Attachment of CO dehydrogenase to the cytoplasmic membrane is limiting the respiratory rate of Pseudomonas carboxydovorans

Abstract Cells of Pseudomonas carboxydovorans from the exponential growth phase revealed the major portion (87%) of CO dehydrogenase attached to the inner aspect of the cytoplasmic membrane. In stationary cells only about half of the total amount of the enzyme remained membrane-bound, and a drop of the CO-oxidizing activity with O₂ was observed. The CO-oxidizing activity with the unphysiological electron acceptor methylene blue, which does not need any contact of the enzyme with the membrane, always exceeded that with O₂. Measurements of respiration rates of extracts with different electron donors in addition to CO suggested that the electron transport chain is not rate-limiting. It is concluded that the electron flow from CO to O₂ in intact cells of P. carboxydovorans is controlled by the amount of CO dehydrogenase attached to a membrane-bound electron acceptor.     

Removal of ferredoxin:NADP+ oxidoreductase from thylakoid membranes, rebinding to depleted membranes, and identification of the binding site

Ferredoxin-NADP+ oxidoreductase associates with thylakoid membranes into two pools of different binding strength that are experimentally distinguished on the basis of resistance to removal by washes in low ionic strength media. The nondenaturing zwitterionic detergent 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid is uniquely able to remove the more tightly bound pool of enzyme, without solubilization of major membrane proteins. The reconstitution of reductase onto depleted thylakoid membranes requires available membrane binding sites and cations, in order of effectiveness trivalent greater than divalent greater than monovalent. The hetero/bifunctional 125I-iodinated Denny-Jaffe cross-linking reagent yields a 54-kDa, covalently cross-linked adduct between ferredoxin-NADP+ oxidoreductase and a component of the thylakoid membrane. Our results show that the more tightly bound pool of enzyme is associated with the 17.5-kDa reductase-binding protein (Vallejos, R. H., Ceccarelli, E., and Chan, R. (1984) J. Biol. Chem. 259, 8048-8051).     

Extraction of chloroplast light effect mediator(s) and reconstitution of light activation of NADP-linked malate dehydrogenase

Washed thylakoid membranes of pea (Pisum sativum var. Little Marvel), on brief exposure to zwittergent, an amphoteric detergent, lost the property of supporting the light activation of stromal NADP-linked malate dehydrogenase. But, these depleted membranes, on reconstitution with dialyzed, high-speed supernatant of the detergent extract, showed marked light activation of the enzyme when assayed in the presence of 2,6-dichlorophenolindophenol-ascorbate. The component of the high-speed supernatant which is required for light activation is sensitive to sulfite and is heat labile. The analysis of the high-speed supernatant on sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed two prominent polypeptides at approximately 18,000 and 36,000 daltons. The surface-specific, chloroglycoluril-mediated iodination of the washed thylakoid membranes revealed that zwittergent had extracted these two polypeptides. The results reveal that the light effect mediator (LEM) is a surface-exposed, tightly bound protein existing in the thylakoid membranes, and that it can be removed by zwitterionic detergent and used in reconstitution studies.     

Carbon monoxide dehydrogenase inhibitor in cell extracts of Pseudomonas carboxydovorans.

Extracts of heterotrophically grown cells of Pseudomonas carboxydovorans were found to contain an inhibitor of carbon monoxide dehydrogenase (CO-DH). The inhibitor activity was not detected in CO-autotrophically grown cells. The inhibitor was extremely stable to heat treatment based on the extent of inhibition of CO-DH activity. The extent of inhibition was proportional to the amount of cell extract added to the reaction mixture. The inhibition was independent of a prior incubation period of the extracts with CO-DH. The inhibitor was precipitable with ammonium sulfate, phenol, and trichloroacetic acid. It was passed through benzoylated dialysis tubing and Amicon ultrafiltration membrane YM2. Denaturing and nondenturing polyacrylamide gel electrophoresis of CO-DH inactivated by inhibitor revealed that the mobilities of native enzyme and subunits were identical to those of active CO-DH. The inhibitor-treated CO-DH retained its original antigenic sites and exhibited enzyme activity upon activity staining. The CO-DH inhibitor of P. carboxydovorans was also active on CO-DHs from Pseudomonas carboxydohydrogena, Acinetobacter sp. strain JC1, and Pseudomonas carboxydoflava.     

