EmCAST: Stabilizing Proteins and Tuning Dynamics with High Precision and Accuracy

项目摘要 没有可靠的方法来稳定蛋白质的高精度。目前可用的方法有 观察到的突变对蛋白质稳定性的影响与预测的影响之间的标准误差范围为1 - 3 千卡/摩尔考虑到蛋白质稳定性对于生物医学应用的重要性，例如保质期和 基于蛋白质的药物的免疫原性，稳定蛋白质的可靠方法的开发， 高精度是关键。为了解决目前知识的不足，我们开发了EmCAST（经验 C-Alpha稳定性工具），并且已经表明它可以使小的三螺旋束乌巴（1）的稳定性加倍， 有四个突变。对于一组八个单重、双重、三重和四重突变变体，其中包含 这四个突变的组合，预测和观察到的稳定性之间的平均误差为0.13 kcal/mol，这是对现有预测稳定突变方法的巨大改进。EmCAST依赖于两个 重要的创新：1）使用从α碳（C++）数据库导出的经验势 所有可能的四残基序列的二面角偏好从2018年发布的 蛋白质数据库和2）选择表面暴露位点用于引入稳定突变。在 建议的工作，我们将证明，EmCAST可以是一个有效的工具，以稳定广泛的蛋白质 折叠，并且它可以用于调节蛋白质构象开关的位置，从而控制蛋白质 功能我们还将发布、维护和升级一个Web服务， 可以轻松地访问和使用这个有价值的工具。 我们将在以下目标的范围内实现这些目标： ·在目标1，纯N-Domain和N-Domain的有理稳定化中，我们将证明EmCAST可以稳定 一组四个额外的螺旋域具有高精度，它也可以应用于稳定 的结构域。这些蛋白质的预测稳定范围为2.5至6千卡/摩尔。 ·在目标2中，混合双/双域和大折叠的稳定化，我们应用EmCAST来稳定集合 四个更复杂的折叠，包括β-螺旋和β-折叠结构，序列长度高达 270个氨基酸。EmCAST预测所选蛋白质的稳定性为3至5 kcal/mol。 ·在目标3，调节环动力学和调整构象开关的位置，我们 将显示EmCAST可应用于蛋白质的交替构象的差异稳定， 允许调节蛋白质功能。

Project Summary There are no reliable methods to stabilize proteins with high accuracy. Currently available methods have standard errors between observed and predicted effects of mutations on protein stability that range from 1 to 3 kcal/mol. Given the importance of protein stability for biomedical applications such as the shelf-life and immunogenicity of protein-based pharmaceuticals, development of reliable methods to stabilize proteins with high accuracy is critical. To address this deficit in current knowledge, we have developed EmCAST (Empirical C-Alpha Stability Tool) and have shown that it can double the stability of a small three helix bundle, UBA(1), with four mutations. For a set of eight single, double, triple, and quadruple mutant variants that contain combinations of these four mutations, the average error between predicted and observed stability was 0.13 kcal/mol, a vast improvement over existing methods to predict stabilizing mutations. EmCAST relies on two important innovations: 1) use of an empirical potential derived from a database of the alpha carbon (C) dihedral angle preferences for all possible four-residue sequences extracted from the 2018 release of the Protein Data Bank and 2) selection of surface-exposed sites for introduction of stabilizing mutations. In the proposed work, we will demonstrate that EmCAST can be an effective tool to stabilize a broad range of protein folds and that it can be used to tune the position of protein conformational switches and hence control protein function. We will also release, maintain, and upgrade a web service so that the protein biochemistry community can readily access and use this valuable tool. We will accomplish these goals in the context of the following Aims: • In Aim 1, Rational Stabilization of Pure  and  Domains, we will show that EmCAST can stabilize a set of four additional helical domains with high accuracy and that it can also be applied to stabilization of -sheet domains. Predicted stabilizations for these proteins range from 2.5 to 6 kcal/mol. • In Aim 2, Stabilization of Mixed / Domains and Large Folds, we apply EmCAST to stabilize a set of four more complex folds that include both -helix and -sheet structure with sequence lengths up to 270 amino acids. EmCAST predicts stabilizations of 3 to 5 kcal/mol for the selected proteins. • In Aim 3, Regulating Loop Dynamics and Tuning the Position of Conformational Switches, we will show that EmCAST can be applied to differential stabilization of alternate conformers of proteins, allowing for tuning of protein function.