Engineering a more thermostable blue light photo receptor Bacillus subtilis YtvA LOV domain by a computer aided rational design method

Abstract

The ability to design thermostable proteins offers enormous potential for the development of novel protein bioreagents. In this work, a combined computational and experimental method was developed to increase the T m of the flavin mononucleotide based fluorescent protein Bacillus Subtilis YtvA LOV domain by 31 Celsius, thus extending its applicability in thermophilic systems. Briefly, the method includes five steps, the single mutant computer screening to identify thermostable mutant candidates, the experimental evaluation to confirm the positive selections, the computational redesign around the thermostable mutation regions, the experimental reevaluation and finally the multiple mutations combination. The adopted method is simple and effective, can be applied to other important proteins where other methods have difficulties, and therefore provides a new tool to improve protein thermostability.

Figure 1. Thermal denaturation of the WT FbFP, the single point mutant N124Y, and the triple mutant N107Y-N124Y-M111F.

The fluorescence intensity of the bound FMN is used to monitor the protein denaturation. As can be seen, the mutants have higher percentages of fluorescence at elevated temperature than WT suggesting mutations increase FbFP thermostability.

Figure 2. Locations of mutated sites exhibiting improved thermostability.

WT residues of the mutated sites are highlighted in yellow and labeled in red. The two subunits are drawn in grey and dark cyan respectively. Residues H22, V25, N107, D109, M111, V120 and N124 are from the dimer interface. The figure was drawn based on FbFP x-ray structure 2PR5 by using Discovery Studio Visualizer program.

Figure 3. Residues in close contact with I120 (A, B), F107 (C, D), F124 (E, F) and F111 (G, H) are labeled where (A, C, E, G) are from subunit 1 and (B, D, F, H) are from subunit 2.

Figure 4. Correlation between the experimental and computational ΔΔG by using FoldX (panel A) and the free energy method (panel B).

The best fit line is ΔΔG_com = 0.233ΔΔG_exp −0.72 kcal/mol in panel A, and ΔΔG_com = 1.38ΔΔG_exp −1.42 kcal/mol in panel B.

Figure 5. Flowchart of designing thermostable FbFP mutants.

Briefly, FoldX followed by FEC (free energy calculation) are used to search for potential thermostable single mutants, from which a dozen are selected for experimental tests. The distribution of thermostable mutants is analyzed to identify the “hot spot”. Then more mutants in the “hot spot” are calculated by FEC and those predicted to be more stable are tested by experiments. Finally all stabilizing mutants are pooled together and multiple mutants are combined to further improve the protein's stability.