Reconstruction And Application Of Saccharopolyspora Erythraea Genome-Scale Metabolic Model Based On Multiomics Data

Saccharopolyspora erythraea(S.erythraea)is the main industrial strain producing macrolide antibiotic erythromycin.The construction of a high-quality Genome-Scale Metabolic Model(GSMM)is of significance for the in-depth understanding of the metabolic mechanism of S.erythraea,thus enabling efficient research on the construction of new strains,the design of synthetic media and the optimization of fermentation processes.The currently published S.erythraea GSMM suffers from incomplete structural descriptions and inadequate integration of multi-omics data,resulting in limited performance and limiting the scope of application.In this paper,we reconstructed the S.erythraea GSMM based on the latest genomic,transcriptomic and metabolomic data of erythromycin industrial high-yielding strain E3,examined its performance comprehensively and applied it to guide the optimization of the n-propanol fed-batch process.The latest genome annotation information of E3 was used to reconstruct and upgrade the GSMM,filling the vacant metabolic pathway reactions in the original model iZZ1342,adding pathways such as the gluconeogenesis and the n-propanol utilization.The original model was manually refined using the latest transcriptomic and metabolomic data to obtain the iJL1426 model of S.erythraea strain E3.The reconstructed iJL1426 model contains 1426 annotated genes,1687 metabolites,and 1858 metabolic reactions.Compared with the previous version model iZZ1342,its number has increased by 10.46%,2.8%and 10.33%respectively.The prediction performance of iJL1426 was inspected using experimental data.The results of the robustness analysis showed that the uptake rates of glucose and oxygen were linearly related to the specific growth rate of the bacteria.The simulation results of carbon and nitrogen source availability are in good agreement with literature reports.Compared with iZZ1342,the accuracy of iJL1426 increased by 14.80%and 12.00%respectively.The validation of the fermentation parameters showed that the model simulations showed the same variability as the experimental data(e.g.specific growth rate of the bacterium(?)and specific synthesis rate of erythromycin(qEry)).The results of 13C metabolic flow comparison showed that the correlation coefficient between simulated and experimental values reached to 0.915.Single knockout simulations showed that 96 genes were essential,mainly distributed in metabolic pathways such as amino acid,purine,pyrimidine,pantothenate and CoA biosynthesis.The double knockout simulations showed that the lethal gene pairs were primarily distributed in the metabolic pathways of amino acids,purines,pyrimidines,ubiquinone and terpenoid backbone biosynthesis.Among the genes that enhance the qEry,the model predictions are relatively consistent with those reported in the literature.Based on the simulation results of iJL1426,the n-propanol fed-batch process was optimized.The qEry was positively related to the uptake rate of n-propanol(qpro)at the initial stage and inversely related to the qpro at a certain level.Based on this finding,four n-propanol fed-batch patterns such as 0,0.025,0.050 and 0.075 g/L/h were designed.The experimental results demonstrated that when the qpro was 0.05 g/L/h,the erythromycin potency was 1442.84?g/mL,which was 45.02%higher than the original culture process,and the qEry was also increased by 30.25%.The above results showed that the iJL1426 model has positive predictive performance,which lays the foundation for an intensive grasp of the metabolic mechanism of S.erythraea.