Inactivation of the Na + -translocating NADH:ubiquinone oxidoreductase from Vibrio alginolyticus by reactive oxygen species Julia Steuber 1 , Miche ` le Rufibach 1 ,Gu¨ nter Fritz 2 , Frank Neese 3, * and Peter Dimroth 1 1 Mikrobiologisches Institut der Eidgeno ¨ ssischen Technischen Hochschule, ETH-Zentrum, Zu ¨ rich, Switzerland; 2 Biochemisches Institut, Universita ¨ tZu ¨ rich, Switzerland; 3 Mathematisch-naturwissenschaftliche Sektion, Fachbereich Biologie, Universita ¨ t Konstanz, Germany The Na + -translocating NADH:quinone oxidoreductase (Na + -NQR) from Vibrio alginolyticus was inactivated by reactive oxygen species. Highest Na + -NQR activity was observed i n anaerobically pre pared membranes that exhi- bited 1 : 1 coupling of NADH oxidation and Q reduction activities (1.6 UÆmg )1 ). Optical a nd EPR s pectroscopy documented the presence of b-type cytochromes, a [2Fe)2S] cluster and an organic radical s ignal in anaerobically pre- pared membranes from V. alginolyticus. It is shown that the [2Fe)2S] cluster previously assigned to the Na + -NQR ori- ginates from the succinate dehydrogenase or the related enzyme fumarate reductase. Keywords: electrochemical sodium gradient; reactive oxygen species; Vibrio;Na + transport; NADH: quinone oxido- reductase. The marine bacterium Vibrio alginolyticus possesses a Na + -translocating NADH:Q oxidoreductase (Na + - NQR) that maintains an electroc hemical sodium gradient required for nutrient uptake and motility [ 1–5]. This respiratory sodium pump contains one Fe–S cluster, noncovalently bound FAD, covalently bound FMN and ubiquinone-8 as redox cofactors [2,6–8] that are likely to participate in N ADH oxidation, e lectron transfer, Na + transport and Q reduction. Sequence analysis of the six subunits of the Na + -NQR from V. alginolyticus (NqrA– NqrF) showed that only the NqrF (or b-) subunit comprises a cysteine motif (Cys69, Cys75, Cys78 and Cys110) required f or ligation of one Fe–S cluster [3,4]. Tightly bound ubisemiquinones were proposed to play a central role during the redox-linked Na + transport by the Na + -NQR [6,9,10]. The translocation of Na + through the protein matrix is thought to be dependent on ion -pair formation of a positively charged Na + cation with a negatively charged ubisemiquinone anion generated in a medium of low dielectric [4,10]. This hypothesis was supported b y t he detection of an NADH-induced radical signal in aerobic samples of the Na + -NQR from V. alginolyticus by EPR spectroscopy [6]. However, th e high concentration of the radical (13 l M ) compared to the [2Fe)2S] cluster (3 l M ), together with the observat ion that the Na + -NQR produced superoxide radicals in the presence of NADH and O 2 , strongly suggested that the overstoichiometric formation of radicals was due to oxidative damage of the enzyme. The identification of the individual redox cofactors of coupled Na + -NQR is a prerequisite to understand the mechanism of r edox-linked Na + transport. Here we show that the Na + -NQR is already inactivated in the membrane-bound state if the cells are disrupted in the presence of dioxygen, resulting in an uncoupling of N ADH oxidation and Q reduction activities. It is also demonstrated that the [2Fe)2S] cluster previously assigned to the Na + -NQR [6] originates from the succinate dehydrogenase or the related fumarate reductase. MATERIALS AND METHODS Preparation of membrane vesicles for EPR spectroscopy V. alginolyticus (DSM 2171 T , Braunschweig, Germany) was grown aerobically in a 300-L fermenter a s described previously [6]. Frozen cells (20 g) were washed with anaerobic e xtraction buffer (10 m M Hepes/KOH, pH 7.5, 5m M Mg 2 SO 4 ,0.2 M K 2 SO 4 ) a nd resuspended in the presence of a trace of DNAse (2–4 mL extraction buffer per g cells). The cells were broken by a single passage through a French press at 80 MPa. The crude extract eluting from the French press was collected under a stream of N 2 .All subsequent manipulations were carried out in the anaerobic chamber under a n atmosphere of N 2 /H 2 (95/5%). 