Prospecting hydrogen production of Escherichia coli by metabolic network modeling

Genome-scale model was applied to analyze the anaerobic metabolism of Escherichia coli. Three different methods were used to find deletions affecting fermentative hydrogen production: flux balance analysis (FBA), algorithm for blocking competing pathways (ABCP), and manual selection. Based on these methods, 81 E. coli mutants possessing one gene deletion were selected and cultivated in batch experiments. Experimental results of H2 and biomass production were compared against the results of FBA. Several gene deletions enhancing H2 production were found. Correctness of gene essentiality predictions of FBA for the selected genes was 78% and 77% in glucose and galactose media, respectively. 33% of the mutations that were predicted by FBA to increase H2 production had a positive effect in experiments. Batch cultivation is a simple and straightforward experimental way to screen improvements in H2 production. However, the ability of FBA to predict the H2 production rate cannot be evaluated by batch experiments. Metabolic network models provide a method for gaining broader understanding of the complicated metabolic system of a cell and can aid in prospecting suitable gene deletions for enhancing H2 production. •Mutations enhancing hydrogen (H2) production were sought using metabolic models.•81 mutations were tested by batch experiment for H2 and biomass production.•Gene essentiality prediction accuracy by FBA in anaerobic batch cultivations was 77%.•Comparison of H2 production rate results to batch experiments was challenged.•Several mutations increasing the H2 production were found.