Applications Of Genome-scale Metabolic Modeling Technologies In Industrial Biotechnology

Genome-scale metabolic models(GEMs)have been playing an important role in systems biology,and have been applied in several fields,such as metabolic engineering,screening of drug targets,metagenomic research etc.In this study,GEMs were applied in industrial biotechnology,including understanding of the phenotypic characteristics under genetic or environmental perturbations,strain design based on in silico simulation and prediction strategies.The following four parts were particularly investigated in this study.Firstly,the relationship between genotype and phenotype of microorganism cells were studied on the basis of in silico simulations with GEMs.Here,the cell growth and succinate production characteristics of the wild-type Escherichia coli and its gene-deficient strains were investigated.The core GEMs of the wild-type E.coli,the ldhA-deficient strain,the pflB-deficient strain,and the ldhA-pflB double deficient strain(i.e.E.coli NZN111)were reconstructed respectively.The non-optimized method Artificial Centering Hit-and-Run(ACHR)was therefore employed to simulate and randomly sample the solution spaces of the four models,respectively.The simulation results showed that the NADH/NAD+ imbalance in E.coli NZN111 under anaerobic conditions probably resulted mainly from the inhibition of pyruvate-formate lyase(PFL).In addition,the enzyme catalyzing reaction NADH16,conversion of ubiquinone-8(q8)to ubiquinol-8(q8h2),was considered as a potential target to be overexpressed to obtain improved cell growth and succinate production in anaerobically growing E.coli NZN 111.Secondly,the design of industrially useful strains was investigated by developing proper simulation frameworks for GEMs.The algorithm,analysis of production and growth coupling(APGC),was successfully developed based on the coupled relationship between biosynthetic network and the metabolic network relating to the desired metabolite.The combination of APGC and previously developed Logical transformation of genome-scale metabolic model(LTM)could identify possible gene amplification targets(genes whose overexpression could probably enhance the production performance).APGC was then employed to identify potential gene amplification targets for the overproduction of succinate in E.coli NZN111 under microaerobic conditions.Therein,four target genes were selected to direct the experimental verification.Genetic manipulation and cultivation experiment results showed that the overexpression of all four genes individually could improve both cell growth and succinate production to some extent.Especially the overexpression of newly found genes gapA and pgk exhibited significant increases in succinate production in E.coli NZN 111.The results from this case study indicated that the developed APGC framework is probably an appropriate method to predict the target genes that should be overexpressed to enhance cell performances.In light of metabolic network of industrial strain Yarrowia lipolytica and latest literatures,the previously developed GEM of Y.lipolytica in our lab iYL619_PCP was updated and reconstructed to obtain new model of iYL_v2.0.The simulation results with iYL_v2.0 indicated that this model could be used to predict more accurately the specific growth rate of Y.lipolytica in glucose minimal medium and the growth on various carbon sources.Furthermore,a comprehensive metabolic network map of iYL₂.0 was constructed by using CellDesigner software for direct visualization and comparison of metabolic reaction fluxes obtained under different conditions.Considering the significant differences in biomass contents of Y.lipolytica cells growing under nitrogen limitation and carbon limitation conditions,the biomass reaction in model iYL_v2.0 was corrected to correspond to two different culture conditions and the whole metabolic network model was further refined to achieve a high quality Constraints-Based Model(CBM),iYL_CN.Model iYL_CN was employed to accurately simulate the experimental specific growth rates and citrate production rates of Y.lipolytica growing under nitrogen or carbon limitation conditions.In the meanwhile,the simulated fluxes and the experimental fluxes from 13C-labeling study showed high Pearson's correlation coefficients.Intensive metabolic studies with model iYL_CN were carried out to provide insights into understanding of metabolic characteristics of Y.lipolytica growing under nutrient limitations.To sum up,the applications of genome-scale metabolic models in understanding and engineering industrial microorganisms,especially E.coli NZN111 and Y.lipolytica,were investigated from multiple angles.Both the development of model prediction frameworks and experimental verifications were intensively performed in this study,which might provide useful insights to promote further researches of metabolic network model based applications as well as the innovative development of the industrial biotechnology technologies.