50 m M EDTA (potassium salt) was added to the crude extract, and the membrane vesicles were sedimented by ultracentrifuga- tion. Membranes were washed once in extraction buffer containing 10 m M EDTA. Subsequently, complexed Mn 2+ was removed by a second and a third washing step with 10 m M Hepes/KOH, pH 7.5 containing 0.2 M K 2 SO 4 . EDTA treatment did not diminish the Na + -NQR activity of the membrane vesicles that were stimulated by Na + Correspondence to J. Steuber, Mikrobiologisches Institut der Eidgeno ¨ ssischen Technischen Hochschule, ETH-Zentrum, Schmelzbergstr. 7, CH-8092 Zu ¨ rich, Switzerland. Fax: + 41 1 6321148, Tel.: + 41 1 6323830, E-mail: fritz-steuber@micro.biol.ethz.ch Abbreviations: Q1, ubiquinone-1; Na + -NQR, Na + -translocating NADH:ubiquinone oxidoreductase; TTFA, 2-thenoyltrifluoro- acetone; MTT, 3-(4,5-dimethyl 2-thiazolyl) 2,5-diphenyl tetrazolium bromide. *Present address: Max-Planck Institut fu ¨ r Strahlenchemie, D-45470 Mu ¨ lheim an der Ruhr, Germany. Dedication: Dedicated to Achim Kro ¨ ger on the occasion of his 65th b irthday. Note: A web site is available at http:// www.micro.biol.ethz.ch (Received 7 November 2001, accepted 8 January 2002) Eur. J. Biochem. 269, 1287–1292 (2002) Ó FEBS 2002 ions by a f actor of t hree (specific Q reductase activity in 100 m M KCl, 0.6 lmolÆmin )1 Æmg )1 ;in100m M NaCl, 2.1 lmolÆmin )1 Æmg )1 ). The membrane v esicles were resus- pended in extraction buffer to a final protein concentration of 42 mg mL )1 . Membrane v esicles (0.3 mL) were mixed with substrates (NADH, succinate, fumarate in H 2 O, potassium salts) or inhibitor (thenoyltrifluoroacetone in ethanol) as indicated and transferred to EPR tubes in t he anaerobic chamber. Analytical methods and spectroscopy Na + was determined by atomic absorption spectroscopy with a Shimadzu AA-646 spectrometer. Protein was deter- mined b y the bicinchoninic acid me thod [11]. If not mentioned otherwise, NADH:Q oxidoreductase assays [8] were performed at 25 °C in sealed cuvettes filled in the anaerobic chamber and flushed with N 2 .Inorderto investigate the influence of dioxygen and NADH on the activity of the Na + -NQR, membrane vesicles were prepared (A) in the presence of dioxygen without additions (B) with 300 UÆmL )1 superoxide dismutase and 40 UÆmL )1 catalase (C) w ith 10 m M pyruvate and 20 UÆmL )1 L -lacate dehy- drogenase (D) under exclusion of dioxygen without further additions. Succinate dehydrogenase activity was followed by the reduction of 3-(4,5-dimethyl 2-thiazolyl) 2,5-diphenyl tetrazolium bromide (MTT) at 560 nm (e 560 ¼ 12.1 m M )1 Æcm )1 ) in the presence of phenazine methosulfate [12]. Visible spectra of membranes were recorded on a Shimadzu UV-3000 spectrophotometer in the difference spectrum mode. X-band EPR spectra were obtained with a Bruker ESP300 spectrometer with peripheral equipment and data handling as previously described [13]. The spectra were simulated with the program EPR [14]. Quantification of the EPR signals was carried out by comparison with a Cu 2+ standard [13]. RESULTS AND DISCUSSION Inhibition and uncoupling of the Na + -NQR activity by dioxygen In the presence of NADH and dioxygen, the purified Na + - NQR from V. alginolyticus was inactivated with a half time of approximately 3 min. This decrease in enzymatic activity correlated with the formation of superoxide radicals, indicating a superoxide-mediated destruction of the enzyme [6]. We reasoned that the Na + -NQR might be already inactivated during the aerobic disruption of the cells due to the presence of NADH i n the crude extract. In order to test this hypothesis, French press cell rupture and preparation of membrane vesicles of V. alginolyticus cells were performed under exclusion of dioxygen. In addition, the effect of enzymes that detoxify reactive oxygen species (superoxide dismutase and catalase) on the Na + -NQR activity during aerobic cell rupture was investigated. In another experiment, the concentration of NADH