Recovery of lens optics and epithelial enzymes after ultraviolet A radiation

PURPOSE

To establish the mechanism by which ultraviolet A (UVA) radiation causes irreversible damage to the eye lens.

METHODS

The authors irradiated 223 bovine lenses in organ culture with 22.4, 33.6, and 44.8 J/cm2 of UVA radiation (365 nm) and studied biochemical and optical properties of the lenses in long-term culture conditions. Each lens tested was placed in a specially designed cell. The lenses were oriented so that the anterior surface faced the incident UVA radiation source, and they were maintained in their cells during irradiation. After irradiation, lens optical quality was monitored throughout the culture period, and lens samples were taken for enzyme analysis.

RESULTS

Full recovery of lens optical damage and activity of the enzymes hexokinase, catalase, and glucose-6-phosphate dehydrogenase in lens epithelium was observed after 8 days in culture after irradiation with 22.4J/cm2. After irradiation with 33.6J/cm2, partial recovery of optical damage was found, and there was between 80% to 90% recovery of the enzyme activity. No recovery of optical and enzyme activity was found after 44.8J/cm2 irradiation.

CONCLUSIONS

Irradiation between 22.4J/cm2 to 33.6J/cm2 of UVA at 365 nm is the minimal level of irradiation that causes irreversible damage to lens enzymes and optics.

Synthesis of pyrroloquinoline quinone in vivo and in vitro and detection of an intermediate in the biosynthetic pathway

In Klebsiella pneumoniae, six genes, constituting the pqqABCDEF operon, which are required for the synthesis of the cofactor pyrroloquinoline quinone (PQQ) have been identified. The role of each of these K. pneumoniae Pqq proteins was examined by expression of the cloned pqq genes in Escherichia coli, which cannot synthesize PQQ. All six pqq genes were required for PQQ biosynthesis and excretion into the medium in sufficient amounts to allow growth of E. coli on glucose via the PQQ-dependent glucose dehydrogenase. Mutants lacking the PqqB or PqqF protein synthesized small amounts of PQQ, however. PQQ synthesis was also studied in cell extracts. Extracts made from cells containing all Pqq proteins contained PQQ. Lack of each of the Pqq proteins except PqqB resulted in the absence of PQQ. Extracts lacking PqqB synthesized PQQ slowly. Complementation studies with extracts containing different Pqq proteins showed that an extract lacking PqqC synthesized an intermediate which was also detected in the culture medium of pqqC mutants. It is proposed that PqqC catalyzes the last step in PQQ biosynthesis. Studies with cells lacking PqqB suggest that the same intermediate might be accumulated in these mutants. By using pqq-lacZ protein fusions, it was shown that the expression of the putative precursor of PQQ, the small PqqA polypeptide, was much higher than that of the other Pqq proteins. Synthesis of PQQ most likely requires molecular oxygen, since PQQ was not synthesized under anaerobic conditions, although the pqq genes were expressed.

Elucidation of the region responsible for EDTA tolerance in PQQ glucose dehydrogenases by constructing Escherichia coli and Acinetobacter calcoaceticus chimeric enzymes

We constructed various chimeric PQQ glucose dehydrogenases (PQQGDHs) from an EDTA-sensitive PQQGDH from Escherichia coli and an EDTA-tolerant PQQGDH from Acinetobacter calcoaceticus by homologous recombination of their structural genes. The EDTA tolerance of the resulting chimeric enzymes was investigated. Our results demonstrated that EDTA tolerance of PQQGDHs can be completely altered by substituting each corresponding region. The EDTA tolerance of A. calcoaceticus PQQGDH is mostly within a region composed of about 90 amino acid residues located between 45 and 56% of the distance from the N-terminal region.

Kinetic studies of gluconate pathway enzymes from Schizosaccharomyces pombe

Glucose dehydrogenase and gluconate kinase which catalyze two-step reactions of the gluconate pathway have been purified from Schizosaccharomyces pombe. Their steady-state kinetic studies were undertaken. The yeast glucose dehydrogenase requires NADP+ as an obligatory coenzyme and mediates the oxidation of D-glucose to D-gluconate via an ordered Bi Bi mechanism with NADP+ as the leading substrate. Kinetic constants for the dehydrogenase reactions have been measured. The yeast gluconate kinase requires Mg2+ as an activator. The phosphorylation catalyzed by the fission yeast gluconate kinase has been studied kinetically at a fixed concentration of Mg2+. The initial velocity and product inhibition results are consistent with a rapid equilibrium random Bi Bi mechanism with the formation of an abortive enzyme-ADP-gluconate complex. Dissociation constants of the two substrates, ATP and D-gluconate from various binary and ternary enzymic complexes, have been determined.

The glucose dehydrogenase-mediated energization of Acinetobacter calcoaceticus as a tool for evaluating its susceptibility to, and defence against, hazardous chemicals

Cells of Acinetobacter calcoaceticus 69-V could be energized by glucose oxidation after the growth on acetate, ethanol, hexanol and benzoate. The velocities of glucose oxidation-driven ATP syntheses were relatively constant in the range from pH 5.4 to 7.5. With decreasing pH values (7.0, 6.0, 5.4) ATP synthesis was inhibited more strongly by the action of 2,4-dinitrophenol and at the same pH value glucose oxidation was nearly unimpaired or inhibited more weakly. This finding is expressed by a decrease of the P/O ratios, indicating the uncoupling of the electron-transport phosphorylation by 2,4-dinitrophenol. The sensitivity towards this uncoupling effect was higher in ethanol-grown cells of Acinetobacter calcoaceticus 69-V than in hexanol- or acetate-grown cells. This increase in sensitivity was accompanied by a decrease of the ratio of saturated (mainly C16:0) to unsaturated (C16:1, C18:1) fatty acids in ethanol-grown cells compared with hexanol-grown ones. The knowledge of such differences in the susceptibility and its molecular background, e.g. possible substrate-induced changes of the fatty acid composition of the cytoplasmic membranes, should help elucidate mechanisms of poisoning by membrane-active hazardous chemicals and develop defence strategies.

