Formation of a cytochrome c–nitrous oxide reductase complex is obligatory for N₂O reduction by Paracoccus pantotrophus

Abstract

Nitrous oxide reductase (N₂OR) catalyses the final step of bacterial denitrification, the two-electron reduction of nitrous oxide (N₂O) to dinitrogen (N₂). N₂OR contains two metal centers; a binuclear copper center, Cu_A, that serves to receive electrons from soluble donors, and a tetranuclear copper-sulfide center, Cu_Z, at the active site. Stopped flow experiments at low ionic strengths reveal rapid electron transfer (k_obs = 150 s⁻¹) between reduced horse heart (HH) cytochrome c and the Cu_A center in fully oxidized N₂OR. When fully reduced N₂OR was mixed with oxidized cytochrome c, a similar rate of electron transfer was recorded for the reverse reaction, followed by a much slower internal electron transfer from Cu_Z to Cu_A (k_obs = 0.1–0.4 s⁻¹). The internal electron transfer process is likely to represent the rate-determining step in the catalytic cycle. Remarkably, in the absence of cytochrome c, fully reduced N₂OR is inert towards its substrate, even though sufficient electrons are stored to initiate a single turnover. However, in the presence of reduced cytochrome c and N₂O, a single turnover occurs after a lag-phase. We propose that a conformational change in N₂OR is induced by its specific interaction with cytochrome c that in turn either permits electron transfer between Cu_A and Cu_Z or controls the rate of N₂O decomposition at the active site.

---

Introduction

Nitrous oxide (N₂O) along with carbon dioxide (CO₂) and methane is one of the three most important greenhouse gases. The largest source of man-made N₂O emissions is agriculturally derived because soluble nitrate (NO₃⁻) and nitrite (NO₂⁻) ions, in the form of artificial fertilizers, are converted by some bacteria and fungi to atmospheric nitrogen in a process known as denitrification, in the course of which N₂O gas, an intermediate, escapes from cells.^1,2 Denitrification requires four specialist enzyme activities to catalyze the sequential conversion of NO₃⁻ → NO₂⁻ → NO → N₂O → N₂.^3,4 Of these the final step of denitrification, the 2-electron reduction of nitrous oxide to dinitrogen is catalysed by nitrous oxide reductase (N₂OR) an enzyme found in bacteria but not apparently in denitrifying fungi.⁵ A detailed investigation of the mechanism of N₂OR would help our understanding of this activity and how it might be harnessed in remedial processes aimed at limiting N₂O emissions.

N₂OR is a homodimer that is located in the bacterial periplasm. Recent X-ray structures of the enzymes from Pseudomonas nautica and Paracoccus denitrificans have revealed that each subunit (M_W = 65 kDa) is organized into two domains^6,7 (Figs. 1a and 1b). One cupredoxin-like domain contains a binuclear copper center known as Cu_A. The Cu_A center can undergo a one-electron redox change and hence has a function similar to that in the well-known aa₃-type cytochrome c oxidases (CcO) where it serves to receive an electron from soluble cytochromes c⁸ (Fig. 1c). For example, in Rhodobacter capsulatus, cytochrome c₂ is essential for electron transfer to N₂OR.⁹ Moreover, studies on Rhodobacter sphaeroides,¹⁰Wolinella succinogenes¹¹ and Paracoccus pantotrophus^12–14 have established the cognate periplasmic cytochromes c as natural electron donors to N₂OR. At least in vitro, the cupredoxin pseudoazurin can also function as an alternative electron donor to N₂OR isolated from P. pantotrophus.¹⁴ In the closely related organism P. denitrificans deletion of the cytochrome c₅₅₀ gene does not compromise its ability to reduce nitrous oxide suggesting that pseudoazurin may also be a natural electron donor to N₂OR in vivo.¹⁵

Fig. 1 Structure of nitrous oxide reductase and its copper clusters. (a) Structure of one subunit of the homodimeric N₂OR. (b) The subunit structure of N₂OR showing the head to tail arrangement that brings the electron storage centre Cu_A to within ca. 10 Å of the catalytic cluster Cu_Z in the other subunit. (c) The molecular structures and redox states of the Cu_A and Cu_Z clusters.

