Production of 5-keto-d-gluconate by acetic acid bacteria is catalyzed by pyrroloquinoline quinone (PQQ)-dependent membrane-bound d-gluconate dehydrogenase

Gluconobacter suboxydans IFO 12528 was selected as the best strain for 5-keto-d-gluconate (5KGA) production by oxidative fermentation. 5KGA was markedly accumulated by the strain during cultivation in a medium containing d-glucose and/or d-gluconate. The resting cells and the membrane fraction also catalyzed 5KGA formation with a minimal formation of 2-keto-d-gluconate (2KGA), an alternative keto-d-gluconate from d-gluconate. The membrane fraction of the organism was confirmed to contain a membrane-bound d-gluconate dehydrogenase (GADH) catalyzing d-gluconate oxidation to 5KGA of which optimum pH and temperature were found at pH 4 and 15°C, respectively. After treating the membrane fraction with EDTA allowing conversion from holo-GADH to the apoenzyme, 5KGA-forming GADH was confirmed to be a pyrroloquinoline quinone (PQQ)-dependent enzyme by the fact that the enzyme activity was restored by the addition of CaCl₂ and PQQ. The 5KGA-forming GADH was totally distinct from 2KGA-forming GADH in which a covalently bound FAD functions as coenzyme. 5KGA-forming GADH was well solubilized from the membrane fraction with n-octyl-β-d-thioglucoside and 5KGA formation was favourably catalyzed at relatively lower temperature, while 2KGA-forming enzyme was solubilized with Triton X-100 and relatively higher temperatures was optimum for 2KGA formation. These results are completely discrepant from the conclusion proposed by Klasen et al. [R. Klasen, S. Bringer-Mayer, H. Sahm, J. Bacteriol., 177, 1995, 2637] claiming that 5KGA was produced by d-gluconate oxidation catalyzed by NADP-dependent cytoplasmic 5KGA reductase from Gluconobacter species at fairly alkaline pH such as 10.