Integration of Quantitative Endometabolomics with Genome-Scale Models to Bypass the Steady-State Assumption

Constraint-based modeling has become a powerful tool in various biological and biomedical fields for understanding cellular metabolism at the systems scale. To date, dynamic simulations of constraint-based models have focused on extracellular metabolic concentrations. However, there are large intracellular metabolite pools in cellular systems, suggesting that changes in those pools can have a significant effect on metabolic flux. Thus, the main assumption and the major limitation of constraint-based modeling is that intracellular metabolites are assumed to be at steady-state. In recent years, the availability and use of quantitative metabolomics has increased, allowing for integration of changes in both extracellular and intracellular metabolite concentrations. In this study, we present a constraint-based modeling workflow and methodology for simulating cellular systems explicitly integrating changes in intracellular and extracellular metabolite profiles, termed unsteady flux balance analysis (uFBA). The uFBA approach is comprised of 1) discretizing metabolite time profiles into metabolic states using principal component analysis, 2) integration of the rate of change of metabolites in that state with the constraint-based model, and 3) reconciling changes in both measured and un-measured metabolites to produce a feasible flux. uFBA is applied to three test cases where such data was available: human red blood cells and platelets stored under blood banking conditions, and Escherichia coli growth in batch culture. Metabolic flux distributions derived from uFBA are substantially different than FBA flux distributions. Globally, the fluxes of a large portion of intracellular reactions were significantly different between the two approaches and nearly all sampled FBA flux vectors had reordering of flux values as compared to sampled uFBA flux vectors. Further, we experimentally validated one particular discrepancy between uFBA and FBA predictions concerning citrate usage in human red blood cells. FBA results indicate that citrate is consumed by red blood cells and utilized to produce malate and fumarate, while uFBA suggest that the production of malate and fumarate can be fully explained by changes in intracellular malate. C13 labeled citrate experiments did not result in an accumulation of labeled malate and fumarate, suggesting that uFBA results were more indicative of actual intracellular fluxes.