Oxidation of Carbon Monoxide in Cell Extracts of Pseudomonas carboxydovorans

Extracts of aerobically, CO-autotrophically grown cells of Pseudomonas carboxydovorans were shown to catalyze the oxidation of CO to CO(2) in the presence of methylene blue, pyocyanine, thionine, phenazine methosulfate, or toluylene blue under strictly anaerobic conditions. Viologen dyes and NAD(P)(+) were ineffective as electron acceptors. The same extracts catalyzed the oxidation of formate and of hydrogen gas; the spectrum of electron acceptors was identical for the three substrates, CO, formate, and H(2). The CO- and the formate-oxidizing activities were found to be soluble enzymes, whereas hydrogenase was membrane bound exclusively. The rates of oxidation of CO, formate, and H(2) were measured spectrophotometrically following the reduction of methylene blue. The rate of carbon monoxide oxidation followed simple Michaelis-Menten kinetics; the apparent K(m) for CO was 45 muM. The reaction rate was maximal at pH 7.0, and the temperature dependence followed the Arrhenius equation with an activation energy (DeltaH(0)) of 35.9 kJ/mol (8.6 kcal/mol). Neither free formate nor hydrogen gas is an intermediate of the CO oxidation reaction. This conclusion is based on the differential sensitivity of the activities of formate dehydrogenase, hydrogenase, and CO dehydrogenase to heat, hypophosphite, chlorate, cyanide, azide, and fluoride as well as on the failure to trap free formate or hydrogen gas in coupled optical assays. These results support the following equation for CO oxidation in P. carboxydovorans: CO + H(2)O --> CO(2) + 2 H(+) + 2e(-) The CO-oxidizing activity of P. carboxydovorans differed from that of Clostridium pasteurianum by not reducing viologen dyes and by a pH optimum curve that did not show an inflection point.     

Role of a small plasmid in the modification of carbon monoxide dehydrogenase in Pseudomonas carboxydovorans

Abstract Soluble fractions prepared from cells of Pseudomonas carboxydovorans bearing a small plasmid (1.76 × 10⁶) exhibited proteolytic activity on the β-subunit of CO dehydrogenase (CO-DH) in plasmid-cured cells of the same strain, implying that the plasmid carries gene(s) for processing the β subunit of the enzyme at the post-translational level. The protease was found to be a constitutive enzyme. It did not hydrolyze the β subunit of CO-DH in Pseudomonas carboxydohydrogena . Analysis of CO-DH after transformation of the cured cells with the small plasmid confirmed that the plasmid plays a role in the modification of the β subunit of CO-DH in P. carboxydovorans .     

Biochemistry and physiology of aerobic carbon monoxide-utilizing bacteria

Abstract The use of CO as a growth substrate by aerobic CO-oxidizing (carboxydotrophic) bacteria requires some features not obvious in other bacteria. These are the presence of the enzyme CO dehydrogenase, a branched respiratory chain with an alternative CO-insensitive terminal oxidase (cytochrome b ₆₅₃) and formation of reduced pyridine nucleotides by a pmf-driven reversed electron transfer. Immunocytochemical localization studies revealed that CO dehydrogenase is attached to the inner aspect of the cytoplasmic membrane of Pseudomonas carboxydovorans . The enzyme is a molybdo iron-sulfur flavoprotein containing bactopterin as the organic portion of the molybdenum cofactor. Recent findings suggest that this novel pterin is universal to eubacterial molybdenum enzymes, whereas molybdopterin is universal to eukaryotic molybdoenzymes.     

Formation of tightly bound forms of annexins V and VI in rabbit skeletal muscle membranes isolated in the presence of barium ions

After isolation of rabbit skeletal muscle membranes in the presence of Ba2+ or Ca2+, significant portions of annexin V and VI tightly bind to membranes and become inaccessible for chelating agents. Tightly bound annexin VI is virtually completely solubilized only after treatment with a buffer supplemented both with EGTA and detergent. The portion of tightly bound annexin V cannot be removed even by extraction with buffer containing both EGTA and detergent. In some cases, tightly bound annexin V or VI is detected even in the control (not treated with cations) membranes, thus indicating the possible formation of tight annexin--membrane complexes in situ. The addition of exogenous cations seems to promote only the accumulation of tightly bound annexins within the cell. After temperature-induced phase separation, annexin V and VI bound to the membranes isolated in the presence of Ba2+ or Ca2+ remains mainly in the aqueous phase, similarly to annexins isolated from the control membranes. Neither annexin partitions into the detergent-enriched phase. This indicates the absence of hydrophobicity change in comparison with the standard EGTA-soluble annexins.     