The second domain comprises a 7-bladed propeller of β-sheets that contains the catalytic site,⁷ a unique [Cu₄S] center, called Cu_Z, Figs. 1a and 1c. The original structural report at 2.9 Å resolution of the enzyme from P. nautica, failed to assign the μ₄ ligand in Cu_Z to inorganic sulfide. A combination of elemental analysis and resonance Raman spectroscopy of a ³⁴S enriched sample showed clearly that sulfur was a ligand to Cu_Z.¹⁶ Recently a higher resolution (1.6 Å) structure of the enzyme from P. denitrificans in which the electron density at the Cu_Z center is better resolved confirms this assignment.⁶ The distance between the Cu_A and Cu_Z centers within a single subunit is greater than 30 Å, a distance that precludes physiologically relevant rates of intra-subunit electron transfer. However, the two subunits are orientated “head to tail” such that the Cu_A center in one subunit lies only 10 Å from the Cu_Z center in the second ensuring that pairs of redox centers in opposite subunits form the catalytically competent unit.

A pre-requisite to understanding the catalytic cycle of N₂OR is the characterization of its redox properties. It is well established that Cu_A is capable of carrying only one electron. However, of the five oxidation levels that can be written for Cu_Z, assuming each copper ion can adopt either +I or +II states, only a diamagnetic state [Cu^II₂Cu^I₂S]⁴⁺ and a S = ½ paramagnetic form [Cu^II₁Cu^I₃S]³⁺ have been well characterized^17,18 (Fig. 1c). Recently reported work¹⁹ established that the enzyme will accept two electrons per subunit when purified under anaerobic conditions but only one when prepared aerobically. The difference between the two forms of the enzyme depends on the state of the Cu_Z center.¹⁹ An optically monitored equilibrium redox titration of the anaerobically prepared form of the enzyme showed that, in the absence of substrate, two electrons are transferred, each in a Nernstian process at +260 mV and +60 mV, the former corresponding to the addition of one electron to Cu_A and, the latter, of one electron to Cu_Z.¹⁹ The enzyme can be prepared with each monomer in a fully oxidized, one-electron reduced (in which only Cu_A is reduced), or two-electron reduced state (in which both Cu_A and Cu_Z have accepted one electron). The one-electron reduced form is prepared with ascorbate, whereas the two-electron reduced state requires exposure to excess dithionite. Under these conditions, one-electron reduction of Cu_Z to the [Cu₃^ICu₁^II]³⁺ state takes about one hour.²⁰ The steady state activities towards N₂O reduction of all these forms of the enzyme, when assayed with HH cytochrome c are similar at 0.034 U mg⁻¹.¹⁴ However, use of the artificial electron donors, benzyl viologen and methyl viologen rather than the natural electron donors, in steady-state kinetic measurements results in increased activity of the enzyme to ca. 5 U mg⁻¹ which increases further over the course of several hours. For the P. denitrificans enzyme we observe a rise from ca. 5 to ca. 50 U mg⁻¹ over 100 minutes. The activity of the Achromobacter cycloclastes N₂OR reached 125 U mg⁻¹ after 8 hours²¹ and in the case of P. nautica enzyme a maximum specific activity of 275 U mg⁻¹ was attained after 60 min.²² During this activation process the EPR signal characteristic of Cu_Z is lost suggesting that the cluster can be reduced by two electrons to the all Cu(I) state, [Cu^I₄S]²⁺.^21,22 Hence, the enzyme can be artificially super-reduced to a state carrying three electrons per subunit, one on Cu_A and two on Cu_Z. However, this activation process is slow and requires a much more powerful reductant than is available in the periplasm of a Gram-negative bacterium.

