Attractor dynamics of network UP states in the neocortex

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• The cerebral cortex traditionally processes sensory input through successive stages.
• Cortical circuits exhibit continuous spontaneous activity and complex interconnections, including feedback loops.
• Neuronal membrane potential spontaneously fluctuates between DOWN and UP states, potentially coordinated across networks.

Purpose of the Study:

• To investigate the dynamics of spontaneous neural activity in the mouse visual cortex.
• To characterize the nature and organization of synchronized UP state transitions.
• To explore the functional implications of these network dynamics for information processing.

Main Methods:

• Utilized two-photon calcium imaging to monitor spontaneous activity.
• Reconstructed the dynamics of up to 1,400 individual neurons in cortical slices.
• Analyzed the spatiotemporal patterns of synchronized UP state transitions.

Main Results:

• Observed synchronized UP state transitions, termed 'cortical flashes,' in spatially organized neuronal ensembles.
• These events involved small groups of neurons and exhibited stereotyped spatiotemporal dynamics.
• Spontaneous activity patterns reflect coordinated neuronal group activity, even without sensory input.

Conclusions:

• Network UP states function as circuit attractors, emergent properties of feedback neural networks.
• These attractors may implement memory states or solve complex computational problems.
• Cortical flashes represent a fundamental mechanism of spontaneous network organization and function.