Modulating the expression of membrane-bound transhydrogenase is the predominant evolutionary strategy to restore redox cofactor balance

NAD(H) and NADP(H) are essential redox cofactors conserved across living systems. Their production and consumption have to be balanced to prevent the retardation of metabolic flux, the perturbation of cell signaling, and causing oxidative damages of DNA and proteins. Disturbance of this homeostasis, however, becomes inevitable when organisms experience oxidative stress, switch growth substrates, or evolve novel biochemical pathways that alter the stoichiometry of NAD(H)/NADP(H). Therefore, mechanisms that buffer redox imbalance are not only crucial to physiological robustness but may promote the evolvability of metabolic systems by smoothing evolutionary transitions. Several redox buffering strategies have been proposed to maintain physiological robustness, including the regulation of isoezymes with different NAD(P) cofactor preferences, rerouting metabolic flux through pathways that recruit distinct redox cofactors, and the activation of transhydrogenases (Sth, PntAB) which directly shuttle hydride between NAD and NADP. Among these I hypothesized the PntAB transhydrogenase having the greatest importance during evolution because (1) it has just one dedicated function; (2) it is shown to mitigate redox imbalance under genetic or environmental perturbations; (3) it is broadly distributed across three domains of life. I tested this hypothesis by laboratory evolution of an Escherichia coli mutant (?zwf edd eda) that was lacking NADPH production during growth on glucose due to the disruption of the oxidative pentose phosphate (PP) pathway. After 1000 generations of propagation in glucose minimal media, twelve replicate populations founded by this Zwf ancestor showed significant improvement in glucose growth rates but exhibited intra- and inter-population variation larger than those founded by wild-type E. coli and evolved in identical conditions. Relative to the Zwf ancestor, phenotypic characterization of the fast-growing isolates (FG, emerged in all 12 populations) and the slow-growing isolates (SG, identified only in 3 populations) revealed a 2-3 fold increased PntAB activity in all FG isolates but no significant change in SG isolates. The parallel phenotype of FG isolates stresses the evolutionary importance of PntAB; yet the emergence of SG isolates and their coexistence with FG isolates in 3 populations suggest the existence of alternative mechanisms to enhance NADPH production. Genome re-sequencing of 5 FG and 2 SG isolates revealed distinct genetic bases of adaptation to NADPH shortage. FG isolates increased their PntAB expression through acquisition of either cis-mutations in the pntAB operon or trans-mutations in the adenylate cyclase gene cyaA. The latter transcriptionally up-regulated pntAB expression through the Crp-cAMP regulon. On the other hand, 2 SG isolates acquired null mutations in a glucose transporter gene ptsG gene, which enabled them to fuel the NADPH pool through acetate/glucose co-utilization. By feeding on acetate secreted by FG isolates coexisting in the same population, SG isolates evolved a novel and unexpected mechanism to generate more NADPH through the citric acid cycle. This study represents the first experimental attempt to address the mechanistic link between robustness and evolvability in metabolic systems. Results here yield molecular insights into the complex redox regulation, confirm the predominant role of the PntAB transhydrogenase in assisting the evolution of metabolic systems, and discover a functional innovation that highlights the evolutionary flexibility of organisms to switch food preference and resolve physiological problems through exploiting ecological opportunities.