Mechanistic studies are required to establish the relevance of this super-reduced state of Cu_Z to enzymatic catalysis and to determine whether it is formed in small amounts under turnover conditions. Few mechanistic studies have been reported to date and these have used the artificial electron donors, benzyl viologen and methyl viologen in steady-state kinetic measurements,^23–27 rather than the natural electron donors that are needed if conclusions are to be drawn about the catalytic cycle. To date the most detailed kinetic study using cytochrome c as the electron donor, has been performed with N₂OR from W. succinogenes.¹¹ However, the kinetic behaviour of N₂OR from this organism may not be typical since N₂OR from W. succinogenes has an additional domain that contains a c-type heme not found in N₂OR from any other source.^26,28

This paper presents the first kinetic experiments on N₂OR carried out under pre-steady-state conditions. We have investigated the reactions of N₂OR in the oxidized, one-electron and two-electron reduced states with HH cytochrome c and N₂O, as well as carrying out single turnover experiments on the enzyme. HH cytochrome c is an efficient electron donor to several other enzymes of denitrification that have been purified from P. pantotrophus, for example, cytochrome cd₁ nitrite reductase.²⁹ Moreover, HH cytochrome c is reported to give a higher steady-state rate of N₂O reduction with N₂OR from P. pantotrophus compared with cytochromes c from other sources.¹⁴ The results of these experiments allow us to propose a model of the catalytic cycle for N₂OR.

Experimental

The preparation of N₂OR from P. pantotrophus has been described elsewhere.¹⁹ For the results reported here, the anaerobic form of the enzyme was used.¹⁹ The experiments described here were carried out with the enzyme in a single redox state generated by incubating N₂OR with 2 mM ferricyanide, ascorbate or dithionite for two hours to form the oxidized, one-electron and two-electron reduced forms, respectively. Excess oxidant or reductant were removed by buffer exchange using Microcon-30 centrifuge filter devices (Millipore). HH cytochrome c, purchased from Sigma and further purified as described elsewhere³⁰ was oxidized with 2 mM ferricyanide, or reduced with 2 mM ascorbate, and removed from excess reactants with Microcon-10 centrifuge filter devices. All operations were performed in a glove box under anaerobic conditions. The methyl viologen-linked N₂O reductase activity was measured spectrophotometrically as described elsewhere.²⁴

UV/Vis stopped flow experiments were performed with an Applied Photophysics Bio-Sequential DX.17 MV spectrophotometer or a High Tech SF 61DX2 instrument in a glove box, both with a 1 cm path length cell. Absorption changes were detected either with a photomultiplier at a single wavelength or a photodiode array from 430 to 1000 nm. In the latter case, a cut off filter at 420 nm was employed in order to avoid photo-oxidation of HH cytochrome c. All solutions for the stopped flow experiments were buffered with 2.5 mM Tris/HCl, pH 7.5 and were made anaerobic by flushing with nitrogen gas. In the single-turnover experiments, either the N₂OR solution or the HH cytochrome c solution was saturated with N₂O. Further details are described in the figure legends. The kinetic data were analyzed with the programs Origin (Microcal) and TableCurve 2D (Jandel). Rate constants were determined by fitting the experimental time courses to exponential or second order functions as appropriate. Non-linear concentration dependence data were fitted to a hyperbolic function by non-linear least squares procedures. Computer simulations of single turnover time courses were performed using the simulation program Gepasi.³¹

Results

The rate of electron transfer between reduced HH cytochrome c and oxidized N₂OR was determined in the absence of N₂O. Cytochrome c oxidation was monitored at 449 nm (Δε = 10.7 mM⁻¹ cm⁻¹) and redox changes in Cu_A were monitored at 830 nm (Δε = 2.0 mM⁻¹ cm⁻¹), a region of the spectrum in which there is no contribution from cytochrome c (Fig. 2). In these experiments no reduction of Cu_Z was observed, presumably because the reduction potential of Cu_Z, at +60 mV, is approximately 200 mV lower than the reduction potentials of either Cu_A or HH cytochrome c.¹⁹ Approximately 20% of the optical density change could not be resolved in our experiments since it occurs within the dead time of the instrument (ca. 2 ms). The magnitude of this absorbance change was determined from control experiments in which either cytochrome c or N₂OR solutions were mixed with anaerobic buffer. The extent of oxidation of cytochrome c was calculated from the absorbance change at 449 nm, corrected for the contributions from Cu_A and from the reaction occurring within the dead time. The amount of cytochrome c that is oxidized is dependent upon the concentration of cytochrome c. The concentration dependence is described by a hyperbolic curve (Fig. 3), consistent with an equilibrium being established between cytochrome c and Cu_A. The apparent reduction potential of the solution at equilibrium was determined from the ratio of the concentrations of oxidized to reduced cytochrome c using eqn. (1):In the best fit to these data, HH cytochrome c was determined to have a reduction potential that is 27 mV lower than that of Cu_A (Fig. 3). This contrasts with the published data which reports that both Cu_A and cytochrome c have the same reduction potential (E′ = +260 mV)^19,24,29,32 and presumably reflects the influence of protein–protein interaction. In the presence of excess reduced HH cytochrome c one electron is transferred from cytochrome c to oxidized N₂OR. The magnitude of the absorbance change at 830 nm indicates that Cu_A accepts one electron simultaneously with the oxidation of one equivalent of reduced cytochrome c. The observed rates of reduction of N₂OR are hyperbolically dependent upon cytochrome c concentration at low (<50 mM NaCl) ionic strengths, and do not increase beyond a limiting value of 150 s⁻¹ (data not shown). At higher salt concentrations the observed rates of electron transfer decrease reflecting the ionic nature of the interaction between HH cytochrome c and N₂OR (Fig. 4). Control-experiments at another concentration but the same Cc/N₂OR ratio showed no changes in the rate constants although these vary slightly from preparation to preparation of the enzyme.

