Trans-Sheet Illumination Microscopy (TranSIM) for decoding whole brain activity at submillisecond temporal resolution

跨片照明显微镜 (TranSIM) 以亚毫秒时间分辨率解码全脑活动

• 批准号: 9552856
• 负责人: KATSUSHI ARISAKA
• 金额: $ 14.65万
• 依托单位: UNIVERSITY OF CALIFORNIA LOS ANGELES
• 依托单位国家: 美国
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-09-01 至 2019-12-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/9552856
• 关键词: Accelerometer; Acoustics; Action Potentials; Algorithms; Animal Model; Area; Behavior; Behavioral; Biological Neural Networks; Brain; Caenorhabditis elegans; Complex; Computer software; Data; Detection; Development; Devices; Disease; Drosophila genus; Electron Microscopy; Electrophysiology (science); Ensure; Four-dimensional; Geometry; Health; Hour; Human; Image; Japan; Knowledge; Larva; Length; Light; Lighting; Lightning; Longevity; Maps; Methodology; Methods; Microscope; Microscopy; Monitor; Motion; Names; Nature; Neurosciences; Optics; Organism; Phototoxicity; Positioning Attribute; Research; Resolution; Sampling; Scanning; Scheme; Sepharose; Slice; Speed; Stimulus; Synapses; System; Testing; Thick; Thinness; Time; Visual; Width; Zebrafish; base; brain volume; design; experimental study; flexibility; fluorescence microscope; in vivo; information processing; innovation; interest; lens; light scattering; millisecond; next generation; optogenetics; public health relevance; relating to nervous system; sensor; temporal measurement; virtual reality; voltage

Today’s knowledge of large-scale neural networks is advancing along two orthogonal directions. Spatial (static, structural) connectomic understanding is achieved through optical and electron microscopy; yielding high spatial resolution with limited or no temporal information. Conversely, electrophysiological methods provide an exceptional temporal understanding of millisecond-order neurodynamic activities in vivo, with restrictions placed on spatial information. Unfortunately, these approaches have historically been rather mutually exclusive and incompatible with each other. This proposal is precisely aimed at breaking the methodological barrier between spatial and temporal observation, through innovative 3D optical scanning concepts which rival the temporal resolution of electrophysiology. The resulting system is ideally suited to imaging genetically expressed voltage- sensitive fluorescent markers. The proposed optical microscope is named TranSIM: Trans-Sheet Illumination Microscope. It is designed to observe brain-wide neurodynamics in model organisms (up to 1 mm3) with ~1 μm resolution in space, and sub- millisecond resolution in time. Unlike a conventional sheet illumination microscope, which illuminates a single x-y focal plane of the detection objective (on the z-axis), the conceived design forms a light sheet in the transverse direction along the z-axis (y-z plane). This sheet of light is then rapidly scanned along the x-axis as it is imaged. Multiple thin z-slices are then collected simultaneously by spatially multiplexing next-generation large-format sCMOS sensors. Regions of Interest, each with slight depth off-set from the focal plane, will map to a segmented part of the sensor, creating an eight-image focal volume in the time of a single frame. As a result, a thick brain volume can be covered by scanning the focal (x-y) plane (~1 mm2) in one direction (in x) at the frame rate of the sCMOS. The fastest sCMOS cameras are currently being jointly developed and optimized for this purpose. A line confocal readout will be implemented electrically by a rolling shutter mode of the sCMOS sensor to minimize contamination of scattering light. With proper Regions of Interest, a large volume (800 x 80 x 1000 μm3) can be scanned at 780 volumes / second; 100 times faster than today’s fastest 3D optical microscopy systems. A smaller volume (100 x 80 x 1000 μm3) can be scanned at an unprecedented rate of 6240 volumes / second, reaching the sub-millisecond temporal resolution of electrophysiological sampling. The illuminative sheet can be formed through the same detective objective for enhanced geometrical flexibility, or by an orthogonal objective (as in conventional horizontal sheet illumination) to minimize phototoxicity. In the case of a multi-view based on four-objective geometry, a large open volume (40 x 20 x 8 mm3) exists between lenses, to observe model organisms (like Zebrafish) under visual and optogenetic stimulation. Such a volume also provides smaller model organisms (C. elegans and Drosophila larvae) with ample space to navigate freely in 3D, while whole brain neural activity is monitored and controlled under various external and internal (optogenetic) stimulations.

如今，大规模神经网络的知识正在沿着两个正交方向发展。空间（静态、结构）连接体的理解是通过光学和电子显微镜实现的；在有限或没有时间信息的情况下产生高空间分辨率，电生理学方法提供了特殊的时间信息。遗憾的是，这些方法在历史上一直是相互排斥且互不兼容的。通过创新的 3D 光学扫描概念进行观察，其时间分辨率可与电生理学的时间分辨率相媲美。由此产生的系统非常适合对基因表达的电压敏感荧光标记进行成像。 所提出的光学显微镜被命名为 TranSIM：跨片照明显微镜，其设计用于观察模型生物体（最大 1 mm3）的全脑神经动力学，空间分辨率约为 1 μm，时间分辨率为亚毫秒。传统的片状照明显微镜，其照亮检测物镜的单个x-y焦平面（在z轴上），概念设计在沿z轴（y-z）的横向上形成光片然后，通过空间复用下一代大幅面 sCMOS 传感器，在成像时沿 x 轴快速扫描该光片，同时收集多个薄 z 切片，每个感兴趣区域都有轻微的深度偏移。从焦平面，将映射到传感器的分段部分，在单帧时间内创建八图像焦体积。因此，可以通过扫描焦（x-y）平面来覆盖厚的大脑体积。 (~1 mm2) 在 sCMOS 帧速率下沿一个方向（x）。 目前正在为此联合开发和优化最快的 sCMOS 相机，通过 sCMOS 传感器的卷帘快门模式以电气方式实现线共焦读出，以最大程度地减少散射光的污染，并具有适当的感兴趣区域（800）。 x 80 x 1000 μm3）可以以 780 体积/秒的速度进行扫描；比当今最快的 3D 光学显微镜系统快 100 倍。体积 (100 x 80 x 1000 μm3) 可以以前所未有的 6240 卷/秒的速率进行扫描，达到电生理采样的亚毫秒时间分辨率。 照明片可以通过相同的探测物镜形成以增强几何灵活性，或者通过正交物镜（如传统的水平片照明）形成以最小化光毒性。在基于四物镜几何形状的多视图的情况下，大的光毒性。镜头之间存在开放体积（40 x 20 x 8 mm3），用于在视觉和光遗传学刺激下观察模型生物（如斑马鱼），这样的体积还提供了较小的模型生物（秀丽隐杆线虫和线虫）。果蝇幼虫）有足够的空间在 3D 中自由导航，同时在各种外部和内部（光遗传学）刺激下监测和控制全脑神经活动。