Controlling Spatio-Temporal Sequences of Neural Activity by Local Synaptic Changes.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Network Dynamics

Background:

• Sequential activity in neural networks underlies behavior.
• Behavioral flexibility necessitates dynamic control over these sequences.
• Mechanisms for controlling and reconfiguring neural sequences are not well understood.

Purpose of the Study:

• To investigate how heterogeneous neural networks can evoke and control sequential activity.
• To identify mechanisms for flexible reconfiguration of neural sequences.
• To explore the computational capabilities of spatially heterogeneous networks.

Main Methods:

• Simulated recurrently connected networks with heterogeneous connectivity and spatial in-degree landscapes.
• Modulated synaptic strength in local neuronal neighborhoods.
• Analyzed the impact of modulation on sequence initiation, termination, extension, gating, and redirection.

Main Results:

• Heterogeneous networks with smooth spatial in-degree landscapes robustly evoke and control sequential activity.
• High-impact locations for sequence control were identified, often coinciding with mid in-degree regions.
• Local modulation of synaptic strength allowed flexible reconfiguration of sequential activity patterns.

Conclusions:

• Spatially heterogeneous networks offer a framework for fast, flexible computations on behavioral timescales.
• These networks enable context-dependent fine-tuning of computations via local modulation.
• The findings suggest a mechanism for dynamic behavioral control within otherwise stable neural pathways.