The physiological function of periplasmic glucose oxidation in phosphate-limited chemostat cultures of Klebsiella pneumoniae NCTC 418

Periplasmic oxidation of glucose into gluconate and 2-ketogluconate in Klebsiella pneumoniae occurs via glucose dehydrogenase (GDH) and gluconate dehydrogenase (GaDH), respectively. Since, as is shown here, in the presence of glucose, gluconate and 2-ketogluconate are not further metabolized intracellularly the physiological function of this periplasmic route was studied. It was found that periplasmic oxidation of glucose could function as an alternative production route of ATP equivalents. Instantaneous activation of either GDH or GaDH reduced the rate of degradation of glucose via glycolysis and the tricarboxylic acid (TCA) cycle in vivo. Furthermore, aerobic, magnesium- and phosphate-limited chemostat cultures with glucose as the carbon source showed high GDH plus GaDH activities in contrast to nitrogen- and sulphate-limited cultures. However, when fructose, which is not degraded by GDH, was the carbon source, specific oxygen consumption rates under these four conditions were essentially the same. The latter observation suggests that high transmembrane phosphate gradients which are supposedly present under phosphate-limited conditions do not cause high energetic demands due to futile cycling of phosphate ions. In addition, dissipation of the transmembrane phosphate gradient of phosphate-limited cells immediately increased the rate of intracellular glucose degradation. It is concluded that under phosphate-limited conditions (i) extensive futile cycling of phosphate ions is absent and (ii) low concentrations of phosphate ions limit intracellular degradation of glucose.(ABSTRACT TRUNCATED AT 250 WORDS)

Absence of glucose uptake by liver microsomes: an explanation for the complete latency of glucose dehydrogenase

The permeability of rat liver microsomes to glucose was investigated in relation to the hexose-6-phosphate dehydrogenase system (EC 1.1.1.47). It was found that glucose-6-phosphate dehydrogenase activity could be assayed with NADP as coenzyme in both untreated and detergent-treated microsomes. However, when glucose was used as substrate, activity was only measurable in detergent-treated microsomes. Moreover, radioactive glucose added to microsomes in a variety of experimental conditions was never taken up by the vesicles. Our results indicate that NADP (or NAD) availability is probably not the reason for the absence of glucose dehydrogenase activity in untreated microsomes but rather membrane impermeability to glucose would account for the complete latency observed. This finding calls for a reevaluation of glucose transport in relation to other enzymes of the endoplasmic reticulum, such as glucose-6-phosphatase.

The crystal structure of glucose dehydrogenase from Thermoplasma acidophilum

BACKGROUND

The archaea are a group of organisms distinct from bacteria and eukaryotes. Structures of proteins from archaea are of interest because they function in extreme environments and because structural studies may reveal evolutionary relationships between proteins. The enzyme glucose dehydrogenase from the thermophilic archaeon Thermoplasma acidophilum is of additional interest because it is involved in an unusual pathway of sugar metabolism.

RESULTS

We have determined the crystal structure of this glucose dehydrogenase to 2.9 A resolution. The monomer comprises a central nucleotide-binding domain, common to other nucleotide-binding dehydrogenases, flanked by the catalytic domain. Unexpectedly, we observed significant structural homology between the catalytic domain of horse liver alcohol dehydrogenase and T. acidophilum glucose dehydrogenase.

CONCLUSIONS

The structural homology between glucose dehydrogenase and alcohol dehydrogenase suggests an evolutionary relationship between these enzymes. The quaternary structure of glucose dehydrogenase may provide a model for other tetrameric alcohol/polyol dehydrogenases. The predicted mode of nucleotide binding provides a plausible explanation for the observed dual-cofactor specificity, the molecular basis of which can be tested by site-directed mutagenesis.

Energetics of alanine, lysine, and proline transport in cytoplasmic membranes of the polyphosphate-accumulating Acinetobacter johnsonii strain 210A

Amino acid transport in right-side-out membrane vesicles of Acinetobacter johnsonii 210A was studied. L-Alanine, L-lysine, and L-proline were actively transported when a proton motive force of -76 mV was generated by the oxidation of glucose via the membrane-bound glucose dehydrogenase. Kinetic analysis of amino acid uptake at concentrations of up to 80 microM revealed the presence of a single transport system for each of these amino acids with a Kt of less than 4 microM. The mode of energy coupling to solute uptake was analyzed by imposition of artificial ion diffusion gradients. The uptake of alanine and lysine was driven by a membrane potential and a transmembrane pH gradient. In contrast, the uptake of proline was driven by a membrane potential and a transmembrane chemical gradient of sodium ions. The mechanistic stoichiometry for the solute and the coupling ion was close to unity for all three amino acids. The Na+ dependence of the proline carrier was studied in greater detail. Membrane potential-driven uptake of proline was stimulated by Na+, with a half-maximal Na+ concentration of 26 microM. At Na+ concentrations above 250 microM, proline uptake was strongly inhibited. Generation of a sodium motive force and maintenance of a low internal Na+ concentration are most likely mediated by a sodium/proton antiporter, the presence of which was suggested by the Na(+)-dependent alkalinization of the intravesicular pH in inside-out membrane vesicles. The results show that both H+ and Na+ can function as coupling ions in amino acid transport in Acinetobacter spp.

Optimization of enzyme ratios in a coimmobilized enzyme reactor for the analysis of D-xylose and D-xylulose in a flow system

A coupled enzyme system for the detection of D-xylose and D-xylulose is presented. The system is based on three consecutive enzymatic steps. The enzymes xylose isomerase (XI), mutarotase (MT), and glucose dehydrogenase (GDH) are coimmobilized on controlled pore glass and packed in a bed reactor. The relative amount of enzymes, i.e., enzyme ratio, plays a critical role in driving the overall reaction, resulting in a system with linear response characteristics and an operational range of several orders of magnitude. Three different enzyme ratios are assayed to achieve maximum conversion efficiencies for xylose and xylulose. The highest enzyme unit ratio assayed, 13.4 of GDH to XI, gave the highest apparent pseudo-first-order rate constant showing the importance of the last enzymatic reaction in the coupled system to make the overall reaction thermodynamically favorable. A pH of 7.0 was found to be an optimum compromise for the multienzyme system. Sensitivity was dependent on NAD+ concentration. The study was carried out in a flow injection system. The optimized reactor has been applied for the catalytic detection of pentoses in flow injection analysis (FIA) and liquid chromatography (LC).

