Paracoccus pantotrophus expresses two nitrate reductases associated with respiratory electron transport, termed NapABC and NarGHI. Both enzymes derive electrons from ubiquinol to reduce nitrate to nitrite. However, while NarGHI harnesses the energy of the quinol/nitrate couple to generate a transmembrane proton gradient, NapABC dissipates the energy associated with these reducing equivalents. In the present paper we explore the nitrate reductase activity of purified NapAB as a function of electrochemical potential, substrate concentration and pH using protein film voltammetry. Nitrate reduction by NapAB is shown to occur at potentials below approx. 0.1 V at pH 7. These are lower potentials than required for NarGH nitrate reduction. The potentials required for Nap nitrate reduction are also likely to require ubiquinol/ubiquinone ratios higher than are needed to activate the H+-pumping oxidases expressed during aerobic growth where Nap levels are maximal. Thus the operational potentials of P. pantotrophus NapAB are consistent with a productive role in redox balancing. A Michaelis constant (KM) of approx. 45 μM was determined for NapAB nitrate reduction at pH 7. This is in line with studies on intact cells where nitrate reduction by Nap was described by a Monod constant (KS) of less than 15 μM. The voltammetric studies also disclosed maximal NapAB activity in a narrow window of potential. This behaviour is resistant to change of pH, nitrate concentration and inhibitor concentration and its possible mechanistic origins are discussed.
Paracoccus pantotrophus (the α-proteobacterium formerly known as Thiosphaera pantotropha and Paracoccus denitrificans GB17) has emerged as the paradigm organism for studies of aerobic nitrate respiration, a process now thought to be widespread among bacteria in which nitrate is used as an electron sink to facilitate redox balancing during oxidative metabolism of reduced carbon substrates [1,2]. In fact P. pantotrophus expresses two nitrate reductases associated with respiratory electron transport, Nap and Nar. These enzymes catalyse the same reaction,
but they are distinguished by their cellular location, biochemistry and pattern of expression (Figure 1). Nar is expressed during anaerobic growth in the presence of nitrate. NarI, the membrane spanning di-haem subunit, acquires the electrons required for nitrate reduction from ubiquinol oxidation. Protons are released to the periplasm and electrons pass along a chain of iron–sulfur clusters in NarH and NarG to the site of nitrate reduction, which is a molybdenum-bis-molybdopterin guanine dinucleotide cofactor (termed Mo[MGD]2) with aspartate co-ordination [3,4]. The disposition of catalytic centres allows Nar to couple cytoplasmic nitrate reduction with ‘periplasmic’ quinol oxidation. As a consequence free energy associated with the ubiquinol/nitrate redox couple is conserved as a transmembrane proton electrochemical gradient .
The respiratory nitrate reductases of Paracoccus pantotrophus
In contrast with the expression of nar, Nap levels in P. pantotrophus are maximal during aerobic growth with the highly reduced carbon source butyrate . Ubiquinol oxidation by Nap releases protons to the periplasm where nitrate reduction also occurs (Figure 1). Energy conservation is precluded and Nap is proposed to dissipate excess reducing equivalents from the ubiquinol pool. Thus Nap is a redox-balancing system that operates to poise the aerobic respiratory electron transport system to maximize growth rate. Nap, like Nar, employs three subunits to couple ubiquinol oxidation with nitrate reduction [1,2]. The membrane anchored tetra-haem cytochrome, NapC, catalyses ubiquinol oxidation. Haem groups in NapB and an iron–sulfur cluster in NapA relay electrons to a cysteine co-ordinated Mo[MGD]2 where nitrate reduction occurs .
