Development of brain-scale neural circuits underlying vertebrate visuomotor transformations

摘要 在发育障碍后恢复视力仍然具有相当大的挑战性，例如先天性 婴儿期眼球震颤，并在受伤后或视网膜变性。这是因为建立的机制 视网膜神经节细胞与其大脑下游靶点之间的功能连接仍然很差 明白了。这种知识差距的部分原因是观察到功能的出现、稳定和 哺乳动物的整个视觉神经回路是不可能维持的。该项目将利用 实验优势斑马鱼幼体为脊椎动物模型系统，研究其功能 作为视觉定向行为基础的保守神经回路的成熟，即视觉运动反应(OMR)。 这将为了解先天性疾病如何发挥作用以及新神经元如何发挥作用奠定基础。 在初始电路开发后添加可以支持健康的视觉处理。最近，我们描述了 将视网膜视觉运动信号转换为运动输出，并表明它需要许多不同的类型 分布在大脑各处的神经元。这些神经元可以根据它们不同的眼睛进行分类-和 特定方向的响应配置文件，它们协作计算视觉场景到底是如何移动的。 令人着迷的是，这种合作支持受精后5天的稳定行为，即使新的神经元 在整个生命周期中都被添加到赛道中。我们将检验中心假设，即在初始形成后，OMR电路 通过在平衡反应类中增加新的神经元来扩展，允许继续执行运动- 引导性行为。在目标1中，我们将测试行为指令表和相关神经的发展 回路受到方向选择性视网膜输入的特定干扰的影响。通过训练递归神经网络， 我们将生成方向选择性视网膜神经节细胞和 下游目标。在目标2中，我们将研究功能神经表征是如何成熟的，我们将 量化个体神经元反应随时间变化的稳定性。通过计算跟踪所有神经元，我们将 直接调查新的神经元功能集成到现有电路的轨迹，并确定如何 功能配置的平衡随着时间的推移而变化，并随着行为的变化而变化。在《目标3》中，将使用全息 光刺激，以考察活动在塑造单个神经元的最终电路角色中的作用。一起， 这些实验将揭示整个运动敏感脊椎动物电路是如何在功能上组装的， 深入了解视网膜神经节细胞及其下游伙伴之间的功能连接 以及关于神经发生的性质和用途。这些结果将为再生治疗策略提供参考 中央视觉处理区域的发育障碍或损伤。

ABSTRACT It remains considerably challenging to restore vision after developmental disturbances, such as congenital infantile nystagmus, and after injury or retinal degeneration. This is because the mechanisms establishing functional connectivity between retinal ganglion cells and their downstream targets in the brain remain poorly understood. This knowledge gap is partly because observing the functional emergence, stabilization, and maintenance of entire visual neural circuits is impossible in mammals. This project will leverage the strategic experimental advantages of the larval zebrafish, a vertebrate model system, to investigate the functional maturation of a conserved neural circuit underlying a visual orienting behavior, the optomotor response (OMR). This will form the basis for understanding how congenital disorders exert their effects and how new neurons added after initial circuit development can support healthy visual processing. Recently, we described the transformation of retinal visual motion signals into motor output and showed that it required many different types of neurons distributed across the brain. These neurons can be classified based on their diverse eye- and direction-specific response profiles, and they collaborate to compute how exactly visual scenes are moving. Fascinatingly, this collaboration supports stable behavior 5 days after fertilization, even though new neurons are added to the circuit throughout life. We will test the central hypothesis that after initial formation, the OMR circuit expands by adding new neurons in balanced response classes, permitting the continued execution of motion- guided behaviors. In Aim 1, we will test how the development of the behavioral repertoire and associated neural circuitry is affected by specific disruption of direction-selective retinal input. By training recurrent neural networks, we will generate predictive models of connectivity between direction-selective retinal ganglion cells and downstream targets. In Aim 2, we will investigate how the functional neural representations mature, and we will quantify the stability of individual neuronal responses over time. By computationally tracking all neurons, we will directly investigate the trajectory of new neuron functional integration into existing circuitry and determine how the balance of functional profiles varies over time and covaries with behavior. In Aim 3, will use holographic photostimulation to examine the role of activity in shaping ultimate circuit role for individual neurons. Together, these experiments will reveal how an entire motion-sensitive vertebrate circuit is functionally assembled, providing insight about the functional connectivity between retinal ganglion cells and their downstream partners and about the nature and utility of neurogenesis. These results will inform regenerative treatment strategies for developmental disorders or injuries to central visual processing areas.