A Connectomic Analysis of a Developing Brain Undergoing Neurogenesis

项目总结/摘要 在发育和成年的过程中，人类大脑建立了由以下组成的神经回路： 成千上万种神经元当新的神经元诞生时，它们被整合到发育中和现有的神经元中。 连接附近神经元和大脑远处神经元的回路。 许多神经系统疾病都与大脑中网络的不当生长有关。然而，我们缺乏基本的 了解神经回路如何随着新神经元的加入而变化。为了解决这一问题，该提案使用 一种新的动物模型，软体动物，Berghiastephanieae，其中有可能构建一个细胞-和 这是几个幼年期和成年期整个大脑的突触水平接线图。利用这些整体 大脑连接体，该项目将跟踪特定神经元，神经回路和整个大脑的变化 随着大脑中神经元数量的增加超过40倍， 神经元将通过交叉标记基因表达来识别，使用多达5个原位组 标记不同mRNA序列的杂交链反应探针。将使用重叠的探针组 这样就可以根据它们的基因模式来区分可单独识别的神经元和神经元类型， 表达与其索马位置和大小相结合。此外，在成年动物中，神经元将被标记 用荧光示踪剂追踪大脑神经。机器学习（ML）将用于 根据所有这些特征对神经元类型进行分类。发育神经元的ML分类 阶段将由人类校正，以增强ML的预测能力。 将构建一系列成年人和青少年四个阶段的大脑连接体。的 大脑将被连续切片。将使用61束扫描对每个30 nm厚的切片进行成像 电镜切片将对齐，所有神经元将自动重建为3D。的 重建将包括所有轴突、树突和突触。同样，人类将校对结果， 纠正ML算法。结果将是五个完整的大脑连接体， 从500个神经元到超过23,000个神经元。 发展系列将进行分析，以测试有关组织和发展的假设， 神经元、神经回路和整个大脑网络。识别出的神经元的神经结构的变化将被 跟踪发展。随着新神经元的加入，将在神经类型之间进行比较。 完整的视觉，嗅觉和运动系统的神经回路将被确定。最后，该项目将 确定枢纽是否围绕最古老的神经元发展，或者网络是否在没有集中的情况下扩展 连接到特定的集线器。这些结果将提供一个前所未有的视角，来观察大脑皮层的突触网络是如何形成的。 整个大脑的神经元随着新神经元的加入而改变。

PROJECT SUMMARY / ABSTRACT Over the course of the development and into adulthood, the human brain builds neural circuits composed of thousands of types of neurons. As new neurons are born, they are incorporated into developing and existing circuits making connections to neurons that are nearby as well as neurons that are in distant parts of the brain. Many neurological conditions are related to the improper growth of networks in the brain. Yet, we lack a basic understanding of how neural circuits change as new neurons join. To address this question, this proposal uses a novel animal model, the mollusc, Berghia stephanieae, in which it is possible to construct a cellular- and synaptic-level wiring diagram of the entire brain at several juvenile stages as well as the adult. Using these whole brain connectomes, the project will track the changes in specific neurons, in neural circuits, and in whole brain networks as the number of neurons in the brain increases by over 40-fold. Neurons will be identified by intersectional labeling of gene expression using sets of up to five in situ hybridization chain reaction probes that label different mRNA sequences. Overlapping sets of probes will used so that individually identifiable neurons and neuron types can be distinguished based on their patterns of gene expression combined with their soma location and size. Additionally, in adult animals, neurons will be labeled using fluorescent tracers applied to nerves emanating from the brain. Machine learning (ML) will be employed to classify neuronal types based on all of these features. ML classifications of neurons across developmental stages will be corrected by humans to enhance the predictive power of the ML. A series of connectomes of the brains of an adult and juveniles from four stages will be constructed. The brain will be serially sectioned. Each of the 30 nm thick sections will be imaged using a 61 beam scanning electron microscope. The sections will be aligned and all neurons will be automatically reconstructed in 3D. The reconstructions will include all axons, dendrites, and synapses. Again, humans will proofread the results to correct the ML algorithm. The result will be five complete brain connectomes spanning from the early juvenile with 500 neurons to the mature adult with over 23,000 neurons. The developmental series will be analyzed to test hypotheses about the organization and development of neurons, neural circuits, and entire brain networks. Changes in neural structure of identified neurons will be tracked over development. Comparisons will be made between neural types as new neurons are added. Complete neural circuitry for visual, olfactory, and motor systems will be determined. Finally, the project will determine whether hubs develop around the oldest neurons or whether the network scales without concentrating connectivity at particular hubs. The results will provide an unprecedented look at how the synaptic networks of neurons across an entire brain change as new neurons are added.