Distributing task-related neural activity across a cortical network through task-independent connections.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Task-related neural activity is widespread in the brain during goal-directed behaviors.
• The synaptic and circuit mechanisms driving these broad activity changes remain largely unknown.

Purpose of the Study:

• To investigate how task-related activity spreads across neuronal populations in a spiking network model.
• To elucidate the role of synaptic interactions in mediating widespread neural activity during decision-making tasks.

Main Methods:

• A spiking neural network model was trained to reproduce motor cortex activity during a decision-making task.
• The model incorporated strong synaptic interactions, independent of the task.
• Optogenetic perturbations were used to validate findings in silico and suggest applicability to cortical networks.

Main Results:

• Task-related activity emerged across the entire network, including untrained neurons.
• Strong, task-independent synapses were identified as the key mediators for spreading activity.
• The network's dynamical state, influenced by these synapses, was crucial for activity propagation.

Conclusions:

• A cortical mechanism involving task-independent strong synapses facilitates the spread of task-related activity.
• This mechanism enables distributed representations of task variables across neuronal populations.
• The findings highlight the importance of network structure and dynamics in neural computation.