Kinetics of electron transfer from NADH to the Escherichia coli nitric oxide reductase flavorubredoxin Joa ˜ o B. Vicente 1 , Francesca M. Scandurra 2 , Joa ˜ o V. Rodrigues 1 , Maurizio Brunori 2 , Paolo Sarti 2 , Miguel Teixeira 1 and Alessandro Giuffre ` 2 1 Instituto de Tecnologia Quı ´ mica e Biolo ´ gica, Universidade Nova de Lisboa, Oeiras, Portugal 2 Department of Biochemical Sciences, CNR Institute of Molecular Biology and Pathology and Istituto Pasteur – Fondazione Cenci Bolognetti, University of Rome ‘La Sapienza’, Italy In humans and other higher organisms, nitric oxide (NO) is produced by the inducible isoform of NO-syn- thase (iNOS) in several cell types, including macro- phages, as part of the immune response to counteract microbial infection [1,2]. NO production is enhanced at the site of infection [2] leading to the formation of highly reactive species, such as peroxynitrite (ONOO – [3]), all of which are cytotoxic towards the invading microbes. As a strategy to evade the host immune attack, pathogenic microorganisms have evolved biochemical pathways to resist to such a stress condition (generally termed ‘nitrosative stress’), and particularly to degrade NO. Many microorganisms express flavohemoglobin [4,5], an enzyme that efficiently catalyzes the oxidation of NO to nitrate (NO 3 – ) in the presence of O 2 , accord- ing to the following reaction: 2NO þ 2O 2 þ NAD(P)H ! 2NO À 3 þ NAD(P) þ þ H þ The flavodiiron proteins (FDPs, originally named A-type flavoproteins [6]), are a different class of micro- bial enzymes that were recently proposed to be involved Keywords flavodiiron proteins; microbial NO detoxification; NADH:rubredoxin oxidoreductase; nitrosative stress; time- resolved spectroscopy Correspondence A. Giuffre ` , Istituto di Biologia e Patologia Molecolari del Consiglio Nazionale delle Ricerche, c ⁄ o Dipartimento di Scienze Biochimiche ‘A. Rossi Fanelli’, Universita ` di Roma ‘La Sapienza’, Piazzale Aldo Moro 5, I-00185 Roma, Italia Fax: +39 06 4440062 Tel: +39 06 49910944 E-mail: alessandro.giuffre@uniroma1.it (Received 22 September 2006, revised 20 November 2006, accepted 21 November 2006) doi:10.1111/j.1742-4658.2006.05612.x Escherichia coli flavorubredoxin (FlRd) belongs to the family of flavodiiron proteins (FDPs), microbial enzymes that are expressed to scavenge nitric oxide (NO) under anaerobic conditions. To degrade NO, FlRd has to be reduced by NADH via the FAD-binding protein flavorubredoxin reduc- tase, thus the kinetics of electron transfer along this pathway was investi- gated by stopped-flow absorption spectroscopy. We found that NADH, but not NADPH, quickly reduces the FlRd-reductase (k ¼ 5.5 ± 2.2 · 10 6 m )1 Æs )1 at 5 °C), with a limiting rate of 255 ± 17 s )1 . The reduc- tase in turn quickly reduces the rubredoxin (Rd) center of FlRd, as assessed at 5 °C working with the native FlRd enzyme (k ¼ 2.4 ± 0.1 · 10 6 m )1 Æs )1 ) and with its isolated Rd-domain (k % 1 · 10 7 m )1 Æs )1 ); in both cases the reaction was found to be dependent on pH and ionic strength. In FlRd the fast reduction of the Rd center occurs syn- chronously with the formation of flavin mononucleotide semiquinone. Our data provide evidence that (a) FlRd-reductase rapidly shuttles electrons between NADH and FlRd, a prerequisite for NO reduction in this detoxi- fication pathway, and (b) the electron accepting site in FlRd, the Rd center, is in very fast redox equilibrium with the flavin mononucleotide. Abbreviations eT, electron transfer; FDP, flavodiiron protein; FlRd, flavorubredoxin; FlRd-reductase, NADH:flavorubredoxin oxidoreductase; FMN, flavin mononucleotide; FMN sq , flavin mononucleotide semiquinone (one electron-reduced); Rd, rubredoxin; Rd-domain, rubredoxin domain of flavorubredoxin; RR, Pseudomonas oleovorans rubredoxin reductase. FEBS Journal 274 (2007) 677–686 ª 2006 The Authors Journal compilation ª 2006 FEBS 677 in NO detoxification, particularly under microaerobic conditions [7]. In the absence of O 2 , FDPs are indeed endowed with NO-reductase activity [8–10], being cap- able of degrading NO most probably to nitrous oxide (N 2 O): 2NO þ 2e À þ 2H þ ! N 2 O þ H 2 O Flavodiiron proteins are widespread among prokaryo- tes [6,11]; based on genomic and functional analysis they were more recently identified also in a restricted number of anaerobic, pathogenic protozoa [11–14]. As a distinctive feature, FDPs are characterized by two structural domains: the N-terminal one, with a metallo- b-lactamase like fold, harboring a nonheme diiron site, and the flavodoxin-like domain with a flavin mononu- cleotide (FMN) moiety [15]. The 3D structure is now available for two FPDs, i.e. the enzyme isolated from Desulfovibrio gigas (originally named rubredoxin:oxy- gen oxidoreductase, ROO [16]), and the one from Moorella thermoacetica [17]. In both cases, the two redox centers (FMN and Fe-Fe) are at a relatively long distance (% 35 A ˚ ), but the enzyme displays a homodimeric assembly in a head-to-tail configuration, bringing the FMN of one monomer in close proximity to the Fe-Fe site of the other monomer. It is therefore likely, though not proven yet, that the dimer is the functional unit of this enzyme, ensuring fast electron equilibration between the redox