Repurposing Thiamine-Dependent Benzaldehyde Lyases as Visible-Light-Driven Radical Acyltransferases

Conspectus Thiamine diphosphate (ThDP)-dependent enzymes are among nature’s most elegant biocatalysts for C–C bond formation and cleavage, typically operating through a two-electron umpolung mechanism. Although a small subset of ThDP enzymes, such as pyruvate: ferredoxin oxidoreductase (PFOR), has been shown to proceed via single-electron radical pathways, this knowledge had not yet been harnessed for asymmetric radical reactions. Inspired by thiamine biocatalysis, chemists have developed biomimetic N -heterocyclic carbene (NHC) complexes since the 1950s, and more recently, established radical NHC catalysis as a powerful platform for radical transformations. In this approach, an NHC-stabilized ketyl radical undergoes cross-coupling with a free C-centered radical. However, achieving enantioselective cross-coupling involving prochiral C-center radicals had remained an unsolved challenge with chiral chemo-NHC catalysts. This difficulty stems from both the highly reactive nature of these open-shell intermediates and the inherent limitation of small organo-NHC complexes in creating a suitable and tunable chiral environment. To address the above challenges, we drew inspiration from nature’s ThDP-dependent biocatalysis and biomimetic chemo-NHC-based radical pathways, developing a synergistic photoredox-/thiamine radical biocatalysis for asymmetric radical acylations. In this system, the thiamine-dependent benzaldehyde lyase from Pseudomonas fluorescens Pf BAL activates aldehyde substrates to form catalytically active enzymatic Breslow intermediates. Visible-light-driven photoredox catalysis enables the concurrent formation of prochiral C-centered radicals and enzyme-bound thiamine-derived ketyl radicals. Through protein engineering, the evolved active site of Pf BAL variant precisely orchestrates the enantioselective cross-coupling of the two radical intermediates. Overall, this combination of chemomimetic and biomimetic strategies repurposes ThDP-dependent Pf BAL into a radical acyltransferase (RAT), expanding the reactivity repertoire of enzymes and overcoming the challenges in stereochemical control of free prochiral C-centered radicals. We first established this synergistic photoredox/thiamine radical catalysis through a decarboxylative radical acylation, in which prochiral C-centered radicals generated from N -(acyloxy)phthalimides under Eosin Y-mediated photoredox conditions were selectively acylated by Pf BAL variants with benzyaldehyde as the acyl donor. Mechanistic studies confirmed the involvement of both ketyl and prochiral radical intermediates. Building on this discovery, we subsequently demonstrated the robustness of thiamine-dependent radical biocatalysis in diverse challenging transformations, including benzylic C( sp 3 )–H bonds acylation, C–C bonds acylation via nitrogen-centered radicals, three-component radical cross-coupling, and the construction of all-carbon quaternary stereocenter bearing minimally differentiated alkyl substituents. Across all these systems, a shared key mechanistic scheme is the cross-coupling of the enzymatic thiamine-derived ketyl radical with a prochiral C-centered radical, which could be assembled as the radical rebound pathway widely observed in P450 enzymes. Protein engineering campaigns and computational studies consistently suggest that residues 480 and 481 create differential steric environments, playing key roles in achieving high enantioselectivity. This Account describes our journey in repurposing Pf BAL from a natural lyase to a versatile radical acyltransferase. The synergistic integration of photoredox catalysis, directed evolution, and enzymology establishes a generalizable platform for controlling free fleeting radical intermediates for enzymatic acylations. Looking forward, deeper mechanistic mimic, artificial intelligence-assisted protein design and engineering promise to further expand the chemical space and applicability accessible to thiamine-dependent enzymes.