Pathway improvement for ethanol production by thermophilic bacteria involving ferredoxin engineering: heterologous expression and in vitro evaluation

Abstract

It is widely recognized that understanding and manipulating redox metabolism is critical for metabolic engineering aimed at achieving high titers and yields, and indeed often more important than manipulation of carbon metabolism. Ferredoxin is a central redox protein in ethanol metabolism in C. thermocellum, where it functions as the electron acceptor for PFOR, as well as the electron donor for the reduction of NADH (through RNF), which is subsequently used in the reduction of acetyl-CoA to ethanol. The electronic and structural properties of ferredoxin influence the metabolic flux network in two ways: 1) the rate of electron transfer in ferredoxin-dependent reactions is partly governed by the geometry and specificity of the transient protein-protein interactions (PPIs) formed between ferredoxin and its partner enzymes; 2) the midpoint potential (E0) of the ferredoxin plays a role in establishing the thermodynamic driving force (G) for a given reaction, which in turn affects the amount of enzyme required to sustain a desired net forward flux. Despite this central role, ferredoxin is likely the least understood electron carrier in C. thermocellum. Moreover, being a protein, it is also distinctive among electron carriers in that it is amenable to protein engineering, allowing us to modify its structural and electronic properties to optimize electron flow toward ethanol. It is proposed a combined biochemical and protein engineering approach aimed at understanding and manipulating ferredoxin to improve ethanol production. In the biochemical approach, we will purify and characterize ferredoxin from both C. thermocellum and T. saccharolyticum. Midpoint potentials will be determined through cyclic voltammetry performed on immobilized ferredoxin in an electrochemical cell. These results, when combined with results from metabolomic experiments, will allow us to determine the ”G for the relevant electron transfer reactions, in order to quantitatively predict modifications of ferredoxin midpoint potential that could improve ethanol flux. The second approach, is to leverage protein engineering to generate two libraries of mutant ferredoxins: one with altered E0 by mutating amino acids in the conserved iron-sulfur cluster-binding domains, and one with altered PPIs by mutating amino acids on the surface and assess their effect on ethanol yield. Since it is not yet possible to predict these properties from primary sequence, we will initially develop a modest library in E. coli based on previous literature, and characterize midpoint potential as described above. A small subset of these mutants with E0's spanning the range identified from the predictions in the first approach will then be transformed into C. thermocellum, and fermentation results analyzed. For the library with altered PPIs, candidates will be initially tested in in vitro assays with PFOR and RNF to identify a subset of mutants that improve activity. These will be transformed into C. thermocellum to assess the impact on ethanol production. The benefit of this approach is that it is based on modifying a single small protein to improve the efficiency of electron flow to ethanol, rather than overexpressing large metabolic enzymes, which can impose an additional metabolic burden on the cell. (AU)