Computational mining and redesign of superoxide dismutase with activity-thermostability improvement

Superoxide dismutase (SOD) is a redox metalloenzyme that serves as a critical defense against oxidative damage caused by reactive oxygen species, with applications across food, cosmetic, and pharmaceutical industries. However, the limited catalytic activity and poor thermostability restrict its effectiveness in industrial processes. Herein, we report a SOD from Deinococcus wulumuqiensis R12 (SOD(R12)) that demonstrates remarkable catalytic activity of 1356.7 U/mg protein, and retains 48.1 % residual activity at 100 °C for 60 min. We further employed computational-guided rational design and combinatorial experiments, resulting in the engineered SOD(R12) variants with improved activity and thermostability. Among these, the D120L variant retains 87.2 % residual activity at 100 °C for 60 min, while the G109H variant shows the highest activity of 3322.5 U/mg protein. Moreover, the E47S variant achieves a specific activity of 1726.9 U/mg protein, and retains 89.4 % residual activity. Molecular dynamics simulations reveal that the D120L variant exhibits a highly rigid structure, while the G109H variant benefits from enhanced electron transfer mediated by in salt bridge modifications. Then, the E47S variant shows a balance between structural flexibility and rigidity. This study demonstrates a robust strategy for boosting enzyme activity and thermostability, offering valuable insights for optimizing SOD performance and guiding future enzyme engineering efforts.