cofactors. The FDP expressed by Escherichia coli contains in addition a rubredoxin-like domain with an iron-sulfur center, fused at the C-terminus of the flavodiiron core; thus this protein was named flavorubredoxin (FlRd) [6]. E. coli FlRd is the terminal component of an elec- tron transport chain (Fig. 1) that involves NADH and flavorubredoxin reductase, a FAD-binding protein of the NAD(P)H:rubredoxin oxidoreductase family. The genes coding for FlRd (norV) and its redox partner FlRd-reductase (norW) form a single dicistronic tran- scriptional unit [18]. In E. coli, the involvement of FlRd in the anaerobic NO detoxification was originally proposed by Gardner et al. [7] on the basis of molecular genetic evidence and confirmed by measuring the NO consumption cat- alyzed by the purified recombinant FlRd [8] and by other bacterial FDPs [9,10]. The protective role of FlRd towards nitrosative stress is further supported by the finding that after exposing E. coli cells to NO under anaerobic conditions, the transcriptional levels of the norVW genes raise considerably and the FlRd protein is promptly expressed [7,19]. It is not clear whether the capability of degrading NO is a common and distinctive feature among all the members of the FDPs family. Every FDP characterized so far seems to be capable of reacting with O 2 as well, though to dif- ferent extents. Moreover, recently it was reported the case of one FDP, the ROO from Desulfovibrio gigas, which in vivo protects from nitrosative stress, but in vitro as purified it consumes O 2 possibly more effi- ciently than NO [20]. Although FDPs might be the targets for novel drugs designed to counteract microbial infection, the informa- tion on the mechanism whereby FDPs degrade NO is as yet very poor. Probably, the active site is the Fe-Fe binuclear center, because substitution of Zn for Fe abolishes the activity [9]. Consistently, we have shown previously that the Rd-domain of FlRd, a genetically truncated version of the enzyme lacking the flavodiiron domain, is unable to catalyze the anaerobic NO degra- dation in the presence of excess reductants [8]. Also based on the redox potentials determined for FlRd [21], it is likely that electrons donated to FlRd enter the enzyme at the [Fe-Cys 4 ] center in the Rd-domain to be subsequently transferred via FMN to the Fe-Fe site where the reaction with NO is expected to occur; how- ever, essentially no information is available as yet on the kinetics of electron transfer to (as well as within) this enzyme. Because the efficiency of NO detoxifica- tion by FlRd (and FDPs in general) clearly depends on the availability of electrons at the site of reaction with NO, this prompted us to use E. coli FlRd as a model to study the kinetics of electron transfer (eT) along the NADH fi FlRd-reductase fi FlRd chain Rd NADH NAD + FAD e NO N 2 O FMN Fe-Fe Flavodiiron Core FlRd-reductase Flavorubredoxin Fig. 1. Schematic representation of the Escherichia coli electron transfer chain coup- ling NADH oxidation to NO reduction. Electron transfer to E. coli flavorubredoxin J. B. Vicente et al. 678 FEBS Journal 274 (2007) 677–686 ª 2006 The Authors Journal compilation ª 2006 FEBS (Fig. 1), which is herein investigated by time-resolved spectroscopy working on the purified recombinant proteins. Results Reduction of flavorubredoxin reductase by NADH The kinetics of the reduction of flavorubredoxin reduc- tase (FlRd-reductase) by NADH was investigated by time-resolved spectroscopy under anaerobic conditions and at 5 °C. Upon mixing with NADH, oxidized FlRd-reductase is fully reduced within 100 ms, as inferred from the absorption bleaching detected in the 400–500 nm range and the absorption increase at % 310 nm (Fig. 2A). Synchronously, a broad band appears at k > 520 nm (thick arrow in Fig. 2A); as previously proposed by Lee et al. [22] for the Pseudo- monas oleovorans rubredoxin reductase (RR), we assign the latter band to the formation of a charge- transfer complex between NAD + and reduced FlRd- reductase. The reduction of FlRd-reductase was followed at 455 and 310 nm (thin arrows in Fig. 2A) at increasing NADH concentrations. At [NADH] < 100 lm,pseudo- first order conditions were not attained and the reaction was thus modeled according to the scheme A+Bfi C. By fitting the experimental time courses to Eqn (1) (Experimental procedures), we estimated a second-order rate constant k ¼ 5.5 ± 2.2 · 10 6 m )1 Æs )1 (inset to Fig. 2A). At [NADH] ‡ 100 lm, i.e., under pseudo-first order conditions, within the experimental error the reaction followed a single exponential time course, proceeding at k¢ ¼ 255 ± 17 s )1 (Fig. 2B). In this [NADH] range, the observed rate constant was independent of [NADH], suggesting a limiting rate for eT within the NADH–FlRd-reductase complex. Under identical experimental conditions, NADPH reduces FlRd-reductase at an % 100-fold slower rate (not shown). 