[Show abstract][Hide abstract] ABSTRACT: The pentaheme cytochrome c nitrite reductase (NrfA) of Escherichia coli is responsible for nitrite reduction during anaerobic respiration when nitrate is scarce. The NrfA active site consists of a hexacoordinate high-spin heme with a lysine ligand on the proximal side and water/hydroxide or substrate on the distal side. There are four further highly conserved active site residues including a glutamine (Q263) positioned 8 A from the heme iron for which the side chain, unusually, coordinates a conserved, essential calcium ion. Mutation of this glutamine to the more usual calcium ligand, glutamate, results in an increase in the K m for nitrite by around 10-fold, while V max is unaltered. Protein film voltammetry showed that lower potentials were required to detect activity from NrfA Q263E when compared with native enzyme, consistent with the introduction of a negative charge into the vicinity of the active site heme. EPR and MCD spectroscopic studies revealed the high spin state of the active site to be preserved, indicating that a water/hydroxide molecule is still coordinated to the heme in the resting state of the enzyme. Comparison of the X-ray crystal structures of the as-prepared, oxidized native and mutant enzymes showed an increased bond distance between the active site heme Fe(III) iron and the distal ligand in the latter as well as changes to the structure and mobility of the active site water molecule network. These results suggest that an important function of the unusual Q263-calcium ion pair is to increase substrate affinity through its role in supporting a network of hydrogen bonded water molecules stabilizing the active site heme distal ligand.
[Show abstract][Hide abstract] ABSTRACT: The periplasmic cytochrome c nitrite reductase (Nrf) system of Escherichia coli utilizes nitrite as a respiratory electron acceptor by reducing it to ammonium. Nitric oxide (NO) is a proposed intermediate in this six-electron reduction and NrfA can use exogenous NO as a substrate. This chapter describes the method used to assay Nrf-catalyzed NO reduction in whole cells of E. coli and the procedures for preparing highly purified NrfA suitable for use in kinetic, spectroscopic, voltammetric, and crystallization studies.
Methods in Enzymology 02/2008; 437:63-77. DOI:10.1016/S0076-6879(07)37004-3 · 2.09 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: NrfB is a small pentahaem electron-transfer protein widely involved in the respiratory reduction of nitrite or nitric oxide to ammonia, processes that provide energy for anaerobic metabolism in many enteric bacteria and also serve to detoxify these reactive nitrogen species. The X-ray crystal structure of Escherichia coli NrfB is presented at 1.74 A (1 A=0.1 nm) resolution. The architecture of the protein is that of a 40 A 'nanowire' in which the five haems are positioned within 6 A of each other along a polypeptide scaffold. During nitrite reduction, the physiological role of NrfB is to mediate electron transfer to another pentahaem protein, NrfA, the enzyme that catalyses periplasmic nitrite or nitric oxide reduction. Protein-protein interaction studies suggest NrfA and NrfB can form a 20-haem NrfA2-NrfB2 heterotetrameric complex.
[Show abstract][Hide abstract] ABSTRACT: The Escherichia coli NapA (periplasmic nitrate reductase) contains a [4Fe-4S] cluster and a Mo-bis-molybdopterin guanine dinucleotide cofactor. The NapA holoenzyme associates with a di-heme c-type cytochrome redox partner (NapB). These proteins have been purified and studied by spectropotentiometry, and the structure
of NapA has been determined. In contrast to the well characterized heterodimeric NapAB systems ofα-proteobacteria, such as
Rhodobacter sphaeroides and Paracoccus pantotrophus, the γ-proteobacterial E. coli NapA and NapB proteins purify independently and not as a tight heterodimeric complex. This relatively weak interaction is
reflected in dissociation constants of 15 and 32 μm determined for oxidized and reduced NapAB complexes, respectively. The surface electrostatic potential of E. coli NapA in the apparent NapB binding region is markedly less polar and anionic than that of the α-proteobacterial NapA, which
may underlie the weaker binding of NapB. The molybdenum ion coordination sphere of E. coli NapA includes two molybdopterin guanine dinucleotide dithiolenes, a protein-derived cysteinyl ligand and an oxygen atom.
