UC Irvine

A key challenge in metabolic engineering is the intracellular control of reductive and free energy. Native organic cofactors NAD(P)H and ATP provide this energy and are generally required in all known bioproduction platforms. While microbes have evolved tightly regulated metabolisms to optimize growth, the need for these highly specific cellular resources complicates the transition to biological manufacturing. High cost and rigid specificity limit the applications of biotechnology and the usability of established biocatalysts. In this work we developed a series of high-throughput (>106), growth-based selection platforms for the rapid customization of important cofactor-dependent biocatalysts for both native cofactors NAD(P)H and for alternative cofactors nicotinamide mononucleotide (NMNH) and pyrophosphate (PPi). Each platform consists of an E. coli strain engineered to obligately link cofactor utilization to growth in defined selection medias. These platforms can serve to expedite protein engineering efforts and provide flexibility for the advancement of industrial biology-based chemical production.The first selection strain was engineered for the aerobic selection of NADPH-dependent enzymes. Thus, it was used in the engineering of two different NADPH-dependent oxygenases, 4-hydroxybenzoate hydroxylase (PobA) and cyclohexanone monooxygenase (CHMO), for altered substrate preference and improved thermostability, respectively. The subsequent selection strain was engineered for the aerobic selection of NADH-dependent enzymes. It was applied in the directed evolution of the strictly NADPH-dependent CHMO to utilize NADH. The next selection characterized addressed a key challenge in engineering oxygen-dependent enzymes, namely their reaction mechanisms are often complex, tightly controlled, and prone to uncoupling. To investigate whether these growth selections were sensitive to the reactive oxygen species formed as byproducts of uncoupling, we used the NADPH-dependent selection strain to improve a highly active, but highly uncoupled P450 enzyme, BM3 mutant GVQ (A74G-F87V-L188Q). While the application of wildtype cofactors may remain essential, their usage ensures a continued interference between artificial pathways and native metabolisms. Introduction of noncanonical cofactors that limit these interactions can provide advanced control of intracellular energy allocation. To engineer preference for these alternative coenzymes, the growth-based platforms were expanded to select for activity with a non-native nicotinamide cofactor, nicotinamide mononucleotide (NMNH) and an atypical free energy source, pyrophosphate (PPi).