Stable heterodimeric nitrate reductases, NarGH and NapAB, are formed from Nar and Nap respectively in the absence of their quinol-oxidizing subunits [8,9]. These soluble enzymes facilitate mechanistic studies of nitrate reduction and we have shown previously that graphite electrodes can substitute for NarI to act as effective electron donors to P. pantotrophus NarGH [10–12]. Nitrate reduction driven by electrodic reduction of NarGH using PFV (protein film voltammetry) is characterized by a Michaelis constant (KM) of approx. 20 μM. This compares favourably with the KM of approx. 13 μM observed when quinol provides electrons to NarGHI, and supports a functional analogy between the graphite electrode and the quinol-oxidizing subunit for studies of nitrate reduction [8,10]. These voltammetric studies offered additional information since they resolved NarGH activity across the electrochemical potential domain. At pH 7 NarGH was shown to be active below approx. 0.2 V and at potentials that should be readily accessible to the quinol pool during anaerobic respiration [E0′ (reduction potential at pH 7) UQ/UQH2 approx. 0.04 V] . Two catalytically competent but kinetically distinct forms of NarGH were also disclosed and proposed to interconvert reversibly through redox transformation of an iron–sulfur cluster or the Mo5+/4+ couple. P. pantotrophus NapAB has not been studied by PFV although it has been the subject of extensive spectroscopic and kinetic analyses which have raised three important questions [9,14–19]. First, NapA is closely related to a family of cytoplasmic nitrate reductases that are involved in anabolic, rather than catabolic, nitrate metabolism and the current view is that NapA evolved from a primordial cytoplasmic assimilatory enzyme . In the case of the cytoplasmic assimilatory nitrate reductases, these enzymes operate in the relatively reducing environment of the cytoplasm using electron donors such as ferredoxin or NAD(P)H that have reduction potentials in the region of −0.3 to −0.4 V. Evolution into a periplasmic respiratory enzyme required that Nap engage with quinolic electron donors having significantly higher reduction potentials. PFV of one assimilatory nitrate reductase, Synecochoccus elongatus NarB, showed that it became active below approx. −0.2 V . Since P. pantotrophus NapA shares 70% sequence similarity with S. elongatus NarB, and all of the residues predicted to be essential for catalysis are conserved between the two enzymes, the question arises as to whether NapAB has evolved to operate in a higher potential domain than NarB and one that is more akin to the distinct, but quinol-dependent, NarGH enzyme?
The second important question about NapAB relates to its apparent affinity for nitrate. In Escherichia coli, NapA activity is associated with anaerobic rather than aerobic metabolism . Studies of NarG and NapA mutants have led to the suggestion that NapA represents a high-affinity nitrate-scavenging system that provides a selective advantage in nitrate-limited environments . Measurements of the Monod constant (KS) yield a value less than 20 μM for nitrate in a bacterial culture expressing napA but mutated in nar. Since NapA is periplasmic it is likely that this KS reflects the KM of the enzyme. However, spectrophotometric studies of nitrate reduction by purified P. pantotrophus NapAB gave a KM of approx. 1000 μM [14,16]. This seems a high KM given that E. coli NapA and P. pantotrophus NapA share 80% sequence similarity, which includes a high degree of conservation for the active-site residues. Thus a reinvestigation of the kinetic properties of purified P. pantotrophus NapAB is warranted. Finally, it is of interest to address whether NapAB exists in catalytically active but kinetically distinct forms that can be reversibly interconverted by a redox ‘switch’. Such behaviour is displayed by NarGH(I) from P. pantotrophus and E. coli, in addition to S. elongatus NarB and Rhodobacter sphaeroides NapAB [10,21,24,25]. PFV is uniquely placed to address this issue, and those raised above, by defining P. pantotrophus NapAB activity as a function of electrochemical potential, substrate concentration and pH. In the present paper we present results from such a study together with an analysis of Nap activity in intact cells of P. pantotrophus.
Samples of the purified periplasmic nitrate reductase, NapAB, were isolated independently from three separate growths of P. pantotrophus strain M6 . Protein purification was by modification of the published method in which the order of the size-exclusion and Ni(II)-IMAC (immobilized metal-ion-affinity chromatography) steps were reversed . The resulting samples were analysed by reducing SDS/PAGE. Gels stained for haem-linked peroxidase activity showed a single band at ∼16 kDa consistent with the presence of NapB. Coomassie-stained gels showed a single dominant band at ∼90 kDa consistent with the presence of NapA and from which the NapAB samples were judged to be >95% pure by densitometry. Purified, air-equilibrated samples had A408nm/A280nm between 0.88 and 0.91 and concentrations were determined spectrophotometrically using ϵ408nm=220 mM−1·cm−1. Purified enzyme was exchanged into 20 mM Hepes and 500 mM NaCl (pH 7.5) containing 20% (v/v) glycerol and stored as aliquots in liquid nitrogen.
Nitrate reduction by Nap in intact cells of P. pantotrophus
Cells of P. pantotrophus strain M6 were suspended in 10 mM phosphate, 30 mM sodium succinate and 10 mM ammonium sulfate (pH 7), at a concentration of 0.5 mg of dry weight·ml−1. The suspension was incubated anaerobically under a nitrogen atmosphere in a 5 ml reaction vessel (at 30 °C) and was stirred continuously. Pulses of nitrate (70 or 35 μM) were introduced to the vessel through stainless steel ports by injection from a Hamilton syringe and the nitrate concentration was monitored with a nitrate electrode.