Two binding sites of inhibitors in NADH: ubiquinone oxidoreductase (complex I). Relationship of one site with the ubiquinone-binding site of bacterial glucose:ubiquinone oxidoreductase

The effect of ten naturally occurring and two synthetic inhibitors of NADH:ubiquinone oxidoreductase (complex I) of bovine heart, Neurospora crassa and Escherichia coli and glucose:ubiquinone oxidoreductase (glucose dehydrogenase) of Gluconobacter oxidans was investigated. These inhibitors could be divided into two classes with regard to their specificity and mode of action. Class I inhibitors, including the naturally occurring piericidin A, annonin VI, phenalamid A2, aurachins A and B, thiangazole and the synthetic fenpyroximate, inhibit complex I from all three species in a partially competitive manner and glucose dehydrogenase in a competitive manner, both with regard to ubiquinone. Class II inhibitors including the naturally occurring rotenone, phenoxan, aureothin and the synthetic benzimidazole inhibit complex I from all species in an non-competitive manner, but have no effect on the glucose dehydrogenase. Myxalamid PI could not be classified as above because it inhibits only the mitochondrial complex I and in a competitive manner. All inhibitors affect the electron-transfer step from the high-potential iron-sulphur cluster to ubiquinone. Class I inhibitors appear to act directly at the ubiquinone-catalytic site which is related in complex I and glucose dehydrogenase.

A novel embryo-specific barley cDNA clone encodes a protein with homologies to bacterial glucose and ribitol dehydrogenase

In order to analyze the genetic programme expressed during the early stages of embryogenesis a cDNA clone bank was constructed from desiccation-tolerant excised barley embryos 18 d after pollination (D. Bartels et al., 1988, Planta 175, 485-492). One of the selected cDNA clones pG31 encodes a transcript of 1300 nucleotides and a protein of 31 kDa, both are specifically expressed in developing embryos and are not detected in other tissues. The expression of the pG31 mRNA is not modulated by the plant hormone cis-abscisic acid but it ceases to be expressed in germinating embryos. The protein sequence deduced from the pG31 transcript shows substantial sequence homologies to bacterial glucose dehydrogenase and ribitol dehydrogenase. Biochemical analysis indicates that glucose dehydrogenase activity is present in protein extracts from embryos 18 d after pollination. This glucose dehydrogenase activity is inhibited by antiserum raised against the recombinant pG31 protein. These findings provide evidence for the discovery of a novel pathway in carbohydrate metabolism acting specifically during embryogenesis.

Topological analysis of quinoprotein glucose dehydrogenase in Escherichia coli and its ubiquinone-binding site

Topological structure of quinoprotein glucose dehydrogenase in the inner membrane of Escherichia coli was determined by constructing protein fusions with alkaline phosphatase or beta-galactosidase. Analysis of the fusions revealed that the dehydrogenase possesses five membrane-spanning segments, and the N-terminal and C-terminal portions resided at the cytoplasmic and periplasmic side of the membrane, respectively. These results agreed with the hydropathy profile based on its primary structure. The topological structure suggests that the predicted binding site of the prosthetic group pyrroloquinoline quinone is located at the periplasmic side and that the amino acid residues corresponding to those that were presumed to interact with ubiquinone in one subunit of mitochondrial NADH dehydrogenase also occur at the periplasmic side. When the purified glucose dehydrogenase and cytochrome o ubiquinol oxidase were reconstituted together with ubiquinone into liposomes, a membrane potential could be generated by the electron transfer at the site of the ubiquinol oxidase but not of the dehydrogenase. These results suggest that glucose dehydrogenase has a ubiquinone reacting site close to the periplasmic side of the membrane, and thus its electron transfer to ubiquinone appears to be incapable of forming a proton electrochemical gradient across the inner membrane of E. coli.

Energy conservation by pyrroloquinoline quinol-linked xylose oxidation in Pseudomonas putida NCTC 10936 during carbon-limited growth in chemostat culture

When grown in carbon source-limited chemostat cultures with lactate or glucose as the carbon and energy source and xylose as an additional source of reducing equivalents. Pseudomonas putida NCTC 10936 oxidized xylose to xylonolactone and xylonate. No other products were formed from this pentose, nor was it incorporated into biomass. The presence of xylose in these cultures resulted in higher Yglucose and Ylactate values as compared to cultures without xylose indicating that biologically useful energy was conserved during the periplasmic oxidation of xylose. As the Y0 values for growth on glucose or on lactate alone were equal to the Y0 values for growth with xylose as co-substrate, it is concluded that for glucose- or lactate-limited growth energy conservation by PQQH2 oxidation is as efficient as by NADH2 oxidation.

Estimates of uridine diphosphate hexoses in erythrocytes: implications for galactosemia

Impurities in a reagent dehydrogenase caused overestimates of erythrocytic uridine diphosphate glucose and accounted for clinically important differences in results between those of one group of investigators using enzymatic methods and those of two other groups using enzymatic methods, high-performance liquid chromatography, and nuclear magnetic resonance. These data have relevance for the current debate regarding the pathophysiologic changes in galactosemia.

Characterization of the gcd gene from Escherichia coli K-12 W3110 and regulation of its expression

DNA sequence and expressional analyses of the gcd gene of Escherichia coli K-12 W3110 revealed that two promoters that were detected were regulated negatively by cyclic AMP and positively by oxygen. Sequence conservation of the gcd gene between E. coli K-12 W3110 and PPA42 suggests that glucose dehydrogenase is required for the E. coli cells, even though it ordinarily exists as an apoprotein.

Microbiosensors for acetylcholine and glucose

Microbiosensors based on carbon and and platinum fibers are described. Carbon fibers were used to construct microelectrodes of 7 microm diameter. Electrochemical operations for pre-electrolysis and measuring were examined for the highly sensitive determination of hydrogen peroxide. A triangular potential (-2 to +2V vs Ag/AgCl) was applied before measuring each pair of double pulses (first pulse: 750 mV; second pulse: 1100 mV). The determination limit was 0.1 microM of hydrogen peroxide. The reproducible determination of hydrogen peroxide is possible even in samples containing albumin protein. The separation of hydrogen peroxide from ascorbic acid is also possible because the oxidation potential of ascorbic acid is different from that of hydrogen peroxide. An acetylcholine microsensor was fabricated by immobilizing acetylcholine esterase and choline oxidase on the carbon fiber by entrapment with poly(vinyl alcohol)-quarternized stilbazole (PVA-SbQ). This sensor gave a linear calibration plot for the range 0.1-1.0 mM with a linear correlation coefficient of 0.9842. Glucose oxidase (GOD) and glucose dehydrogenase (GDH) immobilized cylindrical platinum microelectrodes were fabricated, and their characteristics were evaluated, respectively, by using 1,4-benzoquinone (BQ) and ferricyanide as electron mediators. Each enzyme was immobilized by using PVA-SbQ on a cylindrical microelectrode of 2 microm diameter. A linear range in the calibration curve of the GOD-based glucose microsensor was observed to be wider than that obtained using a disk electrode of 1 mm diameter. The mediated response of the 2 microm glucose sensor was compared with the response resulting from hydrogen peroxide detection. This result showed that a higher response and a wider linear range were observed with highly concentrated mediator. A much higher response of the GDH immobilized 2 microm microelectrode was obtained when not only ferricyanide but also diaphorase was employed to reoxidize the NADH produced by the enzyme reaction of GDH. The GHD-based glucose microsensor was found to be unaffected by the concentration of dissolved oxygen.

