Cell type atlasing of whole human brains using HOLiS: an optimized pipeline for staining, clearing, imaging, and analysis

项目摘要（摘要） 长期以来，全面了解人类大脑中的全脑细胞组织一直是 被认为是理解复杂大脑功能的关键基础，包括我们作为人类的身份。 在这个项目中，我们建议接受这一挑战，并建立一个能够对整个人类进行成像的管道。 脑细胞分辨率。我们相信，我们的组织处理新技术的融合 清晰、超快的3D显微镜和高效的分析将使这个问题变得易于处理和可扩展， 标志着人类大脑研究的新范式。 作为BICCN的一部分，Osten和Wu实验室已经在破译细胞方面取得了重大进展 使用遗传和分子标记的小鼠大脑的类型和三维组织逻辑 与小鼠全脑成像相结合。然而，从小鼠到人类大脑的扩展工作约为2,000倍， 需要技术和概念创新，以提取最大信息并确保高效率。 我们的方法将集中在一个新的人脑优化光片HOLiS显微镜平台开发 希尔曼实验室，其速度，效率和多路复用能力，预计将使蜂窝- 在短短几天之内就可以完成对整个人脑的高分辨率成像。HOLiS平台由一个 吴实验室开发的一种新的优化人脑组织制备方法，名为HuB. Clear， 优化用于高度易处理的5 mm厚的全人脑切片的多重染色和清洁。我们 Osten实验室开发的计算数据分析管道将利用基于深度学习的数据 分析并允许大脑中的每个细胞在整个大脑7T MRI体积内被记录。我们的新人类 脑共同坐标框架将包括基于多重蛋白质谱的细胞多样性分析 和精确的空间特征。在3年项目期间将实现以下基准： 我们的组织制备方法将提供：1）完全组织清除，同时保留形态， 可靠的数据产生和与MRI的整合，以及2）可靠和定量的整体安装免疫标记 具有不同的靶点，以允许在整个大脑中进行多重分子分析。我们的成像技术将 提供：1）足够的分辨率和多路复用能力，以及2）高吞吐量速度， 分析了多个完整的人脑我们的数据分析方法将提供：1）基础设施能力 处理整个人脑成像数据，2）HOLiS优化算法，以提取和解释 丰富的分子和细胞信息与计算机科学的最新进展。 由此产生的管道将很容易扩展和共享，最大限度地提高效益-成本比， 在未来的几年里，我们将有机会对100个甚至1,000个人类大脑进行成像。我们的研究结果将把细胞多样性和 以足够的细胞分辨率在整个人脑中使用分子签名进行形态学分析， 促进进一步的功能研究和跨物种的比较与其他BICCN的协同努力。

Project Summary (Abstract) Gaining a comprehensive understanding of brain-wide cellular organization in the human brain has long been recognized as a critical foundation for understanding complex brain functions, including who we are as humans. In this project we propose to take on this challenge and establish a pipeline capable of imaging the entire human brain at cellular resolution. We believe that the convergence of our novel technologies for tissue processing and clearing, ultra-fast 3D microscopy and highly efficient analysis will make this problem tractable and scalable, marking a new paradigm in human brain research. As part of BICCN, the Osten and Wu labs, have already made significant progress towards deciphering the cell type and three-dimensional organizational logic of the mouse brain, using both genetic and molecular labeling in combination with whole mouse brain imaging. However, scaling efforts ~2,000x from mouse to human brain requires both technical and conceptual innovation to extract maximum information and ensure high efficiency. Our approach will center on a new Human brain Optimized Light-Sheet HOLiS microscopy platform developed by the Hillman lab, whose speed, efficiency and multiplexing capabilities are expected to enable cellular- resolution imaging across the entire human brain in only a few days. The HOLiS platform is complemented by a new optimized tissue preparation method for human brain developed by the Wu lab, named HuB.Clear, optimized for multiplexed staining and clearing of highly tractable 5 mm thick, full human brain slabs. Our computational data analysis pipelines developed by the Osten lab will leverage deep learning-based data analysis and permit every cell in the brain to be registered within the whole brain 7T MRI volume. Our new human brain common coordinate framework will include cellular diversity analysis based on multiplex protein profiling and precise spatial characterization. The following benchmarks will be achieved during the 3-year project: Our tissue preparation method will provide: 1) complete tissue clearing while preserving morphology to allow faithful data production and integration with MRI, and 2) reliable and quantitative whole mount immunolabeling with diverse targets to allow multiplex molecular profiling across the entire brain. Our imaging technology will provide: 1) sufficient resolution and multiplexing capacity, and 2) high-throughput speed to allow exhaustive analyses across multiple whole human brains. Our data analysis methods will provide: 1) infrastructure capable of processing whole human brain imaging data, and 2) algorithms optimized for HOLiS to extract and interpret rich molecular and cellular information with the latest advances in computer science. The resulting pipeline will be easily scalable and sharable with maximized benefit-cost ratio, opening the door to imaging 100's or even 1,000's of human brains in coming years. Our results will link cellular diversity and morphology with molecular signatures across the entire human brain at a sufficient cellular resolution and to facilitate further functional investigations and cross-species comparisons in synergy with other BICCN efforts.