Directed evolution of an integral membrane monooxygenase unlocks its full potential

Chemical synthesis of the API intermediate 2,6-(bis)hydroxymethylpyridine from the natural product lutidine suffers from low atomic efficiency and large consumption of toxic reagents (e.g., potassium permanganate) and solvents. A more sustainable route proceeding at high atom efficiencies and not requiring fossil carbon input during production is catalyzed by recombinant E. coli strains expressing the integral membrane two-component xylene monooxygenase (XylMA) from Pseudomonas putida. However, the XylMA system only directly monohydroxylates lutidine. Oxidation to the diol is only possible after overoxidation of the monohydroylation-product, 6-methyl-2-pyridinemethanol, to 6-(hydroxymethyl)pyridine-2-carbaldehyde and involves a complex four-step intracellular reaction cascade prone to unproductive side reactions leading to the accumulation of biocatalyst-deactivating aldehydic intermediates and dead-end products. In order to achieve direct dihydroxylation, we screened a library of 50,000 XylMA mutants using a customized high-throughput mass spectrometry protocol and identified M116 as key residue in determining XylM substrate specificity. Substitution with M116 with glycine indeed allowed XylMA to directly dihydroxylate lutidine to first 6-methyl-2-pyridinemethanol then to 2,6-(bis)hydroxymethylpyridine. Expression of the M116G XylMA mutant hence enabled E. coli to catalyze the reaction in two instead of four steps thereby reducing metabolic burden and minimizing the escape of dead-end metabolites and intermediates. Scale up the reaction to 450 L using XylMM116GA was straightforward allowing for product concentrations of 17 g L^-1 at a space-time yield of 1.45 g L^-1h^-1. Enzyme engineering proved to considerably increase process efficiency and corroborates the value of recombinant transmembrane hydroxylases for industrial manufacturing by sustainable, highly atom-efficient biocatalysis.