EngramSeq: a new technology for mapping engram networks in mice

我的实验室研究大脑如何编码、存储和使用信息。我的研究项目的一个分支是研究在疾病状态(例如阿尔茨海默病或创伤后应激障碍)下记忆是如何出错的，并由CIHR提供资金。第二个ARM侧重于信息处理的基本基本过程。自从我在2005年开始我的实验室以来，这只手臂一直得到NSERC的支持。我最近完成的NSERC发现基金支持一种开源微型内窥镜的开发和测试，该内窥镜允许在行为自由的小鼠学习和记忆时在它们的大脑中可视化神经元活动。有了这项技术，我们真的可以“按照老鼠的记忆观察它的大脑”。令我欣慰的是，参与这个项目的许多不同的受训人员都在科学领域取得了进一步的成就(Google Deep Mind、两家生物技术初创公司和一名助理教授)。在目前的提案中，我们还将开发一个新的开源工具。Engram Seq将使我们能够可视化特定的记忆“编码”细胞在大脑中的投射位置。大脑可能是宇宙中最复杂的物体之一。人脑包含大约1000亿个神经元，通过大约100万亿个突触连接连接在一起，而小鼠大脑包含大约7000万个神经元。包括记忆在内的基本大脑功能很可能来自神经元之间的连接模式。然而，我们对整个大脑线路图的了解是有限的。为了更好地了解大脑的所有复杂性，重要的是要解开不同神经元投射的巨大纠结。最近，开发了一种技术，该技术利用高通量DNA测序，使数千个单个神经元的投影图谱能够在同一小鼠大脑中进行分析(Kebschull等人，2016年)。这项技术被称为MAPseq(通过测序进行多路投影分析)，它使用独特的RNA“条形码”一次识别数千个神经元的投影模式。尽管MAPseq是一个强大的工具，但它有两个限制，使其无法在我们(以及可能许多)的研究中使用。首先，用于表达条形码的病毒载体是有毒的，会在3天内杀死受感染的神经元。其次，不可能将条形码的表达“定位”到特定的感兴趣神经元。为了解决这些局限性，我们将开发Engram Seq来检查不同个体engram神经元的特定投射。这个多学科项目是培训高素质人才的理想选择，因为学员将获得分子生物学、测序、小鼠脑外科手术、小鼠行为和大数据分析方面的专业知识。重要的是，Engram Seq将是一个多功能工具，不仅可以应用于我们的研究问题(不同的Engram神经元投射到哪里？)，还可以应用于更广泛的神经科学界提出的其他问题。我们的目标是，我们新的开放科学工具也将对整个神经科学研究社区有价值。

My lab studies how the brain encodes, stores and uses information. One arm of my research program examines how memory goes wrong in disease states (e.g., Alzheimer's disease or PTSD) and is funded by CIHR. The second arm focuses on fundamental basic processes underlying information processing. This arm has been continuously supported by NSERC since I began my lab in 2005. My recently completed NSERC Discovery grant supported the development and test-driving of an open-source mini-endoscope that allows neuronal activity to be visualized in the brain of freely-behaving mice as they learn and remember. With this technology, we can literally "watch the brain of a mouse as it remembers". I am gratified that many of the diverse trainees involved in this project have gone on to further careers in science (Google Deep Mind, two biotech start-ups, and an Assistant Professor). In the current proposal, we will also develop a new open-source tool. EngramSeq will allow us to visualize where specific memory "engram" cells project in the brain. The brain is likely one of the most complex objects in the universe. The human brain contains roughly 100 billion neurons linked together by some 100 trillion synaptic connections while the mouse brain contains roughly 70 million neurons. Fundamental brain functions including memory likely derive from the patterns of connections between neurons. However, our knowledge of the overall brain wiring diagram is limited. To better understand the brain in all of its complexity, it is important to unravel the vast tangle of different neuronal projections. Recently, a technique was developed that exploits high-throughput DNA sequencing to allow the projection profile of thousands of individual neurons to be analyzed in the same mouse brain (Kebschull et al., 2016). This technique, called MAPseq (multiplexed analysis of projections by sequencing), uses unique RNA "barcodes" to identify the projection patterns of thousands of neurons at a time. Although MAPseq is a powerful tool, it has two limitations that preclude its use in our (and likely many) studies. First, the viral vector used to express the barcode is toxic and kills infected neurons within 3 days. Second, it is not possible to "target" barcode expression to specific neurons of interest. To address these limitations, we will develop EngramSeq to examine the specific projections of different individual engram neurons. This multidisciplinary project is ideal for training highly-qualified personnel as trainees will gain expertise in molecular biology, sequencing, mouse brain surgery, mouse behaviour, and big data analysis. Importantly EngramSeq will be a versatile tool that can be applied not only to our research questions (where do different engram neurons project?), but also be applied to other questions posed by the broader neuroscientific community. Our goal is that our new open-science tool will also be valuable to the neuroscience research community at large.