In Silico Aided A Systemic Study Of Metabolic Engineering For The Strain Improvement Of Escherichia Coli

In this post-genome era, the abundant "omic" data has made the original reductionism give way to systems biology in the domain of life science. How to combine omics and systemic research ideas to conduct stain improvement is the current focus of metabolic engineering. In this work, we used FBA (Flux Balance Analysis) to study the global and central metabolism of Escherichia coli for in silico strain improvement based on its metabolic pathways, and gene-protein-reaction data.We studied the pathways and genes systemically to identify those that have key effects to increase the yield of succinic acid of E. coli under anaerobic growth conditions. Besides to the common technique used to simulate the metabolic phenotypes of pathway deficiency and gene knockout mutants, we adopted an original method of enlarging the fluxes from wild type strain to simulate the phenotypes of strains which express the corresponding genes. From the simulations based on pathways, we found pathways (ADHEr, PFL, PPC, etc.) which have been validated by publications that have key effects in increasing the yield of succinic acid. From the simulations based on genetic engineering, we found that over-expressing fumarate reductase coded by frdABCD can increase the yield of succinic acid, which has also been validated by publications. Our approach is more effective and comprehensive than the traditional trial-and-error method in guiding strain improvement. Especially, our method can be easily applied to the organisms whose genome information or metabolic network is clear.We also studied the problem of interpreting the fluxes distribution of large-scale metabolic network from the perspective of elementary flux modes (EFMs). We first evaluated the possibility of using the subset of EFMs to reconstruct the fluxes distribution by using the relative error between the reconstructed flux vector and the original one as a sign. In the subsequent analysis, we found that some EFMs have higher frequencies than the other to acquire non-zero weights during the reconstruction process. We believed that these modes can help us better understand the metabolisms of our strains.