Rapamycin, as a macrocyclic polyketide with immunosuppressive, antifungal, and anti-tumor activity produced by Streptomyces hygroscopicus, is receiving considerable attention for its significant contribution in medical field. However, the production capacity of the wild strain is very low. Hereby, a computational guided engineering approach was proposed to improve the capability of rapamycin production. First, a genome-scale metabolic model of Streptomyces hygroscopicus ATCC 29253 was constructed based on its annotated genome and biochemical information. The model consists of 1003 reactions, 711 metabolites after manual refinement. Subsequently, several potential genetic targets that likely guaranteed an improved yield of rapamycin were identified by flux balance analysis and minimization of metabolic adjustment algorithm. Furthermore, according to the results of model prediction, target gene pfk (encoding 6-phosphofructokinase) was knocked out, and target genes dahP (encoding 3-deoxy-D-arabino-heptulosonate-7-phosphate synthase) and rapK (encoding chorismatase) were overexpressed in the parent strain ATCC 29253. The yield of rapamycin increased by 30.8% by knocking out gene pfk and increased by 36.2 and 44.8% by overexpression of rapK and dahP, respectively, compared with parent strain. Finally, the combined effect of the genetic modifications was evaluated. The titer of rapamycin reached 250.8 mg/l by knockout of pfk and co-expression of genes dahP and rapK, corresponding to a 142.3% increase relative to that of the parent strain. The relationship between model prediction and experimental results demonstrates the validity and rationality of this approach for target identification and rapamycin production improvement.
Keywords: Genome-scale metabolic model; Metabolic engineering; Rapamycin; Streptomyces hygroscopicus ATCC 29253; Target prediction.