BackgroundCitramalate, a chemical precursor to the industrially important methacrylic acid (MAA), can be synthesized using Escherichia coli overexpressing citramalate synthase (cimA gene). Deletion of gltA encoding citrate synthase and leuC encoding 3-isopropylmalate dehydratase were critical to achieving high citramalate yields. Acetate is an undesirable by-product potentially formed from pyruvate and acetyl-CoA, the precursors of citramalate during aerobic growth of E. coli. This study investigated strategies to minimize acetate and maximize citramalate production in E. coli mutants expressing the cimA gene.ResultsKey knockouts that minimized acetate formation included acetate kinase (ackA), phosphotransacetylase (pta), and in particular pyruvate oxidase (poxB). Deletion of glucose 6-phosphate dehydrogenase (zwf) and ATP synthase (atpFH) aimed at improving glycolytic flux negatively impacted cell growth and citramalate accumulation in shake flasks. In a repetitive fed-batch process, E. coli gltA leuC ackA-pta poxB overexpressing cimA generated 54.1 g/L citramalate with a yield of 0.64 g/g glucose (78% of theoretical maximum yield), and only 1.4 g/L acetate in 87 h.ConclusionsThis study identified the gene deletions critical to reducing acetate accumulation during aerobic growth and citramalate production in metabolically engineered E. coli strains. The citramalate yield and final titer relative to acetate at the end of the fed-batch process are the highest reported to date (a mass ratio of citramalate to acetate of nearly 40) without being detrimental to citramalate productivity, significantly improving a potential process for the production of this five-carbon chemical.

Metabolically engineered Escherichia coli MEC143 with deletions of the ptsG, manZ, glk, pfkA, and zwf genes converts pentoses such as arabinose and xylose into glucose, with the dephosphorylation of glucose‐6‐phosphate serving as the final step. To determine which phosphatase mediates this conversion, we examined glucose formation from pentoses in strains containing knockouts of six different phosphatases singly and in combination. Deletions of single phosphatases and combinations of multiple phosphatases did not eliminate the accumulation of glucose from xylose or arabinose. Overexpression of one phosphatase, haloacid dehalogenase‐like phosphatase 12 coded by the ybiV gene, increased glucose yield significantly from 0.26 to 0.30 g/g (xylose) and from 0.32 to 0.35 g/g (arabinose). Growing cells under phosphate‐limited steady‐state conditions increased the glucose yield to 0.39 g glucose/g xylose, but did not affect glucose yield from arabinose (0.31 g/g). No single phosphatase is exclusively responsible for the conversion of glucose‐6‐phosphate to glucose in E. coli MEC143. Phosphate‐limited conditions are indeed able to enhance glucose formation in some cases, with this effect likely influenced by the different phosphate demands when E. coli metabolizes different carbon sources.