Computational design and production of levansucrase with increased thermostability

Levansucrase is a carbohydrate-modifying enzyme that catalyzes hydrolysis of sucrose and transfructosylation of a fructosyl unit of sucrose to produce levan and levan-type fructooligosaccharides (LFOs). Due to their various beneficial properties such as low viscosity and non-toxicity. Levan and LFOs have potential applications in many industries such as prebiotics in food industries and anti-tumor agents in pharmaceutical industries. Previous study found that the Y246S mutant of Bacillus licheniformis RN-01 levansucrase (Oligosaccharide Producing Levansucrase, OPL) could effectively produce LFOs, but its thermostability is limited at high temperature. In this study, molecular dynamics (MD) and computational protein design were used to design mutants with higher thermostability than OPL by rigidifying highly flexible residues on the enzyme surface. The results showed that K82, N83, D179, and Q308 were highly flexible, as indicated by their high values of root mean square fluctuation (RMSF) and were suitable for design. Two design approaches were employed. The designed residues were allowed to be 1) amino acids that could potentially form favorable interactions with their neighboring residues or 2) natural amino acids except G, P and C. The design results predicted that three designed mutants (K82H, N83R and Q308S) had less fluctuation of designed residues than that of OPL. The designed mutants with multiple mutations were also created and simulated. The fluctuation of designed residues of the K82H/N83R mutant was less than those of OPL and single mutants. Therefore, four designed mutants such as the K82H, N83R, Q308S and K82H/N83R mutants were experimentally characterized, and their stabilities were measured. Experimental results show that the optimum pH, temperature and product patterns of designed mutants were relatively similar to those of OPL. The K82H/N83R mutant had higher thermostability than OPL with 1.7-fold increase in t1/2, and its Tm was slightly higher than that of OPL. Furthermore, circular dichroism spectrum analysis results suggested that the K82H/N83R mutation did not drastically affect the secondary structures of enzymes. The K82H/N83R mutant may potentially be employed for LFOs production. The developed approach may be beneficial for engineering other carbohydrate-modifying enzymes for thermostability improvement.