Classification of Cortical Neurons by Single Cell Transcriptomics

 描述(申请人提供)：解开哺乳动物大脑的复杂性是当今生物学中最具挑战性的问题之一。神经科学的一个主要目标是了解神经元和非神经元细胞的电路如何处理感觉信息，产生运动，并辅助记忆、情感和认知。要阐明神经回路的特性，需要了解构成这些回路的细胞类型及其在处理和整合信息中的作用。然而，自从一个多世纪前Ramon y Cajal描述了不同的神经细胞类型以来，我们仅仅触及了了解大脑中细胞类型的多样性以及每种细胞类型如何对神经系统功能做出贡献的皮毛。目前对神经元进行分类的方法依赖于一些特征，包括少量基因的差异表达、细胞形态、解剖位置、生理学和连接性--这些重要的描述性特征仍然不足以完全描述或预测组成哺乳动物大脑的大量不同类型的细胞。在这里，我们提出了一套技术，用于识别和分类大脑中存在的无数细胞类型。我们的方法将以来自小鼠体感皮质的第5层锥体细胞作为模型系统。首先，我们将利用DNA测序技术的最新发展来高通量地表征单个第5层神经元上的基因表达谱。这些信息将被用来根据转录组的“指纹”对单个细胞进行分类。其次，被发现定义新发现的神经元亚型的基因将被用来通过Cas9/CRISPR介导的基因组工程获得对这些细胞的遗传访问，以创建转基因报告系。这项技术的发展有望为快速产生多基因小鼠报告菌株开辟一条管道，在这些报告菌株中，特定的神经元亚型通过标记基因的组合进行唯一标记。第三，我们将使用这些基因工程小鼠来确认我们的分类代表了分类神经元的不同功能属性。我们的方法最终可以扩大到对大脑中的细胞类型进行完整的普查，这是使用现代研究工具解剖神经系统功能所急需的资源。

 DESCRIPTION (provided by applicant):Unraveling the complexity of the mammalian brain is one of the most challenging problems in biology today. A major goal of neuroscience is to understand how circuits of neurons and non-neuronal cells process sensory information, generate movement, and subserve memory, emotion and cognition. Elucidating the properties of neural circuits requires an understanding of the cell types that comprise these circuits and their roles in processing and integrating information. However, since the description of diverse neuronal cell types over a century ago by Ramon y Cajal, we have barely scratched the surface of understanding the diversity of cell types in the brain and how each individual cell type contributes to nervous system function. Current approaches for classifying neurons rely upon features including the differential expression of small numbers of genes, cell morphology, anatomical location, physiology, and connectivity - important descriptive properties that nonetheless are insufficient to fully describe or predict the vast number of different cell types that comprise the mammalian brain. Here we propose a suite of technologies for identifying and classifying the myriad cell types present in the brain. Our method will be developed using layer 5 pyramidal cells from mouse somatosensory cortex as a model system. First, we will exploit the latest developments in DNA sequencing technologies to characterize gene expression profiles on single layer 5 neurons at high throughput. This information will be used to classify individual cells based on their transcriptome "fingerprints." Second, genes found to define newly discovered neuronal subtypes will be used to gain genetic access to these cells using Cas9/CRISPR-mediated genome engineering to create transgenic reporter lines. Development of this technology promises to open a pipeline for the rapid generation of multigenic mouse reporter strains in which specific neuronal subtypes are uniquely labeled by combinations of tagged genes. Third, we will use these genetically engineered mice to confirm that our taxonomy represents distinct functional properties of the classified neurons. Our approach can ultimately be scaled up to generate a complete census of cell types in the brain, a critically needed resource for dissecting nervous system function with modern investigative tools.