All solutions were prepared using analytical reagent grade water (Fisher Scientific) and reagents of Analar, or higher, purity. The majority of voltammetric and potentiometric studies were performed in buffer-electrolyte solutions containing 100 mM NaCl and 25 mM of glacial acetic acid, Mes, Pipes, Taps or Ches. The solution pH was adjusted to the desired value with concentrated NaOH or HCl solutions. Where phosphate buffers were used, these were prepared by titration of monobasic and dibasic potassium phosphate solutions to achieve the desired pH. Solutions containing defined concentrations of neomycin-B sulfate, KNO3, NaN3 and KSCN were prepared by dilution of concentrated, aqueous stock solutions with buffer-electrolyte. The pH of the resultant solutions, before and after experimentation, was confirmed to be that desired.
Spectrophotometric assays of nitrate reduction
Assays of NapAB activity were performed anaerobically at room temperature (21 °C) in 3.4 ml of 25 mM Pipes and 100 mM NaCl (pH 7.0) containing 1 mM reduced Methyl Viologen and the desired concentration of nitrate. Dithionite (typically 5 μl of a freshly prepared 17 mg·ml−1 solution) was added to give a stable absorbance of approx. 1 at 600 nm and assays were initiated by the addition of anaerobic NapAB to give a final concentration of 5.4 nM. Rates were calculated using ϵ600nm=13.7 mM−1·cm−1 for reduced Methyl Viologen.
Voltammetry was performed inside a Faraday cage housed in a nitrogen-filled chamber (atmospheric oxygen <2 p.p.m.) as described previously . The glass electrochemical cell contained three electrodes and the sample chamber was maintained at 20 °C. The reference electrode was calomel (saturated KCl) and the working electrodes were pyrolytic graphite edge with a geometric surface area of 0.071 cm2. Immediately prior to each experiment the working electrode surface was lightly abraded with ‘Wet and Dry Abrasive Paper’ of fine grade (English Abrasives and Chemicals) and polished with an aqueous 0.3 μm Al2O3 slurry. After sonication, the electrode was rinsed, dried with a tissue and a few microlitres of ice-cold solution containing 5 μM NapAB and 2 mM neomycin-B sulfate as co-adsorbate were placed on the electrode. After approx. 30 s, excess solution was removed from the electrode which was then immersed in the desired buffer-electrolyte. Potentials are quoted with respect to the SHE (standard hydrogen electrode) following addition of +0.241 V to the measured values.
Quantitative analysis of NapAB voltammetry was restricted to the pH range 5–9 where reproducible behaviour was observed. Catalytic currents were determined after subtraction of a ‘baseline’ film response recorded in the absence of nitrate. When comparing the magnitude of responses from different films this was performed after normalizing the response of each film to that measured in 2 mM nitrate, 2 mM neomycin, 25 mM Mes and 100 mM NaCl (pH 6) at 20 °C with a scan rate of 20 mV·s−1 and an electrode rotation at 3000 rev./min. For kinetic analysis the catalytic currents from a nitrate titration were corrected for a first-order loss of signal magnitude over the time of the experiment and then values of KM and imax (catalytic current magnitude) were determined by regression in the MicroCal Origin program. Lineweaver–Burk and Hanes analyses gave results in good agreement with those from direct fit to the Michaelis–Menten equation. Ecat and Eswitch were defined as the potentials of the maximum and minimum respectively in a plot of the first derivative of the baseline-subtracted catalytic current with respect to applied potential. There was no discernable difference between NapAB PFV recorded in phosphate and Goods buffers for any given pH and nitrate concentration.
Mediated potentiometric titrations were performed in 25 mM Mes and 100 mM NaCl (pH 6.0) using modifications of the method of Dutton et al.  with sodium dithionite as the reductant and potassium ferricyanide as the oxidant. Full details are provided in the Supplementary Information (at http://www.BiochemJ.org/bj/409/bj4090159add.htm). The variation of signal intensity with equilibration potential was fitted to the Nernst equation with the MicroCal Origin program.
Defining catalytic PFV from NapAB
To assess the nitrate reductase activity of purified NapAB 2 μl of an ice-cold sample containing 5 μM enzyme and 2 mM neomycin were placed on a pyrolytic graphite ‘edge’ electrode. After a few seconds excess solution was removed and the electrode placed in buffer-electrolyte. Cyclic voltammetry with rapid electrode rotation showed a featureless response but the introduction of nitrate produced persistent negative currents below approx. 0.1 V (Figure 2). These negative currents reflected electrocatalytic reduction. Since control experiments showed the catalytic currents were not observed in the absence of enzyme or nitrate they were attributed to NapAB catalysed nitrate reduction driven by direct electron transfer from the electrode to NapAB.