in the aerobically prepared crude extract was diminished by the action o f l actate dehydrogenase plus p yruvate. Both the NADH dehydro- genase and the ubiquinone-1 (Q1) reduction activities of membranes were followed in order to determine the degree of uncoupling of the enzyme. The term ÔuncouplingÕ describes the observation that the electrons derived from the oxidation of NADH by the Na + -NQR are not comp- letely transferred to the substrate, Q1. Instead, dioxygen seems to act as electron acceptor at an unspecified site of the enzyme to yield superoxide radicals. The highest specific Q reductase activi ty was foun d in anaerobically prepared membranes t hat exhibited 1 : 1 coupling of NADH dehy- drogenase and Q r eduction activities (1.6 lmolÆminÆmg )1 ; Table 1 ). The lowest Q r eductase activity was observed in membranes p repared aerobically without any a dditions (0.6 lmolÆmin )1 Æmg )1 ). Removal of superoxide or NADH in aerobically prepare d crude extracts and membranes resulted in increased Q reductase activities, but the uncou p- ling o f N ADH oxidation from Q reduction could not be prevented by these measures (Table 1). It is concluded that the Na + -NQR is already inactivated in the membrane- bound state if the cells are disrupted in the presence of dioxygen, resulting in an uncoupling of NADH oxidation and Q reduction activities. The uncoupling of the Na + -NQR activity in membranes by aerobic cell rupture was not restricted to the French press procedure, but was also observed with the more gentle osmotic shock protocol described by Tokuda & Unemoto, with NADH oxidation and Q reduction activities of 2.45 and 0.99 lmolÆmin )1 Æmg )1 , respectively [15]. Characterization of redox cofactors in membranes from V. alginolyticus Based on the results of Table 1, the Na + -NQR is partially inactivated during cell rupture in the presence of dioxygen. Superoxide radicals that are generated during the oxidation of reduced electron carriers by O 2 apparently cause this inhibition. For example, the fumarate reductase from Escherichia coli produces superoxide in air [16]. Optical and EPR spectroscopic investigations were performed to Table 1. Inhibition and uncoupling of the Na + -NQR activities in membranes from V. alginolyticus by dioxygen and NADH. The enzymic activities of membranes were determined a s described in Materials and methods with Q1 as electron acceptor in the presence of NaCl. With membranes prepared un der aerobic conditions, the assay was performed aerobically. With anaerobically prepared membranes, the activity was determined under exclusion of air. ND, not determined. Conditions of cell rupture NADH oxidation (lmolÆmin )1 Æmg )1 ) Q1 reduction (lmolÆmin )1 Æmg )1 ) Ratio of NADH oxidation/ Q1 reduction activity +O 2 1.7 0.6 2.8 +O 2 /superoxide dismutase/catalase 3.2 1.2 2.7 +O 2 /lactate dehydrogenase/pyruvate (4.7) a 0.9 ND –O 2 1.6 1.6 1.0 a The high NADH oxidation activity is due to the presence of residual L -lactate dehydrogenase and pyruvate. 1288 J. Steuber et al. (Eur. J. Biochem. 269) Ó FEBS 2002 characterize the redox enzymes present in V. alginolyticus membranes and to d etermine their redox state. The optical difference spectrum of washed membranes from V. algino- lyticus prepared under exclusion of dioxygen (as isolated minus air-oxidized, Fig. 1) revealed the presence of reduced b-type cytochromes with absorption maxima at 526, 531, 557 and 5 61 nm. No further reduction of the b-t ype cytochromes upon addition of dithionite was achieved. In addition, reduction of air-oxidized membranes with NADH or dithionite did not increase the amount of reduced b-type cytochromes compared to membranes prepared under exclusion of dioxygen (not shown). From their characteristic a-andb-absorption maxima, the redox complexes succinate dehydrogenase, or fumarate reductase (557 nm and 526 nm) [17] and the cytochrome bo-type ubiquinol oxidase (561 nm and 531 nm) [18] were assigned. Based on the extinction coefficient of the succinate d ehydrogenase