A neuronal model of memory - integrative processing in the retrosplenial head direction system
Lead Research Organisation:
University College London
Department Name: Experimental Psychology
Abstract
Understanding memory is one of the most important scientific problems today, because of the devastating effects of memory loss (e.g., in Alzheimer?s disease). Remembering life events is closely linked to navigation, and so memory research has hitherto focused on spatial memory. Spatial memory is complex, however, and progress has been slow. The present proposal aims to develop a research programme studying a simpler version of spatial memory, memory for the landmarks that indicate directional heading and can be used to orient (such as one might orient using the statue of Eros, when emerging from the Piccadilly Underground). To remain oriented during movement, it is necessary to combine learned landmark information with information about self-motion. Both landmark learning and its combination with self-motion signals are thought to take place in a brain region known as retrosplenial cortex (RSC). Consistent with this idea, RSC becomes active when people learn about new environments, and damage to this structure (e.g., following stroke) causes both disorientation and amnesia. The present project will try to understand how landmark memory is formed in RSC, and how it is updated during self-motion. The activity of single RSC neurons will be observed while rats learn about new landmarks, or when familiar landmarks are moved, to try and determine what information the cells ?know? about the landmarks (colour, shape etc) and whether they also know about the spatial position of landmarks, or simply about their identity. Studies of groups of neurons will try to determine the extent to which they influence each other, as opposed to be influenced by the senses (such as vision). The goal is to try and determine whether the ?memory trace? for new landmark learning is located in RSC, and how this interacts with information about movement. If the project is successful, the RSC will be a valuable structure with which to pursue more complex questions about memory formation, such as what chemical signals are involved, and how these processes might be improved, for example with drugs. Thus, this work will not only help elucidate the processing pathways underpinning the sense of direction, but will also contribute more broadly to an understanding of the processes by which memories are formed and stored, and of how amnesia might be treated or even prevented.
Technical Summary
Despite decades of research, the neural basis of memory formation remains little understood. One reason is that the main target of past studies, the hippocampus, is complex, multimodal and anatomically inaccessible. This proposal will study an alternative more tractable model, the head direction (HD) system, a network of peri-hippocampal structures which compute a representation of directional heading. The HD signal is critical for both spatial representation/navigation and memory formation, and is itself is capable of learning. These features make it an exciting candidate for studies of the cellular underpinnings of memory formation, critical for understanding not only memory but also its clinical correlates including Alzheimer?s disease.
This proposal will focus on the largest and most accessible HD area, the rat retrosplenial cortex (RSC), located at the convergence of descending cortical and ascending brainstem streams. We will test the hypothesis that the RSC acquires, computes and subsequently retrieves information about static directional cues (landmarks) which is combined with self-motion cues to provide a continuously updated representation of heading. The research outcome will be an experimental model of mammalian memory formation, plus an understanding of some of its component processes. Additionally, the project may yield a new understanding of other roles of RSC, which ? despite its size and importance in memory ? is little understood.
Using chronic single-neuron recordings from rats foraging in cue-controlled environments, the integrative processes mediated by retrosplenial system will be investigated as follows:
(1) Integration of visual features ? What do HD cells ?know? about visual cues? These experiments will test the hypothesis that RSC HD neurons are sensitive to higher-order visual features such as shape and pattern.
(2) Spatial integration of visual cues ? Are HD cells sensitive to relative spatial location? The spatial position of configurations of landmarks will be varied to determine whether the RSC HD cells are sensitive to spatial as well as visual characteristics of landmark arrays.
(3) Where (in the brain) and how is new landmark information added to the representation? Experiments involving addition of new landmarks will determine whether RSC cells can acquire such information, and attempt (using excitotoxic lesions) to determine whether the location of the memory trace is in the RSC itself.
(4) How are static (landmark) and dynamic (self-motion) inputs integrated? These experiments will explore, using multi-electrode array technology, how ensembles of HD cells act as a collective, and test existing hypotheses about ?attractor dynamics? in memory formation and updating.
This proposal will focus on the largest and most accessible HD area, the rat retrosplenial cortex (RSC), located at the convergence of descending cortical and ascending brainstem streams. We will test the hypothesis that the RSC acquires, computes and subsequently retrieves information about static directional cues (landmarks) which is combined with self-motion cues to provide a continuously updated representation of heading. The research outcome will be an experimental model of mammalian memory formation, plus an understanding of some of its component processes. Additionally, the project may yield a new understanding of other roles of RSC, which ? despite its size and importance in memory ? is little understood.
Using chronic single-neuron recordings from rats foraging in cue-controlled environments, the integrative processes mediated by retrosplenial system will be investigated as follows:
(1) Integration of visual features ? What do HD cells ?know? about visual cues? These experiments will test the hypothesis that RSC HD neurons are sensitive to higher-order visual features such as shape and pattern.
(2) Spatial integration of visual cues ? Are HD cells sensitive to relative spatial location? The spatial position of configurations of landmarks will be varied to determine whether the RSC HD cells are sensitive to spatial as well as visual characteristics of landmark arrays.
(3) Where (in the brain) and how is new landmark information added to the representation? Experiments involving addition of new landmarks will determine whether RSC cells can acquire such information, and attempt (using excitotoxic lesions) to determine whether the location of the memory trace is in the RSC itself.
(4) How are static (landmark) and dynamic (self-motion) inputs integrated? These experiments will explore, using multi-electrode array technology, how ensembles of HD cells act as a collective, and test existing hypotheses about ?attractor dynamics? in memory formation and updating.