PFV of P. pantotrophus NapAB
The most striking feature of the catalytic response displayed by NapAB is the maximum, or ‘peak’, of activity observed at approx. −0.15 V (Figure 2). The peak reflects increased NapAB activity on first lowering the electrode potential followed by a region where activity drops to a constant lower level despite an increased driving force for the reaction that is catalysed. The peak of activity was observed, regardless of scan direction, during at least 30 min of continuous voltammetry and for three independent preparations of NapAB. The peak also persisted during catalysis at pH 5 through to pH 9 and for nitrate concentrations between 5 and 2000 μM. At more alkaline pH the smaller catalytic currents suggested lower enzyme activity. Nevertheless for all conditions investigated the maximum ‘peak’ activity was approx. 2-fold that displayed by NapAB at −0.4 V where the nitrate reduction rate becomes independent of a further increase of driving force (Figure 3A).
Characteristic parameters derived from catalytic PFV of NapAB
Neomycin is often used as a co-adsorbate in PFV. Its influence on the catalytic voltammetry of NapAB was assessed in experiments performed without neomycin present. These produced catalytic waves with shapes, position and initial magnitude in good agreement with those recorded when neomycin was present. However, the magnitude of the response was significantly more persistent in the presence of neomycin (t½∼20 min rather than 8 min in the absence of neomycin) and since this facilitated data collection and analysis, neomycin was included in all further experiments. The catalytic wave shape was also found to be independent of the presence of 140 mM nitrite and variation of the electrode rotation rate between 2000 and 3000 rev./min. These results showed that the response was free from limitations due to product inhibition or the rates of substrate delivery/product removal from the electrode surface. Thus the peak of activity displayed by NapAB reflects reversible switching of the enzyme between catalytically active but kinetically distinct forms on traversing the electrochemical potential domain.
Kinetic analysis of NapAB nitrate reduction
The catalytic response from a NapAB film measured at 5 mV·s−1 was found to superimpose on that measured at 100 mV·s−1 with no sign of a smearing towards lower potentials that would be seen if the catalytic currents were limited by a range of sluggish interfacial electron exchange rates . Thus the kinetic parameters describing NapAB catalysis were obtained from direct analysis of the catalytic current magnitudes (icat) recorded in response to sequential nitrate additions (Figure 2). For each nitrate concentration catalytic current magnitudes were measured at the peak potential (ipeak Ecat) and −0.4 V (i−0.4Vcat) then adjusted for the first-order loss of signal magnitude noted during each experiment. The results were fitted to the Michaelis–Menten equation and values of KM and the maximum catalytic current magnitude (imax) determined for each potential (Figure 4). Catalytic currents could not be converted into turnover numbers since the amount of enzyme giving rise to the voltammetric response is unknown. However, the response of imax to the change of pH (or other experimental variables) quantifies the response of kcat to such perturbations.
Kinetic analysis of catalytic PFV from NapAB
The values of i−0.4Vmax and ipeak Emax showed that NapAB activity was maximal at more acidic pH and decreased in a sigmoidal manner towards higher pH (Figure 3B). This behaviour was confirmed, and its reversibility established, when NapAB-coated electrodes were transferred between solutions of distinct pH in experiments which also demonstrated that the catalytic voltammetry originated from enzyme adsorbed on to the electrode surface. The pH dependence of (i−0.4Vmax and ipeak Emax) were well-described by the equation:
which relates modulation of NapAB activity to a single protonation event with pKa=7.8±0.3 where iacidmax is the maximal catalytic current in the acid limit and ialkmax that in the alkaline limit (Figure 3B).
Inspection of the Michaelis constants for nitrate reduction showed Kpeak EM was smaller than K−0.4VM and that both values increased at higher pH values (Figure 3C). Perhaps more significantly the values of Kpeak EM and K−0.4VM were 23 and 45 (±5) μM respectively at pH 7. These are values considerably below that of approx. 1000 μM previously reported using dithionite-reduced Methyl Viologen as the electron donor so spectrophotometric analysis of the activity of these NapAB samples was also undertaken [9,14,16]. The assays were initiated by addition of NapAB and a rate acceleration, i.e. the appearance of a steeper absorbance versus time trace, was noted during every experiment (see for example the inset to Figure 5). These rate accelerations are significant because they reflect increased activity of NapAB as the Methyl Viologen becomes oxidized and the solution potential rises. The results are consistent with the peak of NapAB activity arising from intrinsic properties of the enzyme rather than the interfacial nature of the voltammetric experiment. This is an interpretation supported by the observation of peaked catalytic waves in PFV at basal plane graphite electrodes (results not shown).
