As the final electron acceptor in the respiratory chain of eukaryotic and many prokaryotic organisms, cytochrome c oxidase (EC 1.9.3.1) catalyzes the reduction of oxygen to water and generates a proton gradient. To test for proton pathways through the oxidase, site-directed mutagenesis was applied to subunit I of the Rhodobacter sphaeroides enzyme. Mutants were characterized in three highly conserved regions of the peptide, comprising possible proton loading, unloading, and transfer sites: an interior loop between helices II and III (Aspl32Asn/Ala), an exterior loop between helices IX and X (His4l]Ala, Asp412Asn, Thr413Asn, Tyr414Phe), and the predicted transmembrane helix VIII (Thr352Ala, Pro358Ala, Thr359Ala, Lys362Met). Most of the mutants had lower activity than wild type, but only mutants at residue 132 lost proton pumping while retaining electron transfer activity. Although electron transfer was substantially inhibited, no major structural alteration appears to have occurred in D132 mutants, since resonance Raman and visible absorbance spectra were normal. However, lower CO binding (70-85% of wild type) suggests some minor change to the binuclear center. In addition, the activity of the reconstituted Aspl32 mutants was inhibited rather than stimulated by ionophores or uncoupler. The inhibition was not observed with the purified enzyme and a direct pH effect was ruled out, suggesting an altered response to the electrical or pH gradient. The results support an important role for the conserved II-I-I loop in the proton pumping process and are consistent with the possibility of involvement of residues in helix VIII and the IX-X loop.Cytochrome c oxidase (EC 1.9.3.1), a key enzyme in aerobic energy metabolism, reduces oxygen to water, yielding substantial energy that drives the formation of a proton gradient; however, the mechanism of coupling between oxygen reduction and proton translocation remains obscure.Recognition of the strong homology between mitochondrial and bacterial enzymes (1, 2) has stimulated the application of molecular genetic tools to the analysis of the oxidase mechanism. The genes for cytochrome c oxidase from Rhodobacter sphaeroides have been cloned, sequenced, deleted, and reintroduced into the bacterium, and sequence comparisons reveal a high degree of homology with the three mitochondrially encoded subunits of mammalian oxidase (3-7). Extensive sitedirected mutagenesis of the largest subunit, COX I, has permitted the assignment of the ligands for the three redox active metal centers, heme a, heme a3, and CUB (7-10), suggesting that all three metal centers are located in COX I toward the outer side of the membrane, while substrate and pumped protons come from the inside (11). Thus some kind of proton channel or relay system is required to convey protons to the site of oxygen reduction, the heme a3-CuB center, and beyond. It is reasonable to look for residues involved in proton pumpingThe publication costs of this article were defrayed in part by page charge payment. This article m...

Cytochrome aa3 of Rhodobacter sphaeroides and cytochrome bo of E. coli are useful models of the more complex cytochrome c oxidase of eukaryotes, as demonstrated by the genetic, spectroscopic, and functional studies reviewed here. A summary of site-directed mutants of conserved residues in these two enzymes is presented and discussed in terms of a current model of the structure of the metal centers and evidence for regions of the protein likely to be involved in proton transfer. The model of ligation of the heme a3 (or o)-CuB center, in which both hemes are bound to helix X of subunit I, has important implications for the pathways and control of electron transfer.

We report the development of a high-yield heterologous expression system for the copper-containing nitrite reductase from a denitrifying variant of Rhodobacter sphaeroides. Typical yields of wild-type protein are 20 mg L-1, which can be fully loaded with copper. Nitrite reductase contains an unusual blue-green Type 1 copper center with a redox/electron transfer function and a nearby Type 2 center where nitrite binds and is reduced to nitric oxide. The wild-type enzyme was characterized by: (1) its blue-green Type 1 optical spectrum; (2) its EPR spectrum showing rhombic character to its Type 1 center and nitrite perturbation to its Type 2 center; (3) its 247-mV Type 1 midpoint potential which is low relative to other Type 1 centers; and (4) its kinetics as measured by both steady-state and stopped-flow methods. The Type 2 copper reduction potential as monitored by EPR in the absence of nitrite was below 200 mV so that reduction of the Type 2 center by the Type 1 center in the absence of nitrite is not energetically favored. The mutation M182T in which the methionine ligand of Type 1 copper was changed to a threonine resulted in a blue rather than blue-green Type 1 center, a midpoint potential that increased by more than 100 mV above that of the wild-type Type 1 center, and a somewhat reduced nitrite reductase activity. The blue color and midpoint potential of M182T are reminiscent of plastocyanin, but the Type 1 cupric HOMO ground-state electronic g value and copper hyperfine properties of M182T (as well as cysteine and histidine ENDOR hyperfine properties; see next paper) were unchanged from those of the blue-green native Type 1 center. His287 is a residue in the Type 2 region whose imidazole ring was thought to hydrogen bond to the Type 2 axial ligand but not directly to Type 2 copper. The mutation H287E resulted in a 100-fold loss of enzyme activity and a Type 2 EPR spectrum (as well as ENDOR spectra; see next paper) which were no longer sensitive to the presence of nitrite.

Present-day knowledge on the regulatory biology of denitrification is based on studies of selected model organisms. These show large variations in their potential contribution to NO, NO, and NO accumulation, attributed to lack of genes coding for denitrification reductases, but also to variations in their transcriptional regulation, as well as to post-transcriptional phenomena. To validate the relevance of these observations, there is a need to study a wider range of denitrifiers. We designed an isolation protocol that identifies all possible combinations of truncated denitrification chains (NO/NO/NO/NO/N). Of 176 isolates from two soils (pH 3.7 and 7.4), 30 were denitrifiers sensu stricto, reducing NO to gas, and five capable of NO reduction only. Altogether, 70 isolates performed at least one reduction step, including two DNRA isolates. Gas kinetics and electron flow calculations revealed that several features with potential impact on NO production, reported from model organisms, also exist in these novel isolates, including denitrification bet-hedging and control of NO/NO/NO accumulation. Whole genome sequencing confirmed most truncations but also showed that phenotypes cannot be predicted solely from genetic potential. Interestingly, and opposed to the commonly observed inability to reduce NO under acidic conditions, one isolate identified as Rhodanobacter reduced NO only at low pH.