The Mo–O bond length is 2.6 Å, which is indicative of a water ligand. The potential range over which the Mo6+ state is reduced to the Mo5+ state in either NapA (between +100 and -100 mV) or the NapAB complex (-150 to -350 mV) is much lower than that reported for
R. sphaeroides NapA (midpoint potential Mo6+/5+ > +350 mV), and the form of the Mo5+ EPR signal is quite distinct. In E. coli NapA or NapAB, the Mo5+ state could not be further reduced to Mo4+. We then propose a catalytic cycle for E. coli NapA in which nitrate binds to the Mo5+ ion and where a stable des-oxo Mo6+ species may participate.
[Show abstract][Hide abstract] ABSTRACT: The periplasmic nitrate reductase of Escherichia coli is important during anaerobic growth in low-nitrate environments. The nap operon encoding this nitrate reductase comprises seven genes including a gene, napF, that encodes a putative cytoplasmic iron-sulphur protein of uncertain subcellular location and function. In this study, N-terminal sequence analysis, cell fractionation coupled with immunoblotting and construction of LacZ and PhoA fusion proteins were used together to establish that NapF is located in the E. coli cytoplasm. A bacterial two-hybrid protein-protein interaction system was used to demonstrate that NapF interacted in the cytoplasm with the terminal oxidoreductase NapA, but that it did not self-associate or interact with other electron-transport components of the Nap system, NapC, NapG or NapH, or with another cytoplasmic component, NapD. NapF, purified as a His(6)-tagged protein, exhibited spectral properties characteristic of an iron-sulphur protein. This protein was able to pull down NapA from soluble extracts of E. coli. A growth-based assay for NapF function in intact cell cultures was developed and applied to assess the effect of mutation of a number of conserved amino acids. It emerged that neither a highly conserved N-terminal double-arginine motif, nor a conserved proline motif, is essential for NapF-dependent growth. The combined data indicate that NapF plays one or more currently unidentified roles in the post-translational modification of NapA prior to the export of folded NapA via the twin-arginine translocation pathway into the periplasm.
[Show abstract][Hide abstract] ABSTRACT: The Escherichia coli K-12 nrf operon promoter can be activated fully by the FNR protein (regulator of fumarate and nitrate reduction) binding to a site
centered at position −41.5. FNR-dependent transcription is suppressed by integration host factor (IHF) binding at position
−54, and this suppression is counteracted by binding of the NarL or NarP response regulator at position −74.5. The E. coli acs gene is transcribed from a divergent promoter upstream from the nrf operon promoter. Transcription from the major acsP2 promoter is dependent on the cyclic AMP receptor protein and is modulated by IHF and Fis binding at multiple sites. We
show that IHF binding to one of these sites, located at position −127 with respect to the nrf promoter, has a positive effect on nrf promoter activity. This activation is dependent on the face of the DNA helix, independent of IHF binding at other locations,
and found only when NarL/NarP are not bound at position −74.5. Binding of NarL/NarP appears to insulate the nrf promoter from the effects of IHF. The acs-nrf regulatory region is conserved in other pathogenic E. coli strains and related enteric bacteria but differs in Salmonella enterica serovar Typhimurium.
Journal of Bacteriology 12/2006; 188(21):7449-56. DOI:10.1128/JB.00975-06 · 2.81 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The plasmid encoded toxin, Pet, is a prototypical member of the serine protease autotransporters of the Enterobacteriaceae. In addition to the passenger and beta-domains typical of autotransporters, in silico predictions indicate that Pet possesses an unusually long N-terminal signal sequence. The signal sequence can be divided into five regions termed N1 (charged), H1 (hydrophobic), N2, H2 and C (cleavage site) domains. The N1 and H1 regions, which we have termed the extended signal peptide region, demonstrate remarkable conservation. In contrast, the N2, H2 and C regions demonstrate significant variability and are reminiscent of typical Sec-dependent signal sequences. Despite several investigations, the function of the extended signal peptide region remains obscure and surprisingly it has not been proven that the extended signal peptide region is actually synthesized as part of the signal sequence. Here, we demonstrate that the extended signal peptide region is present only in Gram-negative bacterial proteins originating from the classes Beta- and Gammaproteobacteria, and more particularly only in proteins secreted via the Type V secretion pathway: autotransporters, TpsA exoproteins of the two-partner system and trimeric autotransporters. In vitro approaches demonstrate that the DNA region encoding the extended signal peptide region is transcribed and translated.