300 400 500 600 700 0.00 0.05 0.10 0.15 Absorbance ΔAbsorbance log k λ (nm) 020406080100 0.00 0.02 0.04 0.06 Δ bArosb n aec Time (ms) 10 20 30 40 50 0.00 0.01 0.02 0.03 0.04 Time (ms) 0 50 100 150 200 250 300 0 100 200 300 400 500 k s(' 1- ) [NADH] ( μ M ) A B C 2 1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 5.8 6.0 6.2 6.4 6.6 6.8 7.0 μ ( M 1/2 ) Fig. 2. Reduction of flavorubredoxin reductase by NADH. (A) Time- resolved absorption spectra collected every 2.56 ms up to 100 ms after mixing oxidized FlRd-reductase with NADH under anaerobic conditions. Concentrations after mixing: [FlRd-reductase] ¼ 7.6 l M; [NADH] ¼ 16.5 l M. Bold line: spectrum of fully oxidized FlRd-reduc- tase (k max ¼ 455 nm, thin arrow). The thick arrow outlines the broad band appearing at k > 520 nm (see text for details). T ¼ 5 °C. Buffer: 50 m M Tris ⁄ HCl, 18% glycerol, pH 8.0. Inset: Time courses of the reaction as measured at [NADH] ¼ 10, 30 and 50 l M (concentrations after mixing), fitted according to Eqn (1) in Experimental procedures. k ¼ 455 nm. (B) Time course of FlRd- reductase reduction probed under pseudo-first order conditions, fol- lowed at 455 nm (line 1) and 310 nm (line 2). Concentrations after mixing: [NADH] ¼ 100 l M; [FlRd-reductase] ¼ 7.6 lM.T¼ 5 °C. Buffer: 50 m M Tris ⁄ HCl, 18% glycerol, pH 8.0. Inset: Observed rate constants measured at three different concentrations of NADH ‡ 100 l M. (C) Ionic strength dependence of the second order rate constant of FlRd-reductase reduction by NADH. Error bar indicates the maximal error observed in this data set. Data were modeled according to the Broensted–Bjerrum equation yielding Z A Z B ¼ )1.3 (see Results and Discussion). T ¼ 5 °C. In these experiments FlRd- reductase was desalted by gel filtration and ionic strength adjusted by addition of KCl to the buffer (5 m M Tris ⁄ HCl, 18% glycerol, pH 8.0). J. B. Vicente et al. Electron transfer to E. coli flavorubredoxin FEBS Journal 274 (2007) 677–686 ª 2006 The Authors Journal compilation ª 2006 FEBS 679 The rate of FlRd-reductase reduction by NADH decreased constantly with increasing ionic strength (Fig. 2C). Data were analyzed according to the Broen- sted–Bjerrum equation, whereby log k is expected to be linearly dependent on the square root of the ionic strength with a slope equal to 2AZ A Z B (A % 0.49 at 5 °C and Z A and Z B are the charges involved). From the data in Fig. 2C we estimated Z A Z B % )1.3, which is consistent with a slight effect of ionic strength on this reaction. Finally, the reduction of FlRd-reductase by NADH was found to be essentially independent of pH in the range 5.0–8.0 (not shown). Reduction of the rubredoxin domain of FlRd by flavorubredoxin reductase The isolated, genetically truncated rubredoxin domain (Rd-domain) of FlRd is characterized in the oxidized state by a typical absorption spectrum (Fig. 3A) that is bleached upon reduction (not shown). The kinetics of the anaerobic reduction of Rd-domain by FlRd-reduc- tase (prereduced by a large excess of NADH) was investigated by stopped-flow spectroscopy. As monitored at 484 nm (arrow in Fig. 3A), the Rd-domain is rapidly (< 1 s) reduced by FlRd-reduc- tase in a concentration-dependent manner, following a single exponential time course (Fig. 3B). This is consis- tent with the fact that FlRd-reductase is kept fully reduced during the whole time course by the excess NADH. After mixing the oxidized Rd-domain with NADH only, i.e., in the absence of FlRd-reductase, no absorbance changes are observed even over several sec- onds (not shown), thus proving that NADH is unable to directly reduce the Rd-domain. When FlRd-reduc- tase is present to shuttle electrons, the observed rate constant for the reduction of the Rd-domain shows a hyperbolic dependence on the FlRd-reductase concen- tration (inset to Fig. 3B). Data were modeled accord- ing to Scheme 1, whereby complex formation between oxidized Rd-domain and reduced FlRd-reductase (k 1 , k )1 ) is associated with intracomplex electron transfer (k 2 ). This is followed by fast dissociation of the part- ners (k 3 ? k 2 ) and re-reduction of oxidized FlRd- reductase by NADH at 255 s )1 , as independently determined (inset to Fig. 2B). As an over-simplifica- tion, in this model intramolecular eT is assumed to be an irreversible process, based on the information that reduction of the Rd-domain by FlRd-reductase is largely favored thermodynamically, according to the redox potentials determined by Vicente et al. [21]. As shown in the inset to Fig. 3B, experimental rates (closed symbols) are suitably fitted by kin- etic simulations (open symbols), by assuming k 1 ¼ 1.3 · 10 7 m )1 Æs )1 , k )1 £ 13 s )1 , k 2 ¼ 300 s )1 and k 3 ‡ 5000 s )1 in Scheme 1. Reduction of FlRd by flavorubredoxin reductase Spectral analysis of FlRd is complex due to the partial overlap of the optical contribution of its redox cofac- tors. Figure 4 shows the absorption spectrum of FlRd in the oxidized state (spectrum A) and after reduction by an excess of NADH in the presence of catalytic amounts of FlRd-reductase (spectrum B). In the visible region, the spectrum of oxidized FlRd is characterized by a broad band centered at 474 nm and a shoulder at 0 200 400 600 0.00 0.02 0.04 0.06 Time (ms) B 300 400 500 600 700 0.000 0.025 0.050 0.075 Absorbance ΔAbsorbance λ (nm) A 0 5 10 15 20 25 0 100 200 300 400 k ( ' s -1 ) [FlRd-Reductase] (μM) Fig. 3. Reduction of the rubredoxin domain of FlRd (Rd-domain) by FlRd-reductase. Oxidized Rd-domain was anaerobically mixed with FlRd-reductase at increasing concentrations, prereduced by excess NADH. Concentrations