K. M. Draths

Benzene‐Free Synthesis of Adipic Acid

Strains of Escherichia coli were constructed and evaluated that synthesized cis,cis‐muconic acid from d‐glucose under fed‐batch fermentor conditions. Chemical hydrogenation of the cis, cis‐muconic acid in the resulting fermentation broth has also been examined. Biocatalytic synthesis of adipic acid from glucose eliminates two environmental concerns characteristic of industrial adipic acid manufacture: use of carcinogenic benzene and benzene‐derived chemicals as feedstocks and generation of nitrous oxide as a byproduct of a nitric acid catalyzed oxidation. While alternative catalytic syntheses that eliminate the use of nitric acid have been developed, most continue to rely on petroleum‐derived benzene as the ultimate feedstock. In this study, E. coli WN1/pWN2.248 was developed that synthesized 36.8 g/L of cis, cis‐muconic acid in 22% (mol/mol) yield from glucose after 48 h of culturing under fed‐batch fermentor conditions. Optimization of microbial cis,cis‐muconic acid synthesis required expression of three enzymes not typically found in E. coli. Two copies of the Klebsiella pneumoniae aroZ gene encoding DHS dehydratase were inserted into the E. coli chromosome, while the K. pneumoniae aroY gene encoding PCA decarboxylase and the Acinetobacter calcoaceticus catA gene encoding catechol 1,2‐dioxygenase were expressed from an extrachromosomal plasmid. After fed‐batch culturing of WN1/pWN2.248 was complete, the cells were removed from the broth, which was treated with activated charcoal and subsequently filtered to remove soluble protein. Hydrogenation of the resulting solution with 10% Pt on carbon (5% mol/mol) at 3400 kPa of H2 pressure for 2.5 h at ambient temperature afforded a 97% (mol/mol) conversion of cis, cis‐muconic acid into adipic acid.

Phosphoenolpyruvate Availability and the Biosynthesis of Shikimic Acid

The impact of increased availability of phosphoenolpyruvate during shikimic acid biosynthesis has been examined in Escherichia coliK‐12 constructs carrying plasmid‐localized aroFFBR and tktAinserts encoding, respectively, feedback‐insensitive 3‐deoxy‐d‐arabino‐heptulosonic acid 7‐phosphate synthase and transketolase. Strategies for increasing the availability of phosphoenolpyruvate were based on amplified expression of E. coli ppsA‐encoded phosphoenolpyruvate synthase or heterologous expression of the Zymomonas mobilis glf‐encoded glucose facilitator. The highest titers and yields of shikimic acid biosynthesized from glucose in 1 L fermentor runs were achieved using E. coli SP1.lpts/pSC6.090B, which expressed both Z. mobilis glf‐encoded glucose facilitator protein and Z. mobilis glk‐encoded glucose kinase in a host deficient in the phosphoenolpyruvate:carbohydrate phosphotransferase system. At 10 L scale with yeast extract supplementation, E. coli SP1.lpts/pSC6.090B synthesized 87 g/L of shikimic acid in 36% (mol/mol) yield with a maximum productivity of 5.2 g/L/h for shikimic acid synthesized during the exponential phase of growth.

Altered Glucose Transport and Shikimate Pathway Product Yields in E.coli

Different glucose transport systems are examined for their impact on phosphoenolpyruvate availability as reflected by the yields of 3‐dehydroshikimic acid and byproducts 3‐deoxy‐d‐arabino‐heptulosonic acid, 3‐dehydroquinic acid, and gallic acid synthesized by Escherichia coli from glucose. 3‐Dehydroshikimic acid is an advanced shikimate pathway intermediate in the syntheses of a spectrum of commodity, pseudocommodity, and fine chemicals. All constructs carried plasmid aroFFBR and tktA inserts encoding, respectively, a feedback‐insensitive isozyme of 3‐deoxy‐d‐arabino‐heptulosonic acid 7‐phosphate synthase and transketolase. Reliance on the native E. coli phosphoenolpyruvate:carbohydrate phosphotransferase system for glucose transport led in 48 h to the synthesis of 3‐dehydroshikimic acid (49 g/L) and shikimate pathway byproducts in a total yield of 33% (mol/mol). Use of heterologously expressed Zymomonas mobilis glf‐encoded glucose facilitator and glk‐encoded glucokinase resulted in the synthesis in 48 h of 3‐dehydroshikimic acid (60 g/L) and shikimate pathway byproducts in a total yield of 41% (mol/mol). Recruitment of native E. coli galP‐encoded galactose permease for glucose transport required 60 h to synthesize 3‐dehydroshikimic acid (60 g/L) and shikimate pathway byproducts in a total yield of 43% (mol/mol). Direct comparison of the impact of altered glucose transport on the yields of shikimate pathway products synthesized by E. coli has been previously hampered by different experimental designs and culturing conditions. In this study, the same product and byproduct mixture synthesized by E. coli constructs derived from the same progenitor strain is used to compare strategies for increasing phosphoenolpyruvate availability. Constructs are cultured under the same set of fermentor‐controlled conditions.

Modulation of Phosphoenolpyruvate Synthase Expression Increases Shikimate Pathway Product Yields in E. coli

Product yields in microbial synthesis are ultimately limited by the mechanism utilized for glucose transport. Altered expression of phosphoenolpyruvate synthase was examined as a method for circumventing these limits. Escherichia coli KL3/pJY1.216A was cultured under fed‐batch fermentor conditions where glucose was the only source of carbon for the formation of microbial biomass and the synthesis of product 3‐dehydroshikimic acid. Shikimate pathway byproducts 3‐deoxy‐d‐ arabino‐heptulosonic acid, 3‐dehydroquinic acid, and gallic acid were also generated. An optimal expression level of phosphoenolpyruvate synthase was identified, which did not correspond to the highest expression levels of this enzyme, where the total yield of 3‐dehydroshikimic acid and shikimate pathway byproducts synthesized from glucose was 51% (mol/mol). For comparison, the theoretical maximum yield is 43% (mol/mol) for synthesis of 3‐dehydroshikimic acid and shikimate pathway byproducts from glucose in lieu of amplified expression of phosphoenolpyruvate synthase.