Reconstruction And Analysis Of Bacillus Megaterium WSH-002Genome-scale Metabolic Model

Bacillus megaterium plays important roles in industry production, environmentprotection，and a great of experimental results of genome sequencing, sporulation mechanismsand enzyme functions were obtained. However, there has been no systematic research on thephysiology in the view of the whole cell. Therefore, in this paper the genome of B.megaterium WSH-002was annotated based on comparative genomics, and B. megateriumspecific biochemical information was also acquired by literature mining. According togene-protein (enzyme)-reacion association, the information above was processed intometabolic model. The model was then transformed into mathematical model and refined byMatlab. We investigated the structure of the model and simulated it. The main results weredescribed as follows:(1) The metabolic models based on SEED annotation, KEGG automatic annotationserver annotation, local amino acid sequence homology alignment, literature mining wereintegrated, and then some biochemical reactions were added or deleted to generate the refinedmetabolic. The results showed that the model had992genes,591enzymes,883metabolitesand1171reactions (excluding exchange reactions). Biochemical reactions were divided into9categories, and reactions in amino acid metabolism were270, and20protein amino acidswere all synthesized.94%of reactions in the model had associated genes, and reactionswithout associated genes mainly transported compounds. The genes specific to eachmetabolic category account for86%of all reactions in the model, and22%genes were fortransport reactions.(2) Simulations were operated according to certain culture conditions, and the modelingresults were compared with experimental results of literatures. With4different substrateuptake rate (mmol/g cell/h), simulations were operated on the medium with glucose as thecarbon source, and in silico specific growth rates were0.105,0.107,0.109and0.378respectively, lower than those in vivo by2%,0%,1%,1%. When simulations were operatedon the medium with pyruvate as the carbon source, in silico specific growth rates were0.093and0.093respectively, lower than those in vivo by1%,1%. When Thermobifida fuscahydrolase was induced, the change of glucose into pyruvate as carbon source decreasedYATPby56%, and increased YNADPH/Xby2%, lined with the variation trend in vivo. In thecentral metabolism, the in silico flux of pentose phopsphate pathway was lower than that invivo, and the in silico fluxes of TCA cycle and glycolysis were higher than those in vivo.Essential genes of B. megaterium WSH-002identified on LB medium were47, and30geneswere essential to Bacillus subtilis, and100%,100%,50%,50%,50%of essential genes ofrespiratory pathway, cell wall biosynthesis, central metabolism, lipid metabolism andnucleotides metabolism were also essential to B. subtilis.(3) According to minimal medium with glucose as carbon source, conditions for simulation were set, and then the simulations were run to indentify compounds that weresynthesized and transported out of the cell. The results showed that there were58transportreactions for these compounds, and transport reactions for amino acids, nucleic acid andorganic acids were most. The compounds identified to be synthesized and transported out ofthe cell included guanine, xanthine, hydroxanthine, nicotinic acid and8amino acids, whichwere reported as the contributor to the growth of Ketogulonicigenium vulgaer and2-KLGproduction. And the model also indicated that B. megaterium might secrete Gly-Cys, Gly-Try,Gly-Leu and pantothenic acid to promote the growth of K. vulgaer and2-KLG production.