In vivo promiscuity and directed evolution of oligomeric beta-decarboxylating dehydrogenases

At the molecular level, natural evolution gave rise a wide variety of proteins endowed with exquisite properties. In this study, we envisioned to contribute a better understanding of the evolutionary mechanisms and structure function relationships of oligomeric enzymes. Beta-decarboxylating dehydrogenases were chosen as a model superfamily. Oligomerization is an advantage for complexifying the properties of a protein such as affording allosterism; however, the impact of oligomerization on the evolutionary mechanisms of proteins is poorly documented. This study will explore the potential of β-decarboxylating dehydrogenases as a platform for directed evolution of oligomeric enzymes using a strategy of activity interconversion between the family members. We first studied the promiscuous activities of D-malate and 3-isopropylmalate dehydrogenases (DmlA and IPMDH) in vivo and showed reciprocal complementation of each enzyme in Escherichia coli strains where the other enzyme is absent. By simulating gene duplication, we demonstrated that heteromeric complex of DmlA is formed in vivo from the interaction of the subunits encoded from the two copies of the gene. Following a directed evolution approach, we successfully generated a NAD dependent E. coli isocitrate dehydrogenase (IDH) from DmlA. We found that a single amino acid substitution (Leu89Ser) endows DmlA with about 60000-fold improvement in IDH activity although the IDH phenotype in vivo is thermosensitive. Co-expression of the wild type DmlA enzyme with the mutant did not result in an increased thermostability indicating that, in this specific case of molecular evolution, there is no paralogous chaperoning from the parental protein in the oligomeric assembly.