Data-Driven Modeling of Signaling Dysregulation in Rheumatoid Arthritis

描述（由申请人提供）：从慢性炎症到纤维化疾病和癌症的人类疾病都以细胞信号通路失调为特征，针对这些通路的治疗方法在治疗癌症和关节炎方面显示出希望。关于单个信号蛋白和“规范”通路的广泛分子数据是可用的，但是从一种细胞类型到另一种细胞类型的信号网络的差异还不太清楚。了解与疾病状态相关的网络水平差异将大大促进新疗法的发展。在类风湿性关节炎（RA）中，新出现的证据表明，常驻成纤维细胞样滑膜细胞（FLS）在疾病进展中起关键作用，但对其失调的信号网络的研究有限。本提案概述了一种综合的实验和计算方法来系统地评估来自正常和患病个体的原代FLS细胞。我将构建FLS信号的预测数据驱动模型，并将信号网络活动与由此产生的细胞反应联系起来。我的具体目标有三个方面：(i)增加我们对FLS细胞在疾病中如何出错的理解，（ii）确定RA标准临床治疗方法对这些细胞的影响，（iii）预测和测试具有高治疗指数潜力的新药物靶点。在我们的初步研究中，我们收集了约15,000个数据点的概要，描述了正常或RA原代细胞（培养）在不同环境刺激下FLS的信号传导。在Aim 1中，我将在多个维度上扩展这一概要，包括从正常和RA患者供体中分离的8种不同的原代人FLS细胞分离物的信号和细胞反应的研究。这将提供对疾病状态与患者之间差异所产生的差异的见解。我还将直接评估在有和没有临床治疗方法的情况下RA的信号传导和反应，以确定在标准治疗模式下持续存在的信号传导节点。在Aim 2中，我将执行多种数据驱动的建模方法，从我们的初步研究和Aim 1中收集的数据中推断出有意义的见解。多元线性回归和偏最小二乘回归分析将特定信号通路的活动与细胞反应联系起来，基于逻辑的建模方法将分别用于生成正常FLS和RA FLS的细胞特异性信号网络模型。在Aim 3中，我将预测和测试用于RA治疗干预的新蛋白靶点。Aim 2中生成的预测模型将用于计算机中的假设检验，并通过实验对有希望的假设进行评估。总的来说，这种实验和计算分析将显著增加我们对类风湿关节炎的理解，并产生可能导致疾病的事件的精确分子模型。此外，它将创造一种综合方法，可用于确定一系列其他人类疾病的治疗干预的新位点。

DESCRIPTION (provided by applicant): Human diseases ranging from chronic inflammation to fibrotic disorders and cancer are characterized by dysregulation of cellular signaling pathways, and therapeutics targeting these pathways have shown promise in treating cancer and arthritis. Extensive molecular data is available on individual signaling proteins and on "canonical" pathways, but differences in signaling networks from one cell type to the next are less well understood. Understanding network-level differences associated with diseased-states would substantially advance the development of novel therapeutics. In rheumatoid arthritis (RA), emerging evidence points to the resident fibroblast-like synoviocytes (FLS) as key players in disease progression, but studies of the signaling networks underlying their dysregulation have been limited. This proposal outlines an integrated experimental and computational approach to systematically evaluate primary FLS cells from normal and diseased individuals. I will construct predictive data-driven models of FLS signaling, and link signaling network activity to the resulting cellular response. My specific goals are three-fold: (i) to increase our understanding into how FLS cells have gone awry in disease, (ii) to determine the effects of standard clinical therapeutics for RA on these cells, and (iii) to predict and test new drug targets with the potentil for high therapeutic index. In our preliminary studies we have colected a compendium of ~15,000 data points describing the signaling of FLS from normal or RA primary cells (in culture) in response to diverse environmental stimuli. In Aim 1 I will expand this compendium in multiple dimensions to include investigation of both signaling and cellular responses in eight different primary human FLS cell isolates from normal and RA patient donors. This will provide insights into differences arising from disease-state vs. patient-to-patient variability. I will also directl evaluate the signaling and responses in the presence and absence of clinical therapeutics for RA to identify signaling nodes that persist in the presence of standard treatment modalities. In Aim 2 I will perform multiple data-driven modeling approaches to infer meaningful insights from data collected in our preliminary studies and in Aim 1. Multilinear regression and partial least squares regression analyses will connect activities of specific signaling pathways with cellular responses, and logic-based modeling approaches will be used to generate cell-specific signaling network models for normal and RA FLS, respectively. In Aim 3 I will predict and test novel protein targets for therapeutic intervention in RA. The predictive models generated in Aim 2 will be used for hypothesis testing in silico, and promising hypotheses will be evaluated experimentally. Collectively, this experimental and computational analysis will significantly increase our understanding of rheumatoid arthritis and generate precise molecular models of events likely to underlie disease. Furthermore, it will create an integrated approach that can be used to identify novel sites for therapeutic intervention in a range of other human diseases.