Fig. 2 Electron-equilibration between HH cytochrome c and Cu_A. In a stopped flow experiment, 3.9 µM oxidized N₂OR monomer was mixed with a 9-fold excess of reduced HH cytochrome c. Time courses at 449 nm and 831 nm, are shown, which monitor the redox state of HH cytochrome c, and Cu_A, respectively. Additionally, single-exponential fits (k_obs = 150 s⁻¹) to the data are shown.

Fig. 3 Dependence of the electron-equilibration between HH cytochrome c and Cu_Aon the relative HH cytochrome c concentration. The number of electrons donated by HH cytochrome c was calculated from the absorbance change at 449 nm (squares) and 551 nm (triangles), and the electrons accepted by Cu_A from the change at 831 nm (circles). These data are fit to eqn. (1) (solid line).

Fig. 4 Dependence of the rate of electron-equilibration between HH cytochrome c and Cu_Aon the ionic strength. 4.5 µM oxidized N₂OR was mixed in a stopped flow experiment with 15 µM reduced HH cytochrome c at different NaCl concentrations in 2.5 mM TrisCl pH 7.5 buffer. Absorbance changes were monitored at 449 nm (squares) and at 831 nm (circles).

In other experiments, oxidized HH cytochrome c was rapidly mixed with two-electron reduced N₂OR in which both Cu_A and Cu_Z are each reduced by a single electron. Under these circumstances we observed a rapid electron transfer between Cu_A and cytochrome c (Fig. 5a) with a rate (150 s⁻¹) very similar to that seen in the cytochrome c oxidation experiment. However, Cu_Z, detected at 691 nm, was oxidized with a 1000-fold slower rate. The consequent slow intra-molecular electron transfer from Cu_Z to Cu_A leads to the slow, rate limited second phase of cytochrome c reduction (Fig. 5b). The rate of the latter phase does not vary with the concentration of HH cytochrome c but does vary slightly from preparation to preparation of N₂OR (k_obs = 0.1–0.4 s⁻¹).

Fig. 5 Reaction of oxidized HH cytochrome c with reduced N₂OR. 13 µM of dithionite-reduced N₂OR was mixed with 21 µM oxidized HH cytochrome c. The absorption changes recorded over a short (a) and a long (b) time scale are shown. Absorption changes were recorded at representative wavelengths: 449 nm (HH cytochrome c), 691 nm (Cu_Z), and 831 nm (Cu_A) are shown.

If either oxidized or one-electron-reduced N₂OR is incubated with N₂O, no changes in the oxidation state of the metal centers can be detected by UV/Vis spectroscopy. Two-electron reduced N₂OR shows only a very slow oxidation of Cu_Z by N₂O in spite of possessing two electrons which are sufficient for the reaction to proceed. However, when an excess of one-electron reduced N₂OR was mixed with reduced HH cytochrome c in the presence of high N₂O concentrations a single turnover occurred which was detected by the simultaneous oxidation of HH cytochrome c and Cu_A (Fig. 6). The time course for the reaction shows a lag-phase at all wavelengths studied, as previously reported in steady-state experiments.¹⁴ With increasing HH cytochrome c concentrations, the turnover is saturated when [HH cytochrome c] > 12 µM with k_obs ≤ 1.5 × 10⁻² s⁻¹ for the slower step. The single turnover rate was found to be independent of the N₂O concentration in the range of 0.5 to 12 mM in these experiments.