The Entner-Doudoroff pathway in Escherichia coli is induced for oxidative glucose metabolism via pyrroloquinoline quinone-dependent glucose dehydrogenase

The Entner-Doudoroff pathway was shown to be induced for oxidative glucose metabolism when Escherichia coli was provided with the periplasmic glucose dehydrogenase cofactor pyrroloquinoline quinone (PQQ). Induction of the Entner-Doudoroff pathway by glucose plus PQQ was established both genetically and biochemically and was shown to occur in glucose transport mutants, as well as in wild-type E. coli. These data complete the body of evidence that proves the existence of a pathway for oxidative glucose metabolism in E. coli. PQQ-dependent oxidative glucose metabolism provides a metabolic branch point in the periplasm; the choices are either oxidation to gluconate followed by induction of the Entner-Doudoroff pathway or phosphotransferase-mediated transport. The oxidative glucose pathway might be important for survival of enteric bacteria in aerobic, low-phosphate, aquatic environments.

Active site similarities of glucose dehydrogenase, glucose oxidase, and glucoamylase probed by deoxygenated substrates

The specificity constants, kcat/KM, were determined for glucose oxidase and glucose dehydrogenase using deoxy-D-glucose derivatives and for glucoamylase using deoxy-D-maltose derivatives as substrates. Transition-state interactions between the substrate intermediates and the enzymes were characterized by the observed kcat/Km values and found to be very similar. The binding energy contributions of individual sugar hydroxyl groups in the enzyme/substrate complexes were calculated using the relationship delta(delta G) = -RT ln [(kcat/KM)deoxy/(kcat/KM)hydroxyl] for the series of analogues. The activity of all three enzymes was found to depend heavily on the 4- and 6-OH groups (4'- and 6'-OH in maltose), where changes in binding energies from 10 to 18 kJ/mol suggested strong hydrogen bonds between the enzymes and these substrate OH groups. The 3-OH (3'-OH in maltose) was involved in weaker interactions, while the 2-OH (2'-OH in maltose) had a very small if any role in transition-state binding. The three enzyme-substrate transition-state interactions were compared using linear free energy relationships (Withers, S. G., & Rupitz, K. (1990) Biochemistry 29, 6405-6409) in which the set of kcat/KM values obtained with substrate analogues for one enzyme is plotted against the corresponding values for a second enzyme. The high linear correlation coefficients (rho) obtained, 0.916, 0.958, and 0.981, indicate significant similarity in transition-state interactions, although the three enzymes lack overall sequence homology.(ABSTRACT TRUNCATED AT 250 WORDS)

Detergent-shock response in enteric bacteria

Our work on bacterial detergent resistance started with the realization that bacteria growing in a sink full of soap must be resistant to the detergents in that soap. We chose sodium dodecyl sulphate (SDS) as a model detergent and decided to see how much SDS the bacterium growing in the sink could tolerate. The research program thus initiated has shown that bacteria such as Enterobacter cloacae can grow in up to 25% SDS and that SDS-shock proteins constitute c. 8% of the proteins synthesized by SDS-grown Escherichia coli. It has also provided explanations why enteric bacteria are oxidase negative, and how pyrroloquinoline quinone (PQQ) enters the periplasmic space. Finally, for E. coli, it has provided evidence for an alternate, phosphate-limited, aquatic life style which places greater emphasis on the Entner-Doudoroff pathway. Detergent resistance is important both medically and ecologically, e.g. entry of pathogens via bile-salt-containing intestinal tracts and biodegradation of detergent-like pollutants such as those resulting from oil spills. Our current research is focused on SDS-induced modifications of the cytoplasmic membrane and the presence of SDS in the periplasm.

GMC oxidoreductases. A newly defined family of homologous proteins with diverse catalytic activities

Sequence comparison of Drosophila melanogaster glucose dehydrogenase, Escherichia coli choline dehydrogenase, Aspergillus niger glucose oxidase and Hansenula polymorpha methanol oxidase indicates that these four diverse flavoproteins are homologous, defining a new family of proteins named the GMC oxidoreductases. These enzymes contain a canonical ADP-binding beta alpha beta-fold close to their amino termini as found in other flavoenzymes. This domain is encoded by a single exon of the D. melanogaster glucose dehydrogenase gene.

Change of the terminal oxidase from cytochrome a1 in shaking cultures to cytochrome o in static cultures of Acetobacter aceti

Acetobacter aceti has an ability to grow under two different culture conditions, on shaking submerged cultures and on static pellicle-forming cultures. The respiratory chains of A. aceti grown on shaking and static cultures were compared, especially with respect to the terminal oxidase. Little difference was detected in several oxidase activities and in cytochrome b and c contents between the respiratory chains of both types of cells. Furthermore, the results obtained here suggested that the respiratory chains consist of primary dehydrogenases, ubiquinone, and terminal ubiquinol oxidase, regardless of the culture conditions. There was a remarkable difference, however, in the terminal oxidase, which is cytochrome a1 in cells in shaking culture but cytochrome o in cells grown statically. Change of the culture condition from shaking to static caused a change in the terminal oxidase from cytochrome a1 to cytochrome o, which is concomitant with an increase of pellicle on the surface of the static culture. In contrast, reappearance of cytochrome a1 in A. aceti was attained only after serial successive shaking cultures of an original static culture; cytochrome a1 predominated after the culture was repeated five times. In the culture of A. aceti, two different types of cells were observed; one forms a rough-surfaced colony, and the other forms a smooth-surfaced colony. Cells of the former type predominated in the static culture, while the cells of the latter type predominated in the shaking culture. Thus, data suggest that a change of the culture conditions, from static to shaking or vice versa, results in a change of the cell type, which may be related to the change in the terminal oxidase from cytochrome a1 to cytochrome o in A. aceti.

