Production of aromatic compounds by metabolically engineered Escherichia coli with an expanded shikimate pathway.

Escherichia coli was metabolically engineered by expanding the shikimate pathway to generate strains capable of producing six kinds of aromatic compounds, phenyllactic acid, 4-hydroxyphenyllactic acid, phenylacetic acid, 4-hydroxyphenylacetic acid, 2-phenylethanol, and 2-(4-hydroxyphenyl)ethanol, which are used in several fields of industries including pharmaceutical, agrochemical, antibiotic, flavor industries, etc. To generate strains that produce phenyllactic acid and 4-hydroxyphenyllactic acid, the lactate dehydrogenase gene (ldhA) from Cupriavidus necator was introduced into the chromosomes of phenylalanine and tyrosine overproducers, respectively. Both the phenylpyruvate decarboxylase gene (ipdC) from Azospirillum brasilense and the phenylacetaldehyde dehydrogenase gene (feaB) from E. coli were introduced into the chromosomes of phenylalanine and tyrosine overproducers to generate phenylacetic acid and 4-hydroxyphenylacetic acid producers, respectively, whereas ipdC and the alcohol dehydrogenase gene (adhC) from Lactobacillus brevis were introduced to generate 2-phenylethanol and 2-(4-hydroxyphenyl)ethanol producers, respectively. Expression of the respective introduced genes was controlled by the T7 promoter. While generating the 2-phenylethanol and 2-(4-hydroxyphenyl)ethanol producers, we found that produced phenylacetaldehyde and 4-hydroxyphenylacetaldehyde were automatically reduced to 2-phenylethanol and 2-(4-hydroxyphenyl)ethanol by endogenous aldehyde reductases in E. coli encoded by the yqhD, yjgB, and yahK genes. Cointroduction and cooverexpression of each gene with ipdC in the phenylalanine and tyrosine overproducers enhanced the production of 2-phenylethanol and 2-(4-hydroxyphenyl)ethanol from glucose. Introduction of the yahK gene yielded the most efficient production of both aromatic alcohols. During the production of 2-phenylethanol, 2-(4-hydroxyphenyl)ethanol, phenylacetic acid, and 4-hydroxyphenylacetic acid, accumulation of some by-products were observed. Deletion of feaB, pheA, and/or tyrA genes from the chromosomes of the constructed strains resulted in increased desired aromatic compounds with decreased by-products. Finally, each of the six constructed strains was able to successfully produce a different aromatic compound as a major product. We show here that six aromatic compounds are able to be produced from renewable resources without supplementing with expensive precursors.