A mismatch-based model for memory reconsolidation and extinction in attractor networks

Area of Science:

• Neuroscience
• Computational Neuroscience
• Cognitive Science

Background:

• Memory reconsolidation and extinction are increasingly studied for clinical relevance.
• Amnestic drugs can disrupt these processes post-reexposure, depending on the protocol.
• Existing research highlights the role of reexposure context and duration.

Purpose of the Study:

• To propose a unified neural network model explaining memory reconsolidation and extinction.
• To investigate the role of representational similarity between learning and reexposure sessions.
• To elucidate the mechanistic basis for how different behavioral protocols yield distinct memory outcomes.

Main Methods:

• Developed a neural network model simulating memory processes.
• Modeled reconsolidation as memory trace updating in the hippocampus.
• Modeled extinction as the formation of a new memory trace.
• Incorporated independent mechanisms for synaptic strengthening and labilization, with protein synthesis inhibition affecting the former.

Main Results:

• The model demonstrates that the similarity between original learning and reexposure contexts dictates whether reconsolidation or extinction occurs.
• Reexposure duration influences the balance between these two memory processes.
• Memory updating during reexposure is crucial for driving reconsolidation.

Conclusions:

• The proposed framework offers a unified mechanistic explanation for experimental findings on memory reconsolidation and extinction.
• Contextual similarity and memory updating are critical factors determining the fate of stored memories.
• The model provides insights into the differential effects of amnestic drugs based on behavioral protocols.