from E. coli [12], the amount of b-type cytochromes associated with succinate dehydrogenase or fumarate redu ctase in V. alginolyticus membranes was 1.3 lmolÆmg protein )1 .The succinate dehydrogenase activity of membranes was 0.2 lmolÆmin )1 Æmg )1 . The presence of a succinate dehy- drogenase or fumarate reductase in membranes from V. alginolyticus was also c onfirmed b y E PR spectroscopy showing the typical resonances of partially reduced center I, the [2Fe)2S] cluster of succinate dehyd rogenase, o r fuma- rate reductase, with g x,y,z ¼ 1.92, 1.93, 2.03 (Fig. 2, traces A and B). N o further r eduction was achieved upon the addition of succinate (Fig. 2, trace B). Furthermore, an organic radical signal at g ¼ 2.005 was detected. Upon addition of NADH to membranes prepared under exclusion of dioxygen, there was a twofold increase in intensity of a near-axial signal with resonances at g x,y ¼ 1.92, 1.93 (Fig. 2, trace C). Addition of 80 m M Na + did not alter the NADH- induced EPR signals measured at 40 K and 2 mW (Fig. 2, trace D ). Note that the residual Na + concentration of washed membrane vesicles was 3–4 m M . Addition of Na + had no obvious effect on the NADH-induced radical s ignal measured at 70 K and 4 lW ( not shown). No additional signals due to [4Fe)4S] clusters were observed at 4 K (see below). These results are very similar to those reported f or membranes from V. alginolyticus prepared in air [6]. In this study, the NADH-induced signal at g % 1.93 was assigned to the [2Fe)2S] cluster of the Na + -NQR, and the strong radical signal to ubisemiquinone bound to the Na + -NQR. However, addition of NADH in the presence of thenoyltri- fluoroacetone (TTFA), a specific inhibitor of succinate dehydrogenase and fumarate reductase, prevented the increase in signal intensity in the g % 1.93 region (Fig. 3). This result clearly d emonstrates that the NADH-induced g % 1.93 EPR s ignal observed by Pfenninger-Li et al.[6] was due to the reduced center I, the [2Fe)2S] cluster in succinate dehydrogenase, or in fumarate reductase. We also investigated the EPR spectrum of the g ¼ 2.005 signal at 70 K and 4 lW in the presence of TTFA and fumarate. Under these conditions, the addition of NADH did not lead to any detectable effect (not shown), suggesting that the radical signal is not due to ubisemiquinone bound to the Na + -NQR, as p roposed previously [6]. The radical might be associated with cofactors of other respiratory complexes, such as t he flavosemiquinone in succinate dehydrogenase or fumarate reductase [17]. TTFA blocks the electron t ransfer between the Fe–S centers of succinate dehydrogenase, or fumarate reductase, and the quinone pool. Both complexes exhibit a lmost identical EPR spec- troscopic properties [17]. As V. alginolyticus is capable of anaerobic growth, the fumarate reductase is likely to be Fig. 1. Optical difference spectrum of membranes from V. alginolyticus. The spectrum shows the difference in absorbance of membranes pre- pared under exclusion of dio xygen (Ôas iso late dÕ) minus the air-oxidized membranes. The protein concentration was 10 mgÆmL )1 . Fig. 2. Electron paramagnetic resonance spectra of membranes from V. alginolyticus. Trace A sho ws the EPR spectrum of anaerobically prepared membranes, as isolated. Trace B, membranes plus 40 m M succinate, 80 m M Na + , 360 m M K + trace C , membranes plu s 8 m M NADH, 4 m M Na + ,450m M K + ; trace D, membranes plus 8m M NADH, 80 m M Na + ,380m M K + ; (final concentrations). The membranes were incubated with substrates for 4–5 min prior to freezing in liquid N 2 . The protein c oncentration was 4 2 mgÆmL )1 .EPR parameters: microwave frequency, 9.652 GHz; microwave power, 2 mW; modulation amplitude, 1 mT; temperature 40 K. Ó FEBS 2002 Na + -translocating NQR from V. alginolyticus (Eur. J. Biochem. 