A novel thymidine-producing strain of Escherichia coli was prepared by genome recombineering. Eleven genes were deleted by replacement with an expression cassette, and 7 genes were integrated into the genome. The resulting strain, E. coli HLT013, showed a high thymidine yield with a low deoxyuridine content. DNA microarrays were then used to compare the gene expression profiles of HLT013 and its isogenic parent strain. Based on microarray analysis, the pyr biosynthesis genes and 10 additional genes were selected and then expressed in HLT013 to find reasonable candidates for enhancing thymidine yield. Among these, phage shock protein A (PspA) showed positive effects on thymidine production by diminishing redox stress. Thus, we integrated pspA into the HLT013 genome, resulting in E. coli strain HLT026, which produced 13.2 g/liter thymidine for 120 h with fedbatch fermentation. Here, we also provide a basis for new testable hypotheses regarding the enhancement of thymidine productivity and the attainment of a more complete understanding of nucleotide metabolism in bacteria.
Thymidine is a commercially useful precursor in the chemical synthesis of various antiviral drugs, including stavudine and zidovudine (azidothymidine), the active ingredient in a formulation for the treatment of AIDS (1, 2). Thymidine, which is composed of 2-deoxyribose and a thymine base, has been produced by employing bacteria belonging to the genera Brevibacterium and Corynebacterium (3, 4). These strains were improved by chemical mutagenesis. Thymidine production in these strains has been further improved by random mutation and selection. A recent report described the production of thymidine by rationally engineered Escherichia coli strains, in which thymidine-biosynthesizing genes were amplified and thymidine-degrading genes were disrupted (5-8). Studies of both constructed E. coli strains and Brevibacterium (and Corynebacterium) spp. have used the same strategies, even though the productivities in those microorganisms were much lower than that in the E. coli recombinant strain (Ͻ500 mg/liter). The first shared strategy was the inactivation of the salvage pathway for preventing the degradation of thymidine. The second strategy was the enhancement of the reduction of nucleoside diphosphate (NDP) to deoxynucleoside diphosphate (dNDP). The third strategy was the enhancement of thymidylate synthase. In E. coli, these strategies could be achieved by deleting deoABCD, tdk, and udp with a rationally designed gene knockout and overexpression of phage T4 NDP reductase subunits and T4 thymidylate synthase (td). To reduce feedback inhibition, we overexpressed T4 thymidylate synthase and T4 NDP reductase subunits, instead of intrinsic enzymes of E. coli. In Brevibacterium (and Corynebacterium) spp., these strategies could be achieved by screening thymidine auxotrophs after chemical mutagenesis, hydroxyurea-resistant mutants, and fluorouraciland trimethoprim-resistant mutants (3, 4).Of the enzymes required for thymidine biosynthesis, NDP reductase and d...