Fig. 6 Single turnover of semi-reduced N₂OR with N₂O and HH cytochrome c. The reaction of 9 µM semi-reduced N₂OR when mixed in a stopped flow apparatus with 7 µM reduced HH cytochrome c and 12 mM N₂O was monitored at 449 nm (circles) and 831 nm (triangles). Simulations of this reaction based on Scheme 1 are shown (solid lines). This started with the semi-reduced form. Steps 1 to 2, 7 to 8 and 8 to 1 were not included. The rates of other steps were identified either from other experiments or determined by best fitting of the data. The outcome was k₂ = 150 s⁻¹, k₋₂ = 150 s⁻¹, k₃ = 3 × 10⁻³ s⁻¹, k₋₃ = 0.15 s⁻¹, k₄ = 1 × 10⁶ s⁻¹ M⁻¹, k₋₄ = 5 s⁻¹, k₅ = 1 × 10⁵ s⁻¹, k₋₅ = 200 s⁻¹, k₆ = k₂, k₋₆ = k₋₂. k₇ and k₈ must be fast since they do not affect the simulation. A rate of 1 × 10⁵ s⁻¹ was used.

When the oxidized form of N₂OR was used, rather than the one-electron reduced form, fast electron equilibration between reduced HH cytochrome c and Cu_A, as described above, was also observed. The rate of turnover, however, was about the same as that seen in experiments using two-electron reduced N₂OR.

A single turnover experiment was also performed in three other ways. HH cytochrome c was pre-mixed with N₂O and rapidly mixed with N₂OR, was pre-mixed with N₂OR and then rapidly mixed with N₂O, and finally N₂OR was pre-mixed with N₂O and rapidly mixed with HH cytochrome c. No differences were observed between the kinetic behaviour under any of these conditions.

Discussion

In this study we have measured the electron transfer rates between cytochrome c, an electron donor that is a homologue of the physiological electron donor cytochrome c₅₅₀, and the Cu_A group of the enzyme, N₂OR. We have also studied the ionic dependence of the interaction between these proteins. Initial studies of the electron transfer process from reduced cytochrome c to N₂OR show that it is capable of reducing only the Cu_A center, in a process that is rate-limited at low ionic strengths by inter-molecular electron-transfer between the proteins (k_obs = 150 s⁻¹) rather than rate of association of the two proteins.

Interestingly, the rate of inter-molecular electron transfer that we observe is low in comparison with the similar reaction between cytochrome c and CcO.^33,34 It has been reported that in the case of CcO a conserved tryptophan residue, localized on the CcO surface close to the Cu_A, is crucial for a fast electron transfer reaction.³⁴ Mutation of this residue substantially lowers the electron transfer rate to values comparable to the cytochrome c–N₂OR reaction. Structural differences between CcO and N₂OR could therefore account for the lower transfer rates to N₂OR. Electron transfer between cytochrome c and N₂OR is, however, still much faster than the observed rate of turnover (described later) and, therefore, not rate limiting. Experimental investigations of electron transfer from reduced N₂OR to oxidized cytochrome c indicate the presence of two kinetic processes. The first of these processes, that is the electron transfer from Cu_A to cytochrome c, occurs at a concentration independent rate of 150 s⁻¹. The close correspondence of this rate with that of the opposed reduction rate is consistent with the essentially identical reduction potentials of the Cu_A and cytochrome c redox centers reported above. The second, slower process involves the intra-molecular electron transfer from the reduced Cu_Z center to the newly oxidized Cu_A center. This reaction occurs with a first order rate constant of 0.1 s⁻¹. Using this rate constant and the equilibrium constant derived from the measured difference in reduction potential between the Cu_Z and Cu_A centers we initially determined the intra-molecular rate of electron transfer from Cu_A to Cu_Z to be approximately 1 × 10⁻⁴ s⁻¹. Since Cu_A is oxidized faster than Cu_Z, the rate of intra-molecular electron transfer between these groups must be rate limiting, rather than the inter-molecular electron transfer to cytochrome c. Moreover, as further reduction of cytochrome c in the second phase occurs at the same rate as the oxidation of Cu_Z, neither the dissociation of the cytochrome c reduced first during the fast phase, nor the association of a second oxidized cytochrome c, which can accept a second electron, can be rate limiting.

