Metabolic fluxes are useful for understanding cellular physiology and identifying ways to alter cell behavior. Metabolic fluxes are determined by incorporating experimental measurements into computational models and calculating the flux distributions which best match experimental observations. These experimental measurements include the (1) nutrient uptake and by-product secretion rates by cells, (2) biomass composition measurements, and (3) GC-MS or NMR measurements of ¹³C labeled metabolites labeled arising from the processing of ¹³C labeled substrates by the metabolic network. In order to quantify metabolic fluxes, experimental data needs to be analyzed by computational models. Metabolic flux analysis uses data on the relative abundance of ¹³C labeled isotopomers (ie. metabolites which differ in the location of ¹³C atoms) to calculate intracellular flux distributions (for review see [1]). Most models only include the central metabolic pathways for an organism, with some recent exceptions [2-4].

We investigated how sensitive flux distributions, calculated by isotopomer models, are to the inclusion of biosynthesis pathways as well as energy and redox balance constraints. We generated a biosynthetic isotopomer model for Escherichia coli, including central and biosynthetic metabolic pathways. Comparisons between model calculations show that inclusion of biosynthetic pathways leads to differences in global optimal solutions, as well as expanded confidence intervals compared to a central model.

We also investigated how balancing energy and redox carrier production and consumption rates, such as ATP, NAD(P) and NAD(P)H affects the calculated flux distributions. We found that incorporating these energy and redox balance constraints did not significantly affect how well the flux distributions fit the experimental data; however, these additional balances eliminate thermodynamically infeasible flux distributions and reduce confidence intervals for central metabolic fluxes. We also evaluated how well different types of data (metabolite uptake & secretion rates and GC-MS analysis of amino acids) eliminate flux distributions, and found that while often not measured in these types of experiments, oxygen uptake rates can significantly restrict flux values through the central metabolic pathways.

References

1. Wiechert W. 2001. ¹³C Metabolic Flux Analysis. Metabolic Engineering. 3: 195-206.

2. Vo, TD; Lim, SK; Lee, WNP; Palsson, BO. 2006. Isotopomer analysis of cellular metabolism in tissue culture: A comparative study between the pathway and network-based methods. Metabolomics. 2(4): 243-256.

3. Vo, TD; Palsson, BO. 2006. Isotopomer analysis of myocardial substrate metabolism: A systems biology approach. Biotechnology and Bioengineering. 95(5): 972-983.

4. Suthers, PF; Burgard, AP; Dasika, MS; Nowroozi, F; Van Dien, S; Keasling, JD; Maranas, CD. 2007. Metabolic flux elucidation for large-scale models using C-13 labeled isotopes. Metabolic Engineering. 9(5-6): 387-405.