269) 1289 present under limiting O 2 concentrations at high cell densities in batch culture. Addition of NADH to mem- branes shifts the redox potential towards more negative values and increases the concentration of ubiquinol compared to succinate-reduce d membranes. Higher quinol concentrations allow the reduction of fumarate reductase that unlike succinate dehydrogenase c annot be completely reduced by succinate [17]. As a consequence, an increase in signal intensity of the reduced center I of fumarate reduc- tase in NADH- compared to succinate-reduced membranes might occur, as shown in F ig. 2 . By s ubtracting the EPR spectrum of succinate-treated membranes from the spec- trum obtained in the presence of NADH, center I of the fumarate reductase (or succinate dehydrogenase) was erro- neously assigned to the Na + -NQR [6]. The amount of center I of succinate dehydrogenase, or fumarate red uctase, in membranes from V. alginolyticus was e stimated by simulation o f the EPR spectrum of NADH-treated membranes using the parameters g x,y,z ¼ 1.92, 1.935, 2.029, width (x,y,z) ¼ 1.2, 1.0, 1.2 mT (Fig. 4). These parameters are based on the EPR spectro- scopic properties of center I, which are remarkably similar in all succinate dehydrogenases and fumarate reductases [17]. The amount of center I (1.0 lmolÆmg protein )1 )in membra nes from V. alginolyticus determined by EPR spectroscopy compares favorably with the amount of b- type cytochromes associated with succinate dehydrogenase, or fum arate reductase, as determined by optical spectros- copy (1.3 lmolÆmg )1 ). Detection of succinate dehydrogenase or fumarate reductase in the partially purified Na + -NQR The NADH-induced g % 1.93 EPR signal that was erro- neously assigned to the membrane-bound Na + -NQR was also observed in the partially purified Na + -NQR obtained by anionic exchange chromatography [6]. This g % 1.93 EPR signal found in the enriched Na + -NQR also origin- ated from reduced center I of succinate dehydrogenase, or fumarate reductase. The presence of fumarate reductase in analiquotoftheNa + -NQR preparation analysed by EPR spectroscopy [6] was confirmed b y immunostaining u sing antiserum raised against the 66-kDa flavoprotein subunit (FrdA) of fumarate reductase from E. coli [17]. Figure 5 shows a Western Blot of the partially purified Na + -NQR analysed by EPR spectroscopy and membranes from E. coli grown on glycerol and fumarate. The antiserum raised against E. coli FrdA detected a polypeptide with an apparent molecular mass of 66 kDa in the Na + -NQR and in the E. coli membranes. A second cross-reactive polypeptide with lower molecular mass in the E. coli membranes probably represents a proteolytic fragmen t of the FrdA subunit [19]. The stoichiometry and localization of the flavin cofactors of the Na + -NQR are a matter of debate [7,20]. A contamination of t he Na + -NQR with FrdA that contains a covalently bound FAD [8a-N(3)-histidyl-FAD] will result in an overestimation of covalently bound flavin in partially purified Na + -NQR preparations. A comparison of the EPR spectra of the Na + -NQR from different purification stages further supports the assumption that the NADH dehydrogenase activity and the Fe–S cluster are properties of two distinct proteins, the Na + -NQR and the fumarate reductase or related succi- nate dehydrogenase. The oxidation of NADH is catalysed by the NqrF subunit o f the Na + -NQR. This subu nit contains a FAD cofactor and a Cys-(X) 5 -Cys-(X) 2 -Cys motif that ligates a Fe–S cluster. N o additional cysteine- rich motifs that i ndicate the presence of F e–S clusters are found on the remaining Nqr subunits, NqrA–NqrE [4]. An increase of the specific NADH d ehydrogenase ac tivity upon purification of the Na + -NQR is therefore expected to be accompanied by an increase in the amount of Fig. 3. Electron paramagnetic resonance spectra of membranes from V. alginolyticus in the presence of TTF A. TTF A is a specific i nhib itor o f succinate dehydrogenase, or fumarate reductase, that prevents the complete reduction of center I. Upper trace, membranes plus 40 m M fumarate, 2 m M TTFA; lower trace, membranes p lus 40 m M fumarate, 2m M TTFA, 8 m M NADH (final concentrations). The protein concentration was 44 mg mL )1 . EPR parameters as in Fig. 2. Fig. 4. Experimental (upper trace) and simulated (lower trace) electron paramagnetic resonance spectrum of center I in NADH-reduced mem- branes from V. alginolyticus. EPR parameters: microwave frequency, 9.652 GH z; microwave power, 0.126 m W; modulation amplitude, 1 mT; temperature 4 K. Simulation parameters: g x,y,z ¼ 1.92, 1.935, 2.029, width (x,y,z) ¼ 1.2, 1.0, 1.2 mT. 1290 J. Steuber et al. (Eur. J. Biochem. 