after mixing: [Rd-domain] ¼ 7.7 l M; [FlRd- reductase] ¼ 0.38, 0.75, 1.5, 3.3, 6.5, 13 or 26 l M; [NADH] ¼ 375 l M.T¼ 5 °C. Buffer: 50 mM Tris ⁄ HCl, 18% glycerol, pH 8.0. (A) Absorption spectrum of 7.7 l M oxidized Rd-domain (k max ¼ 484 nm, arrow). (B) Best fit to single exponential decays of the time courses measured at 484 nm at increasing FlRd-reductase concentrations. Inset: Observed rate constant as a function of FlRd-reductase concentration. Experimental data (closed symbols) were modeled (open symbols) according to Scheme 1, by assuming k 1 ¼ 1.3 · 10 7 M )1 Æs )1 , k )1 £ 13 s )1 , k 2 ¼ 300 s )1 and k 3 ‡ 5000 s )1 . Electron transfer to E. coli flavorubredoxin J. B. Vicente et al. 680 FEBS Journal 274 (2007) 677–686 ª 2006 The Authors Journal compilation ª 2006 FEBS % 560 nm (arrow); this spectrum is contributed by [Fe-Cys 4 ] in the Rd-domain (spectrum C) and by FMN with a possible contribution of the Fe-Fe center (spectrum D). The spectrum of reduced FlRd (Fig. 4, line b) displays a low intensity band centered at % 500 nm, which cannot be directly assigned solely from analyzing these spectra, as it could either result from partially reduced FMN or from the Fe-Fe centre. From these spectra, it is evident that at k > 550 nm the absorption changes are almost exclusively domin- ated by the Rd-domain, making this an adequate wavelength range to monitor redox changes of the Rd centre in the whole enzyme. The kinetics of FlRd reduction was investigated by anaerobically mixing the oxidized protein with FlRd- reductase preincubated with excess NADH. Also in the case of FlRd, no direct reduction by NADH was observed over several seconds. The absolute absorption spectra collected from 2.56 ms to 10 s after mixing are depicted in Fig. 5A, together with the initial spectrum of FlRd in the oxidized state (dotted line); absorption at k < 400 nm is dominated by NADH in excess. The difference spectra are shown in Fig. 5B. The ratio Rd ox :Rd red at each time point was estimated at 560 nm (arrow in Fig. 5B), which allowed us to recon- struct the optical contribution of [Fe-Cys 4 ] (Fig. 5C) to the difference spectra in Fig. 5B. By subtraction we estimated the optical contribution of the FlRd flavodi- iron domain, which is dominated by the FMN moiety (Fig. 5D). Inspection of the latter data reveals the formation of a red flavin semiquinone, characterized by an absorbance increase at % 390 nm and a syn- chronous absorbance decrease at % 450 nm [23]. Summing up, after mixing oxidized FlRd with reduced FlRd-reductase in the presence of an excess of NADH, two events can be deconvoluted: the reduction of [Fe-Cys 4 ] (monitored at 560 nm) and the formation of semiquinone FMN (monitored at 390 nm after sub- traction of the optical contribution of Fe-Cys 4 ). As shown in Fig. 6, both processes appear to be synchron- ous, following a single exponential time course with a rate constant linearly dependent on FlRd-reductase concentration; the calculated apparent second order rate constant is k ¼ 2.4 ± 0.1 · 10 6 m )1 Æs )1 . It should be noted that at the highest concentrations of FlRd- reductase (inset to Fig. 6B), the faster accumulation of flavin mononucleotide semiquinone (FMN sq )is followed by a slower partial decay presumably to 2e-reduced FMN. The effect of ionic strength and pH on the reduction of Rd-domain and FlRd The effect of ionic strength and pH on the reduction of either the Rd-domain or FlRd was also investigated, upon mixing at 20 °C these proteins in the oxidized state with FlRd-reductase prereduced by excess NADH. Figure 7 shows that in the cases of FlRd (A) and Rd-domain (B), the observed rates follow a bell- shaped dependence on ionic strength, with a maximum at around 40–50 mm. As shown in Fig. 8, the kinetics of FlRd reduction was found to be strongly pH dependent with an apparent pK a % 7.3, the asymptotic value at acidic pH being k¢ % 0.04 s )1 . A very similar pH depend- ence was also observed for the reduction of the iso- lated Rd-domain. NADH Rd-D (ox) FlRd-Red (red) + k k 2 + k 3 FlRd-Red (red) Rd-D (ox) FlRd-Red (ox) Rd-D (red) FlRd-Red (ox) Rd-D (red) k -1 1 Scheme 1. 300 400 500 600 700 0.00 0.05 0.10 0.15 Absorbance Wavelength (nm) a b c d Fig. 4. Spectral features of flavorubredoxin (FlRd) and its individual cofactors. Spectrum a: oxidized FlRd (arrow indicates the shoulder at % 560 nm). Spectrum b: reduced FlRd (a few seconds after mix- ing with 0.25 l M FlRd-reductase in the presence of 375 lM NADH). Spectrum c: oxidized Rd-domain. Spectrum d: optical contribution of the oxidized flavodiiron (FMN ⁄ Fe-Fe) domain of FlRd estimated by subtracting spectrum c from spectrum a. Protein concentration: 10 l M. J. B. Vicente et al. Electron transfer to E. coli flavorubredoxin FEBS Journal 274 (2007) 677–686 ª 2006 The Authors Journal compilation ª 2006 FEBS 681 Discussion Flavodiiron proteins (FDPs), expressed in many prok- aryotes [6,11] and in a restricted group of pathogenic amitochondriate protozoa [12–14], are responsible for NO detoxification under anaerobic conditions [7,8], thus helping microbes to survive in NO-enriched microaerobic environments. Because FDPs catalyze the reduction of NO at the level of their nonheme diiron site, their catalytic