Toward causal neuroscience: capture and manipulate emergent neuronal ensembles

描述(由申请人提供)：假设动物的感知、情绪或行为过程是由特定的大脑活动模式或刺激引起的。 神经系统中多个区域的选定神经元群体(集合)。事实上，光学成像和多电极记录技术的最新进展已经开始揭示大量清醒的、行为正常的脊椎动物，如斑马鱼幼体和啮齿动物的异常复杂的动力学。然而，这些可视化/记录实验并不能明确地确定神经元系综的活动与其功能之间的因果关系。对于“因果神经科学”来说，迫切需要一个工具包，使神经科学家不仅能成为观察者，而且还能成为观察到的整体的执行者。开发这样一个工具包的困难在于哺乳动物大脑的复杂性，它包含数十亿个神经元和数万亿个突触。因此，单个神经元可能在不同的时间点参与不同的活动集合。因此，与给定的行为或知觉过程相关的集合是从数百万个神经元之间复杂的相互作用中产生的新特性。因此，集合不太可能是遗传上预先确定的，基于分子或细胞类型的方法对于“标记和操纵”它们是无用的。为了克服这一困难，我们将开发一种由两个关键组件组成的新型工具包：(1)旨在仅在激活的神经元中瞬时和选择性地表达非常不稳定的外来受体的小鼠系，以及(2)仅能感染表达这种外来受体的神经元的无毒、伪型工程病毒。通过这种方式，对伪型病毒的定时注射将使我们能够特异性地和永久性地“捕获”最近激活的整体(即表达病毒受体的整体)。经过改造的病毒可以整合

DESCRIPTION (provided by applicant): It is hypothesized that animals' perceptual, emotional, or behavioral processes are governed/caused by specific patterns of brain activities, or firings of selected populations (ensembles) of neurons across multiple regions in the nervous system. Indeed recent advancements in optic imaging and multi-electrode recording technology have begun to reveal the inordinately complex dynamics of large populations of neurons in awake, behaving vertebrates such as larval zebrafish and rodents. However, these visualization/recording experiments could not definitively establish the causal relationships between the activities of neuronal ensembles and their functions. A toolkit that enables neuroscientists to be not only observers, but also actuators of the observed ensembles is critically needed for "causal neuroscience". The difficulty to developing such a toolkit lies in th complexity of the mammalian brain, which contains billions of neurons and trillions of synapses. Thus, individual neurons are likely to participate in different active ensembles at different time points. Hence the ensembles associated with a given behavioral or perceptual process are emergent properties arising out of the complicated interactions among millions of neurons. Therefore, ensembles are unlikely to be genetically pre-determined, and molecule or cell-type based methods are not useful for "labeling and manipulating" them. To overcome this difficulty, we will develop a novel toolkit consisting of two key components: (1) a mouse line designed to express, transiently and selectively, a very unstable foreign receptor only in activated neurons, and (2) engineered non-toxic, pseudo-typed viruses that can only infect neurons expressing this foreign receptor. In this way, timed-injection of the pseudo-typed viruses will allow us to specifically and permanently "capture" the recently activated ensembles (which are the ones expressing the viral receptor). The engineered viruses can int