CRCNS Research Proposal: Stochastic Processes Driving the Ascending Reticular Activating System

所有的大脑回路，包括那些构成意识和感知的回路，在物理上都嵌入了一个由微小的、蜿蜒的纤维组成的密集网络中。这些纤维是活跃的，并释放出深刻影响大脑的信号分子。然而，单独的纤维没有明确的目的地，似乎在脑组织中随机定向，经常改变方向。这一特性使它们与点对点神经连接有很大不同。由于单个纤维轨迹的不可预测性，到目前为止，大脑研究一直专注于整体纤维的“密度”。这些密度在不同的大脑区域有所不同，并与大脑的正常和变化功能有关。这项研究项目旨在揭示局部的、随机的、个别纤维的决定如何导致大脑区域特定的纤维密度。这项研究还打算确定，在脊椎动物过去2.5亿年的进化过程中，纤维的行为是否从根本上保持不变。这种概念新颖的跨学科方法结合了高分辨率显微镜、自动图像分析(包括机器学习)和随机过程的最新发展。该项目开发的因果模型有望为从单个纤维的动力学预测纤维密度提供必要的工具，并将为基础和应用神经科学中的纤维密度调控奠定理论基础。该项目的数据和分析将用于新设立的跨学科实验室课程，并将加强生物图像信息学的本科生培训计划。该项目自然地将神经生物学、工程学和数学结合在一起；这一点和它的视觉吸引力使其非常适合为K-12学生计划的STEM演讲，以及公开演讲。由于5-羟色胺能纤维存在独特的计算挑战，该项目将有助于图像分析算法的发展。该项目还将为影响包括抑郁症、精神分裂症和自闭症谱系障碍在内的几种精神障碍的“弥漫性”神经传递结构建立严格的理论基础。实际上，脊椎动物大脑中的所有神经过程都发生在密集的纤维矩阵中，释放5-羟色胺、去甲肾上腺素和其他神经递质。这个古老的系统起源于脑干，被称为上升网状激活系统(ARAS)。由于ARAS纤维没有形成明确的投影，并且具有极其曲折的轨迹，它们目前的描述超出了连接学项目的范围，并基于观察到的纤维“密度”。这个跨学科的项目试图重建在大脑中建立和支持ARAS的基本自组织过程。与目前的描述性方法截然不同的是，它假设单个5-羟色胺能纤维的行为可以通过一个三维随机过程来描述，该过程决定了最终的纤维密度(作为一种紧急现象)。脊椎动物的大脑，跨越了大约2.5亿年的进化，将被用来阐明这一过程。单个ARAS纤维将用免疫组织化学(包括组织扩张)显示，并用激光共聚焦扫描显微镜成像。小鼠的长纤维轨迹将使用组织清除技术和光片显微镜的全脑成像进行可视化。将开发一种图像分析算法，以自动检测和跟踪3D空间中的单个纤维轨迹。这些轨迹将被用来建立一个最优的随机模型，利用先进的计算方法。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

All brain circuits, including those that underlie consciousness and perception, are physically embedded in a dense meshwork of microscopic, meandering fibers. These fibers are active and release signaling molecules that profoundly affect the brain. However, individual fibers do not have well-defined destinations and appear to be randomly oriented within brain tissue, frequently changing their direction. This property makes them very different from point-to-point nerve connections. Because of the unpredictability of individual fiber trajectories, brain research to date has focused on overall fiber "densities." These densities vary across brain regions and have been associated with normal and altered functions of the brain. This research project aims to reveal how local, random-like decisions by individual fibers lead to specific fiber densities in brain regions. The research also intends to determine whether the behavior of the fibers has remained fundamentally the same in the last 250 million years of vertebrate evolution. This conceptually-novel, interdisciplinary approach brings together recent developments in high-resolution microscopy, automated image analysis (including machine learning), and stochastic processes. The causal models developed in the project are expected to provide essential tools for the prediction of fiber densities from the dynamics of single fibers and will lay a theoretical foundation for fiber density manipulations in fundamental and applied neuroscience. The project's data and analyses will be used in a newly created interdisciplinary laboratory course and will also strengthen an undergraduate training program in bio-image informatics. The project naturally brings together neurobiology, engineering, and mathematics; this and its visual appeal make it well-suited for planned STEM-oriented presentations for K-12 students, as well as for public talks. Since serotonergic fibers present unique computational challenges, the project will contribute to the development of image analysis algorithms. The project will also build a rigorous theoretical foundation for the structure of "diffuse" neurotransmission that is affected in several mental disorders, including depression, schizophrenia, and autism spectrum disorder.Virtually all neural processes in vertebrate brains take place in a dense matrix of fibers that release serotonin, norepinephrine, and other neurotransmitters. This ancient system originates in the brainstem and is known as the ascending reticular activating system (ARAS). Since ARAS fibers do not form well-defined projections and have extremely meandering trajectories, their current descriptions fall outside the scope of connectomics projects and are based on observed fiber "densities." This interdisciplinary project seeks to reconstruct the fundamental self-organizing process that builds and supports the ARAS in the brain. In a radical departure from current descriptive approaches, it hypothesizes that the behavior of single serotonergic fibers can be described by a three-dimensional stochastic process, which determines the resultant fiber density (as an emergent phenomenon). Vertebrate brains, spanning some 250 million years of evolution, will be used to elucidate this process. Single ARAS fibers will be visualized with immunohistochemistry (including tissue expansion) and imaged with confocal laser scanning microscopy. Long fiber trajectories in mice will be visualized using tissue-clearing techniques and whole-brain imaging with light-sheet microscopy. An image analysis algorithm will be developed to automatically detect and trace individual fiber trajectories in the 3D-space. The trajectories will be used to build an optimal stochastic model, by taking advantage of advanced computational methods.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.