D-fructose dehydrogenase of Gluconobacter industrius: purification, characterization, and application to enzymatic microdetermination of D-fructose.

D-Fructose dehydrogenase was solubilized and purified from the membrane fraction of glycerol-grown Gluconobacter industrius IFO 3260 by a procedure involving solubilization of the enzyme with Triton X-100 and subsequent fractionation on diethylaminoethyl-cellulose and hydroxylapatite columns. The purified enzyme was tightly bound to a c-type cytochrome and another peptide existing as a dehydrogenase-cytochrome complex. The purified enzyme was deemed pure by analytical ultracentrifugation as well as by gel filtration on a Sephadex G-200 column. The molecular weight of the enzyme complex was determined to be about 140,000, and sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed the presence of three components having molecular weights of 67,000 (dehydrogenase), 50,800 (cytochrome c), and 19,700 (unknown function). Only D-fructose was readily oxidized by the enzyme in the presence of dyes such as ferricyanide, 2,6-dichlorophenolindophenol, or phenazine methosulfate. Nicotinamide adenine dinucleotide, nicotinamide adenine dinucleotide phosphate, and oxygen did not function as electron acceptors. The optimum pH of D-fructose oxidation was 4.0. The enzyme was stable at pH 4.5 to 6.0 Stability of the purified enzyme was much enhanced by the presence of detergent in the enzyme solution. Removal of detergent from the enzyme solution facilitated the aggregation of the enzyme and caused its inactivation. An apparent Michaelis constant for D-fructose was observed to be 10(-2) M with the purified enzyme. D-Fructose dehydrogenase was shown to be a satisfactory reagent for microdetermination of D-fructose.     

The effect of triton X-100 on bovine brain synaptic membranes

Bovine brain synaptic membranes which were frozen and then extensively washed showed low affinity [³H]muscimol binding. These membranes contained GABA and calmodulin, apparently tightly bound within the membrane fraction. Membranes which were additionally treated with the detergent Triton X-100 showed high affinity [³H]muscimol binding. These membranes did not appear to contain GABA or calmodulin. Transmission electron microscopy studies demonstrated that the washed membrane fraction contained many synaptosomal and vesicular structures. Triton treatment led to the extensive rupture of these structures. These studies explain the well-reported findings of tightly bound GABA and calmodulin in brain membrane fractions, as being due to the entrapment of these compounds inside sealed membrane-bound structures which are still present after a freezethaw and extensive wash treatment, their complete removal requiring Triton-treatment to rupture the vesicles.     

Relationship between carbon monoxide dehydrogenase and a small plasmid in Pseudomonas carboxydovorans

Abstract Isolation of plasmid DNA followed by plasmid curing was carried out to examine the relationship of plasmid to carbon monoxide dehydrogenase (CO-DH) production in carboxydobacteria. A small plasmid of almost identical size (1.52−1.76 × 10⁶) was present in Pseudomonas carboxydovorans, Azotobacter sp.1, and Azomonas sp.2. Azomonas sp.1 contained two kinds of plasmids (1.5 × 10⁶ and 2.47 × 10⁶). No plasmids were found in Pseudomonas carboxydohydrogena , JC1, and HY1. A plasmid-cured clone of P. carboxydovorans was obtained by growing the cells at 37°C. The cured cell was able to grow CO autotrophically on solid, but not in liquid, medium. CO-DH of the cured cell was active and consisted of three subunits similar to those found in the wild-type enzyme, with the exception that the β subunit of the enzyme was larger than that of the wild-type enzyme. These results suggest that the small plasmids do not carry genes encoding CO-DH but may have gene(s) for processing the β subunit of the enzyme.     

Properties of adenylate cyclase solubilized from rat adrenal membranes: effects of ACTH and other stimulators on solubilization.

