Probing the time scales of perceptual readout with white noise optogenetic inhibition

项目摘要 在视觉引导的行为中，感觉刺激和刺激之间只有几百毫秒的时间。 随后的动作，但感觉运动回路中的尖峰反应开始代表刺激特征和 指导即将采取的行动。由于大脑回路的巨大复杂性，存在很大的不确定性 这些尖峰信号(以及大脑中的哪些区域)对行为有因果关系。我们需要能够解决这些问题的方法 尖峰时刻的波动如何驱动人们的感知和行动。一种用于检查的多功能工具 感觉输入与尖峰或行为之间的关系是白噪声分析。一种随机的、动态的 刺激被提出，这些波动可以被利用来产生对刺激的无偏见的估计 驱动神经元尖峰的特征(尖峰触发的平均值)或驱动的感觉输入周期 知觉决策(心理物理核心)。我们已经将白噪声分析应用于光遗传 刺激。我们将弱光遗传抑制的随机模式传递给行为和视觉区域 计算与感知报告(命中、未命中)相关的平均光遗传信号。我们的初步数据 表明白噪声光遗传抑制迅速产生定义尖峰效应贡献的核 特定神经元群体的特定行为，分辨率为数十毫秒(光遗传 行为内核，obk)。这样精确的定时信息用标准获得是不切实际的 光遗传方法(例如，脉冲或持续抑制)。我们将通过以下方式验证和应用此方法 确定在较低和较高视觉区域中测量的OBK如何依赖于视觉输入和任务需求， 在这样做的过程中，产生了新的观察，将尖峰模式与整个过程中的知觉决定联系起来 视觉大脑。小鼠将被训练在感知任务的阈值下可靠地工作。在第一个实验中， 视觉信息将突然到来(超阈值强度、对比度阶跃)或逐渐到来(接近阈值， 对比度斜率)，我们将把OBK与刺激诱发的尖峰分布的变化联系起来。在 第二个实验，老鼠将被要求对目标方向做出反应，并拒绝干扰。我们会 测量皮质(纹状体和纹外视区)和皮质下(上丘)视觉的OBK 以确定它们各自对任务绩效的贡献的时间进程。一组老鼠将会 还植入了固定线光学电极，以便于同时进行电生理记录和 行为过程中的光遗传干扰。这些数据将与刺激措施的每时每刻的减少有关 和/或任务诱发的峰值和感知报告，并能够比较种群峰值和 好的。这项工作将为大脑皮层和大脑皮质神经元活动期提供前所未有的分辨率。 皮层下的视觉通路，对感知和行为起因果作用。我们的实验将确定 一种强大的方法，用于构建将不同视觉区域中的尖峰动态与可视- 引导行为，这是一种在系统神经学家中得到广泛应用的方法。

Project Abstract During visually guided behaviors, only hundreds of milliseconds elapse between a sensory stimulus and its ensuing action, yet spiking responses across sensorimotor circuits come to represent stimulus features and guide forthcoming actions. Owing to the immense complexity of brain circuits, there is great uncertainty as to which spikes (and in which brain areas) causally contribute to behavior. We need methods capable of resolving how moment-to-moment fluctuations in spiking drive perception and action. A versatile tool for examining relationships between sensory input and spiking or behavior is white noise analysis. A random, dynamic stimulus is presented and these fluctuations can be leveraged to generate unbiased estimates of the stimulus features that drive neuronal spiking (spike-triggered average) or the periods of sensory input that drive perceptual decisions (psychophysical kernel). We have adapted white noise analysis for optogenetic stimulation. We deliver random patterns of weak optogenetic inhibition to visual areas during behavior and calculate the average optogenetic signal associated with perceptual reports (hits, misses). Our preliminary data show that white noise optogenetic inhibition rapidly yields kernels that define the contribution of spiking in specific neuronal populations to specific behaviors, with a resolution of a tens of milliseconds (optogenetic behavioral kernel, OBK). Such precise timing information would be impractical to obtain with standard optogenetic approaches (e.g., pulses or sustained inhibition). We will validate and apply this approach by determining how OBKs measured in lower and higher visual areas depend on visual input and task demands, and in doing so generate new observations that link patterns of spiking with perceptual decisions throughout the visual brain. Mice will be trained to work reliably at threshold in perceptual tasks. In the first experiment, visual information will arrive suddenly (super-threshold intensity, contrast steps) or gradually (near-threshold, contrast ramps) and we will relate OBKs to changes in the distributions of stimulus-evoked spikes. In the second experiment, mice will be required to respond to a target orientation and reject distractors. We will measure OBKs in both cortical (striate and extrastriate visual areas) and subcortical (superior colliculus) visual areas to determine the time course of their respective contributions to task performance. A subset of mice will also be implanted with fixed-wire optrodes to facilitate concurrent electrophysiological recordings and optogenetic perturbations during behavior. These data will relate moment-to-moment reductions in stimulus and/or task-evoked spikes with perceptual reports and enable comparisons between population spiking and OBKs. This work will provide unprecedented resolution into the periods of neuronal activity in cortical and subcortical visual pathways that causally contribute to perception and behavior. Our experiments will establish a powerful method for constructing kernels that relate spiking dynamics in different visual areas to visually- guided behaviors, a method that could find wide application among systems neuroscientists.