Constitutive expression of selected genes from the pentose phosphate and aromatic pathways increases the shikimic acid yield in high-glucose batch cultures of an Escherichia coli strain lacking PTS and pykF.

BACKGROUND: During the last two decades many efforts have been directed towards obtaining efficient microbial processes for the production of shikimic acid (SA); however, feeding high amounts of substrate to increase the titer of this compound has invariably rendered low conversion yields, leaving room for improvement of the producing strains. In this work we report an alternative platform to overproduce SA in a laboratory-evolved Escherichia coli strain, based on plasmid-driven constitutive expression of six genes selected from the pentose phosphate and aromatic amino acid pathways, artificially arranged as an operon. Production strains also carried inactivated genes coding for phosphotransferase system components (ptsHIcrr), shikimate kinases I and II (aroK and aroL), pyruvate kinase I (pykF) and the lactose operon repressor (lacI). RESULTS: The strong and constitutive expression of the constructed operon permitted SA production from the beginning of the cultures, as evidenced in 1 L batch-mode fermentors starting with high concentrations of glucose and yeast extract. Inactivation of the pykF gene improved SA production under the evaluated conditions by increasing the titer, yield and productivity of this metabolite compared to the isogenic pykF+ strain. The best producing strain accumulated up to 43 g/L of SA in 30 h and relatively low concentrations of acetate and aromatic byproducts were detected, with SA accounting for 80% of the produced aromatic compounds. These results were consistent with high expression levels of the glycolytic pathway and synthetic operon genes from the beginning of fermentations, as revealed by transcriptomic analysis. Despite the consumption of 100 g/L of glucose, the yields on glucose of SA and of total aromatic compounds were about 50% and 60% of the theoretical maximum, respectively. The obtained yields and specific production and consumption rates proved to be constant with three different substrate concentrations. CONCLUSIONS: The developed production system allowed continuous SA accumulation until glucose exhaustion and eliminated the requirement for culture inducers. The obtained SA titers and yields represent the highest reported values for a high-substrate batch process, postulating the strategy described in this report as an interesting alternative to the traditionally employed fed-batch processes for SA production.

Metabolic engineering for improving anthranilate synthesis from glucose in Escherichia coli.

BACKGROUND: Anthranilate is an aromatic amine used industrially as an intermediate for the synthesis of dyes, perfumes, pharmaceuticals and other classes of products. Chemical synthesis of anthranilate is an unsustainable process since it implies the use of nonrenewable benzene and the generation of toxic by-products. In Escherichia coli anthranilate is synthesized from chorismate by anthranilate synthase (TrpED) and then converted to phosphoribosyl anthranilate by anthranilate phosphoribosyl transferase to continue the tryptophan biosynthetic pathway. With the purpose of generating a microbial strain for anthranilate production from glucose, E. coli W3110 trpD9923, a mutant in the trpD gene that displays low anthranilate producing capacity, was characterized and modified using metabolic engineering strategies. RESULTS: Sequencing of the trpED genes from E. coli W3110 trpD9923 revealed a nonsense mutation in the trpD gene, causing the loss of anthranilate phosphoribosyl transferase activity, but maintaining anthranilate synthase activity, thus causing anthranilate accumulation. The effects of expressing genes encoding a feedback inhibition resistant version of the enzyme 3-deoxy-D-arabino-heptulosonate-7-phosphate synthase (aroGfbr), transketolase (tktA), glucokinase (glk) and galactose permease (galP), as well as phosphoenolpyruvate:sugar phosphotransferase system (PTS) inactivation on anthranilate production capacity, were evaluated. In shake flask experiments with minimal medium, strains W3110 trpD9923 PTS- and W3110 trpD9923/pJLBaroGfbrtktA displayed the best production parameters, accumulating 0.70-0.75 g/L of anthranilate, with glucose-yields corresponding to 28-46% of the theoretical maximum. To study the effects of extending the growth phase on anthranilate production a fed-batch fermentation process was developed using complex medium, where strain W3110 trpD9923/pJLBaroGfbrtktA produced 14 g/L of anthranilate in 34 hours. CONCLUSION: This work constitutes the first example of a microbial system for the environmentally-compatible synthesis of anthranilate generated by metabolic engineering. The results presented here, including the characterization of mutation in the trpD gene from strain W3110 trpD9923 and the development of a fermentation strategy, establish a step forward towards the future improvement of a sustainable process for anthranilate production. In addition, the present work provides very useful data regarding the positive and negative consequences of the evaluated metabolic engineering strategies.