Understanding the Distributed Control of Flexible Behavior

摘要 我们的大脑有一种非凡的能力，既能以不变的方式产生精确、一致的运动， 环境，并动态调整行为以适应不断变化的环境。几乎所有的日常活动，如 像开车或骑自行车一样，需要这种灵活的多任务处理。失去灵活调整的能力 行为是戏剧性的和衰弱的，如强迫症（OCD）等疾病所证明的 或者说是自闭症谱系障碍的严重形式。虽然大量的工作已经研究了大脑是如何 分布式运动网络在不变的环境中控制一致的运动，这种机制允许 灵活的移动控制仍然是未知的。在这个项目中，我开发了一个行为模型来研究 在小鼠中灵活地产生多种不同的伸展运动。之前的大量工作 特征在于对伸展运动的神经控制提供了一个强大的框架来精确理解 神经对柔韧性的控制我用这个模型来研究分布式控制的灵活运动， 初级运动皮层（M1）、基底神经节（BG）和小脑（CB）。虽然M1-BG和M1-CB网络 虽然先前已经单独研究过，但这三个区域如何相互作用在很大程度上尚未探索。我杠杆（1） M1、BG和CB中的大规模多部位神经生理学，（2）遗传控制的丘脑操纵， M1-BG和M1-CB网络，（3）多区域递归神经网络模型，以及（4）强迫症小鼠模型 和自闭症谱系障碍，显示出行为上的可解释性，以揭示大脑分布的基本原则， 运动网络控制着运动控制的灵活性，并阐明了这些机制是如何在运动控制中发生功能障碍的。 大脑紊乱

ABSTRACT Our brains have the remarkable ability to both produce precise, consistent movements in an invariant environment and dynamically adjust behavior to a changing environment. Almost all our everyday actions, such as driving a car or riding a bike, necessitate such flexible multi-tasking. Losing the ability to flexibly adjust behavior is dramatic and debilitating, as evidenced by disorders such as obsessive-compulsive disorder (OCD) or profound forms of autism spectrum disorder (ASD). While considerable work has examined how the brain’s distributed motor network controls consistent movements in an invariant environment, the mechanisms that allow for flexibility in movement control remain unknown. In this project, I develop a behavioral model to study the flexible production of multiple distinct reaching movements in mice. The extensive previous work that has characterized the neural control of reaching movements provides a powerful framework to precisely understand the neural control of flexibility. I use this model to investigate the distributed control of flexible movements across the primary motor cortex (M1), basal ganglia (BG), and cerebellum (CB). While the M1-BG and M1-CB networks have been previously investigated in isolation, how all three regions interact is largely unexplored. I leverage (1) large-scale multi-site neurophysiology in M1, BG, and CB, (2) genetically controlled thalamic manipulations of M1-BG and M1-CB networks, (3) multi-region recurrent neural network models, and (4) mouse models of OCD and ASD that display behavioral inflexibility to uncover fundamental principles by which the brain’s distributed motor network governs flexibility in movement control and shed light on how these mechanisms dysfunction in brain disorders.