Modeling memory consolidation during posttraining periods in cerebellovestibular learning

Area of Science:

• Neuroscience
• Computational Neuroscience
• Motor Control

Background:

• Long-term depression (LTD) at parallel fiber-Purkinje cell (PF-PC) synapses is a proposed mechanism for cerebellar motor learning.
• Emerging evidence suggests multiple plasticity mechanisms contribute to memory formation in the cerebellum.
• Cerebellar circuits integrate sensory and motor information for adaptive control.

Purpose of the Study:

• To investigate the roles of different synaptic plasticity mechanisms in cerebellar motor learning.
• To model the gain adaptation of the optokinetic response (OKR) using a minimal cerebellar circuit.
• To differentiate the contributions of PF-PC and mossy fiber-vestibular nuclear neuron (MF-VN) synapses to memory formation.

Main Methods:

• Formulation of a simplified computational model of the cerebellum based on known anatomy and physiology.
• Implementation of both long-term depression (LTD) and long-term potentiation (LTP) at PF-PC and MF-VN synapses.
• Simulation of OKR gain adaptation in response to training paradigms.

Main Results:

• The model generated short-term memory of OKR gain at PF-PC synapses after training, which decayed within a day.
• Long-term memory of OKR gain was established at MF-VN synapses through daily training repetition, persisting for days.
• Memory consolidation was observed during post-training periods, not during active training.
• Spaced training proved more effective for long-term memory formation than massed training.
• Simulated OKR changes mirrored experimental findings in wild-type and gene-manipulated mice, including vestibulo-ocular reflex (VOR) alterations.

Conclusions:

• Cerebellar motor learning involves distinct temporal dynamics of memory storage at different synaptic sites.
• PF-PC synapses primarily support short-term adaptation, while MF-VN synapses are crucial for long-term memory consolidation.
• The model provides a framework for understanding cerebellar memory mechanisms and predicting the effects of pharmacological or genetic manipulations.