These reactions were all studied in the absence of substrate, N₂O. Next we followed the reaction of the enzyme with N₂O in three different oxidation levels, oxidized, one-electron and two-electron reduced. Simple addition experiments indicate that none of the forms, oxidized, one-electron or two-electron reduced, are capable of reducing the substrate, N₂O at a significant rate. Hence, neither states of Cu_Z, [Cu^II₂Cu^I₂S]⁴⁺ or [Cu^II₁Cu^I₃S]³⁺, can react with N₂O even in the presence of excess chemical reductant. Moreover, ligand exchange studies indicate that N₂O does not bind tightly with one-electron reduced N₂OR.

In contrast a single turnover of the one-electron reduced form of N₂OR with substrate N₂O is observed if cytochrome c is used as the electron donor. The kinetics of oxidation of both cytochrome c and Cu_A are similar and both show a lag-phase (Fig. 6). No absorption change in the range from 430 to 1000 nm could be associated with any intermediate species during this lag-phase nor could we detect reduction of Cu_Z during this period. This might simply be due to very low concentrations of the intermediate species. The insensitivity displayed by the reaction kinetics to the order of mixing suggests that the association of the substrate is not rate limiting, a fact confirmed by the lack of dependence of the rate on the concentration of N₂O.

We have combined all of these findings to construct a possible scheme for the action of N₂OR with its soluble redox partner proteins (Scheme 1). There are two important features of this scheme. First, we envisage N₂O binding to the Cu_Z center when it is in the [Cu^II₁Cu^I₃S]³⁺ state within the one-electron reduced (Cu_A oxidized-Cu_Z reduced) form of the enzyme that forms at low concentrations during turnover (intermediate 4 of Scheme 1). Secondly, we propose that cytochrome c must form a complex with N₂OR to allow transfer of the second electron from Cu_A to Cu_Z to form the super-reduced [Cu^I₄S]²⁺ state that can reduce the substrate. Note that the one-electron reduced form would not be populated in the presence of dithionite. As a consequence we propose that the two-electron reduced enzyme formed in the reaction with dithionite is probably of no biological significance. Initial computer simulations employing Scheme 1 and the intra-molecular electron transfer rate from Cu_A to Cu_Z, determined from the independently measured redox potential for Cu_Z, yielded time courses for turnover which closely matched the form of the experimental data. However, the amplitude of the theoretical time courses only matched those of the experimental data if the intra-molecular electron transfer rate was 3 × 10⁻³ s⁻¹. In simple terms, the rate calculated from the redox potential data is too small to account for the rate of N₂O reduction determined experimentally. In order to account for the observed rate of N₂O reduction the redox potential of the Cu_Z centre in the enzyme during turnover must be 150 mV rather than 60 mV. This discrepancy could arise in two ways. Firstly, the reported redox potential for the Cu_Z centre was determined from the oxidation of the two-electron reduced enzyme in the absence of cytochrome c and so the value measured may not be appropriate for the enzyme during turnover. Alternatively as we have described above, evidence exists that the enzymic activity of N₂OR is affected by complex formation with cytochrome c. Whether this gating function derives directly from some conformationally coupled change in the structural environment of Cu_Z simultaneous with complex formation, or whether the redox potential change is consequent on the formation of a semi-reduced conformation cannot be determined at this time.

Scheme 1 Proposed catalytic cycle of N₂OR in the presence of reductant, cytochrome c, and substrate N₂O. The rate for the forward reaction of the internal ET (k₂) was calculated from k₋₂ and K_eq, based on the experimental midpoint potentials of Cu_A and Cu_Z in the absence of HH cytochrome c. It is assumed that the second one-electron reduction of Cu_A has a similar rate to the first.