[Show abstract][Hide abstract] ABSTRACT: Escherichia coli K-12 strains expressing either NarU or NarK as the only nitrate transport protein are both able to support nitrate-dependent anaerobic growth. The narK gene is highly expressed during anaerobic growth in the presence of nitrate, consistent with a role for NarK in nitrate transport coupled to nitrate reduction by the most active nitrate reductase encoded by the adjacent narGHJI operon. The physiological role of NarU is unknown. Reverse transcriptase PCR experiments established that, unlike the monocistronic narK gene, narU is co-transcribed with narZ as the first gene of a five-gene narUZYWV operon. The narK and narU genes were fused in-frame to a myc tag: the encoded fusion proteins complemented the nitrate-dependent growth defect of chromosomal narK and narU mutations. A commercial anti-Myc antibody was used to detect NarK and NarU in membrane fractions. During anaerobic growth in the presence of nitrate, the quantity of NarU-Myc accumulated during exponential growth was far less than that of NarK-Myc, but NarU was more abundant than NarK in stationary-phase cultures in the absence of nitrate. Although the concentration of NarU-Myc increased considerably during the post-exponential phase of growth, NarK-Myc was still more abundant than NarU-Myc in stationary-phase bacteria in the presence of nitrate. In chemostat competition experiments, a strain expressing only narU had a selective advantage relative to a strain expressing only narK during nutrient starvation or very slow growth, but NarK(+) bacteria had a much greater selective advantage during rapid growth. The data suggest that NarU confers a selective advantage during severe nutrient starvation or slow growth, conditions similar to those encountered in vivo.
[Show abstract][Hide abstract] ABSTRACT: The transcription factor FNR, the regulator of fumarate and nitrate reduction, regulates major changes as Escherichia coli adapts from aerobic to anaerobic growth. In an anaerobic glycerol/trimethylamine N-oxide/fumarate medium, the fnr mutant grew as well as the parental strain, E. coli K12 MG1655, enabling us to reveal the response to oxygen, nitrate, and nitrite in the absence of glucose repression or artifacts because of variations in growth rate. Hence, many of the discrepancies between previous microarray studies of the E. coli FNR regulon were resolved. The current microarray data confirmed 31 of the previously characterized FNR-regulated operons. Forty four operons not previously known to be included in the FNR regulon were activated by FNR, and a further 28 operons appeared to be repressed. For each of these operons, a match to the consensus FNR-binding site sequence was identified. The FNR regulon therefore minimally includes at least 103, and possibly as many as 115, operons. Comparison of transcripts in the parental strain and a narXL deletion mutant revealed that transcription of 51 operons is activated, directly or indirectly, by NarL, and a further 41 operons are repressed. The narP gene was also deleted from the narXL mutant to reveal the extent of regulation by phosphorylated NarP. Fourteen promoters were more active in the narP+ strain than in the mutant, and a further 37 were strongly repressed. This is the first report that NarP might function as a global repressor as well as a transcription activator. The data also revealed possible new defense mechanisms against reactive nitrogen species.
[Show abstract][Hide abstract] ABSTRACT: Consistent with its role as a nitric oxide (NO)-detoxifying globin in Campylobacter jejuni, Cgb (Campylobacter globin) expression is strongly and specifically induced following exposure to nitrosative stress, suggesting a previously unrecognized capacity for NO-related stress sensing in this food-borne pathogen. In this study, Fur and PerR have been eliminated as major regulators of cgb, and NssR (Cj0466), a member of the Crp-Fnr superfamily, has been identified as the major positive regulatory factor that controls nitrosative stress-responsive expression of this gene. Accordingly, disruption of nssR resulted in the abolition of inducible cgb expression, which was restored by a complementing chromosomal insertion of the wild-type gene with its indigenous promoter at a second location. The NssR-deficient mutant was more sensitive to NO-related stress than a cgb mutant and this phenotype most likely arises from the failure of these cells to induce other NO-responsive components in addition to Cgb. Indeed, analysis of global gene expression, by microarray and confirmatory real-time polymerase chain reaction (PCR) in the wild type and nssR mutant, not only confirmed the dependence of inducible cgb expression on NssR, but also revealed for the first time a novel NssR-dependent nitrosative stress-responsive regulon. This regulon of at least four genes includes Cj0465c, a truncated globin. Consistent with NssR being a Crp-Fnr superfamily member, an Fnr-like binding sequence (TTAAC-N(4)-GTTAA) was found upstream of each gene at locations -40.5 to -42.5 relative to the centre of the binding sites and the transcription start point. Site-directed mutagenesis confirmed that this cis-acting motif mediates the nitrosative stress-inducible expression of cgb.
