Self-avoidance dominates the selection of hippocampal replay

Area of Science:

• Neuroscience
• Computational Neuroscience

Background:

• The brain generates spontaneous neural activity sequences, including hippocampal replay, without external input.
• Hippocampal replay sequences during immobility represent past experiences or future predictions, but their generation mechanism is unclear.

Purpose of the Study:

• To investigate the principles governing the selection and temporal organization of hippocampal replay sequences.
• To elucidate the role of cortical input in shaping hippocampal replay dynamics.

Main Methods:

• Recording from large ensembles of hippocampal place cells in freely behaving rats.
• Utilizing optogenetic manipulation of cortical input to alter neural activity.
• Employing a symmetry-breaking attractor model to simulate sequence generation.

Main Results:

• Hippocampal replay strongly avoids recently traversed paths, prioritizing future trajectories immediately after movement cessation.
• Replay sequences exhibit self-avoidance on short timescales, differing from the avoidance of past behavioral trajectories.
• A later temporal window (seconds into stopping) shows dominance of past trajectory replay.
• Optogenetic perturbations confirmed that medial entorhinal cortex input is crucial for the later replay of past experiences.

Conclusions:

• Hippocampal replay selection is governed by a self-avoidance principle, with temporal dynamics influenced by movement cessation.
• A modified attractor model, incorporating medial entorhinal cortex input, accurately predicts the observed temporal organization of replay.
• Medial entorhinal cortex plays a critical role in maintaining a memory trace that biases hippocampal replay, providing insight into sequence generation.