ABSTRACT: Many phenylalanine- and tyrosine-producing strains have used plasmid-based overexpression of pathway genes. The resulting strains achieved high titers and yields of phenylalanine and tyrosine. Chromosomally engineered, plasmid-free producers have shown lower titers and yields than plasmid-based strains, but the former are advantageous in terms of cultivation cost and public health/environmental risk. Therefore, we engineered here the Escherichia coli chromosome to create superior phenylalanine- and tyrosine-overproducing strains that did not depend on plasmid-based expression. Integration into the E. coli chromosome of two central metabolic pathway genes (ppsA and tktA) and eight shikimate pathway genes (aroA, aroB, aroC, aroD, aroE, aroGfbr , aroL, and pheAfbr ), controlled by the T7lac promoter, resulted in excellent titers and yields of phenylalanine; the superscript "fbr" indicates that the enzyme encoded by the gene was feedback resistant. The generated strain could be changed to be a superior tyrosine-producing strain by replacing pheAfbr with tyrAfbr A rational approach revealed that integration of seven genes (ppsA, tktA, aroA, aroB, aroC, aroGfbr , and pheAfbr ) was necessary as the minimum gene set for high-yield phenylalanine production in E. coli MG1655 (tyrR, adhE, ldhA, pykF, pflDC, and ascF deletant). The phenylalanine- and tyrosine-producing strains were further applied to generate phenyllactic acid-, 4-hydroxyphenyllactic acid-, tyramine-, and tyrosol-producing strains; yield of these aromatic compounds increased proportionally to the increase in phenylalanine and tyrosine yields.IMPORTANCE Plasmid-free strains for aromatic compound production are desired in the aspect of industrial application. However, the yields of phenylalanine and tyrosine have been considerably lower in plasmid-free strains than in plasmid-based strains. The significance of this research is that we succeeded in generating superior plasmid-free phenylalanine- and tyrosine-producing strains by engineering the E. coli chromosome, which was comparable to that in plasmid-based strains. The generated strains have a potential to generate superior strains for the production of aromatic compounds. Actually, we demonstrated that four kinds of aromatic compounds could be produced from glucose with high yields (e.g., 0.28 g tyrosol/g glucose).