269) Ó FEBS 2002 [2Fe)2S] cluster. However, purification of the Na + -NQR led to a loss of the Fe–S EPR signal (Table 2). In the Na + -NQR obtained after gel filtration that exhibited the highest NADH dehydrogenase activity, no Fe–S cluster could be detected by EPR spectroscopy. We monitored the s uccinate d ehydrogenase activity in fractions enriched in NADH:quinone oxidoreductase during the purification of the Na + -NQR according t o [6]. In p arallel with a decrease in signal intensity o f the Fe–S cluster, there was a decrease in the specific succinate dehydrogenase activity (Table 2). No succinate dehydrogenase activity was observed in t he Na + -NQR purified by gel filtration. These data clearly show that the Fe–S cluster observed in fractions enriched in Na + -NQR [6] is identical to center I of fumarate reductase (or s uccinate dehydrogenase). CONCLUSIONS To our knowledge, this study demonstrates for the first time that a purification procedure under strict exclusion of dioxygen will be an absolute prerequisite to obtain a realistic picture of the coupling o f NADH:Q oxidoreduction to Na + transport by the Na + -NQR from V. alginolyticus. The succinate dehydrogenase and/or fumarate reductase are major respiratory complexes present in V. alginolyticus membranes that impede the d etection of the Fe–S cluster of the Na + -NQR in the membrane-bound state by EPR spectroscopy. ACKNOWLEDGEMENTS This work was supported by a grant from the commission of research, ETH, to J. S. We t hank S. Albracht, University of Amsterdam, for valuable discussions, and P. Kroneck, for using the EPR facilities at the University of Konstanz. Antiserum against the E. coli fumarate reductase flavoprotein subunit was a generous gift from G. Cecchini, VA Medical Center, San Francisco. REFERENCES 1. Tokuda, H. (1993) The Na + cycle in Vibrio alginolyticus.InAlkali Cation Transport Systems in Prokaryotes (Bakker, E.P., ed.), pp. 125–138. CRC Press, Boca Raton. 2. Unemoto, T. & Hayashi, M. (1993) Na + -translocat in g N ADH- quinone reductase of marine and halophilic bacteria. J. Bioenerg. Biomembr. 25, 385–391. 3. Rich, P.R., Meunier, B. & Ward, F.B. (1995) Predicted structure and possible ionmotive mechanism of the sodium-linked NAD- ubiquinone oxidoreductase of Vibrio alginolyticus. FEBS Lett . 375, 5–10. 4. Steuber, J. (2001) Na + translocation by bacterial NADH:quinone oxidoreductases: an extension to the complex I-family of primary redox pumps. Biochim. Biophys. Acta 1505, 45–56. 5. Nakayama, Y., Hayashi, M. & Unemoto, T. (1998) Identification of six subunits constituting Na + -translocating NADH-quinone reductase from the marine Vibrio alginolyticus. FEB S Lett. 422, 240–242. Fig. 5. Detection of fumarate reductase in the Na + -NQR analysed by EPR spectroscopy. The arrow indicates the FrdA flavoprotein subunit of fumarate reductase present in E. coli membranes (10 lg) and in the Na + -NQR (10 lg) after Q-Sepharose chromatography. Table 2. Succinate dehydrogenase activity and amount of reduced [2Fe)2S] cluster (center I) detected in the Na + -NQR at different p urifica tion stage s. The Na + -NQR was purified according to [6]. The nearly axial g % 1.93 EPR signal of a reduced [2Fe)2S] c luster (center I of succinate dehydrogenase or fumarate reductase) p resent in the Na + -NQR was quantified as described previously [6]. ND, not detected. Purification step NADH dehydrogenase activity (lmolÆmin )1 Æmg )1 ) Succinate dehydrogenase activity (lmolÆmin )1 Æmg )1 ) Amount of reduced [2Fe)2S] (l M spin concentration) Membranes 1.2 0.2 40 DEAE Sephacel 14 0.05 2.8 Q Sepharose 88 0.02 1.4 Superdex 200 120 0 ND Ó FEBS 2002 Na + -translocating NQR from V. alginolyticus (Eur. J. Biochem. 