Additionally, the question remains as to how N₂O is activated by the enzyme in the two-electron reduced state when the two electrons required to reduce the substrate are shared between two redox sites, Cu_A^Red and Cu_Z^Red, about 10 Å apart.^6,7 One possibility is that a one-electron intermediate [N₂O]⁻¹ is formed allowing a second electron to be transferred from Cu_A^Red to an intermediate Cu_Z–[N₂O]⁻¹. The second possibility is that reduced Cu_Z can be further reduced in the presence of both substrate (N₂O) and cytochrome c to a super-reduced, all cuprous state, [Cu^I₄S]²⁺. This state would be expected to be diamagnetic (Cu^I = d¹⁰, S = 0) and lacking the intense charge-transfer bands of the partially oxidized redox-states of Cu_Z.²¹ We favor a role for this species in our reaction scheme even though it would be formed transiently at very low concentrations and would be difficult to detect spectroscopically.

The participation of super-reduced Cu_Z would be consistent with recent reports that indicate that this species is found in the enzyme after prolonged reduction with either methyl-or benzyl-viologen and that it will directly reduce N₂O.^21,22 In steady-state assays using these artificial electron donors the rate of turnover is significantly enhanced if the reaction is initiated with fully reduced N₂OR,²¹ although, in the oxidizing environment of the bacterial periplasm there are not any known redox proteins with a sufficiently low reduction potential (> −400 mV) that can serve as electron donors to reduce fully the Cu_Z center.

It has previously been shown that steady-state rates observed with cytochrome c, or pseudoazurin, as electron donor (ca. 0.04 s⁻¹) are much lower than in experiments with viologens (1–10 s⁻¹) and depend on the electron donating protein.^11,14 It was, therefore, concluded that electron donation is probably the rate-limiting step in the catalytic cycle of N₂OR. Direct analysis of this reaction does not support this conclusion. In fact, the electron donation, including association and dissociation of cytochrome c, is fast compared to the turnover. However, we have found a slow internal electron transfer of about 0.1 s⁻¹ in the exergonic direction from Cu_Z to Cu_A. In order to explain the fast turnover with viologens one must assume that these electron donors bypass the electron-transfer path used with cytochrome c and that the decomposition of the substrate N₂O is not rate limiting just as in the case of the cytochrome c–N₂O turnover.

This argues that a conformational change in N₂OR is induced by its specific interaction with cytochrome c which in turn permits electron transfer between Cu_A and Cu_Z to allow formation of the super-reduced [Cu^I₄S]²⁺ form of Cu_Z. The postulated complex has not yet been confirmed by independent experiments. This model will form the basis for further experimentation aimed at understanding why intra-molecular electron transfer in N₂OR appears to be so highly controlled in turnover.

Acknowledgements

We thank the UK Research Councils EPSRC and BBSRC for support of this work through grants 83/B08413 and 83/B11958.

