Structure-based design of protein function

描述（由申请人提供）：我的实验室开发的基于计算的、基于结构的蛋白质设计方法的最新进展，成功地预测了急剧改变受体蛋白配体结合特异性的突变。利用这些设计技术，大肠杆菌周质结合蛋白（PBP）超家族的几个成员通常结合糖或氨基酸，已被转化为具有高亲和力和特异性的识别化学上不同配体的受体。荧光和电化学报告基团的引入使得工程PBPs可以用作无试剂光学或生物电子生物传感器。受体也可以重新引入大肠杆菌，在那里它们控制合成信号转导途径，介导对非自然细胞外化学信号的转录激活反应。这些早期结果非常令人鼓舞，并表明可以考虑各种生物功能的计算设计。然而，为了使这种能力成为现实，有必要进一步发展计算设计技术，并将其扩展到处理越来越复杂的结合位点，如蛋白质-蛋白质和蛋白质- dna相互作用。以高精确度和复杂性设计蛋白质的能力在生物医学上有许多应用。我建议进一步开发和实验测试用于操纵蛋白质分子识别的计算设计技术，使用特定的生物医学相关应用作为设计目标，指导将要设计的受体系统和配体的选择。因此，下面确定的任务旨在具有明确的实际应用，说明计算蛋白质工程在生物医学科学中的潜在广泛用途，同时探索有关蛋白质分子识别的基本科学问题。Aims 1-3的重点是开发具有显著改变配体结合特性的受体。这些课程探索了临床科学（目标1）、药理学（目标2）和细胞生物学（目标3）的不同应用。从基础科学的角度来看，它们将使我们能够开发工程结合位点的技术，并探索除PBPs之外的支架（目的2）。目标4旨在将设计技术扩展到更复杂的系统。

DESCRIPTION (provided by applicant): Recent advances in computational, structure-based protein design methods, developed in my laboratory, successfully predict mutations that drastically alter the ligand-binding specificity of receptor proteins. Using these design techniques, several members of the E. coli periplasmic binding protein (PBP) superfamily that normally bind sugars or amino acids have been converted into receptors that recognize chemically diverse ligands with high affinity and specificity. Introduction of fluorescent and electrochemical reporter groups has permitted the engineered PBPs to be used as reagentless optical or bioelectronic biosensors. The receptors also can be re-introduced into E. coli where they control synthetic signal transduction pathways that mediate transcriptional activation response to non-natural, extracellular chemical signals. These early results are highly encouraging and suggest that the computational design of a wide variety of biological functions can be contemplated. However, in order for this capability to become a reality, it is necessary to further develop the computational design techniques, and to extend them to dealing with binding sites of increasing complexity, such as protein-protein and protein-DNA interactions. The ability to engineer proteins with a high degree of precision and sophistication has numerous biomedical applications. I propose to further develop and experimentally test the computational design techniques for manipulating molecular recognition in proteins, using specific, biomedically relevant applications as design targets that guide the choice of receptor systems and ligands that will be engineered. The tasks identified below are therefore intended to have clear practical applications, illustrating the potential wide-ranging utility of computational protein engineering to the biomedical sciences, while at the same time exploring basic scientific questions regarding molecular recognition in proteins. Aims 1-3 are focused on the development of receptors with drastically altered ligand-binding properties. These explore different applications in clinical science (Aim 1), pharmacology (Aim 2), and cell biology (Aim 3). From a basic science point of view, they will allow us to develop the techniques for engineering binding sites, and to explore scaffolds other than the PBPs (Aim 2). Aim 4 is intended to extend the design technique to much more complex systems.