Stereocontrol in Reductive Aminases through Steric Modification of Residues Around Substrate and Cofactor Binding Pockets

Enzymatic reductive amination catalysed by imine reductases (IREDs) and reductive aminases (RedAms) provides green, stereoselective, and direct access to 2° and 3° chiral amines. Given existing interests in exploiting RedAms for commercial-scale synthesis of active pharmaceutical ingredients, developing efficient enantiodivergent RedAm systems would boost biocatalysis prospects in early drug discovery. To develop adaptable enantiodivergent RedAm systems, residues/motifs that can serve as handles for stereocontrol in this enzyme family need to be identified. In this work, we have identified residues at both the cofactor and substrate binding pockets in a fungal reductive aminase (MaRedAm) and a bacterial imine reductase (AoIRED) that allow stereocontrol through steric modification of the side chains. In MaRedAm, steric bulk removal as in W33A and R35A at the cofactor binding pocket improved (R)-selectivity towards the synthesis (R)-rasagiline with up to 95% e.e., whereas W211A mutation at the substrate binding pocket inverted the selectivity of MaRedAm to yield (S)-rasagiline (42% e.e.). Similarly, varying steric bulk at N241 in AoIRED achieved enantiodivergency as N241L/H/F/Y variants produce (S)-configured 1-methyl-1,2,3,4,-tetrahydroisoquinoline (92->99% e.e. compared to 85% e.e. (S) achieved by the wild-type), whereas N241A/S/C mutations yield the (R)-configured amine (up 40% e.e.). Atomistic insights obtained from molecular dynamics (MD) simulations revealed that steric modulation at the cofactor binding pocket (as in MaRedAM W33A and W35A) causes the reorganisation and downsizing of the active site. In contrast, the W211A mutation at the substrate binding pocket showed this pocket opening up to allow flexibility for the rotation of the substrate. Our work lays an important foundation for speedy rational engineering of stereoselectivity in IREDs/RedAms.