[Show abstract][Hide abstract] ABSTRACT: Expression from the Escherichia coli nrf operon promoter is activated by the anaerobically triggered transcription factor, FNR, and by the nitrate/nitrite ion-controlled response regulators, NarL or NarP, but is repressed by the IHF and Fis proteins. Here, we present in vitro studies on the nrf promoter, using permanganate footprinting to measure open complex formation, and DNase I footprinting to monitor binding of the different regulators and the interactions between them. Our results show that open complex formation is completely dependent on FNR and is enhanced by NarL, but is repressed by IHF or Fis. NarL counteracts repression by IHF but is unable to alter repression by Fis. These results suggest mechanisms by which nrf promoter activity is modulated by the different factors. Expression from the nrf promoter is known to be repressed in rich media, especially in the presence of glucose, but the molecular basis of this is not understood. Here, we show that this catabolite repression is relieved by mutations that weaken the DNA site for Fis, improve the DNA site for FNR or improve the promoter -10 or -35 elements. Hence, Fis protein is a major factor responsible for catabolite repression at the nrf promoter, and Fis can override activation by FNR and NarL or NarP.
[Show abstract][Hide abstract] ABSTRACT: The genome of Mycobacterium tuberculosis H37Rv includes a homologue of the CRP/FNR (cAMP receptor protein/fumarate and nitrate reduction regulator) family of transcription regulators encoded by Rv3676. Sequencing of the orthologous gene from attenuated Mycobacterium bovis Bacille Calmette-Guérin (BCG) strains revealed point mutations that affect the putative DNA-binding and cNMP-binding domains of the encoded protein. These mutations are not present in the published sequences of the Rv3676 orthologues in M. bovis, M. tuberculosis or Mycobacterium leprae. An Escherichia coli lacZ reporter system was used to show that the M. tuberculosis Rv3676 protein binds to DNA sites for CRP, but this DNA binding was decreased or abolished with the Rv3676 protein counterparts from BCG strains. The DNA-binding ability of the M. tuberculosis Rv3676 protein was decreased by the introduction of base changes corresponding to the BCG point mutations. Conversely, the DNA binding of the BCG Rv3676 proteins from BCG strains was restored by removing the mutations. These data show that in this reporter system the point mutations present in the Rv3676 orthologue in BCG strains render its function defective (early strains) or abolished (late strains) and suggest that this protein might be naturally defective in M. bovis BCG strains. This raises the possibility that a contributing factor to the attenuation of BCG strains may be an inability of this global regulator to control the expression of genes required for in vivo survival and persistence.
[Show abstract][Hide abstract] ABSTRACT: Escherichia coli can reduce nitrite to ammonium via a 120-kDa decaheme homodimeric periplasmic nitrite reductase (NrfA) complex. Recent structure-based spectropotentiometric studies are shedding light on the catalytic mechanism of NrfA; however, electron input into the enzyme has not been addressed biochemically. This study reports the first purification of NrfB, a novel 20-kDa pentaheme c-type cytochrome encoded by the nrfB gene that follows the nrfA gene in many bacterial nrf operons. Analyses by gel filtration demonstrated that NrfB purifies as a decaheme homodimer. Analysis of NrfB by UV-visible and magnetic circular dichroism spectroscopy demonstrates that all five NrfB ferric heme irons are low spin and are most likely coordinated by two axial histidine ligands. Spectropotentiometry revealed that the midpoint redox potentials of five ferric hemes were in the low potential range of 0 to -400 mV. Analysis by low temperature EPR spectroscopy revealed signals that arise from two classes of bis-His ligated low spin hemes, namely a rhombic trio at g(1,2,3) = 2.99, 2.27, and 1.5 that arises from two hemes in which the planes of histidine imidazole rings are near-parallel and a large g(max) signal at g = 3.57 that arises from three hemes in which the planes of the histidine imidazole rings are near-perpendicular. NrfB was also overexpressed as a recombinant protein, which had similar spectropotentiometric properties as the native protein. Reconstitution experiments demonstrated that the reduced decaheme NrfB dimer could serve as a direct electron donor to the oxidized decaheme NrfA dimer, thus forming a transient 20-heme [NrfB](2)[NrfA](2) electron transfer complex.
