Editing the Neural Basis of Perception

项目摘要： 神经计算需要数千个相互关联且通常在遗传上相似的神经元的协调努力 神经元这些神经元形成物理上混合的网络和子网络， 并加强感官知觉或选择运动动作。已知这样的共活性系综是 优选地互连，并且可以表示具有独特的神经处理的功能元件。 属性，如模式完成和合奏之间的竞争。在这份提案中，我将获得 机械的理解如何合奏的共同活动的神经元相互作用，通过探测的功能， 单个和成组的神经元。我将研究：锥体细胞的集合 在视觉任务中与其他锥体细胞和局部抑制神经元相互作用，这些集合如何影响 运动行为，以及特定的整体如何对来自其他皮层区域的信息做出反应。尽管 潜在的重要性，合奏皮层编码，这些群体的混合性质，使他们 特别难学。虽然传统的光遗传学技术可以操纵基因识别， 在一个区域的神经元，他们无法选择性地操纵混合的神经元，不同的只是他们的神经元。 功能特性重要的是，新的多光子光遗传学技术开始允许操纵光遗传学。 细胞的选择仅取决于其活性，但此类技术需要进一步发展。在K99阶段 这个建议，我将继续我的训练，通过发展新的光学系统的多光子 刺激并通过使用这些技术来理解皮层功能。通过结合这些新的光学 我已经开发了一种新的用于体内多光子使用的视蛋白技术，现在我已经 能够以前所未有的精度写入或编辑许多神经元的神经活动。通过 在视觉感知过程中改变整体活动，我将确定个体的因果贡献 神经元以及神经元群体对感觉编码的影响。在R00阶段，通过在 运动皮层我将解开这些群体的行为影响，探索运动合奏中发挥的作用， 运动动作，并研究神经元如何跨模态相互作用。编辑和监视活动的能力 神经子网络是获得感知和行动的机械理解的关键。的 我们从这个提议中得出的结论将有助于描述所有信息是如何在皮层中呈现的，但是， 只有通过先进的技术才能达到。

Project Summary: Neural computation requires the coordinated effort of thousands of interrelated and often genetically similar neurons. These neurons form physically intermingled networks and subnetworks that act together to amplify and strengthen sensory perceptions or select motor action. Such co-active ensembles are known to be preferentially interconnected, and may represent a functional element of neural processing with unique properties, such as pattern completion and competition between ensembles. In this proposal I will gain a mechanistic understanding of how ensembles of co-active neurons interact by probing the function of individual and groups of neurons in an awake mouse. I will examine: how ensembles of pyramidal cells interact with other pyramidal cells and local inhibitory neurons in a visual task, how these ensembles influence motor behavior, and how specific ensembles respond to information from other cortical areas. Despite the potential importance of ensembles in cortical coding, the intermixed nature of these groups has made them particularly hard to study. While conventional optogenetic techniques can manipulate genetically identified neurons in a region, they are incapable of selectively manipulating intermingled neurons that differ only by their functional properties. Critically, new multiphoton optogenetic techniques are beginning to allow manipulation of cells chosen by their activity alone, however such techniques require further development. In the K99 phase of this proposal, I will continue my training through the development of novel optical systems for multiphoton stimulation and through use of these technique understand cortical function. By combining these new optical techniques with novel opsins designed for in vivo multiphoton use that I have already developed, I now have the ability to write in or edit neural activity across many neurons with a precision never before possible. By altering ensemble activity during visual perception I will determine the causal contributions of individual neurons as well as populations of neurons to sensory coding. In the R00 phase, through manipulations in motor cortex I will unravel the behavioral impact of these groups, probing the role motor ensembles play in motor action, and study how neurons interact across modalities. The ability to both edit and monitor the activity of neural subnetworks is critical to gaining a mechanistic understanding of perception and action. The conclusions we draw from this proposal will help to describe how all information is presented in the cortex, but can only be reached with advanced techniques.