Protein Engineering via Combinatorial Fragment Pairing

描述（由申请人提供）：用于合成化学的定制生物催化剂设计是生物技术研究的一个非常热门的领域。酶的吸引力不仅是因为它们在很少或不需要能量输入的情况下实现了显著的速率提高，而且它们还避免了与大多数常规催化剂相关的有毒废物问题。此外，酶在区域特异性和立体特异性方面的精确性对制药和精细化学品制造商具有重大意义。定制生物催化剂设计的定向进化方法的进展一直很活跃，特别是在过去的五年里。然而，酶功能的许多潜在的非常有用的改变是不可访问的，因为没有有效的技术存在重塑活性位点的总体特征。目前的定向进化方法学在特异性（即，将选择压力精确地集中在所寻求的功能性上）方面表现出色，它们利用该特异性搜索邻近起始结构的序列空间区域，但是为了对结合口袋几何形状进行实质性改变，有必要首先对序列空间的大面积区域进行采样，以期鉴定用于优化的合理起始点。这关键的第一步的成功取决于两个因素：文库大小和选择阈值。所提出的技术利用基于互补蛋白片段配对的创新组合方法，以极低的亲和力选择阈值访问前所未有的大小的文库。此外，所提出的方法避免了配体固定化，因此与活性位点盖子的体外进化完全兼容，因此经常发现对天然酶的能力至关重要。

DESCRIPTION (provided by applicant): Custom biocatalyst design for use in synthetic chemistry is a very hotly pursued area of biotechnological research. Not only are enzymes attractive because they achieve remarkable rate enhancements with little or no requirement for energy input, but they also avoid the toxic waste problem associated with most conventional catalysts. Furthermore, the precision with which enzymes act, both in terms of regio- and stereo-specificity, is of substantial interest to pharmaceutical and fine-chemical manufactures. Progress in directed evolution approaches to custom biocatalyst design has been brisk, especially in the past five years. However, many potentially very useful alterations to enzyme function are inaccessible because no effective technology exists for reshaping gross features of the active site. Current directed evolution methodologies excel in the specificity (i.e. focusing selective pressure precisely on the sought functionality) with which they search regions of sequence space proximal to the starting structure, but to make substantial changes to binding pocket geometry it is necessary to first sample a vast area of sequence space in the hope of identifying reasonable starting points for optimization. Success in this crucial first step is dependent upon two factors: library size and threshold of selection. The proposed technology utilizes an innovative combinatorial approach, based on pairing of complementary protein fragments, to access libraries of unprecedented size with an extremely low threshold for affinity selection. Furthermore, the proposed methodology obviates ligand immobilization and thus is fully compatible with in vitro evolution of the active site lids so often found to be critical for natural enzyme proficiency.