Every few seconds, a sleeping rat’s hippocampus briefly replays a past experience, firing the same sequence of neurons that fired when it ran through a maze earlier that day. These split-second “non-local” reactivations are thought to be how the brain consolidates memories and plans future routes, but the neural machinery behind them remains poorly understood. This project aims to build a detailed, neuron-level model of how the hippocampus and cortex talk to each other during these events, how the replayed information updates existing brain maps, and what chemical signals control when reactivations occur. The research is fundamental science. It will not produce a therapy or device in the short term. But the hippocampus is one of the first brain regions damaged in Alzheimer’s disease, and memory replay is known to break down in schizophrenia and epilepsy. A mechanistic understanding of how healthy circuits perform these non-local computations could eventually reveal exactly where and how those circuits fail in disease—and point toward targets for intervention.
View original technical description
The overall goal of this proposal is to build a neural-level understanding of how non-local cortico-hippocampal communication mediates memory consolidation and spatial computations. The well-studied network of spatially modulated neurons in the hippocampus and associated regions provides the pre-eminent cellular-model of memory for events and places. However, the activity of these neurons mainly encodes local information, that is, the current configuration of an animal in its environment. Work conducted by us, and others, have identified transient reactivations of hippocampal neurons and cortical counterparts as a key mechanism supporting systems consolidation and spatial planning. These brief ‘non-local’ events provide a means by which remembered experiences can gradually update memory networks, equally they are theorised to support the calculations necessary for route planning. Our aims are: 1) to understand how hippocampus and cortex interacts during reactivations; 2) determine how reactivated information affects existing representations; 3) precisely define the spatial computations that guide navigation; 4) investigate how neuromodulation controls the occurrence of reactivations. To this end, our approach is to combine computational modelling, machine learning, and state-of-the-art experimental techniques. Developing a basic understanding of these processes opens the way to understand how they fail in disease and may ultimately deliver tremendous therapeutic benefits.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know