We have solubilized adenylate cyclase in a relatively stable form from rat adrenal membranes. The solubilized enzyme elutes on a column of Sepharose 4BR as a distinct peak with a higher molecular weight than the soluble fractions which bind 125I-ACTH. Both the soluble and membrane bound enzymes are activated by NaF and Gpp(NH)p, and both have similar affinities for MgATP. While the membrane bound enzyme is activated similarly by either Mg2+ or Mn2+, the soluble enzyme is more fully activated by Mn2+. Pretreatment of adrenal membranes with NaF or Gpp(NH)p before the addition of detergent enhances recovery of soluble enzyme activity, while recovery of activity in the unsolubilized membrane pellet is unchanged. In contrast, addition of ACTH prevents solubilization of the enzyme and greatly increases its recovery in the pellet. This observation is consistent with the theory that action of the hormone on a receptor subunit leads to an association between the receptor and a catalytic subunit. Such an association might make it more difficult to remove the enzyme from the surrounding lipid matrix of the membrane.     

Purification, reconstitution, and steady-state kinetics of the trans-membrane 17 beta-hydroxysteroid dehydrogenase 2

Human membrane 17 beta-hydroxysteroid dehydrogenase 2 is an enzyme essential in the conversion of the highly active 17beta-hydroxysteroids into their inactive keto forms in a variety of tissues. 17 beta-hydroxysteroid dehydrogenase 2 with 6 consecutive histidines at its N terminus was expressed in Sf9 insect cells. This recombinant protein retained its biological activity and facilitated the enzyme purification and provided the most suitable form in our studies. Dodecyl-beta-D-maltoside was found to be the best detergent for the solubilization, purification, and reconstitution of this enzyme. The overexpressed integral membrane protein was purified with a high catalytic activity and a purity of more than 90% by nickel-chelated chromatography. For reconstitution, the purified protein was incorporated into dodecyl-beta-D-maltoside-destabilized liposomes prepared from l-alpha-phosphatidylcholine. The detergent was removed by adsorption onto polystyrene beads. The reconstituted enzyme had much higher stability and catalytic activity (2.6 micromol/min/mg of enzyme protein with estradiol) than the detergent-solubilized and purified protein (0.9 micromol/min/mg of enzyme protein with estradiol). The purified and reconstituted protein (with a 2-kDa His tag) was proved to be a homodimer, and its functional molecular mass was calculated to be 90.4 +/- 1.2 kDa based on glycerol gradient analytical ultracentrifugation and chemical cross-linking study. The kinetic studies demonstrated that 17 beta-hydroxysteroid dehydrogenase 2 was an NAD-preferring dehydrogenase with the K(m) of NAD being 110 +/- 10 microM and that of NADP 9600 +/- 100 microM using estradiol as substrate. The kinetic constants using estradiol, testosterone, dihydrotestosterone, and 20 alpha-dihydroprogesterone as substrates were also determined.     

Glyceraldehyde-3-phosphate dehydrogenase of rat erythrocytes has no membrane component

In the course of studying mammalian erythrocytes we noted prominent differences in the red cells of the rat. Analysis of ghosts by sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis showed that membranes of rat red cells were devoid of band 6 or the glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (D-glyceraldehyde-3-phosphate: NAD+ oxidoreductase (phosphorylating), EC 1.2.1.12). Direct measurements of this enzyme showed that glyceraldehyde-3-phosphate dehydrogenase activity in rat erythrocytes was about 25% of that in human cells; all of the glyceraldehyde-3-phosphate dehydrogenase activity in rat erythrocytes was within the cytoplasm and none was membrane bound; and in the human red cell, about 1/3 of the enzyme activity was within the cytoplasm and 2/3 membrane bound. The release of glyceraldehyde-3-phosphate dehydrogenase from fresh rat erythrocytes immediately following saponin lysis was also determined using the rapid filtration technique recently described. The extrapolated zero-time intercepts of these reactions confirmed that, in the rat erythrocyte, none of the cellular glyceraldehyde-3-phosphate dehydrogenase was membrane bound. Failure of rat glyceraldehyde-3-phosphate dehydrogenase to bind to the membranes of the intact rat erythrocyte seems to be due to cytoplasmic metabolites which interact with the enzyme and render it incapable of binding to the membrane.     

