High-speed volumetric imaging of neural activity throughout the living brain

为了理解大脑是如何计算的，我们需要了解电路中的单个神经元是如何整合的 将它们的大量输入转化为输出信号，以及它们如何共同工作来编码感官输入或 执行一个动作指令。电路和神经元是三维的， 延伸到数百或数千微米。因此，了解它们的运作需要 在3D中以捕捉所有活动的图像速率以突触和细胞分辨率监测它们的活动 事件动物行为对这一目标提出了一系列挑战。现有的3D成像技术 由于体积成像速度不足、脑运动引起的图像伪影以及复杂的 硬件和软件实现。这些限制阻止了它们被生物学所采用 实验室，仍然是神经科学研究的技术障碍。成功完成我们的提案 将克服这些局限性，并对神经科学研究产生深远影响。我们最近开发了一个 贝塞尔聚焦扫描技术（BEST），易于集成到现有的双光子显微镜中， 抗运动伪影，并已实现30赫兹，突触分辨体积成像， 稀疏标记的神经元群体在各种各样的模式生物。在这一建议中，结合 借助显微镜学家、生物学家和数据科学家的专业知识，我们建议进一步优化BEST， 能够高速、高通量和高分辨率地记录稀疏和 密集标记的回路遍布整个大脑。我们的目标是记录苍蝇的全脑活动在>10 Hz和小鼠中在~ 2 Hz下的穿透皮层体积成像。通过将BEST与显微内窥镜检查相结合， 我们的目标是实现30 Hz的突触分辨率体积显微内窥镜成像，并利用它来研究 小鼠脑深埋核团的结构和功能可塑性。通过纠正大脑诱发的 光学像差，自适应光学将使最佳保持突触分辨率在整个 小鼠皮层易于采用，BEST已被集成到多个双光子荧光 显微镜在世界各地的实验室。随着用户群的不断扩大， 优化项目将立即使广泛的实验室受益，使他们能够研究 在整个活体大脑中以前所未有的高时空分辨率测量体积神经活动。

To understand how the brain computes, we need to understand how individual neurons in a circuit integrate their numerous inputs into output signals, as well as how they work together to encode a sensory input or execute a motor command in a behaving animal. Circuits and neurons are three-dimensional (3D) and can extend over hundreds or thousands of microns. Therefore, understanding their operations requires monitoring their activity at both synaptic and cellular resolution in 3D at image rates that capture all activity events. Behaving animals present a host of challenges to this goal. Existing 3D imaging technologies suffer from insufficient volume imaging speed, brain-motion-induced image artifacts, as well as complex hardware and software implementation. These limitations have prevented their adoption by biology laboratories and remain a technical barrier for neuroscience research. Successful completion of our proposal will overcome these limitations and profoundly impact neuroscience research. We recently developed a Bessel focus scanning technology (BEST) that is easily integrated into existing two-photon microscopes, resistant to motion artifacts, and have already achieved 30-Hz, synapse-resolving volumetric imaging of sparsely labelled neuronal populations in a wide variety of model organisms. In this proposal, combining the expertise of microscopists, biologists, and data scientists, we propose to further optimize BEST to enable high-speed, high-throughput, and high-resolution volumetric activity recording of both sparsely and densely labelled circuits throughout the living brain. We aim to record whole-brain activity in the fly at >10 Hz and through-cortex volume imaging in the mouse at ~2Hz. By combining BEST with microendoscopy, we aim to achieve synaptic-resolution volumetric microendoscopic imaging at 30 Hz and use it to study structural and functional plasticity in deeply buried nuclei of the mouse brain. By correcting brain-induced optical aberrations, adaptive optics will enable BEST to maintain synapse resolution throughout the entire mouse cortex. Easily adoptable, BEST has already been integrated into multiple two-photon fluorescence microscopes in laboratories worldwide. With a continuously expanding user base, the proposed optimization project will immediately benefit a wide range of laboratories, allowing them to study volumetric neural activity at unprecedented high spatiotemporal resolution throughout the living brain.