Investigation of the Synaptic Molecular Network using Multiplexed Imaging

项目总结 突触分子网络是一个由数百种蛋白质组成的复杂、紧密相互作用的系统，形成了 学习、记忆和其他大脑功能的基础。它是许多神经学上被破坏的场所， 神经退行性疾病和精神疾病，是小分子精神疾病的一个行动重点 治疗。理解这个网络和支配它的规则对于理解分子是必要的。 脑部疾病的病因学，以及精神病药物的合理设计。为了实现这一目标，我们建议应用 PRISM是最近开发的一种多路成像工具，它可以在 单突触解析，以构建和验证相互依赖的因果预测模型 谷氨酸能突触的蛋白质、它们的亚单位组成和激活状态。我们将扩大 可用于PRISM的目标曲目，并在跨 各种化学环境。这将有两个目的：测量突触的许多快照 分子网络被拉向不同的方向，这是后续模型学习所必需的，并深入 包括抗抑郁药在内的扰动对下游突触生化影响的表征 不同的班级。我们还将结合对突触活性的敏感的活体钙成像和后续的棱镜检查 对相同突触的测量。然后我们将使用这些数据来构建一个贝叶斯网络模型 这些测量的概率分布之间的因果依赖关系。这个贝叶斯网络将 关于节点之间的因果关系和扰动某些目标的下游影响的产量预测， 我们随后将对其进行测试。这项研究的结果将是一个强大的预测性模型，将多达30 蛋白质水平、亚单位组成、磷酸化状态和突触活性的测量。这款车型将 提供统一的上下文，以将特定突触参与者之间交互的机械性细节集成到 对突触作为一个整体的整体理解。它还将提供对系统级影响的预测 化学扰动，潜在地为一种全新的体外筛选模式铺平了道路 精神治疗。

PROJECT SUMMARY The synaptic molecular network is a complex, tightly interacting system of hundreds of proteins that forms the basis for learning, memory, and other brain functions. It is a disrupted locus of many neurological, neurodegenerative, and psychiatric disorders, and is a focal point of action for small molecule psychiatric treatments. Understanding this network and the rules that govern it is necessary for understanding the molecular etiology of brain diseases, and for the rational design of psychiatric drugs. To achieve this, we propose to apply PRISM, a recently developed multiplexed imaging tool which allows in-situ measurements of many proteins at single-synapse resolution, to construct and validate a causal, predictive model of interdependencies among proteins of the glutamatergic synapse, their subunit composition, and activation state. We will expand the repertoire of targets available for PRISM and measure these targets in a large population of synapses across a variety of chemical environments. This will serve two purposes: measurement of many ‘snapshots’ of the synaptic molecular network pulled in different directions, which is necessary for subsequent model learning, and in-depth characterization of the downstream synaptic biochemical effects of perturbations that include antidepressants of different classes. We will also combine sensitive live calcium imaging of synapse activity with subsequent PRISM measurements of the same synapses. We will then use these data to construct a Bayesian network model of causal dependencies between the probability distributions of these measurements. This Bayesian network will yield predictions about causal connections between nodes and downstream effects of perturbing certain targets, which we will subsequently test. The result of this study will be a powerful, predictive model connecting up to 30 measures of protein levels, subunit compositions, phosphorylation states, and synapse activity. This model will provide a unifying context to integrate mechanistic details of interactions between specific synaptic actors into a holistic understanding of the synapse as a whole. It will also provide predictions about system-level effects of chemical perturbations, potentially paving the way for an entirely novel modality for in vitro screening of psychiatric treatments.