Reconciling Barseq Datasets for Z. Mobilis Using a Genome-Scale Metabolic Model

Zymomonas mobilis is an industrially relevant, gram negative, ethanologen known for very high glycolytic fluxes, high ethanol production, and low biomass yields. Development of engineered strains of Z. mobilis has been relatively slow, and future efforts stand to benefit from being informed and directed by updated genome-scale metabolic models. Here we present iZM4_472, an updated model of Z. mobilis ZM4, and use it to analyze and reconcile BarSeq datasets from pooled mutant fitness experiments [Deutschbauer et al. Journal of bacteriology (2014)]. iZM4_472 contains 757 metabolic and transport reactions (of which 611 have GPR associations), 472 genes, and 632 unique metabolites, making it the largest and best annotated model of Z. mobilis to date. Data from reported pooled mutant experiments was used to assess the accuracy of gene essentiality predictions and identify genes associated with gap-filled reactions in the model. Several discrepancies between gene essentiality predictions and experimental results were caused by gene duplication, polar effects, or mis-mapped barcodes in the mutant library; thus, highlighting potential challenges with applying these high-throughput datasets to improve metabolic models. Correlations between transposon mutant fitness across 492 experiments were used to identify gene candidates for gap-filled reactions involved in histidine biosynthesis. Additional genes for reactions involved in tyrosine, biotin, ubiquinone, and pyridoxal 5’-phosphate synthesis were identified and confirmed through a combination of E. coli mutant complementation experiments, and our previously developed model-enabled gene search (MEGS) approach [Pan et al. Journal of Biological Chemistry (2017)]. Finally, flux coupling analysis of iZM4_472 was used to analyze the fitness scores of all 492 experiments in the reported BarSeq datasets to identify metabolic modules where mutant phenotypes were poorly correlated. These modeling efforts help pinpoint where additional metabolic knowledge gaps of Z. mobilis metabolism reside.