EAGER: Inferring Activity From Anatomy in Neuronal Cultures

在突触水平绘制整个大脑图谱的新兴技术将很快产生完整的神经解剖图谱，但活动仅与回路间接相关，在如何使用解剖图谱推断活动方面留下了很大的空白。在面对整个大脑的eb级规模之前，有必要开发出从更小、更简单但仍然完整的模型系统的解剖学中推断活动的方法。神经细胞培养（体外）是高度简化的系统，显示出复杂的动态活动，可以在空间活动方面进行监测，并且可以通过化学调节参数。重要的是，培养物是小而完整的网络，其中每个物理连接都可以被映射，并在单个细胞水平上监测活动。因此，对神经网络中物理连接和活动如何相互关联的解释将不会被人工制品或其他利用活体动物或脑组织的实验方法所遇到的限制所混淆（例如，活体脑切片有数千个断开的连接）。神经细胞培养作为一种基础，在此基础上开发出后来应用于全脑的方法，神经细胞培养本身就是一种令人感兴趣的模型系统。在一个远离神经科学的独立发展中，对“活跃物质”（自主代理的相互作用）的研究提出了关于静态状态如何变得活跃并可能同步的新原理。本项目旨在利用细胞培养、神经记录和连接组学等最新工具，将新颖的理论视角与简化的生物学相关实验结合起来。该项目的目标是产生一个明确的物理模型，其中神经元系统的三个关键要素被连接起来：来自2D神经细胞网络的功能记录；串行电子显微镜连通性测量；神经系统动力学的明确理论建模。最根本的问题是：我们能从解剖学上推断出活动吗？本研究的重点是神经状态之间的动态转换，包括同步。癫痫是一种同步性疾病，其中一名共同研究者的主要研究活动是儿童癫痫的临床研究。关于（时间）过渡到癫痫发作的基本理解很少，我们希望对模型系统（参数驱动）过渡的理解可能是有用的。模型系统在生物学和物理学中很重要。我们希望建立一个框架来分析神经细胞培养将有助于正常化的调查，否则将断开。pi将与芝加哥州立大学的电子显微镜项目合作，这是一所历史上为少数民族服务的大学。科罗拉多州立大学的学生将参与数据分析，这既是他们显微镜技术培训的一部分，也是研究的全面合作伙伴。pi特别提出了这样一个问题：拥有一个可以在没有完全同步的情况下自发启动的平衡网络意味着什么？一个人能控制从一个通用动态阶段到另一个通用动态阶段的转换吗？在这样的转变中是否存在突现的空间和时间尺度？具有长期和短期相关性的网络之间是否存在定性差异？这项工作的目的是建立一个框架，可以在未来应用于越来越多的来自不同大脑区域和其他离体实验平台（如活体脑切片/类器官）的已发表的连接组数据集，并为全脑的大规模连接组分析提供信息。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Emerging technologies to map whole brains at the synaptic level will soon produce complete maps of neural anatomy, but activity is only indirectly related to circuits, leaving large gaps in how we use anatomical maps to infer activity. Before facing the exabyte scales of whole brains, it is necessary to develop methods to infer activity from anatomy in smaller, simpler, but still complete model systems. Neural cell cultures (in vitro) are highly simplified systems that show complex dynamical activity, can be monitored in terms of spatial activity, and can have parameters tuned by chemistry. Critically, cultures are small, complete networks where every physical connection can be mapped and activity monitored at the single cell level. Thus, interpretations on how the physical wiring and activity in neural networks are correlated will not be confounded by artifacts or limitations encountered by other experimental methods that utilize living animals or brain tissue (e.g. living brain slices have thousands of severed connections.). As well as a substrate on which to develop methods later to be applied to whole brains, neural cell cultures are of interest as model systems in their own right. In a separate development far from neuroscience, research on 'active matter' - the interactions of autonomous agents - has suggested new principles about how quiescent states become active, and potentially synchronize. This project aims to bring together the novel theoretical perspective with simplified but biologically relevant experiments, using the latest tools of cell culture, neural recording, and connectomics. The goal of the project is to produce an explicit physical model where the three key elements of neuronal systems are joined up: functional recording from a 2D neural cell network; connectivity measurement through serial electron microscopy; explicit theoretical modelling of the dynamics of the neural system. The fundamental question is: Can one infer activity from anatomy? This research focuses on dynamical transitions between neural states, including synchrony. Epilepsy is a disease of synchrony and one of the co-investigators has his principal research activity in clinical investigations of pediatric epilepsy. There is little fundamental understanding about the (temporal) transition to seizure and we hope that understanding in a model system a (parameter driven) transition could be useful. Model systems are important in biology and physics. We hope that establishing a framework to analyze neural cell cultures will help normalize investigations which would otherwise be disconnected. The PIs will work with the electron microscopy program at Chicago State, a historically minority serving university. CSU students will be engaged in data analysis both as a component of their training in microscopy techniques and as full partners in the research.The PIs specifically ask: What does it mean to have a balanced network that can spontaneously fire without complete synchrony? Can one control the transitions from one generic dynamical phase to another? Is there emergent spatial and temporal scaling at such a transition? Are there qualitative differences between networks with long- and short-range correlations? This work is intended to build a framework that can be applied in the future to the growing number of published connectomic datasets derived from different brain regions and other ex vivo experimental platforms such as living brain slices/organoids and inform the analysis of large scale connectomics in whole brains.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.