Abstractβ‐elemene is one of the most commonly used antineoplastic drugs in cancer treatment. As a plant‐derived natural chemical, biologically engineering microorganisms to produce germacrene A to be converted to β‐elemene harbors great expectations since chemical synthesis and plant isolation methods come with their production deficiencies. In this study, we report the design of an Escherichia coli cell factory for the de novo production of germacrene A to be converted to β‐elemene from a simple carbon source. A series of systematic approaches of engineering the isoprenoid and central carbon pathways, translational and protein engineering of the sesquiterpene synthase, and exporter engineering yielded high‐efficient β‐elemene production. Specifically, deleting competing pathways in the central carbon pathway ensured the availability of acetyl‐coA, pyruvate, and glyceraldehyde‐3‐phosphate for the isoprenoid pathways. Adopting lycopene color as a high throughput screening method, an optimized NSY305N was obtained via error‐prone polymerase chain reaction mutagenesis. Further overexpression of key pathway enzymes, exporter genes, and translational engineering produced 1161.09 mg/L of β‐elemene in a shake flask. Finally, we detected the highest reported titer of 3.52 g/L of β‐elemene and 2.13 g/L germacrene A produced by an E. coli cell factory in a 4‐L fed‐batch fermentation. The systematic engineering reported here generally applies to microbial production of a broader range of chemicals. This illustrates that rewiring E. coli central metabolism is viable for producing acetyl‐coA‐derived and pyruvate‐derived molecules cost‐effectively.

Co-localizing biochemical processes is a great strategy when expressing the heterologous metabolic pathway for product biosynthesis. The RNA scaffold is a flexible and efficient synthetic compartmentalization method to co-localize the enzymes involved in the metabolic pathway by binding to the specific RNA, binding domains fused with the engineered enzymes. Herein, we designed two artificial RNA scaffold structures�0D RNA scaffolds and 2D RNA scaffolds�using the reported aptamers PP7 and BIV-Tat and the corresponding RNA-binding domains (RBDs). We verified the interaction of the RBD and RNA aptamer in vitro and in vivo. Then, we determined the efficiencies of these RNA scaffolds by co-localizing fluorescent proteins. We employed the RNA scaffolds combined with the enzyme fusion strategies to increase the metabolic flux involved in the enzymes of the mevalonate pathway for mevalonate and isoprene production. Compared with the no RNA scaffold strain, the mevalonate levels of the 0D RNA scaffolds and 2D RNA scaffolds increased by 84.1% (3.13 ± 0.03 g/L) and 76.5% (3.00 ± 0.09 g/L), respectively. We applied the 0D RNA scaffolds for increasing the isoprene production by localizing the enzymes involved in a heterologous multi-enzyme pathway. When applying the RNA scaffolds for co-localizing the enzymes mvaE and mvaS, the isoprene production reached to 609.3 ± 57.9 mg/L, increasing by 142% compared with the no RNA scaffold strain. Our results indicate that the RNA scaffold is a powerful tool for improving the efficiencies of the reaction process in the metabolic pathway.

Saccharomyces cerevisiae has been extensively used as a convenient synthetic biology chassis to reconstitute fungal polyketide biosynthetic pathways. Despite progress in refactoring these pathways for expression and optimization of the yeast production host by metabolic engineering, product yields often remain unsatisfactory. Such problems are especially acute when synthetic biological production is used for bioprospecting via genome mining or when chimeric fungal polyketide synthases (PKSs) are employed to produce novel bioactive compounds. In this work, we demonstrate that empirically balancing the expression levels of the two collaborating PKS subunits that afford benzenediol lactone (BDL)-type fungal polyketides is a facile strategy to improve the product yields. This is accomplished by systematically and independently altering the copy numbers of the two plasmids that express these PKS subunits. We applied this plasmid copy number engineering strategy to two orphan PKSs from genome mining where the yields of the presumed BDL products in S. cerevisiae were far too low for product isolation. This optimization resulted in product yield improvements of up to 10-fold, allowing for the successful isolation and structure elucidation of new BDL analogues. Heterocombinations of these PKS subunits from genome mining with those from previously identified BDL pathways led to the combinatorial biosynthesis of several additional novel BDL-type polyketides.