Spectrophotometric analysis of NapAB nitrate reduction
Analysis of the initial velocities from the spectrophotometric assays yielded KM=112±12 μM and a turnover number of 58±1 s−1 (Figure 5). Similar analysis for the steepest gradient in each assay gave a KM of 176±23 μM and turnover number of 104±3 s−1. Thus the specificity constant (kcat/KM) for NapAB nitrate reduction is in the order of 6×105 M−1·s−1. The KM values obtained with the spectrophotometric assay were slightly higher than those measured by PFV for the same enzyme sample, but the results are within acceptable variation given the physical and chemical distinctions between the two assay formats. The reason for the discrepancy between the present and previous spectrophotometric analyses is unclear but most importantly the KM values defined in the present study through two independent methods bring the kinetic description of P. pantotrophus NapAB into line with those for related enzymes [14,21,25].
To attempt to correlate the KM values determined for the purified enzyme with that of the enzyme when it is operating in the intact cell with physiological electron donors, Nap nitrate reduction was monitored in whole cell suspensions of P. pantotrophus strain M6. This strain is a Tn5 insertion mutant deficient in the membrane-bound nitrate reductase Nar so that Nap is the only functional respiratory nitrate reductase . In wild-type cells nap expression is confined to aerobic growth on highly reduced carbon substrates. However, P. pantotrophus M6 also carries a point mutation in the nap promoter that leads to deregulation of the nap operon so that it is expressed under both anaerobic and aerobic conditions . Since Nap is the only functional respiratory nitrate reductase in P. pantotrophus M6, nitrate reduction by Nap in intact cells can be monitored by measuring nitrate disappearance from anaerobic cell suspensions using a nitrate electrode. The response of the nitrate electrode to nitrate was approximately linear in the 0–70 μM range. This is illustrated in Figure 6 where two pulses of nitrate (70 and 35 μM) were added to an anaerobic P. pantotrophus strain M6 cell suspension. The electron source was succinate, the catabolism of which results in electron input to the quinol pool via succinate dehydrogenase and NADH dehydrogenase with electrons then leaving the quinol pool via NapC. The maximum rate of nitrate reduction was 60 nmol·min−1·mg of cells−1 and this was sustained until the nitrate concentration approached 15 μM. At this point the nitrate reduction rate decelerated (Figure 6). This indicated that the KS of the cells for nitrate was <15 μM and similar to that estimated for intact cells of Rhodobacter capsulatus (<3 μM) and E. coli (<10 μM) expressing nap [23,30]. Since nitrate does not need to be transported across the cytoplasmic membrane for reduction by Nap to occur it is likely that the KS of the cells reflects the KM of Nap itself and for P. pantotrophus NapAB the low micromolar value of KS is certainly in line with the PFV results.
Nitrate reduction by intact cells of P. pantotrophus strain M6
Inhibition of NapAB nitrate reduction
Thiocyanate and azide have been shown to inhibit NapAB, so nitrate titrations at a number of inhibitor concentrations were analysed at −0.09 and −0.40 V to assess the behaviour of the high- and low-potential enzyme forms respectively, see for example Figure 7 [14,16]. Lineweaver–Burk plots showed that thiocyanate acted as a competitive inhibitor at both potentials by increasing KM without affecting imax (Figure 7A and Supplementary Figure 1 at http://www.BiochemJ.org/bj/409/bj4090159add.htm). Previous work has reported the dissociation constant (Kd) for the thiocyanate–NapAB interaction as 4±0.6 mM . The Kd values calculated from the present study are in good agreement with this, 6 (±3) and 3 (±2) mM at −0.09 and −0.40 V respectively, and show a similar affinity of thiocyanate for the high- and low-potential forms of NapAB . Thus the ratio of inhibited to uninhibited enzyme is defined by the ratio of thiocyanate to nitrate regardless of electrode potential and the wave shapes parallel those seen during nitrate titrations in the absence of thiocyanate, e.g. compare Figure 2 with Figure 7(A). In contrast, azide induces a change of catalytic wave shape (Figure 7B). A peak of activity is retained but inhibition of the high-potential form is more marked than inhibition of the low-potential form. Lineweaver–Burk plots indicate mixed inhibition at −0.09 and −0.40 V since both KM and imax were sensitive to azide concentration (Figure 7B and Supplementary Figure 2 at http://www.BiochemJ.org/bj/409/bj4090159add.htm). The plots failed to define a single intercept that would indicate complete (linear) inhibition [31,32]. This is consistent with the approach of imax to a finite but non-zero value at the higher azide concentrations used in the present study (Supplementary Figure 2) and suggests the presence of a catalytically competent NapAB–azide complex.
