运用单细胞全脑示踪技术解析小鼠初级视觉皮层（V1）的投射规律

Visual signal contains rich information about the external environment. The capability of efficient distributing relevant information to the downstream brain areas is a key feature of primary visual cortex (V1). Previous work has identified the targeting areas of V1, but how information flows between them is still not clear, because neighboring neurons within a brain region can exhibit broad heterogeneity in both projection patterns and function properties. Identified long-range projecting areas can receive similar or distinct information from the same source region. Therefore reconstruction of single neurons is crucial for revealing how neural signals are organized and distributed to target areas. We developed a platform to reliably label, image and reconstruct single neurons across the entire brain and a suite of computational tools to process such large-scale image data. It enables us to quantitatively analyze the diversity of single-neuron projection patterns in mouse primary visual cortex. We will reveal the rules how visual information is broadcasted to the downstream brain areas, such as (1) the relationship between several higher visual areas, (2) the relationship within and between motor related and memory related downstream areas; (3) the rules of distributing visual information between visual related and non-related areas; and (4) identify striatum projecting neurons in the visual cortex and their full projecting patterns. Those findings are key steps towards the fundamental circuit mechanism of high brain functions like perception, cognition and memory formation. Finding answer to those questions will shed light on the causes and possible treatment of neuronal and psychiatric disorders.

视觉信号中包含了大量关于外界环境的信息，将相应的视觉信号合理地分配到下游脑区是保证视觉信息被高效处理和利用的基础。但是我们对初级视觉皮层（V1）的投射规律缺乏全面的量化分析，无法将视觉信息的分配原则和接受视觉信号的下游脑区的功能联系起来。由于现有方法无法定量分析V1的投射规律，我们在前期建立了定量分析全脑范围单细胞长程连接多样性的平台和针对大容量三维图像堆的数据处理流水线。采用该方法，我们能定量分析Ｖ１单细胞长程投射的多样性，揭示视觉信息在已知的长程连接中的分流规律。我们将着重分析：（1）视觉信息在多个高级视觉皮层间分配的原则；（2）流向行动相关和认知记忆相关的区域时的分流规律；（3）在视觉高级皮层和行动、记忆相关脑区之间分配的原则；（4）识别Ｖ１投射到纹状体的通路，并分析该通路与其他目标区域的关系。这对揭示产生感知、形成记忆、控制行动的神经环路机制有重要意义。

大脑的一个重要特征就是能够非常高效地对海量的感觉信息进行即时处理，这依赖于大脑强大的计算功能和高效率的信息分流的方式。单个神经元作为大脑功能和结构的基本单位，其本身的轴突长程连接构成了跨脑区信息传递的独立通道，因此，单个神经元投射连接的多样性决定了大脑信息传递可能存在的多种分配模式。本项目利用高通量的全脑单细胞形态的成像和分析平台，取得了约120个小鼠初级视觉皮层的单细胞的轴突形态，对约600个神经元的多脑区投射模式进行了定量统计分析，确定了大脑皮层存在的多种信息分流的规律：一对一独立输出和一对多广播式输出的共同存在，首次明确了后者中随机性和有序性两类广播模式包含的组合规律。这对建立类脑神经网络计算模型提供了非常重要的新思路和实验支持。同时利用在项目执行过程中建立的全脑图像重构和分析平台，对小鼠阿尔兹海默症模型中的胆碱能神经元和神经纤维分布和形态在全脑范围内进行了系统分析，发现了在发病早期的新的病理特征，为开发针对认知功能缺陷的新疗法提供新靶点。