efficiency clearly depends on the availability of reducing equivalents at this bimetallic site. In E. coli NADH is the source of these electrons, which are then transferred to FlRd via FlRd-reductase ([15], Fig. 1). The results herein presented show that E. coli FlRd-reductase is highly specific for NADH, that acts as a very efficient electron donor (k ¼ 5.5 ± 2.2 · 10 6 m )1 Æs )1 ,at5°C) contrary to NADPH. This specificity can be possibly understood based on the protein engineering studies on glutathione reductase [24] and dihydrolipoamide dehydrogenase [25] from E. coli, which are specific for NADPH and NADH, respectively. Sequence analyses and homology modeling of FlRd-reductase (not shown) suggest: (a) the presence of the residues competent to form H-bonds with the ribose 2¢-OH and 3¢-OH groups of NADH, and (b) the absence of a nest of positively charged residues to stabilize the extra phosphate group in NADPH. In the present study, we have observed several anal- ogies between E. coli FlRd-reductase and the rubre- doxin reductase (RR) from Pseudomonas oleovorans 300 400 500 600 700 0.00 0.05 0.10 0.15 sbAo nabrec λ (nm) 300 400 500 600 700 0.00 0.02 0.04 0.06 0.08 0.10 Δ bAosabrnec λ (nm) B A Rd FMN Fe-Fe Rd FMN Fe-Fe Rd FMN Fe-Fe Rd FMN Fe-Fe time 300 400 500 600 700 -0.04 -0.02 0.00 0.02 0.04 Δ r os bAbn a ec λ (nm) 300 400 500 600 700 0.00 0.02 0.04 0.06 0.08 0.10 Δ bAbroscnae λ (nm) D C Rd FMN Fe-Fe FMN Fe-Fe Fig. 5. Reduction of flavorubredoxin (FlRd) by FlRd-reductase. (A) Absolute spectra collected after mixing oxidized FlRd with FlRd-reductase prereduced by excess NADH. Concentrations after mixing: [FlRd] ¼ 10 l M; [FlRd-reductase] ¼ 0.25 lM; [NADH] ¼ 375 lM. Spectra acquired in a logarithmic time mode, from 2.56 ms up to 10 s (arrow depicts the direction of the absorption changes with time). Buffer: 50 m M Tris ⁄ HCl, 18% glycerol, pH 8.0. T ¼ 5 °C. (B) Difference spectra obtained by subtracting the final spectrum in (A) (t ¼ 10 s) from the remain- ders. Arrow depicts 560 nm as a suitable wavelength to monitor the redox changes of the [Fe-Cys 4 ] centre in FlRd. (C) Optical contribution of the Rd-domain to the difference spectra depicted in (B). These spectra were reconstructed by estimating the Rd ox :Rd red ratio at every time point from the absorption changes detected at 560 nm along the reaction [arrow in (B)]. (D) Optical contribution of the flavodiiron domain estimated by subtracting the contribution of the Rd-domain (C) from the difference spectra depicted in (B). Spectra reveal the forma- tion of flavin red semiquinone, as indicated by the increase at 390 nm (arrow). Electron transfer to E. coli flavorubredoxin J. B. Vicente et al. 682 FEBS Journal 274 (2007) 677–686 ª 2006 The Authors Journal compilation ª 2006 FEBS [22]. The latter is a FAD-binding protein sharing a significant amino acid sequence similarity with E. coli FlRd-reductase (27% identity, 50% similarity); its phy- siological role is to shuttle electrons between NADH and rubredoxin, the electron donor of a membrane bound diiron x-hydroxylase required for the hydroxy- lation of alkanes [26]. Comparing E. coli FlRd-reduc- tase and P. oleovorans RR, we notice that: (a) the FAD moiety accepts the two electrons from NADH as a single kinetic step with no evidence for flavin radical accumulation; (b) at saturating NADH concentrations (> 100 lm), flavin is reduced at comparable limiting rates [255 ± 17 s )1 in FlRd-reductase (inset Fig. 2B), to be compared with 180–190 s )1 measured for RR]; (c) upon reduction by NADH, a charge transfer com- plex with NAD + is formed, identified by a broad absorption band at k > 520 nm (Fig. 2A in the pre- sent study to be compared with Fig. 4 in [22]). Based on the structural and functional similarities with P. oleovorans RR, it may be expected that the physiological function of E. coli FlRd-reductase is to shuttle electrons between NADH and the Rd center in FlRd, as originally proposed by Gomes et al. [15]. Consistently, we observed that NADH is unable to directly reduce [Fe-Cys 4 ] in the Rd-domain, either iso- lated or as part of FlRd, unless FlRd-reductase is pre- sent to catalyze this eT process (Figs 3B and 6A). In the latter case, the Rd center is promptly reduced by FlRd-reductase and this reaction was found to be highly dependent both on pH and ionic strength (Figs 7 and 8). Namely, we found that the reaction (a) speeds up at alkaline pH (apparent pKa % 7.3), a find- ing that appears physiologically relevant as in the cyto- sol of E. coli, where FlRd-reductase and FlRd are found, pH is % 7.5 and (b) displays a bell-shaped B 0.0 0.2 0.4 0.6 0.8 1.0 0.00 0.01 0.02 0.03 0.04 Δ sbAo ecnabr Time (s) A 560 nm Rd 390 nm FMN sq 0.0 0.2 0.4 0.6 0.8 1.0 0.00 0.02 0.04 0.06 Δ ecnabrosbA Time (s) 0 20 40 60 80 100 120 0 5 10 15 20 25 30 0 20 40 60 80 100 [FlRd-Reductase] (μM) 'k s( 1- ) Fig. 6. Kinetics of electron transfer between FlRd-reductase and fla- vorubredoxin. Concentrations after mixing: [FlRd] ¼ 10 l M; [FlRd- reductase] ¼ 0.25, 0.5, 1.5, 2.3, 3.4, 5, 11.5, 17.5 and 26 l M; [NADH] ¼ 375 l M.T¼ 5 °C. Buffer: 50 mM Tris ⁄ HCl, 18% gly- cerol, pH 8.0. Data collected after anaerobically mixing oxidized FlRd with increasing concentrations of FlRd-reductase prereduced by excess NADH. Observed rate constants obtained by fitting to single exponential decays the absorption changes collected at 560 nm (A) and at 390 nm (B). 