Neural mechanisms of spatial and episodic memory

Episodic memory is a crucial function and its loss devastating. Research into its neural mechanisms is hindered by the knowledge gap between the molecular biology and pharmacology of neurons and synapses and behaviour/symptoms. To begin closing this gap I have developed a neural-level computational model of remembering the spatial context of an event. This model, including formation and use of place, grid, boundary and head-direction cells in spatial memory, will be tested, refined and extended to episodic memory, by experiments in humans and mice performing similar mnemonic tasks, such as virtual-reality navigation. Optogenetic, 2-photon microscopy and electrophysiological experiments in mice, will be combined with fMRI, MEG and intracranial recordings in humans to acquire the data across neuronal, circuits, systems, and behavioural levels, needed to test the model. Model predictions will be tested in patients with memory disorders. Key goals will be to test the model’s main neural-level mechanisms: how information can either be stored as a new memory or trigger reconstruction of an entire previous event, i.e. hippocampal pattern separation/completion; how egocentric perception and imagery interface with long-term allocentric representations, i.e. coordinate transformations in retrosplenial cortex; and how movement-related inputs update spatial representations in memory, mental imagery and planning.