Comparative analysis of L-sorbose dehydrogenase by docking strategy for 2-keto-L-gulonic acid production in Ketogulonicigenium vulgare and Bacillus endophyticus consortium

Improving the yield of 2-keto-l-gulonic acid (2-KGA), the direct precursor of vitamin C, draws more and more attention in industrial production. In this study, we try to increase the 2-KGA productivity by computer-aided selection of genes encoding l-sorbose dehydrogenases (SDH) of Ketogulonicigenium vulgare. First, six SDHs were modeled by docking strategy to predict the binding mode with co-factor PQQ. The binding energy between SSDA1-H/SSDA1-L and PQQ was the highest, followed by SSDA3/SSDA2. The binding energy between SSDA1-P/SSDB and PQQ was the lowest. Then, these genes were overexpressed, respectively, in an industrial strain K. vulgare HKv604. Overexpression of ssda1-l and ssda1-h enhanced the 2-KGA production by 7.89 and 12.56 % in mono-cultured K. vulgare, and by 13.21 and 16.86 % when K. vulgare was co-cultured with Bacillus endophyticus. When the engineered K. vulgare SyBE_Kv000116013 (overexpression of ssda1-p) or SyBE_Kv000116016 (overexpression of ssdb) was co-cultured with B. endophyticus, the 2-KGA production decreased significantly. The docking results were in accordance with the experimental data, which indicated that computer-aided modeling is an efficient strategy for screening more efficient enzymes.