Quantitative aspects of glucose metabolism by Escherichia coli B/r, grown in the presence of pyrroloquinoline quinone

Escherichia coli B/r was grown in chemostat cultures under various limitations with glucose as carbon source. Since E. coli only synthesized the glucose dehydrogenase (GDH) apo-enzyme and not the appropriate cofactor, pyrroloquinoline quinone (PQQ), no gluconate production could be observed. However, when cell-saturating amounts of PQQ (nmol to mumol range) were pulsed into steady state glucose-excess cultures of E. coli, the organisms responded with an instantaneous formation of gluconate and an increased oxygen consumption rate. This showed that reconstitution of GDH in situ was possible. Hence, in order to examine the influence on glucose metabolism of an active GDH, E. coli was grown aerobically in chemostat cultures under various limitations in the presence of PQQ. It was found that the presence of PQQ indeed had a sizable effect: at pH 5.5 under phosphate- or sulphate-limited conditions more than 60% of the glucose consumed was converted to gluconate, which resulted in steady state gluconate concentrations up to 80 mmol/l. The specific rate of gluconate production (0.3-7.6 mmol.h-1.(g dry wt cells)-1) was dependent on the growth rate and the nature of the limitation. The production rate of other overflow metabolites such as acetate, pyruvate, and 2-oxoglutarate, was only slightly altered in the presence of PQQ. The fact that the cells were now able to use an active GDH apparently did not affect apo-enzyme synthesis.

A single amino acid substitution changes the substrate specificity of quinoprotein glucose dehydrogenase in Gluconobacter oxydans

Gluconobacter oxydans contains pyrroloquinoline quinone-dependent glucose dehydrogenase (GDH). Two isogenic G. oxydans strains, P1 and P2, which differ in their substrate specificity with respect to oxidation of sugars have been analysed. P1 can oxidize only D-glucose, whereas P2 is also capable of the oxidation of the disaccharide maltose. To investigate the nature of this maltose-oxidizing property we cloned the gene encoding GDH from P2. Expression of P2 gdh in P1 enables the latter strain to oxidize maltose, indicating that a mutation in the P2 gdh gene is responsible for the change in substrate specificity. This mutation could be ascribed to a 1 bp substitution resulting in the replacement of His 787 by Asn.

Preparation and kinetic properties of 5-ethylphenazine-glucose-dehydrogenase-NAD+ conjugate, a semisynthetic glucose oxidase

5-Ethylphenazine-glucose-dehydrogenase-NAD+ conjugate (EP(+)-GlcDH-NAD+) was prepared by linking both poly(ethylene glycol)-bound 5-ethylphenazine and poly(ethylene glycol)-bound NAD+ to glucose dehydrogenase. The average number of the ethylphenazine moieties bound/enzyme subunit was 0.8, and that of the NAD+ moieties was 1.2. This conjugate is a semisynthetic enzyme having glucose oxidase activity using oxygen or 3-(4,5-dimethyl-2-thiazolyl)-2, 5-diphenyl-2H-tetrazolium bromide (MTT) as an electron acceptor. When the concentration of oxygen or MTT is varied, the oxidase activity fits the Michaelis-Menten equation with the following values of the kinetic constants: for the system with oxygen, the turnover number per subunit is 0.40 s-1 and Km for oxygen is 1.57 mM; and for the system with MTT, the turnover number is 0.11 s-1 and Km for MTT is 0.072 mM. The catalytic cycle of the semisynthetic oxidase has two catalytic steps: reduction of the NAD+ moiety by the active site of the glucose dehydrogenase moiety and oxidation of the NADH moiety by another catalytic site of the ethylphenazine moiety. The apparent intramolecular rate constants of these steps were estimated, and the values are as follows: 0.39 s-1 for the reductions of the NAD+ moiety, 2.2 s-1 and 0.12 s-1 for the oxidation of the NADH moiety in the systems with oxygen and with MTT, respectively, and 3.2 s-1 and 0.18 s-1 for the reduction of the ethylphenazine moiety in the systems with oxygen and with MTT, respectively. On the bases of these results, the following three rate-acceleration mechanisms of the semisynthetic glucose oxidase are discussed: high effective concentration, intramolecular coupling of successive catalytic reactions, and multiple connection between the two kinds of the catalytic sites.

Mutants of Escherichia coli producing pyrroloquinoline quinone

In glucose minimal medium a PTS- strain of Escherichia coli [delta (ptsH ptsI crr)] could grow slowly (doubling time, d = 10 h). When the population reached 5 x 10(6) to 2 x 10(7) cells ml-1, mutants growing rapidly (d = 1.5 h) appeared and rapidly outgrew the initial population. These mutants (EF mutants) do not use a constitutive galactose permease for glucose translocation. They synthesize sufficient pyrroloquinoline quinone (PQQ) to yield a specific activity of glucose dehydrogenase (GDH) equivalent to that found in the parent strain grown in glucose minimal medium supplemented with 1 nM-PQQ. Membrane preparations containing an active GDH oxidized glucose to gluconic acid, which was also present in the culture supernatant of EF strains in glucose minimal medium. Glucose utilization is the only phenotypic trait distinguishing EF mutants from the parent strain. Glucose utilization by EF mutants was strictly aerobic as expected from a PQQ-dependent catabolism. The regulation of PQQ production by E. coli is discussed.

Nutritional complementation of oxidative glucose metabolism in Escherichia coli via pyrroloquinoline quinone-dependent glucose dehydrogenase and the Entner-Doudoroff pathway

Two glucose-negative Escherichia coli mutants (ZSC113 and DF214) were unable to grow on glucose as the sole carbon source unless supplemented with pyrroloquinoline quinone (PQQ). PQQ is the cofactor for the periplasmic enzyme glucose dehydrogenase, which converts glucose to gluconate. Aerobically, E. coli ZSC113 grew on glucose plus PQQ with a generation time of 65 min, a generation time about the same as that for wild-type E. coli in a defined glucose-salts medium. Thus, for E. coli ZSC113 the Enter-Doudoroff pathway was fully able to replace the Embden-Meyerhof-Parnas pathway. In the presence of 5% sodium dodecyl sulfate, PQQ no longer acted as a growth factor. Sodium dodecyl sulfate inhibited the formation of gluconate from glucose but not gluconate metabolism. Adaptation to PQQ-dependent growth exhibited long lag periods, except under low-phosphate conditions, in which the PhoE porin would be expressed. We suggest that E. coli has maintained the apoenzyme for glucose dehydrogenase and the Entner-Doudoroff pathway as adaptations to an aerobic, low-phosphate, and low-detergent aquatic environment.

