In silico aided metabolic engineering of Streptomyces roseosporus for daptomycin yield improvement

In silico metabolic network models are valuable tools for strain improvement with desired properties. In this work, based on the comparisons of each pathway flux under two different objective functions for the reconstructed metabolic network of Streptomyces roseosporus, three potential targets of zwf2 (code for glucose-6-phosphate hydrogenase), dptI (code for alpha-ketoglutarate methyltransferase), and dptJ (code for tryptophan oxygenase) were identified and selected for the genetic modifications. Overexpression of zwf2, dptI, and dptJ genes increased the daptomycin concentration up to 473.2, 452.5, and 489.1 mg/L, respectively. Furthermore, co-overexpression of three genes in series resulted in a 34.4% higher daptomycin concentration compared with the parental strain, which ascribed to the synergistic effect of the enzymes responsible for daptomycin biosynthesis. Finally, the engineered strain enhanced the yield of daptomycin up to 581.5 mg/L in the fed-batch culture, which was approximately 43.2% higher than that of the parental strain. These results demonstrated that the metabolic network based on in silico prediction would be accurate, reasonable, and practical for target gene identification and strain improvement.