This study aimed to improve the hydrolase activity of the well-characterised bacterial sialidase from Micromonospora viridifaciens. The enzyme and its mutated versions were produced in Bacillus subtilis and secreted to the growth medium. Twenty amino acid positions in or near the active site were subjected to site-saturation mutagenesis and evaluated on the artificial sialidase substrate 2-O-(p-nitrophenyl)-α-d-N-acetylneuraminic acid and on the natural substrate casein glycomacropeptide. A considerably higher fraction of the mutants exhibited increased activity on the artificial substrate compared with the natural one, with the most proficient mutant showing a 13-fold improvement in kcat/Km. In contrast, no mutants displayed more than a 2-fold increase in activity on the natural substrate. To gain further insight into this important discrepancy, we analysed the stability of mutants using the PoPMuSiC software, a property that also correlates with the potential for introducing chemical variation, after validating the method with a set of experimental stability estimates. We found a significant correlation between improved hydrolase activity on the artificial substrate and reduced apparent stability. Together with the minor improvement on the natural substrate this shows an important difference between naturally evolved functionality and new laboratory functionality. Our results suggest that when engineering sialidases and potentially other proteins towards non-natural substrates that are not optimized by natural evolution, major changes in chemical properties are advantageous, and these changes tend to correlate with decreased stability, partly explaining commonly observed trade-offs between stability and proficiency.
Bibliographical note© The Author 2015
- Micromospora viridifaciens
- Protein evolution
- Stability–activity trade-off