Long Synfire Chains Emerge by Spike-Timing Dependent Plasticity Modulated by Population Activity

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Precisely timed neuronal activity sequences are common in various brain regions and species.
• Synfire chains are a leading model for explaining these sequential neuronal activities.
• The self-organization and emergence of synfire chains in initially unstructured networks remain poorly understood.

Purpose of the Study:

• To investigate the conditions under which synfire chains can emerge via self-organization in sparse random networks.
• To explore the role of spike-timing dependent plasticity (STDP) modulated by global activity in chain formation.
• To analyze the properties and potential applications of self-organized synfire chains.

Main Methods:

• Simulations of sparse random networks using both binary threshold neurons and conductance-based leaky integrate-and-fire (LIF) neurons.
• Implementation of a learning rule based on spike-timing dependent plasticity (STDP) modulated by global population activity.
• Theoretical analysis of the learning rule to determine the optimal length of emerging synfire chains.

Main Results:

• Long synfire chains spontaneously emerge in sparse random networks with the proposed STDP learning rule.
• The learning rule promotes efficient neuron reuse, allowing participation in multiple chains, consistent with experimental findings.
• Sparse networks prevent short, cyclic chains and demonstrate that specific synapse formation is not required for chain emergence.
• Theoretical predictions of chain length were validated in simulated LIF neuron networks.

Conclusions:

• STDP modulated by global activity is a viable mechanism for the self-organization of long synfire chains in sparse neural networks.
• The emergent synfire chains exhibit properties like neuron reuse and optimal length, supporting their biological relevance.
• The developed model offers a framework for understanding sequence formation in the brain and proposes a novel one-shot memory system for neuronal activity sequences.