Preliminary studies on quinoprotein glucose dehydrogenase under extreme conditions of temperature and pressure

The kinetics of the reduction of the quinoprotein glucose dehydrogenase by substrate were studied as a function of 3 parameters: pressure (1-1000 bar), temperature (down to -25 degrees C) and solvent (water and 40% dimethyl sulfoxide, DMSO) using a high-pressure low-temperature stopped-flow apparatus. A 2-step formation of the reduced enzyme by its substrate (xylose), was observed. A rapid equilibrium described by the constant K1 was followed by a slower process described by the constants k2 and k-2. By using the transition state theory, the thermodynamic quantities delta V (activation volumes) were determined for these various kinetics constants under different experimental conditions. The results are discussed in terms of conformational change and solvation effect on the protein shell, and compared with results obtained for other systems as the 2-step formation of horseradish peroxidase compound I.

Continuous Regeneration of NAD(H) Covalently Bound to a Cysteine Genetically Engineered into Glucose Dehydrogenase

We introduced a cysteine residue on the surface of glucose dehydrogenase from Bacillus subtilis using site-directed mutagenesis. To this mutant, an NAD-analogue was covalently attached by a disulphide bridge so that it was active intramolecularly. The glucose dehydrogenase-cys44-NAD complex, which contained one reactive NAD molecule per subunit of glucose dehydrogenase, was operated together with lactate dehydrogenase in a coupled enzymatic regeneration of NAD(H) in a hollow fiber reactor. L-lactate and gluconic acid were continuously produced from pyruvate and D-glucose, respectively, with a turnover number of 45 cycles per minute for each NAD molecule. The total turnover per coenzyme was 135,000 for the first 2.5 days.

Glucose dehydrogenase from Bacillus subtilis expressed in Escherichia coli. I: Purification, characterization and comparison with glucose dehydrogenase from Bacillus megaterium

Escherichia coli containing the Bacillus subtilis glucose dehydrogenase gene on a plasmid (prL7) was used to produce the enzyme in high quantities. Gluc-DH-S was purified from the cell extract by (NH4)2SO4-precipitation, ion-exchange chromatography and Triazine-dye chromatography to a specific activity of 375 U/mg. The enzyme was apparently homogenous on SDS-PAGE with a subunit molecular mass of 31.5 kDa. Investigation of Gluc-DH-S was performed for comparison with the corresponding properties of Gluc-DH-M. The limiting Michaelis constant at pH 8.0 for NAD+ is Ka = 0.11 mM and for D-glucose Kb = 8.7 mM. The dissociation constant for NAD+ is Kia = 17.1 mM. Similar to Gluc-DH-M, Gluc-DH-S is inactivated by dissociation under weak alkaline conditions at pH 9.0. Complete reactivation is attained by readjustment to pH 6.5. Ultraviolet absorption, fluorescence and CD-spectra of native Gluc-DH-S, as well as fluorescence- and CD-backbone-spectra of the dissociated enzyme were nearly identical to the corresponding spectra of Gluc-DH-M. The aromatic CD-spectrum of dissociated Gluc-DH-S was different, representing a residual ellipticity of tryptophyl moieties in the 290-310 nm region. Density gradient centrifugation proved that this behaviour is due to the formation of inactive dimers in equilibrium with monomers after dissociation. In comparison to Gluc-DH-M, the kinetics of inactivation as well as the time-dependent change of fluorescence intensity at pH 9.0 of Gluc-DH-S showed a higher velocity and a changed course of the dissociation process.

Direct evidence of the Entner-Doudoroff pathway operating in the metabolism of D-glucosamine in bacteria

Pseudomonas fluorescens (Migula) (IFO 14808) has both a membrane-bound PQQ-dependent D-glucose (D-Glc) dehydrogenase [EC 1.1.99.17] [which also acts on D-glucosamine (D-GlcN)] and a PLP-dependent D-glucosaminate (D-GlcNA) dehydratase [EC 4.2.1.26]. Further, these two enzymes were induced when D-GlcN was added to the culture medium. However, D-glucosamine-6-phosphate (D-GlcN-6-P) isomerase [EC 5.3.1.10], another enzyme involved in the metabolism of D-GlcN, was only present at a low level in this bacterium. The bacterium was able to grow in a minimal medium containing D-GlcN or D-GlcNA as the sole source of carbon and nitrogen. Intact cells of P. fluorescens (Migula) converted D-GlcN to D-GlcNA and then to 2-keto-3-deoxy-D-gluconate (KDGA). These results demonstrate that D-GlcN is metabolized via D-GlcNA to KDGA in P. fluorescens (Migula) (Entner-Doudoroff pathway). In contrast, Enterobacter cloacae(IFO 13535) and Agrobacterium radiobacter (IAM 1526) have significant amounts of D-GlcN-6-P isomerase with low levels of the D-Glc dehydrogenase and D-GlcNA dehydratase. Further, only the isomerase activity was induced on the addition of D-GlcN to the culture medium. These results demonstrate that there is a new route (Entner-Doudoroff pathway), i.e., in addition to the known one (Embden-Meyerhof pathway), for the metabolism of D-GlcN in bacteria and one of the two routes is predominant in the each of bacteria examined.

Glucose dehydrogenase, glucose-6-phosphate dehydrogenase and hexokinase in liver of rainbow trout (Salmo gairdneri). Effects of starvation and temperature variations

1. Activities of trout liver glucose dehydrogenase (GDH, EC 1.1.1.47) and glucose-6-phosphate dehydrogenase (G6PD, EC 1.1.1.49) were increased after a sudden drop in water temperature, but not in long-time cold acclimated as compared with warm acclimated trout. 2. Possibly, the activities of GDH and G6PD were temporarily increased in connection with metabolic adaptation to the lower temperature. 3. The activities of GDH and G6PD were not changed by the stress of handling. 4. Partially purified trout liver GDH has a lower activation energy with glucose than with glucose-6-phosphate as substrate, and the Km (glucose) decreases with decreasing assay temperature. 5. At low temperatures, the activity of trout liver GDH with glucose as substrate may be comparable to that of glucose-6-phosphate. 6. Partially purified beef liver GDH has a high activation energy with glucose as substrate, and the Km (glucose) does not change with the assay temperature. 7. Hexokinase (HK, EC 2.7.1.1) and GDH activities were unchanged when trout were deprived of food for 4 weeks. Apparently, the trout liver glucose utilization did not adapt to the starvation.

Glycolysis and Entner-Doudoroff pathways in Halobacterium halobium: some new observations based on 13C NMR spectroscopy

13C NMR was used to study glucose metabolism in intact cells of Halobacterium halobium. Spectra of glucose grown cells incubated with [1-13C] glucose indicate the presence of gluconate as the initial product. The existence of glycolytic pathway is also indicated. In the extracts of these cells an NADP dependent glucose dehydrogenase was detected. Galactose grown cells failed to metabolise glucose but exhibited glucose dehydrogenase activity although about 20-50% less than that for glucose grown cells. Possible explanations of these experiments are discussed.