0 100 200 300 400 0 1 2 3 4 5 'k s( 1- ) μ (mM) 0 1 2 3 4 5 'k s( 1- ) A B Rd FMN Fe-Fe Rd FMN Fe-Fe Rd Fig. 7. Effect of ionic strength. Ionic strength dependence of the rate constants observed for the anaerobic reduction by FlRd-reduc- tase of FlRd (A) or the isolated Rd-domain (B). Concentrations after mixing: 8.5 l M FlRd, 2 lM FlRd-reductase, 375 lM NADH (A) or 10.5 l M Rd-domain, 0.5 lM FlRd-reductase, 375 lM NADH (B). T ¼ 20 °C. Rd-domain and FlRd were previously desalted and equil- ibrated with 5 m M Tris ⁄ HCl, 18% glycerol, pH 7.6, by gel permea- tion chromatography. Ionic strength was then adjusted by addition of KCl to the buffer. The dashed lines are merely shown to repre- sent the observed bell-shaped behavior. J. B. Vicente et al. Electron transfer to E. coli flavorubredoxin FEBS Journal 274 (2007) 677–686 ª 2006 The Authors Journal compilation ª 2006 FEBS 683 dependence on ionic strength, a fairly common feature for interprotein electron transfer [27], with maximum rate at around 40–50 mm. Under optimal eT conditions (pH ¼ 8.0 and l % 40 mm), kinetic data could be modeled according to Scheme 1 (Fig. 3B); we observe that FlRd-reductase and the Rd domain form a tight complex rapidly (k % 1 · 10 7 m )1 Æs )1 ; K d £ 1 lm), followed by an intra- complex eT (from FAD to [Fe-Cys 4 ]) proceeding at a limiting rate of % 300 s )1 . With the whole FlRd, elec- trons donated by FlRd-reductase enter the protein at the Rd center (with an apparent k of 2.4 · 10 6 m )1 Æs )1 ) and clearly re-equilibrate with FMN, leading to forma- tion of FMN sq (detected by spectral analysis detailed in Fig. 5). Such a finding is consistent with the reduc- tion potentials of FMN ⁄ FMN sq and [Fe 3+ ) Cys 4 ] ⁄ [Fe 2+ ) Cys 4 ] being very similar (E 0 ¼ )40 mV and )60 mV, respectively [21]). These two events were observed to proceed synchronously even at the highest FlRd-reductase concentration, thus strongly suggesting that [Fe-Cys 4 ] and FMN are in very fast redox equilib- rium (Scheme 2). It is interesting that also in the flavocytochrome P450BM3 (from Bacillus megaterium), the flavin semiquinone shuttles one electron at a time to the heme active site, whereas the fully (two electron) reduced flavin contributes to inactivation of the enzyme [28]. The lack of a UV-visible spectral fingerprint for the Fe-Fe site hampers the detection of this site’s prompt reduction via FMN sq . However, we notice that if the reduction of the Fe-Fe site was not occurring synchro- nously with [Fe-Cys 4 ] and FMN 1e-reduction, only two electrons would quickly equilibrate within FlRd. Because the redox potentials of both [Fe 3 ± Cys 4 ] ⁄ [Fe 2 ± Cys 4 ] and FMN ⁄ FMN sq are similar [21], in the absence of other effects the observed apparent rate constant for the reduction of [Fe-Cys 4 ] and FMN should be approximately two-fold smaller than that measured with the isolated Rd-domain (accepting only one electron). Taking this into account, it is relevant that the second order rate constants for eT from FlRd-reductase to isolated Rd-domain (% 1 · 10 7 m )1 Æs )1 ) or FlRd (2.4 · 10 6 m )1 Æs )1 ) actually differ by a factor significantly greater than two. This leads us to hypothesize that electrons entering FlRd at the Rd center quickly equilibrate also with the Fe-Fe site via FMN sq . In conclusion, we have thoroughly investigated the eT kinetics to flavorubredoxin, the crucial enzyme in the E. coli anaerobic NO-detoxification pathway. We found that FlRd-reductase acts as an efficient electron shuttle between NADH and the [Fe-Cys 4 ] center of FlRd, where electrons quickly equilibrate intramolecu- larly with FMN sq and most probably Fe-Fe, to become available for the reduction of NO to N 2 O. Experimental procedures Materials NADH, glucose oxidase and catalase were purchased from Sigma (St. Louis, MO). The concentration of NADH in stock solutions was determined spectrophotometrically using the extinction coefficient e 340nm ¼ 6.2 mm )1 Æcm )1 . Unless otherwise specified, experiments were performed at 5 °Cin50mm Tris ⁄ HCl, 18% (v ⁄ v) glycerol, pH 8.0. The low temperature was chosen in order to slow down the reactions that were otherwise too fast to be time-resolved. Glycerol was used to enhance the stability of purified FlRd Rd FMN Fe-Fe Rd FMN Fe-Fe Rd 5.5 6.0 6.5 7.0 7.5 8.0 0 1 2 3 4 pH 'k F(lRd) s( 1- ) 0 5 10 15 ' k R ( -doD a mn i ) s( 1- ) Fig. 8. Effect of pH. Rate constants obtained by measuring the anaerobic reduction of FlRd (closed symbols) or Rd-domain (open symbols) by FlRd-reductase, at different pH values. Concentrations after mixing: 8.5 l M FlRd or 10.5 lM Rd-domain, 2 lM FlRd-reduc- tase, 375 l M NADH. T ¼ 20 °C. Buffer: 5 mM Tris ⁄ HCl, 18% gly- cerol, pH 7.6. FlRd and Rd-domain were previously desalted and equilibrated in 5 m M Tris ⁄ HCl, 18% glycerol, pH 7.6, by gel per- meation chromatography, whereas NADH and FlRd-reductase were diluted into concentrated buffers (100 m M) at different pH values. Ionic strength % 145 m M after mixing. Scheme 2. Electron transfer to E. coli flavorubredoxin J. B. Vicente et al. 684 FEBS