Zarifi, Niayesh2025-02-212025-02-212025-02-21http://hdl.handle.net/10393/50207https://doi.org/10.20381/ruor-30941The design of tailor-made enzymes for new-to-nature reactions holds significant industrial potential. Despite advancements in de novo enzyme design, designed enzymes often exhibit catalytic efficiencies far lower than their natural counterparts, necessitating directed evolution to enhance their activities. This thesis aims to improve enzyme design methods to enable the creation of efficient enzymes without reliance on directed evolution by addressing key challenges in current approaches. Firstly, existing methods focus solely on the active site and do not include distal sites, whose contribution to catalysis is not well understood. In Chapter 2, we demonstrate that while active-site mutations alone can create ordered, efficient active sites, distal mutations can reorganize surface loops and widen tunnels to the active site, facilitating substrate binding and product release. The second key challenge is the availability of suitable scaffolds for enzyme design. In Chapter 3, we investigate minimal, de novo, and functionless TIM barrels as versatile templates customizable for any reaction. Using physics-based and machine-learning tools, we modified a simple TIM barrel and introduced structural segments to enclose the active site, creating a functional Kemp eliminase. The third challenge we address is the design of efficient enzymes for multistep reactions, which requires stabilizing multiple states along the reaction pathway. In Chapter 4, we present the first instance of multistate enzyme design, where we stabilized multiple chemical and conformation states to create de novo Michaelases. Our results suggest that to design efficient artificial enzymes, a holistic approach is required wherein the design procedure includes the whole reaction coordinate. This thesis advances computational enzyme design by unravelling the impact of distal mutations, evaluating the utility of machine-learning tools for enzyme design, and incorporating multistate approaches to design enzymes for multistep reactions. Our work lays the foundation for the design of highly efficient enzymes for diverse applications.enenzyme designThesis