Principles of operation of a neural learning circuit

项目摘要 我们可以学习并成功地回忆起面孔、事件、语言、概念、地点、事实， 恐惧或奖励，以及熟练地移动我们的运动效应器所需的运动命令。 几十年的科学研究已经指出，突触可塑性的作用是大脑的基本货币， 学习和记忆的能力。虽然我们对管理变化的规则有相当详细的了解， 在突触强度和神经元内在膜兴奋性的改变中，我们对如何 这些可塑性变化导致行为学习和记忆仍处于婴儿期。行为学习是一种 一个完整的神经学习回路的涌现特性，其中可塑性的位点和机制是 嵌入式如果不了解电路层面的可塑性效应，我们就无法真正理解 学习和记忆。可以说，运动适应是我们最有机会了解 由于感官刺激和适应性之间的微妙关系， 行为小脑已被证明是对运动学习至关重要的大脑结构，并提供了一个 神经位点开始勾勒出控制学习的电路规则。我们的目标是利用 小脑回路的保守细胞结构，以确定支持神经系统的运作原则， 学习电路更普遍。在这个奖项的指导阶段，我们将专注于一个良好的描述， 小脑依赖行为：追求方向学习。即使在发生了一个单一的运动后 如果在这个任务中出现错误，大脑会从错误中学习，试图在下一次试验中将错误降到最低。上 目的，我们将表征驱动这种运动记忆的错误依赖性获取的信号， 小脑回路中的特定突触，被认为有助于大量的行为运动学习。 在第二次瞄准时，我们将记录整个小脑回路。我们的目标是描述 小脑回路中的单个元件和突触有助于行为适应，包括 对导致行为学习的可塑性位点的限制，并允许关于 学习发生在小脑皮质之前、内部或下游。在独立阶段， 我们将再次记录在不同的小脑学习任务中的完整小脑回路：扫视 适应使用依赖于不同小脑区域的适应行为，我们可以开始解剖 广泛适用于小脑学习的电路级原则。总之，我们的结果将提供 行为学习的第一个回路级规则。这些结果应该具有广泛的影响， 其他的学习和记忆系统，所有这些都是以驱动行为的复杂电路存在的。

PROJECT SUMMARY We can learn and successfully recall faces, events, language, concepts, places, facts, things that were frightening or rewarding, and the movement commands required to skillfully move our motor effectors. Decades of scientific research have pointed to the role of synaptic plasticity as the basic currency of the brain’s ability to learn and remember. While we have a fairly detailed understanding of the rules that govern changes in synaptic strength and modifications of a neuron’s intrinsic membrane excitability, our understanding of how these plastic changes lead to behavioral learning and memory is still in its infancy. Behavioral learning is an emergent property of a complete neural learning circuit in which the sites and mechanisms of plasticity are embedded. Without an understanding of the effects of plasticity at the circuit-level, we cannot truly understand learning and memory. Arguably, motor adaptation is the domain where we have best chance to understand the circuit-level rules that govern learning, due to the exquisite relationship between sensory stimuli and adaptive behavior. The cerebellum has been shown to be the brain structure crucial for motor learning, and provides a neural locus to begin to outline the circuit rules that govern learning. Our goal is to leverage the highly conserved cytoarchitecture of the cerebellar circuit to identify the principles of operation that underpin neural learning circuits more generally. During the mentored phase of this award, we will focus on a well-described cerebellar-dependent behavior: pursuit direction learning. Even after the occurrence of a single movement error in this task, the brain learns from the mistake, attempting to minimize the error in the next trial. In the first aim, we will characterize the signals that drive the error-dependent acquisition of this motor memory at a specific synapse in the cerebellar circuit, which is thought to contribute to the bulk of behavioral motor learning. During the second aim, we will record from the complete cerebellar circuit. Our goal is to describe how individual elements and synapses in the cerebellar circuit contribute to behavioral adaptation, including constraints on the site(s) of plasticity that cause behavioral learning and allowing conclusions about the extent to which learning occurs before, inside, or downstream of the cerebellar cortex. During the independent phase, we will again record from the complete cerebellar circuit during a different cerebellar learning task: saccadic adaptation. Using an adaptive behavior that relies on a different cerebellar region, we can begin to dissect the circuit-level principles that generalize broadly across cerebellar learning. Together, our results will provide the first circuit-level rules that underlie behavioral learning. These results should have broad implications across other learning and memory systems, all of which exist as complex circuits that drive behavior.