[Show abstract][Hide abstract] ABSTRACT: Expression from the Escherichia coli nir promoter is co-dependent on both the FNR protein (an anaerobically triggered transcription activator) and NarL or NarP proteins (transcription activators triggered by nitrite and nitrate). We found previously that FNR binds to a site centred at position - 41.5 at the nir promoter, but that FNR-dependent activation is repressed by IHF binding to a site centred at position -88 (IHF I) and Fis binding to sites centred at positions -142 (Fis I) and +23 (Fis II). Here, we have studied the binding of purified IHF, Fis and FNR to the nir promoter in vitro. Our results show that the nir promoter contains a second IHF site at position -115 (IHF II) and a third Fis site at position -97 (Fis III), and that IHF, Fis and FNR can bind together to form multiprotein complexes. Surprisingly, IHF binding at the IHF II site increases FNR-dependent activation by decreasing the repression mediated by IHF and Fis binding at the other sites. In previous work, we found that NarL or NarP activates the nir promoter by binding to a site centred at position -69.5 and counteracting the repressive effects of IHF and Fis. We now show that NarL can displace IHF bound at the IHF I site, but IHF is unable to displace bound NarL. We suggest that NarL interferes with IHF binding at the nir promoter by distorting the minor groove at its target site, and we argue that the resulting activation by NarL results from remodelling of the local nucleoprotein structure to facilitate FNR-dependent transcription.
[Show abstract][Hide abstract] ABSTRACT: Nap (periplasmic nitrate reductase) operons of many bacteria include four common, essential components, napD, napA, napB and napC (or a homologue of napC ). In Escherichia coli there are three additional genes, napF, napG and napH, none of which are essential for Nap activity. We now show that deletion of either napG or napH almost abolished Nap-dependent nitrate reduction by strains defective in naphthoquinone synthesis. The residual rate of nitrate reduction (approx. 1% of that of napG+ H+ strains) is sufficient to replace fumarate reduction in a redox-balancing role during growth by glucose fermentation. Western blotting combined with beta-galactosidase and alkaline phosphatase fusion experiments established that NapH is an integral membrane protein with four transmembrane helices. Both the N- and C-termini as well as the two non-haem iron-sulphur centres are located in the cytoplasm. An N-terminal twin arginine motif was shown to be essential for NapG function, consistent with the expectation that NapG is secreted into the periplasm by the twin arginine translocation pathway. A bacterial two-hybrid system was used to show that NapH interacts, presumably on the cytoplasmic side of, or within, the membrane, with NapC. As expected for a periplasmic protein, no NapG interactions with NapC or NapH were detected in the cytoplasm. An in vitro quinol dehydrogenase assay was developed to show that both NapG and NapH are essential for rapid electron transfer from menadiol to the terminal NapAB complex. These new in vivo and in vitro results establish that NapG and NapH form a quinol dehydrogenase that couples electron transfer from the high midpoint redox potential ubiquinone-ubiquinol couple via NapC and NapB to NapA.