Journal 274 (2007) 677–686 ª 2006 The Authors Journal compilation ª 2006 FEBS and FlRd-reductase in solution, an effect already documen- ted for Pseudomonas oleovorans rubredoxin reductase (RR) and rubredoxin [22,29]. Anaerobic conditions were achieved by N 2 -equilibration of the buffers and by scavenging residual contaminant oxygen using glucose oxidase (17 unitsÆmL )1 ), glucose (2 mm) and catalase (130 unitsÆmL )1 ). Escherichia coli flavorubredoxin (FlRd), flavorubredoxin reductase (FlRd-reductase) and a truncated version of FlRd consisting of the only rubredoxin domain (Rd-domain) were overexpressed in E. coli, purified as previously des- cribed [8,15] and stored at )80 °C until use. The concentra- tion of oxidized FlRd-reductase and Rd-domain was determined spectrophotometrically using the extinction coeffi- cients e 455nm ¼ 12 mm )1 Æcm )1 and e 484nm ¼ 7mm )1 Æcm )1 , respectively. The protein concentration of FlRd was deter- mined by the bicinchoninic acid method [30], iron and FMN contents were quantitated as in [31] and [32], respect- ively. As purified, FlRd contained the expected amount of iron (% 3 Fe per monomer), but substoichiometric FMN (0.5–0.6 instead of 1 FMN per monomer), pointing to par- tial loss of flavin during the purification procedure or incomplete incorporation of flavin during expression. Absorption spectroscopy and data analysis UV ⁄ visible static spectra were recorded by using a Shimadzu (Tokyo, Japan) spectrophotometer (UV-1603). Stopped-flow experiments were carried out with a thermostated instrument (DX.17 MV, Applied Photophysics, Leatherhead, UK) equipped with either a monochromator or a diode-array (light path ¼ 1 cm). When the instrument was used in the multiwavelength mode (diode-array), time-resolved absorp- tion spectra were recorded with an acquisition time of 2.56 ms per spectrum and a wavelength resolution of 2.1 nm. Typically, three independent traces were collected to be averaged before analysis. Kinetic data were analyzed by nonlinear least-squares regression analysis using the software matlab (MathWorks, South Natick, NA). The reaction of FlRd-reductase with NADH, whenever it was not probed under pseudo-first order conditions, was modeled according to a scheme of the type A þ B ! k C. The apparent second order rate con- stant k was thus obtained by fitting the experimental time courses to the equation: ln B 0 ðA 0 À xÞ A 0 ðB 0 À xÞ  ¼ ktðA 0 À B 0 Þð1Þ where A 0 and B 0 are the initial concentrations of A and B and x is the amount of A and B reacted at the time t. The time courses of both Rd-domain and FlRd reduction by FlRd-reductase were fitted to single exponential decays. In the case of FlRd, observed rate constants were linearly dependent on the concentration of the FlRd-reductase and the apparent second-order rate constant was thus estimated by linear regression analysis. The observed rate constants for Rd-domain reduction showed a hyperbolic dependence on the concentration of FlRd-reductase. In this case, the apparent second-order rate constant was estimated by kin- etic simulations performed using the software facsimile (AEA Technology, Didcot, UK). Acknowledgements This work was partially supported by Ministero dell’Istruzione, dell’Universita ` e della Ricerca of Italy (PRIN ‘Meccanismi molecolari e aspetti fisiopatologici dei sistemi bioenergetici di membrana’ and FIRB RBAU01F2BJ to P.S.), by Fundac¸ a ˜ o para a Cieˆ ncia e Tecnologia of Portugal (project grant POCTI ⁄ 2002 ⁄ BME ⁄ 44597 to M.T. and PhD grants SFRH ⁄ BD ⁄ 9136 ⁄ 2002 to J.B.V. and SFRH ⁄ BD ⁄ 14380 ⁄ 2003 to J.V.R.), and by Consiglio Nazionale delle Ricerche of Italy and Gabinete de Relac¸ o ˜ es Internacionais da Cieˆ ncia e do Ensino Superior of Portugal (to A.G. and M.T.). References 1 MacMicking J, Xie QW & Nathan C (1997) Nitric oxide and macrophage function. Annu Rev Immunol 15, 323–350. 2 Bogdan C (2001) Nitric oxide and the immune response. Nat Immunol 2, 907–916. 3 Beckman JS & Koppenol WH (1996) Nitric oxide, superoxide, and peroxynitrite: the good, the bad, and ugly. Am J Physiol 271, C1424–C1437. 4 Poole RK, Anjum MF, Membrillo-Hernandez J, Kim SO, Hughes MN & Stewart V (1996) Nitric oxide, nitrite, and Fnr regulation of hmp (flavohemoglobin) gene expression in Escherichia coli K-12. J Bacteriol 178, 5487–5492. 5 Gardner PR, Gardner AM, Martin LA & Salzman AL (1998) Nitric oxide dioxygenase: an enzymic function for flavohemoglobin. Proc Natl Acad Sci USA 95, 10378–10383. 6 Wasserfallen A, Ragettli S, Jouanneau Y & Leisinger T (1998) A family of flavoproteins in the domains Archaea and Bacteria. Eur J Biochem 254, 325–332. 7 Gardner AM, Helmick RA & Gardner PR (2002) Fla- vorubredoxin, an inducible catalyst for nitric oxide reduction and detoxification in Escherichia coli. J Biol Chem 277, 8172–8177. 8 Gomes CM, Giuffre A, Forte E, Vicente JB, Saraiva LM, Brunori M & Teixeira M (2002) A novel type of nitric-oxide reductase. Escherichia coli flavorubredoxin. J Biol Chem 277 , 25273–25276. 9 Silaghi-Dumitrescu R, Coulter ED, Das A, Ljungdahl LG, Jameson GN, Huynh BH & Kurtz DM Jr (2003) A flavodiiron protein and high molecular weight J. B. Vicente et al. Electron transfer to E. coli flavorubredoxin FEBS Journal 274 (2007) 677–686 ª 2006 The Authors Journal compilation ª 2006 FEBS 685 rubredoxin from Moorella thermoacetica with nitric oxide reductase activity. Biochemistry 42, 2806–2815. 