基于periostin基因互作网络的抗心肌纤维化中药成分筛选及靶标识别

Anti-cardiac fibrosis is one of the key therapeutic strategies for the treatment of various cardiovascular diseases. Therefore, development of drugs targeting cardiac fibrosis is promising. It has been proved that many herbal medicines display definite therapeutic effects on preventing cardiac fibrosis. However, there are very few studies that systemically explore the underlying effective ingredients and their corresponding targets. In this research, we aim to build an integrated model for systematic pursuit of anti-cardiac fibrosis compounds and their targets based on a tissue-specific molecular network. Firstly, we develop a heart-specific periostin-based functional gene interaction network using a data-driven Bayesian methodology that integrates thousands of diverse experiments spanning tissue and disease states. Secondly, data mining technology is used to collect potential medical herbs against cardiac fibrosis in the literature. Then, these potential herbs are investigated using a mouse model with cardiac fibrosis induced by myocardial infarction. Thirdly, PCR array is applied to determine mRNA expression changes of all genes in the periostin-based gene networks affected by anti-cardiac fibrosis herbs. And then, a gene expression inference method is used to predict the pharmacological target space of anti-cardiac fibrosis herbs. Finally, three ligand-target interaction prediction methods are integrated to discern the binding ability between chemical ingredients of anti-cardiac fibrosis herbs and the pharmacological target space. Then, a network convergence method is used to integrate compound-target dataset and periostin-based gene network to build a compound-target-effect network. Those key compound–target pairs are extracted by analysis of the compound-target-effect network. Biochemical and cell experiments are used to investigate the interaction of these key compound–target pairs and the change of downstream pathways. This strategy integrating different types of technologies is expected to help create new opportunities for development of drugs targeting cardiac fibrosis.

抗心肌纤维化是改善各种心血管疾病的关键治疗策略之一，但其药物研发以针对相关靶标的零散研究为主，缺乏系统性。本项目以抗心肌纤维化的中药为研究对象，探索基于组织特异性病理分子网络的抗心肌纤维化中药成分筛选研究新模式。利用数据驱动的贝叶斯方法构建心脏组织特异性periostin基因相互作用网络；采用数据挖掘在文献中收集抗心肌纤维化潜在中药，以小鼠心梗后心肌纤维化为模型，筛选得到抗心肌纤维化中药；利用PCR列阵测定中药对periostin基因网络表达谱的影响，结合基因表达推演法预测靶标；联合配体-靶标相互作用预测方法识别中药成分和靶标之间的直接结合能力，利用网络融合方法融合成分-靶标数据集与periostin基因网络建立成分-靶标-效应网络，通过网络分析挑选关键成分-靶标组合，进行后续实验验证。本项目的完成有望识别一批抗心肌纤维化中药成分及靶标，建立一套适宜于中药抗心肌纤维化活性成分筛选的新方法。

以心肌纤维化为特征的心脏组织异常重构是多种心血管疾病发展到一定阶段的核心病理改变，是心功能及恶性心血管事件的重要病理基础。因此，抑制甚至逆转心肌纤维化过程，已经成为治疗心血管疾病的重要途径之一。在临床实践工作中，中西医在抗心肌纤维化领域均取得了明确的疗效，但目前成果仍不能满足临床需要，大部分靶向心肌纤维化的药物在临床转化中均遇到一定的困难。本项目提出抗心肌纤维化药物筛选的新思路，该思路以中药为研究对象，围绕心肌纤维化强效标志物periostin，建立组织特异性病理分子网络，并以此为基础筛选抗心肌纤维化中药效应成分。①成功构建心肌纤维化的多重分子网络，包括以periostin基因为核心的心肌纤维化功能网络，心脏组织特异的以periostin基因为中心的PPI（protein-proteininteraction）网络和心脏组织特异的以periostin基因为中心的基因共表达网络。这些网络作为标签，代表心肌纤维化病理过程，用于抗心肌纤维化药物的筛选和病理机制研究；②建立两种适用于中药的靶标预测方法：转录谱推演法和中药-靶标相互作用网络(HTINet)方法，从中药标签数据包括转录组、化学成分、中药症状等数据出发，可高效预测中药的效应靶标；③鉴于炎性免疫过程和心肌纤维化的紧密联系，我们从304份免疫细胞状态变化的转录谱中捕获了608个免疫标签（signatures）代表炎性免疫过程。通过该标签数据，成功发现一批中药和中药单体可通过改善炎性免疫治疗心肌纤维化；④在上述方法的基础上，成功建立了潜在抗心肌纤维化中药及其效应成分库，已成功识别了34个潜在抗心肌纤维化中药及其效应成分，实验验证了中药复方芪参颗粒和3个中药单体红景天甙、连翘苷、五味子酯甲的抗心肌纤维化效应。本项目的研究成果在抗心肌纤维化药物发现领域具有重要意义和应用价值，有望开发成为潜在的抗心肌纤维化药物，为预防和治疗心肌纤维化开辟新的路径，为临床选择有效的治疗手段提供重要的科学依据。