Cell type-specific connectome predicts distributed working memory activity in the mouse brain

Area of Science:

• Neuroscience
• Computational Biology
• Cognitive Science

Background:

• Advances in connectomics and neurophysiology enable whole-brain mechanism studies.
• Working memory is crucial for internally holding and processing information without sensory input.

Purpose of the Study:

• To develop a large-scale model of the multiregional mouse brain to study working memory.
• To investigate how connectome data and interneuron density influence working memory coding.

Main Methods:

• Constructed a mesoscopic model of the mouse brain using interareal cortical connection data.
• Incorporated a macroscopic gradient of parvalbumin-expressing interneuron density.
• Analyzed cell type-specific graph measures to predict activity patterns and subnetworks.

Main Results:

• Working memory coding is distributed but modular, influenced by long-range, cell type-specific targeting and interneuron density.
• Identified a core subnetwork essential for memory maintenance.
• The model exhibited multiple attractor states, representing self-sustained internal brain states.

Conclusions:

• The developed model provides a framework for interpreting large-scale brain activity recordings during cognitive tasks.
• Emphasizes the critical role of cell type-specific connectomics in understanding brain function.
• Suggests that distributed yet modular coding, shaped by specific neuronal populations, underlies working memory.