Generating compatible translation initiation regions for heterologous gene expression inEscherichia coliby exhaustive periShine-Dalgarno mutagenesis. Human glutathione reductase cDNA as a model

Adaptation of eucaryotic cDNA to heterologous expression was studied by mutating the translation initiation (TI) region upstream (mTI) and downstream (MTI) of the start codon. In the mTI subregion the 8 bases flanking the invariant Shine-Dalgarno motif GG-AG were mutagenized exhaustively, while the MTI subregion was subjected to random silent mutations at the wobble positions. The quality of a given TI sequence was judged on the basis of expressed enzyme activity. Low-yield and high-yield mutants of both TI subregions were selected and recombined systematically. The analysis of these double cartridges gave the following results: 1. As a rule, an unfavourable MTI subregion can be compensated for by mutations in the mTI subregion and vice versa. 2. The compatibility between mTI and MTI subregion is explainable at least in part by a low interaction tendency; a delta G(o)'-value of -10.7 kcal/mol appears to be a physical threshold for heterologous cDNA expression. 3. On the basis of periShine-Dalgarno mutations, the expression yield for different cDNA sequences could be increased by 1 to 2 orders of magnitude. One of these sequences encoded delta(1-15)human glutathione reductase, a mutant lacking the flexible N-terminal extension of the protein. In conclusion, to study and overcome TI region-based expression problems it is worthwhile to start out with a versatile vector containing exhaustive mutations in the periShine-Dalgarno sequences; as a rule the coding MTI subregion can be kept unchanged.