Identification of novel glycoside hydrolases via whole genome sequencing of for production of various minor ginsenosides.

In this study, many bacterial strains were screened for the production of minor ginsenosides, but based on conversion competence among the strains, the strain BS26 has the good ginsenoside-transforming ability. Therefore, the strain BS26 was selected for complete genome sequence analysis to determine the target (glycoside hydrolase) functional genes. Whole genome analysis of strain BS26 showed 43 glycoside hydrolase genes in total. To determine the target functional gene, 12 sets of six different glycoside hydrolases (3 set of β-glucosidase; 3 set of trehalase; 3 set of arabinofuranosidase; 2 set of xylosidase; and one set of each α-galactosidase and α-fucosidase, respectively) were selected and cloned in BL21 (DE3) using the pGEX4T-1 vector and were characterized. Among these 12 sets of clones, only one, β-glucosidase (BglNg-767), showed ginsenoside conversion ability. The BglNg-767 comprised 767 amino acids and belonged to glycoside hydrolase family 3 (GH3). The recombinant GST-BglNg-767 was capable of altering the ginsenosides Rb, Rd, and gypenoside XVII (Gyp-XVII) to F; Rb to C-O; Rb to C-Mx1, and Rc to C-Mc1. Besides, complete genome sequence analysis of strain BS26 also indicates 30 endopeptidase genes, which may be responsible for self-hydrolysis of the proteins. Therefore, using SDS-PAGE analysis, we predict that the difference between the molecular weight of the expressed protein (around 90 kDa) and the predicted amino-acid sequence (102.7 kDa) is due to self-hydrolysis of the proteins.