Multi-regional neural circuit dynamics underlying short-term memory

点击翻译按钮获取中文摘要

Abstract The focus of this BRAIN Initiative funding opportunity is to use “innovative approaches to understand how circuit activity gives rise to a specific behavior”. Cognitive behaviors arise from collective interactions of multiple brain systems. Yet, for most cognitive processes, we do not yet know which brain areas are involved and how multi-regional interactions mediate specific cognitive processes. This gap in knowledge arises because separate parts of the brain are studied individually, yet, the brain circuits driving behavior vary from one behavior to another. The goal of this proposal is to establish a working example of how brain-wide activity dynamics collectively generate one cognitive behavior. We address this question by studying how a mouse flexibly generate a volitional movement based on short-term memory. Neurons in multiple parts of the brain, including the frontal cortex, thalamus, midbrain, and cerebellum respond robustly during this short-term memory and causally contribute to the behavior. Taking advantage of this opportunity to establish how activity distributed across multiple brain systems orchestrates one coherent behavior, in this proposal, we will use newly developed experimental frameworks to analyze the underlying neural circuitry at brain-wide scale and establish causal relationships between specific activity patterns and behavior. First, we will use brain-wide loss-of-function screen, high-density silicon probe recording, and anatomical techniques to produce multi- modal maps of core neural substrates of the short-term memory. The outcome datasets will be put into standardized brain coordinates, making it possible to link the functional data to existing connectional and gene expression atlases. Next, we will use simultaneous recordings and spatiotemporally-precise perturbations to probe multi-regional interactions underlying the observed activity patterns and relationships to behavior. Finally, we will build multi-regional models that offer interpretable description of the behaviorally-relevant dynamics and relate them to underlying circuit connectivity. The outcome will disambiguate competing models of how information distributed over multiple brain regions is coordinated during cognitive processes, how information is dynamically routed and gated. The experimental, analysis, and modeling approaches will be broadly useful for analyzing distributed circuits driving behavior, as is the focus of multiple collaborative U19 grants. All the data and code will be published in the well document Neurodata Without Border (NWB) format.