Flavin-dependent halogenases catalyze enantioselective olefin halocyclization

Abstract

Halocyclization of alkenes is a powerful bond-forming tool in synthetic organic chemistry and a key step in natural product biosynthesis, but catalyzing halocyclization with high enantioselectivity remains a challenging task. Identifying suitable enzymes that catalyze enantioselective halocyclization of simple olefins would therefore have significant synthetic value. Flavin-dependent halogenases (FDHs) catalyze halogenation of arene and enol(ate) substrates. Herein, we reveal that FDHs engineered to catalyze site-selective aromatic halogenation also catalyze non-native bromolactonization of olefins with high enantioselectivity and near-native catalytic proficiency. Highly selective halocyclization is achieved by characterizing and mitigating the release of HOBr from the FDH active site using a combination of reaction optimization and protein engineering. The structural origins of improvements imparted by mutations responsible for the emergence of halocyclase activity are discussed. This expansion of FDH catalytic activity presages the development of a wide range of biocatalytic halogenation reactions.

Fig. 1. Expanding the scope of FDH catalysis.

a Arene/enol (R = OH) halogenation and b a simplified scheme for olefin halocyclization.

Fig. 2. Optimization of FDH-catalyzed bromolactonization.

a Selected yield and enantioselectivity data for bromolactonization of substrate 1. b Optimization of halocyclization reaction conditions. Reaction mixtures contained 1 mM substrate, 5 equiv. NaBr, 5 mol% FDH, and a cofactor regeneration system comprising a flavin reductase, a glucose dehydrogenase, and glucose. Glutathione (1 mM), catalase, and optimized buffers were used as described in the supporting information. Product assay yields (hereafter, “yields”) and selectivities are the average of triplicate measurements determined by LC/MS using p-bromoanisole internal standard. ^aReaction conducted on 15 mg scale.

Fig. 3. Halolactonization substrate scope of evolved RebH variants.

Reaction mixtures contained 1 mM substrate, 5 equiv. NaBr or 100 equiv. NaCl, 5 mol% FDH, and a cofactor regeneration system comprising a flavin reductase, a glucose dehydrogenase, and glucose. Glutathione (1 mM), catalase, and optimized buffers were used as described in the supporting information. Product yields and selectivities are the average of triplicate measurements determined by LC/MS using the internal standards indicated in the SI.

Fig. 4. Structural origins of improved halocyclase activity in evolved FDHs.

a Effects of mutations in evolved FDHs on halocyclase activity. Mutations are listed relative to the parent in the previous row. Yields and selectivities are the average of duplicate measurements determined by LC/MS using p-bromoanisole internal standard. b Docking poses for the cationic intermediate generated upon bromination of 1 in the structure of RebH variant 3-LSR (left) and 3-LSR F111S (right). Key active site residues, including K79 and F/S111 are shown in yellow and a surface rendering of several residues is provided to illustrate the space created by F111S. The K79_ε-amino-Br and C_benzyl-O_carboxylate distances are shown. ^aReaction conducted without added glutathione.