The role of futile cycles in the energetics of bacterial growth

In this contribution we describe the occurrence of futile cycles in growing bacteria. These cycles are thought to be active when organisms contain two uptake systems for a particular nutrient (one with a high, the other with a low affinity for its substrate). The high-affinity system is responsible for uptake of the nutrient, some of which is subsequently lost to the medium again via leakage through the low-affinity-system. A special futile cycle is caused under some growth conditions by the extremely rapid diffusion of ammonia through bacterial membranes. When the ammonium ion is taken up via active transport, the couple NH3/NH4+ will act as an uncoupler. This is aggravated by the chemical similarity of the potassium and the ammonium ion, which leads to ammonium ion transport via the Kdp potassium transport system when the potassium concentration in the medium is low. Other examples of futile cycles, such as those caused by the production of fatty acids by fermentation, are briefly discussed.

Evidence for electron transfer via ubiquinone between quinoproteins D-glucose dehydrogenase and alcohol dehydrogenase of Gluconobacter suboxydans

Gluconobacter suboxydans contains membrane-bound D-glucose and alcohol dehydrogenases (GDH and ADH) as the primary dehydrogenases in the respiratory chain. These enzymes are known to be quinoproteins having pyrroloquinoline quinone as the prosthetic group. GDH reduces an artificial electron acceptor, ferricyanide, in the membrane, but not after solubilization with Triton X-100, while ADH can react with the electron acceptor even after solubilization and further purification. In this study, it has been shown that the ferricyanide reductase activity of GDH is restored by adding the supernatant solubilized with Triton X-100 to the residue, and also by incorporation of purified ADH into the membranes of an ADH-deficient strain. G. suboxydans var. alpha. In addition, the ferricyanide reductase activity of GDH was reconstituted in proteoliposomes from GDH, ADH, and ubiquinone-10. Thus, the results indicated that the electron transfer from GDH to ferricyanide was mediated by ubiquinone and ADH. The data also suggest that GDH and ADH transfer electrons mutually via ubiquinone in the respiratory chain.

The role of magnesium and calcium ions in the glucose dehydrogenase activity of Klebsiella pneumoniae NCTC 418

Magnesium-limited chemostat cultures of Klebsiella pneumoniae NCTC 418 with 20 microM CaCl2 in the medium showed a low rate of gluconate plus 2-ketogluconate production relative to potassium- or phosphate-limited cultures. However, when the medium concentration of CaCl2 was increased to 1 mM, the glucose dehydrogenase (GDH) activities also increased and became similar to those observed in potassium- or phosphate limited cultures. It is concluded that this is due to Mg2+ and Ca2+ ions being involved in the binding of pyrroloquinoline quinone (PQQ) to the GDH apoenzyme. There seems to be an absolute requirement of divalent cations for proper enzyme functioning and in this respect Ca2+ ions could replace Mg2+ ions. The high GDH activity which has been found in cells grown under Mg2(+)-limited conditions in the presence of higher concentrations of Ca2+ ions, is compatible with the earlier proposal that GDH functions as an auxiliary energy generating system involved in the maintenance of high transmembrane ion gradients.

Enzymatic properties of isozymes and variants of glucose dehydrogenase from Bacillus megaterium

Three glucose dehydrogenases (GlcDH) from Bacillus megaterium, GlcDH-I, GlcDH-II and GlcDH-IWG3, were purified from Escherichia coli cells harboring one of the hybrid plasmids, pGDK1, pGDK2 and pGDA3, respectively, pGDK1 and pGDK2 contain two isozyme genes, gdhI and gdhII, respectively, from B. megaterium IAM 1030 and pGDA3 contains an isozyme gene from B. megaterium IWG3; GlcDH-IWG3 is a variant of GlcDH-I. GlcDH-I and GlcDH-II have similar pH/activity profiles and the profile for GlcDH-IWG3 is identical to that of GlcDH-I. The pH/stability profiles of these enzymes show that GlcDH-IWG3 is the most stable enzyme in the acidic region, while GlcDH-II is the most stable in the alkaline region, and GlcDH-I is the most unstable throughout the entire pH range examined. As for thermostability, GlcDH-II is the most resistant against heat inactivation at pH 6.5. The values of the first-order rate constant for heat inactivation at 50 degrees C are 0.27 min-1, 0.05 min-1 and 0.11 min-1 for GlcDH-I, GlcDH-II and GlcDH-IWG3, respectively. Kinetic studies show that these enzymes have similar kinetic constant values except that there are some differences in Kia for NAD(P) and Ka (the limiting Michaelis constant) for NAD; the values of the ratio of Kia for NAD and NADP are 11,340 and 8.7 for GlcDH-I, GlcDH-II and GlcDH-IWG3, respectively. GlcDH-I and GlcDH-IWG3 have very similar substrate specificities and GlcDH-II has a slightly higher specificity for D-glucose and 2-deoxy-D-glucose than the others. The results are discussed on the basis of the amino acid substitutions between the enzymes.

Purification and characterization of glucose dehydrogenase from the thermoacidophilic archaebacterium Thermoplasma acidophilum

Glucose dehydrogenase was purified to homogeneity from the thermoacidophilic archaebacterium Thermoplasma acidophilum. The enzyme is a tetramer of polypeptide chain Mr 38,000 +/- 3000, it is catalytically active with both NAD+ and NADP+ cofactors, and it is thermostable and remarkably resistant to a variety of organic solvents. The amino acid composition was determined and compared with those of the glucose dehydrogenases from the archaebacterium Sulfolobus solfataricus and the eubacteria Bacillus subtilis and Bacillus megaterium. The N-terminal amino acid sequence of the Thermoplasma acidophilum enzyme was determined to be: (S/T)-E-Q-K-A-I-V-T-D-A-P-K-G-G-V-K-Y-T-T-I-D-M-P-E.