269) 1291 6. Pfenninger-Li, X.D., Albracht, S.P.J., Belzen, R.V. & Dimroth, P. (1996) The NADH:ubiquinone oxidoreductase of Vibrio algino- lytic us: p urificatio n, properties and recon stitution of the N a + pump. Biochemistry 35, 6233–6242. 7. Nakayama, Y., Yasui, M., Sugahara, K., Hayashi, M. & Unemoto, T . (2000) Covalently b ound flavin i n the NqrB an d NqrC subunits of Na + -translocating NADH-quinone reductase from Vibrio alginolyticus. FEBS Lett. 474, 165–168. 8. Steuber, J., Krebs, W. & Dimroth, P. (1997) The Na + -trans- locating NAD H:ubiquinon e o xidoreductase from Vibrio algino- lytic us: redox states of the FAD prosthetic group and mechanism of Ag + inhibition. Eur. J. Biochem. 249, 770–776. 9. Hayashi, M. & Unemoto, T. (1984) Characterization of the Na + -dependent respiratory chain NADH:quinone oxidoreduc- tase of the marine bacterium, Vibrio alginolyticus,inrelationtothe primary Na + pump. Biochim. Biophys. Acta 767, 470–478. 10. Dimroth, P. (1997) Primary so dium io n translocating enzym es. Biochim. Biophys. A cta 1318, 11–51. 11. Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gartner, F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson, B.J. & Klenk, D.C. (1985) Measurement of protein using bicinchoninic acid. Anal. Biochem. 150, 76–85. 12. Kita, K., Vibat, C.R.T., Meinhardt, S., Guest, J.R. & Gennis, R.B. (1989) One-step purification from Escherichia coli of complex II (succinate: ubiquinone oxidoreductase) associated with succi- nate-reducible cytochrome b 556 . J. Biol. Chem. 264, 2672–2677. 13. Neese, F., Zumft, W.G., Antho line, W.E. & K roneck , P.M .H. (1996) The purple mixed-valence Cu A center in nitrous oxide reductase: EPR of the copper-63-, copper-65-, and both copper-65 and [ N-15 ÆÆ ] h istidine-enriched e nzyme and a m olecul ar orbital interpretation. J. Am. Chem. Soc. 118, 8692–8699. 14. Neese, F. (1995) The program EPR. Quant. Chem. Progr. Exch. Bull. 136,5. 15. Tokuda, H. & Unemoto, T. (1984) Na + is translocated at NADH:quinone oxidoreductase segment in the respiratory ch ain of Vibrio alginolyticus. J. Biol. Chem. 259, 7785–7790. 16. Imlay, J.A. (1995) A metabolic enzyme that rapidly produces superoxide, fumarate reductase of Escherichia coli. J. Biol. Chem. 270, 19767–19777. 17. Ackrell, B.A.C., Johnson, M.K., Gunsalus, R.P. & Cecchini, G. (1992) Structure a nd function of succinate dehydrogenase and fumarate reductase. Chemistry and Biochemistry of Flavoenzymes (Mu ¨ ller, F., ed.), pp. 229–297. CRC Press, Boca Raton. 18. Miyoshi-Akiyama, T., Hayashi, M. & U nemoto, T. (1993) Purification and properties of cytochrome bo-type ub iquinol oxidase from a marine bacterium Vibrio alginolyticus. Biochim. Biophys. Acta 1141, 283–287. 19. Luna-Chavez, C., Iverson, T.M., Rees, D.C. & C ecchini, G. (2000) Overexpression, purification, and crystallization of the membrane- bound fumarate reductase from Escherichia coli. Prot. Expres. Purif. 19, 188–196. 20. Zhou, W., Bertsova, Y.V., Feng, B., Tsatsos, P., Verkhovskaya, M.L., Gennis, R.B., Bogachev, A.V. & Barquera, B. (1999) Sequencing and preliminary characterizatio n of the N a + -trans- locating NADH:ubiquinone oxidoreductase from Vibrio harveyi. Biochemistry 38, 16246–16252. 1292 J. Steuber et al. (Eur. J. Biochem. 269) Ó FEBS 2002 . uncoupling of the Na + -NQR activity by dioxygen In the presence of NADH and dioxygen, the purified Na + - NQR from V. alginolyticus was inactivated with a half time of approximately 3 min. This. Sequence analysis of the six subunits of the Na + -NQR from V. alginolyticus (NqrA– NqrF) showed that only the NqrF (or b-) subunit comprises a cysteine motif (Cys69, Cys75, Cys78 and Cys110) required. Inactivation of the Na + -translocating NADH:ubiquinone oxidoreductase from Vibrio alginolyticus by reactive oxygen species Julia Steuber 1 , Miche ` le Rufibach 1 ,Gu¨