Optimized synthesis of L-sorbose by C5-dehydrogenation of D-sorbitol with Gluconobacter oxydans

The optimization of L-sorbose synthesis by regiospecific dehydrogenation of D-sorbitol using Gluconobacter oxydans is reported. The current L-sorbose production processes that are based on G. oxydans and other bacterial strains are suboptimal as to yield and rate of L-sorbose synthesis. One reason for these problems is the toxicity that is induced by the substrate D-sorbitol when used in concentrations of >10% (w/v). This phenomenon significantly limits the potentials of L-sorbose production from an industrial point of view. The goal of this study was to develop a fast production process that yields L-sorbose in stoichiometric amounts starting from D-sorbitol concentrations that exceed 10% (w/v). A gradual improvement of the inoculum build-up procedure, culture medium composition, and process parameters ultimately led to a theoretically maximal L-sorbose productivity (200 g L-1 of L-sorbose from 200 g L-1 of D-sorbitol in 28 h of fermentation) using a Gluconobacter oxydans mutant strain that was selected under conditions of substrate inhibition. Because the D-sorbitol/L-sorbose bioconversion is used to mass-produce vitamin C, the procedure reported here will contribute to a more efficient and more economic synthesis of vitamin C.