Quinoprotein D-glucose dehydrogenase of the Acinetobacter calcoaceticus respiratory chain: membrane-bound and soluble forms are different molecular species

Acinetobacter calcoaceticus is known to contain soluble and membrane-bound quinoprotein D-glucose dehydrogenases, while other oxidative bacteria contain the membrane-bound enzyme exclusively. The two forms of glucose dehydrogenase were believed to be the same enzyme or interconvertible forms. Previously, Matsushita et al. [(1988) FEMS Microbiol. Lett 55, 53-58] showed that the two enzymes are different with respect to enzymatic and immunological properties, as well as molecular weight. In the present study, we purified both enzymes and compared their kinetics, reactivity with ubiquinone homologues, and immunological properties in detail. The purified membrane-bound enzyme had a molecular weight of 83,000, while the soluble form was 55,000. The purified enzymes exhibited totally different enzymatic properties, particularly with respect to reactivity toward ubiquinone homologues. The soluble enzyme reacted with short-chain homologues only, whereas the membrane-bound enzyme reacted with long-chain homologues including ubiquinone 9, the native ubiquinone of the A. calcoaceticus. Furthermore, the two enzymes were distinguished immunochemically; the membrane-bound enzyme did not cross-react with antibody raised against the soluble enzyme, nor did the soluble enzyme cross-react with antibody against the membrane-bound enzyme. Thus, each glucose dehydrogenase is a molecularly distinct entity, and the membrane-bound enzyme only is coupled to the respiratory chain via ubiquinone.

Reversible thermal inactivation of the quinoprotein glucose dehydrogenase from Acinetobacter calcoaceticus. Ca2+ ions are necessary for re-activation

The soluble form of the homogeneous quinoprotein glucose dehydrogenase from Acinetobacter calcoaceticus is reversibly inactivated at temperatures above 35 degrees C. An equilibrium is established between active and denatured enzyme, this depending on the protein concentration and the inactivation temperature used. Upon thermal inactivation the enzyme dissociates into the prosthetic group pyrroloquinoline quinone and the apo form of glucose dehydrogenase. After inactivation at 50 degrees C active enzyme is re-formed again at 25 degrees C. Ca2+ ions are necessary for the re-activation process. The velocity of re-activation depends on the protein concentration, the concentration of the prosthetic group pyrroloquinoline quinone and the Ca2+ concentration. The apo form of glucose dehydrogenase can be isolated, and in the presence of pyrroloquinoline quinone and Ca2+ active holoenzyme is formed. Even though native glucose dehydrogenase is not inactivated in the presence of EDTA or trans-1,2-diaminocyclohexane-NNN'NH-tetra-acetic acid, Ca2+ stabilizes the enzyme against thermal inactivation. Two Ca2+ ions are found per subunit of glucose dehydrogenase. The data suggest that pyrroloquinoline quinone is bound at the active site via a Ca2+ bridge. Mn2+ and Cd2+ can replace Ca2+ in the re-activation mixture.

Factors relevant in bacterial pyrroloquinoline quinone production

Quinoprotein content and levels of external pyrroloquinoline quinone (PQQ) were determined for several bacteria under a variety of growth conditions. From these data and those from the literature, a number of factors can be indicated which are relevant for PQQ production. Synthesis of PQQ is only started if synthesis of a quinoprotein occurs, but quinoprotein synthesis does not depend on PQQ synthesis. The presence of quinoprotein substrates is not necessary for quinoprotein and PQQ syntheses. Although the extent of PQQ production was determined by the type of organism and quinoprotein produced, coordination between quinoprotein and PQQ syntheses is loose, since underproduction and overproduction of PQQ with respect to quinoprotein were observed. The results can be interpreted to indicate that quinoprotein synthesis depends on the growth rate whereas PQQ synthesis does not. In that view, the highest PQQ production can be achieved under limiting growth conditions, as was shown indeed by the much higher levels of PQQ produced in fed-batch cultures compared with those produced in batch experiments. The presence of nucleophiles, especially amino acids, in culture media may cause losses of PQQ due to transformation into biologically inactive compounds. Some organisms continued to synthesize PQQ de novo when this cofactor was administered exogenously. Most probably PQQ cannot be taken up by either passive diffusion or active transport mechanisms and is therefore not able to exert feedback regulation on its biosynthesis in these organisms.

Physiological significance and bioenergetic aspects of glucose dehydrogenase

The regulation of the PQQ-linked glucose dehydrogenase in different organisms is reviewed. It is concluded that this enzyme functions as an auxiliary energy-generating mechanism, because it is maximally synthesized under conditions of energy stress. It is now definitively established that the oxidation of glucose to gluconate generates metabolically useful energy. The magnitude of the contribution of the oxidation of glucose to gluconate via this enzyme to the growth yield of organisms such as Acinetobacter calcoaceticus is not yet clear.

Cloning of the genes encoding the two different glucose dehydrogenases from Acinetobacter calcoaceticus

Glucose dehydrogenase (GDH) is a PQQ dependent bacterial enzyme which converts aldoses to their corresponding acids. A. calcoaceticus contains two different PQQ dependent glucose dehydrogenases designated GDH-A which is active in vivo and GDH-B of which only in vitro activity can be shown. We cloned the genes coding for the two GDH enzymes. The DNA sequences of both gdh genes were determined. There is no obvious homology between gdhA and gdhB. Both GDH enzymes oxidize D-glucose in vitro but disaccharides are specific GDH-B substrates and 2-deoxyglucose is specifically oxidized by GDH-A.

Quinoprotein D-glucose dehydrogenases in Acinetobacter calcoaceticus LMD 79.41: purification and characterization of the membrane-bound enzyme distinct from the soluble enzyme

Acinetobacter calcoaceticus is known to contain soluble and membrane-bound quinoprotein D-glucose dehydrogenases while other oxidative bacteria such as Pseudomonas or Gluconobacter contain only membrane-bound enzyme. The two different forms were believed to be the same enzyme or interconvertible. Present results show that the two different forms of glucose dehydrogenase are distinct from each other in their enzymatic and immunological properties as well as in their molecular size. The soluble and membrane-bound glucose dehydrogenases were separated after French press-disruption by repeated ultracentrifugation, and then purified to nearly homogeneous state. The soluble enzyme was a polypeptide of 55 Kdaltons, while the membrane-bound enzyme was a polypeptide of 83 Kdaltons which is mainly monomeric in detergent solution. Both enzymes showed different enzymatic properties including substrate specificity, optimum pH, kinetics for glucose, and reactivity for ubiquinone-homologues. Furthermore, the two enzymes could be distinguished immunochemically; the membrane-bound enzyme is cross-reactive with an antibody raised against membrane-bound enzyme purified from Pseudomonas but not with antibody elicited against the soluble enzyme, while the soluble enzyme is not cross-reactive with the antibody of membrane-bound enzyme. Data also suggest that the membrane-bound enzyme functions by linking to the respiratory chain via ubiquinone though the function of the soluble enzyme remains unclear.