Transomics data-driven, ensemble kinetic modeling for system-level understanding and engineering of the cyanobacteria central metabolism

In silico kinetic modeling is an essential tool for rationally designing metabolically engineered organisms based on a system-level understanding of their regulatory mechanisms. However, an estimation of enzyme parameters has been a bottleneck in the computer simulation of metabolic dynamics. In this study, the ensemble-modeling approach was integrated with the transomics data to construct kinetic models. Kinetic metabolic models of a photosynthetic bacterium, Synechocystis sp. PCC 6803, were constructed to identify engineering targets for improving ethanol production based on an understanding of metabolic regulatory systems. A kinetic model ensemble was constructed by randomly sampling parameters, and the best 100 models were selected by comparing predicted metabolic state with a measured dataset, including metabolic flux, metabolite concentrations, and protein abundance data. Metabolic control analysis using the model ensemble revealed that a large pool size of 3-phosphoglycerate could be a metabolic buffer responsible for the stability of the Calvin-Benson cycle, and also identified that phosphoglycerate kinase (PGK) is a promising engineering target to improve a pyruvate supply such as for ethanol production. Overexpression of PGK in the metabolically engineered PCC 6803 strain showed that the specific ethanol production rate and ethanol titers at 48 h were 1.23- and 1.37-fold greater than that of the control strain. PGK is useful for future metabolic engineering since pyruvate is a common precursor for the biosynthesis of various chemicals.