10 Silaghi-Dumitrescu R, Ng KY, Viswanathan R & Kurtz DM Jr (2005) A flavo-diiron protein from Desulfovibrio vulgaris with oxidase and nitric oxide reductase activ- ities. Evidence for an in vivo nitric oxide scavenging function. Biochemistry 44, 3572–3579. 11 Saraiva LM, Vicente JB & Teixeira M (2004) The role of the flavodiiron proteins in microbial nitric oxide detoxification. Adv Microb Physiol 49, 77–129. 12 Andersson JO, Sjogren AM, Davis LA, Embley TM & Roger AJ (2003) Phylogenetic analyses of diplomonad genes reveal frequent lateral gene transfers affecting eukaryotes. Curr Biol 13 , 94–104. 13 Sarti P, Fiori PL, Forte E, Rappelli P, Teixeira M, Mastronicola D, Sanciu G, Giuffre A & Brunori M (2004) Trichomonas vaginalis degrades nitric oxide and expresses a flavorubredoxin-like protein: a new pathogenic mechanism? Cell Mol Life Sci 61, 618–623. 14 Loftus B, Anderson I, Davies R, Alsmark UC, Samuel- son J, Amedeo P, Roncaglia P, Berriman M, Hirt RP, Mann BJ, et al. (2005) The genome of the protist para- site Entamoeba histolytica. Nature 433, 865–868. 15 Gomes CM, Vicente JB, Wasserfallen A & Teixeira M (2000) Spectroscopic studies and characterization of a novel electron-transfer chain from Escherichia coli invol- ving a flavorubredoxin and its flavoprotein reductase partner. Biochemistry 39, 16230–16237. 16 Frazao C, Silva G, Gomes CM, Matias P, Coelho R, Sieker L, Macedo S, Liu MY, Oliveira S, Teixeira M et al. (2000) Structure of a dioxygen reduction enzyme from Desulfovibrio gigas. Nat Struct Biol 7 , 1041–1045. 17 Silaghi-Dumitrescu R, Kurtz DM Jr, Ljungdahl LG & Lanzilotta WN (2005) X-ray crystal structures of Moor- ella thermoacetica FprA. Novel diiron site structure and mechanistic insights into a scavenging nitric oxide reductase. Biochemistry 44, 6492–6501. 18 da Costa PN, Teixeira M & Saraiva LM (2003) Regula- tion of the flavorubredoxin nitric oxide reductase gene in Escherichia coli: nitrate repression, nitrite induction, and possible post-transcription control. FEMS Micro- biol Lett 218, 385–393. 19 Justino MC, Vicente JB, Teixeira M & Saraiva LM (2005) New genes implicated in the protection of anae- robically grown Escherichia coli against nitric oxide. J Biol Chem 280 , 2636–2643. 20 Rodrigues R, Vicente JB, Felix R, Oliveira S, Teixeira M & Rodrigues-Pousada C (2006) Desulfovibrio gigas flavodiiron protein affords protection against nitrosative stress in vivo. J Bacteriol 188, 2745–2751. 21 Vicente JB & Teixeira M (2005) Redox and spectroscopic properties of the Escherichia coli nitric oxide-detoxifying system involving flavorubredoxin and its NADH-oxidiz- ing redox partner. J Biol Chem 280, 34599–34608. 22 Lee HJ, Basran J & Scrutton NS (1998) Electron trans- fer from flavin to iron in the Pseudomonas oleovorans rubredoxin reductase-rubredoxin electron transfer com- plex. Biochemistry 37, 15513–15522. 23 Ghisla S & Edmondson DA (2001) Flavin Coenzymes. In Encyclopedia of Life Sciences. Nature Publishing Group. John Wiley & Sons, Chichester. 24 Mittl PR, Berry A, Scrutton NS, Perham RN & Schulz GE (1994) Anatomy of an engineered NAD-binding site. Protein Sci 3, 1504–1514. 25 Bocanegra JA, Scrutton NS & Perham RN (1993) Crea- tion of an NADP-dependent pyruvate dehydrogenase multienzyme complex by protein engineering. Biochemis- try 32, 2737–2740. 26 McKenna EJ & Coon MJ (1970) Enzymatic omega- oxidation. IV. Purification and properties of the omega- hydroxylase of Pseudomonas oleovorans. J Biol Chem 245, 3882–3889. 27 McLendon G & Hake R (1992) Interprotein Electron Transfer. Chem Rev 92, 481–490. 28 Warman AJ, Roitel O, Neeli R, Girvan HM, Seward HE, Murray SA, McLean KJ, Joyce MG, Toogood H, Holt RA, Leys D, Scrutton NS & Munro AW (2005) Flavocytochrome P450 BM3: an update on structure and mechanism of a biotechnologically important enzyme. Biochem Soc Trans 33, 747–753. 29 Perry A, Tambyrajah W, Grossmann JG, Lian LY & Scrutton NS (2004) Solution structure of the two-iron rubredoxin of Pseudomonas oleovorans determined by NMR spectroscopy and solution X–ray scattering and interactions with rubredoxin reductase. Biochemistry 43, 3167–3182. 30 Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ & Klenk DC (1985) Measurement of pro- tein using bicinchoninic acid. Anal Biochem 150, 76–85. 31 Fischer DS & Price DC (1964) A Simple Serum Iron Method Using the New Sensitive Chromogen Tripyri- dyl-S-Triazine. Clin Chem 10, 21–31. 32 Susin S, Abian J, Sanchez-Baeza F, Peleato ML, Abadia A, Gelpi E & Abadia J (1993) Riboflavin 3¢- and 5¢ -sul- fate, two novel flavins accumulating in the roots of iron-deficient sugar beet (Beta vulgaris). J Biol Chem 268, 20958–20965. Electron transfer to E. coli flavorubredoxin J. B. Vicente et al. 686 FEBS Journal 274 (2007) 677–686 ª 2006 The Authors Journal compilation ª 2006 FEBS . Kinetics of electron transfer from NADH to the Escherichia coli nitric oxide reductase flavorubredoxin Joa ˜ o B. Vicente 1 ,. depends on the availability of electrons at the site of reaction with NO, this prompted us to use E. coli FlRd as a model to study the kinetics of electron transfer