Mimicking the secondary structures of proteins and peptides to modulate Protein-Protein Interactions

人类基因组编码大约20,000种不同的蛋白质，形成了一个潜在的蛋白质-蛋白质相互作用(PPI)的巨大网络，这一网络只有部分特征。蛋白质-蛋白质相互作用体(涉及PPI的网络)是细胞内通路的中心，并决定任何特定蛋白质在生理学中的作用。这种PPI的调节有望阐明生物系统中的关键分子机制，并对理解蛋白质如何发挥细胞功能至关重要。该计划的核心是开发模拟多肽的分子，即能够模拟多肽结构的分子，同时有利地调整它们的不良性质，并利用计算科学来分析这种支架的动力学，并优先合成化合物。1.以多样性为导向合成富含热点的PPI抑制物库对PPI的解剖结构进行了细致的分析，从而能够表征参与识别的分子特征。我们将利用最近的发现和我们在固相支持合成方面的专业知识，开发一个新型模拟多肽的PPI调节剂文库，旨在揭示与热点中过多代表的侧链相同的分子识别元件。这些元素将被放置在支架上，这些支架提供适合PPI化学空间的化学空间和形状，这些空间富含已知PPI抑制剂中过度丰富的分子识别元素。我们进行了彻底的电子分析，得到了一类在拓扑结构和分子描述符方面具有适当轮廓的化合物。2.用三维预测方法合理地设计和合成了PPI界面的二级结构。结构生物学的最新进展使我们能够在分子水平上阐明复杂的大分子组件的结构，并表征蛋白质之间的相互作用界面。利用一种创新而强大的三维构象预测方法，结合分子动力学、独立成分分析(TICA)和本课题组最近开发的高斯混合隐马尔可夫模型(GM HMM)，我们将合理地设计和合成模仿两种感兴趣蛋白质之一的PPI界面上的表位和二级结构。这一涉及化学、生物学和计算科学的多学科研究计划有望在PPI目标上提供比标准HTS文库更高的命中率。计算科学和实验评估协会将就获得高度选择性和高效的PPI调节器所需的关键相互作用提供见解。文库有望成为一套通用的化学生物学工具，用于表型分析，以破译关键的PPI途径。

The human genome encodes for approximately 20,000 different proteins creating a vast network of potential protein-protein interactions (PPI), which is only partially characterized. The protein-protein interactome (network involving PPIs) is central to intracellular pathways and determines any given protein's role in physiology. Modulation of such PPI promises to shed light on the key molecular mechanisms within a biological system and is crucial to understanding how proteins perform cellular functions. At the heart of the program is the development of peptidomimetic molecules, i.e., molecules capable of mimicking the structure of peptides while advantageously adjusting their undesirable properties and the exploitation of computational sciences to analyze the dynamics of such scaffolds and to prioritize compounds to synthesize. 1. Diversity-oriented synthesis of hot-spot enriched PPI inhibitor library Meticulous analysis of the anatomy of PPIs has allowed to characterize the molecular features involved in recognition. We will take advantage of recent findings and our expertise in solid phase-support synthesis to develop a library of novel peptidomimetic PPI modulators designed to expose the same molecular recognition elements as over-represented side chains in hot-spots. These elements will be on scaffolds that provide a chemical space and shape appropriate for the PPI chemical space enriched with molecular recognition elements over-abundant in known PPI inhibitors. We performed a thorough in silico analysis to produce a class of compounds that possess the proper profile in terms of topology and molecular descriptors 2. Rational design and synthesis of secondary structures at PPI interface using 3D prediction approach. Recent advances in structural biology now allow us to elucidate, at the molecular level, the structure of complex macromolecular assemblies and characterize the interactions interface between proteins. Using an innovative and powerful 3D conformation prediction approach combining molecular dynamics, independent component analysis (tICA), and a Gaussian Mixture Hidden Markov Model (GM HMM) recently developed in our group, we will rationally design and synthesize epitope and secondary structures at PPI interfaces mimicking one of the two proteins of interest. This multidisciplinary research program at the interface of chemistry, biology, and computational science is expected to give superior hit rates compared to standard HTS libraries on PPI targets. Association of computational sciences and experimental evaluation will provide insights on the key interactions needed to access highly selective and efficient PPI modulators. Libraries are expected to become a versatile chemical biology set of tools used in phenotypic assays to decipher critical PPI pathways.