Plasma-membrane-bound malate dehydrogenase activity in maize roots

Summary Plasma membranes were isolated and purified from 14-day-old maize roots (Zea mays L.) by two-phase partitioning at a 6.5% polymer concentration, and compared to isolated mitochondria, microsomes, and soluble fraction. Marker enzyme analysis demonstrated that the plasma membranes were devoid of cytoplasmic, mitochondrial, tonoplast, and endoplasmic-reticulum contaminations. Isolated plasma membranes exhibited malate dehydrogenase activity, catalyzing NADH-dependent reduction of oxaloacetate as well as NAD⁺-dependent malate oxidation. Malate dehydrogenase activity was resistant to osmotic shock, freeze-thaw treatment, and salt washing and stimulated by solubilization with Triton X-100, indicating that the enzyme is tightly bound to the plasma membrane. Malate dehydrogenase activity was highly specific to NAD⁺ and NADH. The enzyme exhibited a high degree of latency in both right-side-out (80%) and inside-out (70%) vesicle preparations. Kinetic and regulatory properties with ATP and P_i, as well as pH dependence of plasma-membrane-bound malate dehydrogenase were different from mitochondrial and soluble malate dehydrogenases. Starch gel electrophoresis revealed a characteristic isozyme form present in the plasma membrane isolate, but not present in the soluble, mitochondrial, and microsomal fractions. The results presented show that purified plasma membranes isolated from maize roots contain a tightly associated malate dehydrogenase, having properties different from mitochondrial and soluble malate dehydrogenases.Abbreviations FCR ferricyanide reductase - MDH malate dehydrogenase     

Characterization of a CO: heterodisulfide oxidoreductase system from acetate-grown Methanosarcina thermophila.

During the methanogenic fermentation of acetate by Methanosarcina thermophila, the CO dehydrogenase complex cleaves acetyl coenzyme A and oxidizes the carbonyl group (or CO) to CO2, followed by electron transfer to coenzyme M (CoM)-S-S-coenzyme B (CoB) and reduction of this heterodisulfide to HS-CoM and HS-CoB (A. P. Clements, R. H. White, and J. G. Ferry, Arch. Microbiol. 159:296-300, 1993). The majority of heterodisulfide reductase activity was present in the soluble protein fraction after French pressure cell lysis. A CO:CoM-S-S-CoB oxidoreductase system from acetate-grown cells was reconstituted with purified CO dehydrogenase enzyme complex, ferredoxin, membranes, and partially purified heterodisulfide reductase. Coenzyme F420 (F420) was not required, and CO:F420 oxidoreductase activity was not detected in cell extracts. The membranes contained cytochrome b that was reduced with CO and oxidized with CoM-S-S-CoB. The results suggest that a novel CoM-S-S-CoB reducing system operates during acetate conversion to CH4 and CO2. In this system, ferredoxin transfers electrons from the CO dehydrogenase complex to membrane-bound electron carriers, including cytochrome b, that are required for electron transfer to the heterodisulfide reductase. The cytochrome b was purified from solubilized membrane proteins in a complex with six other polypeptides. The cytochrome was not reduced when the complex was incubated with H2 or CO, and H2 uptake hydrogenase activity was not detected; however, the addition of CO dehydrogenase enzyme complex and ferredoxin enabled the CO-dependent reduction of cytochrome b.     

Catalytic mechanism and interactions of NAD+ with glyceraldehyde-3-phosphate dehydrogenase: correlation of EPR data and enzymatic studies

