Engineering Escherichia Coli Based On Synthetic Biology For Production Of Flavonoid Scaffolds

Flavonoids have become one of important functional nutritional products due to its biochemical properties that are useful for treating many human diseases. However, their widespread availability and application are currently limited by inefficiencies and high-cost in extraction from plant sources. This project investigated the microbial production of flavonoid scaffold to be modified for functional products. By employing strategies and tools of synthetic biology such as de novo pathway design, multivariate modular metabolic engineeering, anti-sense RNA and CRISPRi systems, flavonoid scaffolds were successfully synthesized in Escherichia coli from glucose. The established innovative strategies and elucidated mechanisms in this study would offer new perspectives to overcome the limitation and challenge of microbial production of other important pharmaceutical compounds. Major results achieved with this research are highlighted below:(1) Flavonoid scaffolds cannot be gained by native metabolic pathways from renewable and cheap substrates. Firstly, it was necessary to construct synthetic pathways for efficiently producing these compounds based on novel tools and strategies. The L-phenylalanine overproducing strains were constructed by overexpressing 3-deoxy-D-arabinoheptulosonate-7-phosphate(DAHP) synthase(aro Fwt) and feed-back resistant chorismate mutase/prephenate dehydratase(CM/PDT: phe Afbr). The L-tyrosine overproducing strains were constructed by overexpressing feed-back resistant 3-deoxy-D-arabinoheptulosonate 7-phosphate synthase(aro Gfbr) and chorismate mutase/prephenate dehydrogenase(CM/PDH: tyr Afbr). PAL/TAL from Rhodotorula glutinis, 4CL from Petroselinum crispum, CHS from Petunia X hybrida, CHI from Medicago sativa were chose to construct the main flavonoid pathway. The recombinant malonate assimilation pathway from Rhizobium trifolii was utilized to increase the supply of malonyl-Co A. The constructed strain was able to produce 2.2 mg/L(2S)-pinocembrin and 4.5 mg/L(2S)-naringenin.(2) After obtaining the synthetic pathway, multivariate modular metabolic engineering was introduced to optimize the overall pathway. The overall pathway was divided into four modules: Module I consisted of phenylalanine/tyrosine synthetic pathway; Module II consisted of TAL and 4CL; Module III consisted of CHS and CHI; Module IV consisted of mat B and mat C. Four plasmids with different gene copy numbers were employed to optimally balance four pathway modules by varying modular expression level. It was demonstrated that efficient conversion of aromatic amino acid to flavonoid scaffold is the limiting factor. Hence, modular pathway engineering strategies were employed again to alleviate this bottleneck. The pathway from aromatic amino acid to flavonoid scaffold was divided into three modules and the bottleneck pathway was balanced by modifying plasmid gene copy number and promoter strenghs. The optimum strains were capable of producing 84.2 mg/L(2S)-pinocembrin and 105.1 mg/L(2S)-naringenin.(3) Malonyl-Co A is a limiting factor for the synthesis of flavonoids, as this cofactor is often used for producing phospholipids and fatty acids. Tuning of malonyl-Co A into heterologous pathways by traditional gene knock-out strategy resulted in growth retardation and even cell death. Here, antisense RNA(as RNA) was employed to fine-tune the fatty acid pathway to balance the demands on malonyl-Co A for target-product synthesis and cell health. The relationship between sequence and function for as RNA was investigated. We found that the gene-silencing effect of as RNA could be tuned by directing as RNA to different positions in the 5’-UTR of the target gene. Based on this, the activity of as RNA was quantitatively tailored to balance the need for malonyl-Co A in the flavonoid scaffold production and cell growth. Appropriate inhibitory efficiency of this as RNA system improved the(2S)-pinocembrin and(2S)-naringenin production titer to 250.1 mg/L and 390.2 mg/L, respectively.(4) To further increase the supply of intracellular malonyl-Co A in Escherichia coli, a clustered regularly interspaced short palindromic repeats(CRISPR) interference system was constructed for fine-tuning central metabolic pathways to efficiently channel carbon flux toward malonyl-Co A. Using synthetic sg RNA to silence candidate genes, genes that could increase the intracellular malonyl-Co A level by over 223.3% were used as target genes. The efficiencies of repression of these genes were tuned to achieve appropriate levels so that the intracellular malonyl-Co A level was enhanced without significantly altering final biomass accumulation(the final OD600 decreased by less than 10.0%). Based on the results, multiple gene silencing was successful in approaching the limit of the amount of malonyl-Co A needed to produce(2S)-pinocembrin and(2S)-naringenin. By coupling the genetic modifications to cell growth, the combined effects of these genetic perturbations increased(2S)-pinocembrin and(2S)-naringenin titer to 260.1 mg/L and 421.6 mg/L.(5) Current microbial production of flavonoid scaffold always used two separate culture steps to alleviate the metabolic burden. Briefly, strains were first cultured in rich media to a target density. After that, cells were collected and transferred to minimal media to produce flavonoid scaffold. Hence, current microbial production was limited on a laboratory scale and one-step fermentation become necessary for large-scale fermentation. Firstly, culture species, conditions and components were optimized in shake flasks to find a suitable medium. Further culture medium optimization was conducted in 3L-fermentor based on glucose concentration and p H value. By exploring the effects of different culture p H on microbial fermentation, it was found that high p H values favored upstream pathway expression or catalysis, while low p H values favored downstream pathway expression or catalysis. Based on this theory, a two-stage p H control strategy was proposed. The p H was controlled at 7.0 during 0-10 h, and was shifted to 6.5 after 10 h. Finally, the titers of(2S)-naringenin and(2S)-pinocembrin were increased to 261.3 mg/L and 150.3 mg/L, respectively.