Abstract
<title>Abstract</title> <p> Ancestral sequence reconstruction (ASR) is an effective semirational strategy for enzyme engineering that integrates evolutionary information with bioinformatic analysis. Here, ASR was combined with structure-guided mutagenesis using FuncLib and experimental validation to improve the catalytic performance and stability of the reductive aminase from <italic>Asp</italic> ergillus oryzae ( <italic>Asp</italic> RedAm). Phylogenetic analysis of <italic>Asp</italic> RedAm and its homologs enabled the reconstruction of ancestral variants, including Anc43, Anc74, and Anc75, and guided the identification of key substrate-binding pockets and catalytically relevant residues. Targeted mutagenesis was then performed to generate improved variants. Among the resulting enzymes, Anc74_7 and Anc75_20 exhibited markedly enhanced catalytic activity and thermostability relative to <italic>Asp</italic> RedAm. In particular, their specific activities increased by up to 1.6- and 2.2-fold, respectively. Molecular docking together with kinetic analysis suggested that these improvements arose from active-site reconfiguration and an optimized substrate-binding microenvironment. Collectively, these results demonstrate the utility of integrating ASR with structure-guided engineering for the development of high-performance biocatalysts and provide insight into enzyme functional evolution and rational design. </p>