Inhibition of NapAB nitrate reduction
Correlation of NapAB catalytic voltammetry with cofactor reduction potentials
The catalytic wave defines windows of potential over which the activity of NapAB is modulated, up or down, as the potential is varied. The steepest point on each flank of the peak quantifies the potential of that part of the wave, Ecat for the high-potential flank and Eswitch for the low-potential flank (Figure 2). At 2 mM nitrate Eswitch varied linearly with pH and by −15 mV per pH unit (Figure 3D). Ecat showed a more complex pH-dependence and one that indicated protonation of an oxidized form of the enzyme. If it is assumed that Ecat reflects the properties of a single redox couple within NapAB its pH-dependence can be described by the equation:
where Eacid is Ecat in the acid limit, pKa describes the properties of ionizable site and n is the number of electrons associated with the redox event. The overlapping contributions to the catalytic wave shape preclude unambiguous assignment of the electron stoichiometry associated with Ecat, therefore n=1 was assumed to gain an approximation of pKa and a good-fit to the data was found with pKa=7.2±0.1 and Eacid=0.3±0.01 V. At 5 μM nitrate the values of Ecat and Eswitch were higher by approx. 50 and 100 mV respectively than those at 2 mM nitrate. However, the pH-dependence of both Ecat and Eswitch at 5 μM nitrate was similar to their respective behaviours at 2 mM nitrate. These results suggest that the ionizing residues responsible for the modulation of Ecat and Eswitch are accessible in both the free enzyme and Michaelis complex and may not form integral parts of the nitrate reduction site.
A comparison of Ecat, Eswitch and the reduction potentials of centres within NapAB can provide a basis for considering possible origins of its distinctive catalytic properties. Previous studies have determined haem reduction potentials at pH 7.4 and the [4Fe–4S]2+/1+ reduction potential at pH 7.0 (Table 1). To complement these values and assess the properties of NapAB used in the present study we performed potentiometric titrations at pH 6 where NapAB is most active (Supplementary Figures 3 and 4 at http://www.BiochemJ.org/bj/409/bj4090159add.htm). Haem reduction was monitored by UV-visible absorption spectroscopy and reduction of the [4Fe–4S]2+ cluster was followed by EPR spectroscopy. The results are summarized in Table 1 where it is seen that the haems have higher reduction potentials than the [4Fe–4S] cluster and that each of these redox transitions occurs in the range of potentials where modulations of NapAB activity are detected. In an attempt to assess active-site redox chemistry we also measured EPR spectra of our samples at 66 K where signals arising from Mo5+ species are visible [14,19]. As in previous studies, a number of signals were detected, none of which had an intensity that accounted for more than approx. 30% of molybdenum in the sample despite full molybdenum loading .
|Cofactor||Reduction potential (mV)||pH||Spectroscopic method||Reference|
|[4Fe–4S]2+/1+||−124±20||6.0||EPR||The present study|
|Haems||−18±20 77±20||7.0||UV-visible absorption|||
|Haems||−30±20||6.0||UV-visible absorption||The present study|
Bacterial nitrate reductases divide into two structural groups, those for which the protein ligand to the molybdenum ion originates from an aspartate residue, the Nar group, and those for which it arises from a cysteine residue, the Nap and Nas enzymes . The Nap and Nar enzymes receive electrons from the quinol pool which for P. pantotrophus is comprised of ubiquinol and ubiquinone. In the present study we see that NapAB activity is turned on below approx. 0.15 V and so, like NarGH, its operational potentials appear well-matched to the properties of the ubiquinol electron donor (Figure 8). Interesting comparisons can be made with the assimilatory nitrate reductase (NarB) from S. elongatus. NarB activity is switched on below approx. −0.2 V and at potentials well-matched to the NAD+/NADH couple (E0′∼−0.32 V) via which electrons from sugar oxidation are passed to cytoplasmic reductive metabolism . Thus the activities of these enzymes appear to be tuned to reflect their source of electrons rather than the immediate molybdenum co-ordination sphere. In light of this it is interesting to note that NapA is thought to have evolved from the cytoplasmic eubacterial assimilatory enzymes . In this evolutionary view the primordial ‘cytoplasmic’ enzyme became fused to a TAT (twin-arginine targeting) signal peptide that allowed its translocation to the periplasm and extra proteins, NapB and NapC, were recruited to allow communication with quinol. If our observations to date are generally applicable this evolution was also accompanied by a positive shift of enzyme-operating potential.
