Gluconobacter species rapidly perform limited oxidations of a large number of different polyhydroxy alcohols (polyols). These oxidations are catalyzed by constituitively synthesized plasma-membrane bound dehydrogenases. These bacteria would have to expend much energy to constituitively synthesize a separate substrate-specific dehydrogenase for each polyol substrate oxidized. Therefore, it is my hypothesis that Gluconobacter possess a single polyol dehydrogenase that oxidizes many different polyols. To test this hypothesis, a membrane-bound sorbitol dehydrogenase was isolated and tested for its ability to oxidize a wide range of substrates. This enzyme was removed from the membrane fractions with Triton X-100 and fractionated from other membrane proteins by anion and cation-exchange and hydrophobicinteraction chromatographies. This procedure resulted in a 36-fold enrichment of the enzyme and a 31% recovery. The isolated enzyme showed one protein band after non-denaturing polyacrylamide electrophoresis (PAGE) and three polypeptides after SDS-PAGE. Only the 67 kDa subunit had catalytic activity. The 46 kDa subunit was a C-type cytochrome. The isolated enzyme oxidized all 8 polyols tested, but did not oxidize mono-, di-, and cyclic-alcohols, aldehydes, carboxylic acids, or mono-, di-, and oligo-saccharides. Therefore, I propose that this enzyme is a polyol dehydrogenase (PDH). The isolated PDH complex showed optimal sorbitol oxidation from pH 5 to 6 at 40°C, and contained pyrroloquinoline quinone (PQQ) as its prosthetic group. Apo-PDH could be created by salt treatment and the holoenzyme reconstituted with authentic PQQ in the presence of two species of divalent cations. The c-type cytochrome of the PDH complex was not reduced by the substrate alone, but it was reduced by substrate if either CoQ₁ or the artificial electron acceptor methylphenazonium methosulfate (MPMS) were present. It is my hypothesis that, in vivo, the electrons removed from the substrate are passed from the PQQ prosthetic group of the catalytic subunit to CoQ₁₀ in the plasma membrane, and then to the electron transport chain via the cytochrome c subunit of the PDH complex. When the PDH complex is removed from the membranes with detergent, the CoQ₁₀ is likely disassociated from the enzyme, but can be replaced with MPMS when assayed with artificial electron acceptors. / Ph. D.