Circuit Mechanisms of Information Processing and Storage in Brain Slices

University Of Wisconsin-Madison, Madison WI

Investigators

Linked publications, trials & patents

Abstract

? DESCRIPTION (provided by applicant): The nervous system operates through the generation and transformation of patterns of electrical activity distributed sparsely through a vast array of interconnected neurons. These patterns encode information ultimately arising from sensory inputs. Within this framework the storage of information by the nervous system can be viewed as an imprinting process that enables a network to recapitulate patterns of activity stored by prior experience. The canonical operation of pattern completion is one of the most fundamental forms of network operation envisaged by neuroscientists, and this concept has had a major influence on efforts to understand cognitive function. Pattern completion results when a representation has been stored by selective strengthening of synapses between participating neurons. When a subsequent event activates only a subset of these neurons, representing a part of this pattern, this activity can then spread to the entire set of neurons through the newly strengthened synapses to reconstruct the original pattern. These ideas have been developed within two very different disciplines, mathematical/computer modeling of neural networks and experimental studies of behavior. Until recently these ideas had not been studied in an intact neural circuit, leaving important hypotheses about neural network function without direct experimental tests. The present study will use voltage imaging to test hypotheses about pattern completion in hippocampal slices. Voltage imaging generates maps of electrical activity distributed through a slice, and these maps represent 'patterns' of electrical activity. By quantification of image similarity using methods derived from digital image processing and information theory, comparisons of these maps provide a rigorous experimental test of pattern completion. Recent work from this laboratory laid the foundation for this experimental approach and demonstrated that a pattern of activity can be stored in the CA3 region of a hippocampal slice by long-term potentiation (LTP). Subsequent partial inputs then retrieved the complete pattern. The present project will develop this in vitro approach for the study of information storage and recall. Aim 1 will evaluate neuromodulators known to influence both LTP and behavior to determine whether acetylcholine and norepinephrine receptors can modify information storage and pattern completion in the hippocampal CA3 region. Aim 2 will address temporal aspects of information storage, first testing the role of input timing (spike-timing dependent plasticity), and then investigating the complementary issue of persistence of the information trace in relation to the decay of different forms of LTP. Aim 3 will use a genetically-encoded voltage sensor to evaluate information stored during an animal's experience. This Aim will also evaluate the hypothesis of pattern completion at the level of single cells and synapses. This work will provide novel experimental tests for computational models of neural network function, and advance our understanding of the mechanisms by which neural circuits store, recall, and process information.