Schematic representation of the normalized activities of P. pantotrophus NapAB (thick line) and NarGH (thin line) together with potentials that may be accessed through modulation of the ubiquinol (UQH2) content (grey) of the cytoplasmic membrane at pH 6 and 7
When the periplasmic respiratory system of P. pantotrophus is considered as a whole it is notable that the three denitrification enzymes with periplasmic, or periplasmic facing, active sites are coupled to electron transfer at a point downstream of the cytochrome bc1 complex . Thus the cytochrome cd1 nitrite reductase, the integral membrane nitric oxide reductase and the nitrous oxide reductase accept electrons from either reduced cytochrome c550 or pseudoazurin (E0′∼0.25 V) generated by the cytochrome bc1 complex in protonmotive processes. Here we find that the operating potential of NapAB is sufficiently positive to allow it to couple into the respiratory chain at the level of the quinol pool but not sufficiently high to allow it to operate upstream of cytochrome c550 and pseudoazurin. Thus NapAB activity cannot be coupled into the cytochrome bc1-complex protonmotive activity.
Closer inspection of the activity-potential profiles for P. pantotrophus NapAB and NarGH in 4 μM nitrate and at pH 6 and 7 shows that lower potentials are required to activate NapAB nitrate reduction (Figure 8). Indeed, lower potentials are required to activate NapAB than NarGH in like-for-like comparisons and although it is difficult to know the environments in which each nitrate reductase will operate some further observations can be made if these general properties are related to potentials that may be imposed by the UQ/UQH2 pool (Figure 8). From a thermodynamic perspective any reduction of the quinol-pool would appear to be sufficient to drive nitrate reduction by NarGH. This seems a sensible scenario given that Nar is deployed to catalyse the first step of denitrification and drive energy conservation during anaerobic respiration in the presence of nitrate. By contrast NapAB will only become active when higher quinol levels are generated, consistent with the role of NapAB in dissipating excess reducing equivalents from the cytoplasmic membrane during aerobic respiration [1,2]. Unregulated electron flux through Nap would be detrimental to the cell as it would result in an unnecessary waste of redox energy.
Three energy conserving, H+-translocating oxidases are present during aerobic respiration by P. pantotrophus when nap is expressed . The cytochrome aa3- and cytochrome cbb3-type oxidases receive electrons from UQH2 via soluble cytochromes and the cytochrome bc1 complex with the consequence that six protons are moved across the cytoplasmic membrane for each UQH2 oxidized. In contrast, the cytochrome bo3-type oxidase reacts directly with UQH2 to move four protons across the membrane for each UQH2 oxidized. Oxygen reduction by the cytochrome bc1 complex-dependent oxidases was shown to occur when the UQH2 content of P. pantotrophus membranes rose above approx. 15% and it is likely that turnover of the cytochrome bc1 complex and hence these oxidases will become limited by UQ availability at very high UQH2/UQ ratios . The cytochrome bo3-type quinol oxidase became active when the UQH2 content exceeded 25%. Thus our detection of significant (>25%) NapAB nitrate reductase activity at potentials that correlate with a UQH2 content of approx. 25% is entirely consistent with the role of NapAB in productive redox balancing. In the presence of highly reducing carbon substrates the levels of NapAB and the cytochrome bo3-type quinol oxidase are up-regulated and this genetic regulation appears to be complemented by the biochemistry of the enzymes to provide opportunities for a rapid, finely tuned response to change of metabolism/respiratory pathway.
The KS and KM reported in the present study for P. pantotrophus NapAB nitrate reduction place the apparent affinity of this enzyme for its substrate in line with those of homologous enzymes [21,23,25]. The ability of NapAB to convert reversibly between two catalytically competent but kinetically distinct forms on traversing the electrochemical potential domain is also a property shared with other Mo[MGD]2-enzymes [10,21,24,25,34]. However, the conditions under which this behaviour is displayed and the mechanistic implications depend on the identity of the enzyme. In one mechanism the Mo5+/4+ couple is proposed to act as a redox ‘switch’, converting the enzyme between kinetically distinct forms [10,24,25,34]. Here, the higher-activity enzyme form present at more positive potentials arises from substrate (proton or nitrate) binding to the Mo5+ state. Increasingly negative potentials and/or lower substrate concentrations favour Mo4+ formation prior to substrate binding and catalysis proceeds via a slower pathway which results in attenuated catalytic rates. Thus higher substrate concentrations are predicted to favour substrate binding to Mo5+ and produce an increasingly sigmoidal catalytic wave shape that will ultimately reflect reduction of the Mo5+/4+ couple in substrate-bound enzyme. Such behaviour is seen in the pH-dependence of the catalytic PFV from E. coli dimethyl sulfoxide reductase and the nitrate dependence of the catalytic responses from S. elongatus NarB and NarGH(I) of E. coli and P. denitrificans. By contrast the P. pantotrophus NapAB wave shape is remarkably resilient to change of pH and nitrate concentration.
