A versatile approach for highly multiplexed, high-resolution imaging of endogenous molecules

项目摘要 理解大脑复杂结构的探索已经变得更具挑战性，因为 近年来，脑细胞之间的分子异质性程度变得很明显。在地图中绘制大脑 细节将需要在高分辨率成像中融入大量分子信息。当前 成像方法受到可区分检测通道数量的限制，因此更大程度的 多路传输需要反复剥离和重新应用探头。这些方法降低了组织的完整性 并损害了敏感度，并且没有解决多路传输的另一个主要挑战-- 蛋白质和RNA的标记方法以及同时检测这两种方法的需要。我们建议 一种新的成像方法，连续切片平行免疫/荧光原位杂交(SpiFISH)， 其核心策略是将样本在物理上细分为比 神经元胞体。每个部分都被视为单独的样本进行标记和成像，因此数百个 离散标记实验可以在给定的神经元上并行执行。这种方法是基于超薄的 切片，但不同于现有的超薄切片方法，如电子显微镜(EM)和阵列 断层扫描，SpiFISH不使用EM包埋树脂。无树脂干扰，灵敏的免疫标记 和RNA检测是可能的。每个部分都被单独标记和成像，因此任何给定的单元都可以 在优化的条件下，用许多不同的抗体和RNA探针标记。各节是 货架稳定，因此可以跨时间甚至跨实验室构建大型数据集。该方法允许 技术和标签的多路复用，因此同一样本可以用于多个成像和 染色平台。该项目的目标是开发健壮、可重现的协议和工作流 通过不同规模的数据分析准备样品。这将包括整个啮齿动物的小样本 智能和简化的方法，通过全自动数据收集和分析实现完全手动。

Project Summary The quest to understand the brain’s complex structure has become more challenging as the high degree of molecular heterogeneity among brain cells has become evident in recent years. Mapping the brain in detail will require incorporating large amounts of molecular information into high-resolution imaging. Current imaging methods are limited by the number of distinguishable detection channels, so greater degrees of multiplexing entail repeated cycles of stripping and reapplying probes. These methods degrade tissue integrity and impair sensitivity, and do not address the other major challenge of multiplexing- incompatibility between protein and RNA labeling methods and the need to compromise both for simultaneous detection. We propose a novel imaging approach, Serial-section parallel immuno/ Fluorescence In Situ Hybridization (SpiFISH), whose core strategy is to physically subdivide specimens into sections two orders of magnitude smaller than a neuronal cell body. Each section is treated as a separate sample for labeling and imaging, so hundreds of discrete labeling experiments can be performed in parallel on a given neuron. The method is based on ultrathin sectioning, but unlike existing ultrathin sectioning methods such as electron microscopy (EM) and array tomography, SpiFISH does not use EM embedding resins. Without resin interfering, sensitive immunolabeling and RNA detection are possible. Each section is labeled and imaged separately, so that any given cell can be labeled with many different antibodies and RNA probes under conditions optimized for each. Sections are shelf-stable, so large datasets can be built up across time and even across laboratories. The method allows multiplexing of techniques as well as labels, so the same sample can be used with multiple imaging and staining platforms. The goal of this project is to develop robust, reproducible protocols and workflows from sample preparation through data analysis across scales. This will include small samples through whole rodent brains and streamlined methods for fully manual through fully automated data collection and analysis.