Abstact

The functionally versatile (β/α)(8) barrel scaffold was used to migrate triosephosphate isomerase (TPI) to thiamin phosphate synthase (TPS) activity, two enzymes that share the same fold but catalyze unrelated reactions through different mechanisms. The high sensitivity of the selection methodology was determinant to succeed in finding proteins with the desired activity. A combination of rational design and random mutagenesis was used to achieve the desired catalytic migration. One of the parallel directed evolution strategies followed resulted in TPI derivatives able to complement the thiamin phosphate auxotrophy phenotype of an Escherichia coli strain deleted of thiE, the gene that codes for TPS. Successive rounds of directed evolution resulted in better complementing TPI variants. Biochemical characterization of some of the evolved TPI clones demonstrated that the K(m) for the TPS substrates was similar to that of the native TPS; however and in agreement with the very slow complementation phenotype, the k(cat) was 4 orders of magnitude lower, indicating that substrate binding played a major role on selection. Interestingly, the crystal structure of the most proficient variant showed a slightly modified TPI active site occupied by a thiamin-phosphate-like molecule. Substitution of key residues in this region reduced TPS activity, strongly suggesting that this is also the catalytic site for the evolved TPS activity. The presence of the TPS reaction product at the active site explains the fast inactivation of the enzyme observed. In conclusion, by combining rational design, random mutagenesis and a very sensitive selection, it is possible to achieve enzymatic activity migration.