It is difficult to relate the activities seen in the present study to active-site redox chemistry in the absence of reduction potentials for a significant population of the NapAB sample and as an aside it is worth noting that the failure to detect significant Mo5+ EPR intensities in potentiometric studies may indicate that much of the sample moves between Mo6+ and Mo4+ over a narrow potential window. However, preferential nitrate binding to E. coli NarGH in the Mo5+ state was also inferred from voltammetry in the presence of the competitive inhibitor azide . Azide inhibited the high-potential enzyme form with a greater potency than that displayed for the low-potential enzyme form and the catalytic wave shape changed from peaked to sigmoidal. This was taken to reflect the higher affinity of azide for Mo5+ than Mo4+ expected from electrostatic considerations and it was proposed that nitrate would behave similarly. Thiocyanate is a competitive inhibitor of P. pantotrophus NapAB and previous spectroscopic studies have provided evidence for thiocyanate co-ordination to the molybdenum . However, the NapAB catalytic wave shape was unaffected by the presence of thiocyanate providing no indication that thiocyanate discriminates between the catalytically active oxidation states. It could be argued that thiocyanate as a soft ligand is less discriminating than azide in its affinity for molybdenum oxidation states. However, azide, a partial mixed inhibitor of NapAB, produces a clear change in catalytic wave shape suggests that the peak of activity may not simply arise from preferential substrate binding to Mo5+ in this enzyme.
Alternative switching mechanisms can be proposed from more generic consequences of redox modulation within an enzyme. Here it is worth noting that peaked catalytic waves are not specific to molybdenum enzymes and they have been observed in multi-centred reductases with flavin- and haem-containing active sites [35–38]. For these enzymes attenuation of activity has been associated with conformational change driven by active-site reduction and the consequences of reduction at redox centres remote from the active site. For NapAB such changes could be driven by redox chemistry at the active site or the [4Fe–4S] cluster whose reduction potential lies in the vicinity of Eswitch. Substantial conformational change is unlikely to account for the modulated rate since this may be expected to occur at a slower rate than catalysis such that increasing the voltammetric scan rate would produce an increasingly sigmoidal wave as the attenuation is ‘tuned-out’. This was not observed for NapAB where the entire catalytic response was compromised at scan rates above approx. 100 mV·s−1. Extended X-ray absorption fine structure of ferricyanide-oxidized and dithionite-reduced P. pantotrophus NapAB has suggested a lower thiocyanate co-ordination of Mo4+ than Mo6+ that could result in modulation of the catalytic rate . Alternatively the activity may be modulated by movement of an amino acid and/or a change of pKa that affects proton and/or electron transfer events associated with catalysis. Such perturbations could have an impact on catalysis via sites that are not immediately involved in nitrate co-ordination and it is interesting that the presence of azide, a partial inhibitor of NapAB, perturbs the wave shape for nitrate reduction.
In conclusion, we have presented results that demonstrate the high apparent affinity of P. pantotrophus NapAB towards its substrate nitrate. The mechanistic details that underpin the peak of activity displayed by this enzyme remain to be defined. However, it is apparent that the operational potentials of NapAB are well-placed to allow for effective redox balancing by dissipation of excess reducing equivalents during anaerobic growth of P. pantotrophus.
We are grateful to Ann Reilly (School of Biological Sciences, University of East Anglia, Norwich, U.K.) for NapAB purification, Dr Myles Cheesman (School of Chemical Sciences and Pharmacy, University of East Anglia, Norwich, U.K.) for assistance with EPR spectroscopy and to Gemma Kemp for fruitful discussions. A. J. G. was funded by EPSRC (Engineering and Physical Science Research Council) through DTA (Doctoral Training Account) and financial support for the project was provided from the U.K. BBSRC (Biotechnology and Biological Sciences Research Council; grants 83/17233 and 83/13842) and a JIF (Joint Infrastructure Funding) award (062178).