Subsets of Single Neurons Predict Ensemble Activity and Memory Choices

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Identifying critical network components is challenging across disciplines.
• In neuroscience, pinpointing specific neurons influencing behavior is complex due to the brain's scale.

Purpose of the Study:

• To identify neuronal subsets in the CA1 region that significantly contribute to ensemble activity.
• To investigate the role of these identified subsets in spatial memory and behavior.

Main Methods:

• Utilized compressed sensing-inspired techniques to analyze neuronal ensemble activity.
• Recorded neuronal firing rates in male rats during spatial memory and cue-approach tasks.
• Identified predictive neuronal subsets based on co-varying firing rates with overall ensemble activity.

Main Results:

• Identified specific neuronal subsets whose activity asymmetrically predicted other units' activity.
• Excluding these subsets did not affect current location decoding but impaired past/future location decoding.
• These findings suggest the predictive subsets encode non-local information.

Conclusions:

• The identified predictive subsets may represent a hierarchical and sparse coding scheme in the CA1 region.
• Further study of these neuronal sub-populations could illuminate fundamental brain computational processes.
• This research provides a novel method for dissecting complex neural network functions.