[Show abstract][Hide abstract] ABSTRACT: The Neisseria gonorrhoeae genome encodes a homologue of the Escherichia coli FNR protein (the fumarate and nitrate reductase regulator). Despite its similarity to E. coli FNR, the gonococcal FNR only partially complemented an E. coli fnr mutation. After error-prone PCR mutagenesis of the gonococcal fnr gene, we identified four mutant fnr derivatives carrying the same S18F substitution, and we showed that the mutant FNR could activate transcription from a range of class I and class II FNR-dependent promoters in E. coli. Prompted by the similarities between gonococcal and E. coli FNR, we made changes in gonococcal fnr that created substitutions that are equivalent to previously characterized substitutions in E. coli FNR. First, our experiments showed that cysteine, C116, in the gonococcal FNR, equivalent to C122 in E. coli FNR, is essential, presumably because, as in E. coli FNR, it binds to an iron-sulfur center. Second, the L22H and D148A substitutions in gonococcal FNR were made. These changes are equivalent to the L28H and D154A changes in E. coli FNR, which had been shown to increase FNR activity in the presence of oxygen. We show that the effects of these substitutions in gonococcal FNR are distinct from those of the S18F substitution. Similarly, substitutions in the putative activating regions of gonococcal FNR were made. We show that the activity of gonococcal FNR in E. coli can be increased by transplanting certain activating regions from E. coli FNR. The effects of these substitutions are additive to those due to S18F. From these data, we conclude that the effects of the S18F substitution in gonococcal FNR are distinct from the effects of the other substitutions. S18 is immediately adjacent to one of three N-terminal cysteine residues that coordinate the iron-sulfur center, and thus the S18F substitution is most likely to stabilize this center. Support for this came from complementary experiments in which we created the S24F substitution in E. coli FNR, which is equivalent to the S18F substitution in gonococcal FNR. Our results show that the S24F substitution changes the activity of E. coli FNR and that the changes are distinct from those due to previously characterized substitutions.
Journal of Bacteriology 09/2003; 185(16):4734-47. DOI:10.1128/JB.185.16.4734-4747.2003 · 2.81 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Paracoccus pantotrophus grown anaerobically under denitrifying conditions expressed similar levels of the periplasmic nitrate reductase (NAP) when cultured in molybdate- or tungstate-containing media. A native PAGE gel stained for nitrate reductase activity revealed that only NapA from molybdate-grown cells displayed readily detectable nitrate reductase activity. Further kinetic analysis showed that the periplasmic fraction from cells grown on molybdate (3 microM) reduced nitrate at a rate of V(max)=3.41+/-0.16 micromol [NO(3)(-)] min(-1) mg(-1) with an affinity for nitrate of K(m)=0.24+/-0.05 mM and was heat-stable up to 50 degrees C. In contrast, the periplasmic fraction obtained from cells cultured in media supplemented with tungstate (100 microM) reduced nitrate at a much slower rate, with much lower affinity (V(max)=0.05+/-0.002 micromol [NO(3)(-)] min(-1) mg(-1) and K(m)=3.91+/-0.45 mM) and was labile during prolonged incubation at >20 degrees C. Nitrate-dependent growth of Escherichia coli strains expressing only nitrate reductase A was inhibited by sub-mM concentrations of tungstate in the medium. In contrast, a strain expressing only NAP was only partially inhibited by 10 mM tungstate. However, none of the above experimental approaches revealed evidence that tungsten could replace molybdenum at the active site of E. coli NapA. The combined data show that tungsten can function at the active site of some, but not all, molybdoenzymes from mesophilic bacteria.
[Show abstract][Hide abstract] ABSTRACT: Nitric oxide is a key element in host defense against invasive pathogens. The periplasmic cytochrome c nitrite reductase (NrfA) of Escherichia coli catalyzes the respiratory reduction of nitrite, but in vitro studies have shown that it can also reduce nitric oxide. The physiological significance of the latter reaction in vivo has never been assessed. In this study the reduction of nitric oxide by Escherichia coli was measured in strains active or deficient in periplasmic nitrite reduction. Nrf(+) cells, harvested from cultures grown anaerobically, possessed a nitric-oxide reductase activity with physiological electron donation of 60 nmol min(-1) x mg dry wt(-1), and an in vivo turnover number of NrfA of 390 NO* s(-1) was calculated. Nitric-oxide reductase activity could not be detected in Nrf(-) strains. Comparison of the anaerobic growth of Nrf(+) and Nrf(-) strains revealed a higher sensitivity to nitric oxide in the NrfA(-) strains. A higher sensitivity to the nitrosating agent S-nitroso-N-acetyl penicillamine (SNAP) was also observed in agar plate disk-diffusion assays. Oxygen respiration by E. coli was also more sensitive to nitric oxide in the Nrf(-) strains compared with the Nrf(+) parent strain. The results demonstrate that active periplasmic cytochrome c nitrite reductase can confer the capacity for nitric oxide reduction and detoxification on E. coli. Genomic analysis of many pathogenic enteric bacteria reveals the presence of nrf genes. The present study raises the possibility that this reflects an important role for the cytochrome c nitrite reductase in nitric oxide management in oxygen-limited environments.