References

1. G. P. Robertson, E. A. Paul and R. R. Harwood, Science, 2000, 289, 1922–1925.
2. N. Takaya, M. A. Catalan-Sakairi, Y. Sakaguchi, I. Kato, Z. Zhou and H. Shoun, Appl. Environ. Microbiol., 2003, 69, 3152–3157.
3. B. C. Berks, S. J. Ferguson, J. W. B. Moir and D. J. Richardson, Biochim. Biophys. Acta, 1995, 1232, 97–173.
4. D. J. Richardson and N. J. Watmough, Curr. Opin. Chem. Biol., 1999, 3, 207–219.
5. H. Shoun and T. Tanimoto, J. Biol. Chem., 1991, 266, 11078–11082.
6. T. Haltia, K. Brown, M. Tegoni, C. Cambillau, M. Saraste, K. Mattila and K. Djinovic-Carugo, Biochem. J., 2003, 369, 77–88.
7. K. Brown, M. Tegoni, M. Prudencio, A. S. Pereira, S. Besson, J. J. Moura, I. Moura and C. Cambillau, Nat. Struct. Biol., 2000, 7, 191–195.
8. B. C. Hill, J. Bioenerg. Biomembr., 1993, 25, 115–120.
9. D. J. Richardson, L. C. Bell, A. G. McEwan, J. B. Jackson and S. J. Ferguson, Eur. J. Biochem., 1991, 199, 677–683.
10. M. Itoh, K. Matsuura and T. Satoh, FEBS Lett., 1989, 251, 104–108.
11. C.-S. Zhang and T. C. Hollocher, Biochim. Biophys. Acta, 1993, 1142, 253–261.
12. I. Matchová and I. Kucera, Biochim. Biophys. Acta, 1991, 1058, 256–260.
13. F. C. Boogerd, H. W. van Verseveld and A. H. Stouthamer, FEBS Lett., 1980, 113, 279–284.
14. B. C. Berks, D. Baratta, D. J. Richardson and S. J. Ferguson, Eur. J. Biochem., 1993, 212, 467–476.
15. R. J. Van Spanning, C. Wansell, N. Harms, L. F. Oltmann and A. H. Stouthamer, J. Bacteriol., 1990, 172, 986–996.
16. T. Rasmussen, B. C. Berks, J. Sanders-Loehr, D. M. Dooley, W. G. Zumft and A. J. Thomson, Biochemistry, 2000, 39, 12753–12756.
17. P. Chen, S. DeBeer George, I. Cabrito, W. E. Antholine, J. J. Moura, I. Moura, B. Hedman, K. O. Hodgson and E. I. Solomon, J. Am. Chem. Soc., 2002, 124, 744–745.
18. V. S. Oganesyan, T. Rasmussen, S. Fairhurst and A. J. Thomson, Dalton Trans., 2004, 996–1002.
19. T. Rasmussen, B. C. Berks, J. N. Butt and A. J. Thomson, Biochem. J., 2002, 364, 807–815.
20. J. Riester, W. G. Zumft and P. M. Kroneck, Eur. J. Biochem., 1989, 178, 751–762.
21. J. M. Chan, J. A. Bollinger, C. L. Grewell and D. M. Dooley, J. Am. Chem. Soc., 2004, 126, 3030–3031.
22. S. Ghosh, S. I. Gorelsky, P. Chen, I. Cabrito, J. J. Moura, I. Moura and E. I. Solomon, J. Am. Chem. Soc., 2003, 125, 15708–15709.
23. J. K. Kristjansson and T. C. Hollocher, J. Biol. Chem., 1980, 255, 704–707.
24. C. L. Coyle, W. G. Zumft, P. M. Kroneck, H. Korner and W. Jakob, Eur. J. Biochem., 1985, 153, 459–467.
25. S. W. Snyder and T. C. Hollocher, J. Biol. Chem., 1987, 262, 6515–6525.
26. S. Teraguchi and T. C. Hollocher, J. Biol. Chem., 1989, 264, 1972–1979.
27. K. Sato, A. Okubo and S. Yamazaki, J. Biochem., 1999, 125, 864–868.
28. J. Simon, O. Einsle, P. M. Kroneck and W. G. Zumft, FEBS Lett., 2004, 569, 7–12.
29. C. D. Richter, J. W. Allen, C. W. Higham, A. Koppenhofer, R. S. Zajicek, N. J. Watmough and S. J. Ferguson, J. Biol. Chem., 2002, 277, 3093–3100.
30. D. L. Brautigan, S. Ferguson-Miller and E. Margoliash, Methods Enzymol., 1978, 53, 128–164.
31. P. Mendes, Comput. Appl. Biosci., 1993, 9, 563–571.
32. M. Prudencio, A. S. Pereira, P. Tavares, S. Besson, I. Cabrito, K. Brown, B. Samyn, B. Devreese, J. Van Beeumen, F. Rusnak, G. Fauque, J. J. Moura, M. Tegoni, C. Cambillau and I. Moura, Biochemistry, 2000, 39, 3899–3907.
33. L. M. Geren, J. R. Beasley, B. R. Fine, A. J. Saunders, S. Hibdon, G. J. Pielak, B. Durham and F. Millett, J. Biol. Chem., 1995, 270, 2466–2472.
34. K. Wang, Y. Zhen, R. Sadoski, S. Grinnell, L. Geren, S. Ferguson-Miller, B. Durham and F. Millett, J. Biol. Chem., 1999, 274, 38042–38050.

---

Footnote

† Based on the presentation given at Dalton Discussion No. 8, 7–9th September 2005, University of Nottingham, UK.

---