Engineering, Predicting, and Understanding Cofactor Specificity in Ketol-Acid Reductoisomerases

The branched-chain amino acid biosynthesis pathway has attracted considerable interest for the production of these valuable nutrients, and also as a source of potential biofuels.  However, the NADPH-dependence of the enzyme which catalyzes the second step, ketol-acid reductoisomerase (KARI), has been a barrier to the large-scale fermentative implementation of this pathway.  After successfully engineering NADH-preference in E. coli KARI by directed evolution, we developed a general recipe for engineering it in any KARI, and demonstrated it on a representative set.  Subsequently, we used the insight gained from these engineering efforts to analyze thousands of experimentally uncharacterized KARI genes and predict ones that could utilize NADH.  From these, we expressed and characterized four KARIs that preferred NADH and four that were bispecific for the two cofactors, the first such enzymes in a class previously believed to be exclusively NADPH-dependent.  Crystal structures of two of these enzymes show insertions into the specificity-determining β2αB-loop, giving insight into the molecular evolution of cofactor specificity.