Brain-wide Neuronal Circuit Mapping with X-ray Nano-Holography

项目摘要 该提案的目标是开发基于同步加速器的X射线成像技术，以实现高分辨率成像 大脑神经回路的结构全面绘制全脑回路目前还不可行，即使是在小的 哺乳动物模型系统，因为光学显微镜（LM）缺乏足够的分辨率和电子显微镜（EM） 不能应用于大容量。利用新的第四代同步加速器前所未有的品质 来源于欧洲同步加速器，我们将开发X射线纳米全息（XNH）成像技术，用于大规模 脑回路的成像利用光源相干性和亮度的改进，我们将改进 成像分辨率允许直接可视化神经元之间的突触连接，并开发成像协议 其允许在典型的束线实验内对厘米级电路体积进行成像。我们将联合收割机非- 使用EM和LM成像技术进行破坏性XNH，以严格和定量地验证XNH的准确性， 基于电路重构。然后，我们将使用这个相关的工作流程来研究远程之间的关系， 感觉输入，局部突触微电路和单神经元活动，研究后部电路如何 顶叶皮层（PPC）支持知觉决策。最后，我们将XNH电路映射应用于整个 皮质半球，并利用基于深度学习的机器视觉算法来获得皮质半球的综合图谱 连通性。这个图谱将在原则上解决所有长距离连接皮层区域之间的单轴突 分辨率，贷款洞察不同的皮层区域如何实现专门的功能，以及如何分布的皮层 网络支持认知，并受到精神疾病的影响。

Project Summary This proposal's objective is to develop synchrotron-based X-ray imaging technologies to enable high-resolution imaging of brain-wide neuronal circuits. Comprehensively mapping brain-wide circuits is not currently feasible, even in small mammalian model systems, because light microscopy (LM) lacks sufficient resolution and electron microscopy (EM) cannot be applied over large volumes. Leveraging the unprecedented qualities of the new 4th generation synchrotron source at the European Synchrotron, we will develop X-ray nano-holography (XNH) imaging techniques for large-scale imaging of brain circuits. Taking advantage of improvements in source coherence and brightness, we will improve imaging resolution to allow direct visualization of synaptic connections between neurons, and develop imaging protocols that allow imaging of centimeter-scale circuit volumes within a typical beamline experiment. We will combine non- destructive XNH with EM and LM imaging techniques to rigorously and quantitatively validate the accuracy of XNH- based circuit reconstruction. We will then use this correlative workflow to study the relationships between long-range sensory inputs, local synaptic micro-circuitry, and single-neuron activity, investigating how circuits in the posterior parietal cortex (PPC) support perceptual decision-making. Lastly, we will apply XNH circuit-mapping over an entire cortical hemisphere, and utilize deep-learning based machine vision algorithms to obtain a comprehensive atlas of cortical connectivity. This atlas will in principle resolve all long-range connections between cortical areas at single-axon resolution, lending insight into how distinct cortical areas achieve specialized function, and how distributed cortical networks support cognition and are affected by psychiatric disorders.