers used a combination of electrophysiology and behavioral tasks in rats.

  • Electrophysiology: They recorded neural activity (local field potentials and single-unit spiking) from the primary motor cortex (M1) and the dorsolateral striatum (DLS) simultaneously.
  • Behavioral Task: Rats were trained on a reach-to-grasp task over several days.
  • Pharmacological Intervention: They infused an NMDA receptor antagonist (AP5) or saline into the DLS after training sessions to investigate the role of striatal NMDA receptor activation in offline plasticity.
  • Connectivity Analysis:
    • LFP Coherence: They measured the synchronization of LFP signals between M1 and DLS in the 4-8 Hz (theta) frequency band during NREM sleep to quantify functional connectivity.
  • Spike-Timing Relationships: They used a "basic spike jitter" method to identify pairs of M1 and DLS neurons with consistent short-latency spike-timing relationships, indicating direct functional connections.
  • Cross-Correlations: They analyzed cross-correlations of spiking activity to measure corticostriatal transmission strength.
  • Sleep Stage Classification: Behavioral states (wake, NREM, REM) were classified based on LFP power and movement.
  • NREM Rhythm Analysis: They identified and analyzed NREM sleep rhythms like slow oscillations (SOs) and sleep spindles.

Datasets

Not discussed in the paper. The data generated in this study is available on Dryad.

Key Insights and Ideas

  • Sleep is not just downtime; it's learning time: The study strongly suggests that NREM sleep is a critical period for consolidating motor skills by strengthening corticostriatal connections.
  • NMDA receptors in the striatum are key for offline learning: Blocking these receptors after training significantly impairs skill learning, highlighting their role in sleep-dependent plasticity.
  • Corticostriatal connectivity strengthens offline: The increase in functional connectivity between M1 and DLS primarily occurred during periods away from training (sleep), not during active practice.
  • Sleep spindles are the maestros of corticostriatal plasticity: These specific NREM sleep events, especially when temporally linked to slow oscillations, appear to be uniquely capable of mediating the strengthening of connections between the motor cortex and striatum.
  • Preservation of learned connections: Short-latency spike-timing relationships between M1 and DLS neurons were specifically preserved during the post-training period in sleep-spindle-modulated pairs, suggesting a mechanism for maintaining learned pathways.

Key Results

  • Blocking striatal NMDA receptors with AP5 infusions after training significantly reduced the day-to-day improvement in the velocity profile of the reach-to-grasp task compared to saline infusions.
  • 4-8 Hz LFP coherence between M1 and DLS increased predominantly offline (between training sessions) as rats learned the task. This offline increase was strongly correlated with behavioral improvements (reach velocity profile correlation).
  • The ability to predict DLS neural trajectories from M1 spiking activity during action execution increased with training, and this predictive ability was correlated with pre-training NREM LFP coherence.
  • Corticostriatal transmission strength, measured by short-latency spike-timing relationships, was significantly higher during NREM sleep compared to wakefulness.
  • Transmission strength was greatest during sleep spindles compared to slow oscillations or delta waves, suggesting spindles are particularly important for plasticity.
  • Short-latency spike-timing relationships were preserved in spindle-modulated neuron pairs during the post-training period, and this preservation was correlated with overnight increases in LFP coherence.
  • The temporal proximity of sleep spindles to preceding slow oscillations (SOs) influenced corticostriatal connectivity, with "SO-nested" spindles showing a greater impact on transmission strength.

Limitations and Future Work

  • Functional vs. Structural Changes: The study primarily measured functional connectivity (LFP coherence, spike-timing). Future work is needed to determine if these functional changes correspond to structural changes in synaptic strength.
  • NMDA Infusion Specificity: While blocking striatal NMDA receptors disrupted learning, the authors note that AP5 infusions might also affect spontaneous striatal activity during NREM. Further research is needed to clarify this.
  • Direction of Communication: The study focused on M1-to-DLS communication. Future work could explore how the direction of communication between cortex and striatum evolves during learning.
  • Therapeutic Potential: The authors suggest that understanding how NREM rhythms regulate memory consolidation could inform therapeutic strategies for maladaptive behaviors linked to the corticostriatal network, such as addiction.
  • Synaptic Homeostasis: The findings on preserved connections during sleep spindles need to be reconciled with the broader "synaptic homeostasis hypothesis," which suggests sleep generally downscales synapses. Future research could explore how NREM rhythms differentially impact synaptic downscaling versus preservation.

Summary updated

Abstract

The strength of cortical connectivity to the striatum influences the balance between behavioral variability and stability. Learning to consistently produce a skilled action requires plasticity in corticostriatal connectivity associated with repeated training of the action. However, it remains unknown whether such corticostriatal plasticity occurs during training itself or 'offline' during time away from training, such as sleep. Here, we monitor the corticostriatal network throughout long-term skill learning in rats and find that non-rapid-eye-movement (NREM) sleep is a relevant period for corticostriatal plasticity. We first show that the offline activation of striatal NMDA receptors is required for skill learning. We then show that corticostriatal functional connectivity increases offline, coupled to emerging consistent skilled movements, and coupled cross-area neural dynamics. We then identify NREM sleep spindles as uniquely poised to mediate corticostriatal plasticity, through interactions with slow oscillations. Our results provide evidence that sleep shapes cross-area coupling required for skill learning.
BibTeX
@inproceedings{lemke2021coupling,
  author = {Stefan Lemke and Dhakshin Ramanathan and David Darevksy and Daniel Egert and Joshua D. Berke and Karunesh Ganguly},
  title = {{Coupling between motor cortex and striatum increases during sleep over long-term skill learning}},
  booktitle = {eLife},
  year = {2021},
  doi = {10.7554/elife.64303},
}
Figure