The FuncLib section of the CSFp strategy was used to further evolve the mutant F308G Nine residues close to the substrate in the binding pocket were selected for diversification Figure 3A The sequenc

The FuncLib module in the CSFp strategy was utilized to further enhance the mutant F308G by selecting nine residues in close proximity to the substrate in the binding pocket for diversification (Figure 3A). A sequence space for D-Lac was generated, resulting in 1000 variants (Table S2), of which only the top 50 with the best energy were considered (Figure 3B). A 30 ns MD simulation was performed for each of these 50 variants, and ΔMM/GBSA was calculated for each variant relative to WT (Figure 3C). Subsequently, 10 variants with ΔMM/GBSA were selected for gene synthesis and validation to explore the diversity of multipoint combinatorial mutations while reducing the screening intensity (Figure 3D). However, the results showed that half of the 10 combination mutants were completely inactivated, and the optimal mutant activity only increased by 0.4-fold relative to WT, which was significantly lower than that of F308G (Figure 3E). These results suggest that the epistasis effect of mutations leads to unsatisfactory outcomes from FuncLib. To eliminate the negative impact of poor mutations and prevent the loss of excellent mutants from the sequence space, each residue of all the combinatorial mutants in Figure 3D was mutated in combination with F308G, resulting in a total of 13 double mutants (Figure 3F). While five of the 13 double mutants lost activity completely, the mutants N96S/F308G, A271N/F308G, and A271D/F308G exhibited 13-16-fold higher activity than WT. Moreover, N96S and A271N/D were mutated in combination with F308G, resulting in the mutants N96S/A271D/F308G and N96S/A271N/F308G, which displayed a 9- and 17-fold increase in activity, respectively (Figure 3G)