还原性酮胺化酶顺序催化酮和胺脱水缩合成亚胺，并不对称还原亚胺生成手性仲胺。其底物特异性取决于酶对两步反应中“双重”底物（酮与胺/亚胺离子）的识别。但由于缺乏酮胺化酶顺序识别“双重”底物的分子机制研究，分子改造其底物特异性难度较大，是制约其应用的重要因素。我们前期克隆得到来源于米曲霉的酮胺化酶ZJUT212，解析了裸蛋白的晶体结构；底物谱研究显示其对短链芳香族酮胺底物强烈偏好，对脂肪族底物的活力较低；动力学分析揭示限速步骤在第一步。本项目拟解析该酶与不同类型底物结合的复合物结构，结合分子动力学模拟，以“静动结合”的手段建立其底物识别模型，解析该酶在识别双重底物过程中底物结合中心的结构变化及关键影响因素，揭示其在催化两步反应过程中的结构动态转换机制，并通过定向进化调控其底物特异性，实现底物谱的拓展。本研究有助于增加对酮胺化酶结构与功能关系的认识，同时为理性改造其底物特异性提供新思路和新方法。

Reductive aminase (RA) sequentially catalyze the condensation of ketones and amines into imines and asymmetrically reduce imines to form chiral secondary amines. The substrate selectivity lies in the "dual" substrate recognition of the two-step reaction, i.e. first the ketone and amine of the step and the imine ion of the second step. However, the molecular mechanism of this enzyme sequentially recognizes the substrate is not clear, resulting in a narrow substrate range. Previously, we cloned an aminase ZJUT212 from Aspergillus oryzae and solved its structure in apo form; but its substrate spectrum showed strong preference for short-chain aromatic ketones and amines than aliphatic compounds; dynamic analysis reveals rate-limiting step is from ketone to imine. In this project, a "Static (crystal structure) and Dynamic (MD simulation)" method is used to establish substrate recognition model for analyze of the changes in the conformational dynamics and domains of the active site and the key influencing factors in the process of identifying the substrates. Furthermore, the substrate recognition model analysis will reveal the changes in the conformational dynamics of active site, key-motifs and key-domains, which deciphers the structure-dynamic-transformation mechanism during the sequential reaction process. Moreover, the obtained rules and principles can guide protein engineering to regulate substrate specificity and to extend its substrate spectrum. This study can increase the understanding of structure-function relationships of RA and provide a new insight into rational engineering of its substrate specificity.