Biocatalysis is valued for its role in sustainable production of biofuels, fine chemicals, and pharmaceutical compounds, benefiting from chemoselectivity, enantioselectivity, and mild reaction conditions. Industrial biocatalysts like hydrolases possess desirable traits such as stability, catalytic efficiency, and broad substrate specificity, enabling effective synthetic transformations. However, the limitations of aqueous systems, particularly for hydrophobic substrates, impede enzyme applications. Non-aqueous media, such as organic solvents and ionic liquids, offer a potential solution for biocatalysis, but challenges arise including inefficient substrate binding, solubility issues, and enzyme inactivation caused by these media.

Enzyme engineering is crucial in academia and industry for developing biocatalysts in various fields ^[1]. However, predicting the effects of amino acid substitutions on enzyme function remains challenging ^[2]. To accelerate enzyme evolution, rational design approaches are used, focusing on redesigning enzymes through targeted libraries based on sequence-function relationships. In this presentation, I discuss the challenges and opportunities of engineering enzymes for catalysis in non-conventional media. By combining in silico methods and directed evolution, we have discovered principles at the molecular level that enable efficient operation of redesigned enzymes in organic solvents ^[3] and ionic liquids ^[4] . This provides insights into structure-function relationships, general design principles, and successful strategies for stabilizing enzymes in non-conventional media for sustainable catalysis.