Efficient transferase engineering for SAM analog synthesis using combinatorial library design and high-throughput...

Alkylation reactions are of great importance in chemical synthesis and in the pharmaceutical industry to modulate the function of bioactive compounds. As conventional alkylation strategies often suffer from limited chemo- and regioselectivity, highly selective biocatalytic alkylation reactions using S-adenosyl-ʟ-methionine (SAM)-dependent methyltransferases (MTs) and SAM analogs represent a promising avenue to overcome these challenges. Recent studies have shown that anion MTs can synthesize SAM from S-adenosyl-ʟ-homocysteine (SAH) and readily available methyl donors such as iodomethane and methyl tosylate. The optimization of these transferases through classical directed evolution has successfully extended this approach to the (re)generation of SAM analogs from "off-the-shelf" alkylating agents, though this has so far been largely limited to iodoethane. Here, we present an efficient engineering approach to obtain highly active enzyme variants toward diverse haloalkanes. A combinatorial mutant library targeting multiple active site residues in parallel was designed solely based on sequence information as well as modeled protein structure and dynamics. While leveraging the natural sequence diversity at the targeted residues, the mutant library was generated using synthetic DNA fragments. A high-throughput LC/MS-based screening allowed the efficient identification of enzyme variants exhibiting orders-of-magnitude higher activity in the formation of several SAM analogs compared to the wild-type enzyme. We believe that the protocol described herein holds promise for engineering MTs as well as other classes of enzymes.