Perdeuterated spin label (DSL) analogs of NAD+, with the spin label attached at either the C8 or N6 position of the adenine ring, have been employed in an EPR investigation of models for negative cooperativity binding to tetrameric glyceraldehyde-3-phosphate dehydrogenase and conformational changes of the DSL-NAD+-enzyme complex during the catalytic reaction. C8-DSL-NAD+ and N6-DSL-NAD+ showed 80 and 45% of the activity of the native NAD+, respectively. Therefore, these spin-labeled compounds are very efficacious for investigations of the motional dynamics and catalytic mechanism of this dehydrogenase. Perdeuterated spin labels enhanced spectral sensitivity and resolution thereby enabling the simultaneous detection of spin-labeled NAD+ in three conditions: (1) DSL-NAD+ freely tumbling in the presence of, but not bound to, glyceraldehyde-3-phosphate dehydrogenase, (2) DSL-NAD+ tightly bound to enzyme subunits remote (58 A) from other NAD+ binding sites, and (3) DSL-NAD+ bound to adjacent monomers and exhibiting electron dipolar interactions (8-9 A or 12-13 A, depending on the analog). Determinations of relative amounts of DSL-NAD+ in these three environments and measurements of the binding constants, K1-K4, permitted characterization of the mathematical model describing the negative cooperativity in the binding of four NAD+ to glyceraldehyde-3-phosphate dehydrogenase. For enzyme crystallized from rabbit muscle, EPR results were found to be consistent with the ligand-induced sequential model and inconsistent with the pre-existing asymmetry models. The electron dipolar interaction observed between spin labels bound to two adjacent glyceraldehyde-3-phosphate dehydrogenase monomers (8-9 or 12-13 A) related by the R-axis provided a sensitive probe of conformational changes of the enzyme-DSL-NAD+ complex. When glyceraldehyde-3-phosphate was covalently bound to the active site cysteine-149, an increase in electron dipolar interaction was observed. This increase was consistent with a closer approximation of spin labels produced by steric interactions between the phosphoglyceryl residue and DSL-NAD+. Coenzyme reduction (DSL-NADH) or inactivation of the dehydrogenase by carboxymethylation of the active site cysteine-149 did not produce changes in the dipolar interactions or spatial separation of the spin labels attached to the adenine moiety of the NAD+. However, coenzyme reduction or carboxymethylation did alter the stoichiometry of binding and caused the release of approximately one loosely bound DSL-NAD+ from the enzyme. These findings suggest that ionic charge interactions are important in coenzyme binding at the active site.     

The first step in polyethylene glycol degradation by sphingomonads proceeds via a flavoprotein alcohol dehydrogenase containing flavin adenine dinucleotide.

Several Sphingomonas spp. utilize polyethylene glycols (PEGs) as a sole carbon and energy source, oxidative PEG degradation being initiated by a dye-linked dehydrogenase (PEG-DH) that oxidizes the terminal alcohol groups of the polymer chain. Purification and characterization of PEG-DH from Sphingomonas terrae revealed that the enzyme is membrane bound. The gene encoding this enzyme (pegA) was cloned, sequenced, and expressed in Escherichia coli. The purified recombinant enzyme was vulnerable to aggregation and inactivation, but this could be prevented by addition of detergent. It is as a homodimeric protein with a subunit molecular mass of 58.8 kDa, each subunit containing 1 noncovalently bound flavin adenine dinucleotide but not Fe or Zn. PEG-DH recognizes a broad variety of primary aliphatic and aromatic alcohols as substrates. Comparison with known sequences revealed that PEG-DH belongs to the group of glucose-methanol-choline (GMC) flavoprotein oxidoreductases and that it is a novel type of flavoprotein alcohol dehydrogenase related (percent identical amino acids) to other, so far uncharacterized bacterial, membrane-bound, dye-linked dehydrogenases: alcohol dehydrogenase from Pseudomonas oleovorans (46%); choline dehydrogenase from E. coli (40%); L-sorbose dehydrogenase from Gluconobacter oxydans (38%); and 4-nitrobenzyl alcohol dehydrogenase from a Pseudomonas species (35%).     

Purification of the 45 kDa, membrane bound NADH dehydrogenase of Escherichia coli (NDH-2) and analysis of its interaction with ubiquinone analogues

The NADH:ubiquinone reductase (NDH-2) of Escherichia coli was expressed as a His-tagged protein, extracted from the membrane fraction using detergent and purified by chromatography. The His-tagged NDH-2 was highly active and catalyzed NADH oxidation by ubiquinone-1 at rates over two orders of magnitude higher than previously reported. The purified, His-tagged NDH-2, like native NDH-2, did not oxidize deamino-NADH. Steady-state kinetics were used to analyze the enzyme's activity in the presence of different electron acceptors. High V(max) and low K(m) values were only found for hydrophobic ubiquinone analogues, particularly ubiquinone-2. These findings strongly support the notion that NDH-2 is a membrane bound enzyme, despite the absence of predicted transmembrane segments in its primary structure. The latter observation is in agreement with possible evolutionary relation between NDH-2 and water-soluble enzymes such as dihydrolipoamide dehydrogenase. There is currently no clear indication of how NDH-2 binds to biological membranes.