Influence of internal state on communication in distributed neuronal circuits

Summary/Abstract, Project 1 Our responses to the world around us are controlled by diverse aspects of our complex internal states. For example, we are more likely to take action when we are more vigilant and engaged, and we are more likely to give particular interpretations to our percepts when we have prior expectations about our environment. Understanding the neural basis of such internal state changes is important for unraveling the basic mechanisms of flexible behavior in mammals and for understanding the etiology of disorders of state such as autism. Here we propose to investigate the neural mechanisms underlying two types of internal state changes: spontaneous fluctuations in engagement and goal-directed changes in perceptual bias. The team is part of the International Brain Laboratory, an established consortium that has developed a standardized mouse decision-making task and standardized methods for training, neural measurement, and data analysis, along with a working, scalable infrastructure for sharing data. We will test the novel hypothesis that behavioral differences across these states result from alterations in the structure of information flow between brain regions. Specifically, we hypothesize that disengaging from a task dampens propagation of specific dimensions of population activity to downstream structures, and that changing bias to favor one choice over another rotates the dimensions of information propagation across the brain. To investigate these hypotheses, we will take advantage of our recent development of technology for recording neural activity at large scale and of algorithms that quantify behavioral states and multi-dimensional communication patterns between brain regions. We will first simultaneously record large, dense populations of neurons from key sets of brain regions using Neuropixels 2.0 probes and systematically characterize the dimensionality and magnitude of correlations between these regions. Then, we will quantify how these correlation patterns depend on internal state, using novel algorithmic quantification of spontaneous state transitions during the standardized and high-throughput behavioral task that has already been established by the International Brain Laboratory. Finally, we will develop and apply a new class of analysis methods designed to measure the interactions between three or more simultaneously recorded brain regions to identify whether one region gates or modulates the multi-dimensional communication between the other regions, thus discovering putative controller regions that direct the flow of information. This project will deliver the first systematic characterization of multi-dimensional communication patterns across cortical and subcortical regions; tests of new hypotheses about information routing in the brain; algorithms that quantify the relationships between large populations of neurons; and a large-scale openly shared dataset of neural activity during flexible behavior across the mouse brain.