During denitrification, the production and consumption of nitric oxide (NO), an obligatory and freely diffusible intermediate, must be tightly regulated in order to prevent accumulation of this highly reactive nitrogen oxide. Sequencing upstream of norCB, the structural genes for NO reductase, in the denitrifying bacterium Rhodobacter sphaeroides 2.4.3, we have identified a gene, designated nnrR, which encodes a protein that is a member of the cyclic AMP receptor family of transcriptional regulators. Insertional inactivation of nnrR prevents growth on nitrite, as well as the reduction of nitrite and NO, but has no effect on reduction of nitrate or photosynthetic growth. By using nirK-lacZ and norB-lacZ fusions, we have shown that NnrR is a positive transcriptional regulator of these genes. nnrR is expressed at a low constitutive level throughout the growth of R. sphaeroides 2.4.3. These results show that NnrR is not a global regulator but is instead a regulator of genes whose products are directly responsible for production and reduction of NO. Evidence is also presented suggesting that an NnrR homolog may be present in the nondenitrifying bacterium R. sphaeroides 2.4.1. The likely effector of NnrR activity, as determined on the basis of work detailed in this paper and other studies, is discussed.

Nitrite reductase catalyzes the reduction of nitrite to nitric oxide, the first step in denitrification to produce a gaseous product. We have cloned the gene nirK, which encodes the copper-type nitrite reductase from a denitrifying variant of Rhodobacter sphaeroides, strain 2.4.3. The deduced open reading frame has significant identity with other copper-type nitrite reductases. Analysis of the promoter region shows that transcription initiates 31 bases upstream of the translation start codon. The transcription initiation site is 43.5 bases downstream of a putative binding site for a transcriptional activator. Maximal expression of a nirK-lacZ construct in 2.4.3 requires both a low level of oxygen and the presence of a nitrogen oxide. nirK-lacZ expression was severely impaired in a nitrite reductase-deficient strain of 2.4.3. This suggests that nirK expression is dependent on nitrite reduction. The inability of microaerobically grown nitrite reductase-deficient cells to induce nirK-lacZ expression above basal levels in medium unamended with nitrate demonstrates that changes in oxygen concentrations are not sufficient to modulate nirK expression.The availability of fixed nitrogen is often a major factor controlling the biological productivity of ecosystems. During the cycling of nitrogen in the biosphere, a significant sink for fixed nitrogen is denitrification, the reduction of nitrate to gaseous forms of nitrogen, primarily nitrogen gas (22). Gaseous forms of nitrogen are unavailable for use by the majority of organisms. Because denitrification is a process that results in the conversion of fixed forms of nitrogen to gaseous forms, it can have a significant impact on the productivity of an ecosystem. For example, nitrate concentrations in the ocean have been found to be a major factor limiting biological productivity (19). Denitrification is the major sink for ocean nitrate, establishing a direct link between denitrification and biological productivity in the marine environment (6). This has been dramatically demonstrated in recent studies that have shown that decreases in the rate of denitrification during the last glacial maximum may have increased the productivity of the ocean enough to lower the partial pressure of carbon dioxide in the atmosphere (1, 9).Denitrification is a respiratory process in which bacteria utilize nitrate and other inorganic oxides of nitrogen as alternate electron acceptors when oxygen concentrations are limiting. In the first step of denitrification, nitrate is reduced to nitrite (12). This reaction is not unique to denitrification, however, since it occurs during ammonification and assimilatory nitrate reduction. The next step in denitrification is the reduction of nitrite to nitric oxide (NO). This reaction is catalyzed by nitrite reductase and is the defining reaction of denitrification, since it produces the first gaseous intermediate (38). Moreover, nitric oxide-producing nitrite reductases are associated only with denitrification (12). There are two classes of nitrite reductase...

Electron nuclear double resonance (ENDOR) of protons at Type 2 and Type 1 cupric active sites correlates with the enzymatic pH dependence, the mutation of nearby conserved, nonligating residues, and electron transfer in heterologously expressed Rhodobacter sphaeroides nitrite reductase. Wild-type enzyme showed a pH 6 activity maximum but no kinetic deuterium isotope effect, suggesting protons are not transferred in the rate-limiting step of nitrite reduction. However, protonatable Asp129 and His287, both located near the Type 2 center, modulated enzyme activity. ENDOR of the wild-type Type 2 center at pH 6.0 revealed an exchangeable proton with large hyperfine coupling. Dipolar distance estimates indicated that this proton was 2.50-2.75 or 2.25-2.45 A from Type 2 copper in the presence or absence of nitrite, respectively. This proton may provide a properly oriented hydrogen bond to enhance water formation upon nitrite reduction. This proton was eliminated at pH 5.0 and showed a diminished coupling at pH 7.5. Mutations of Asp129 and His287 reduced enzyme activity and altered the exchangeable proton hyperfine spectra. Mutation of Asp129 prevented a pH-dependent change at the Type 1 Cys167 ligand as observed by Cys C(beta) proton ENDOR, implying there is a Type 2 and pH-dependent alteration of the Type 1 center. Mutation of the Type 1 center ligand Met182 to Thr and mutation of Asp129 increased the activation energy for nitrite reduction. Involvement of both the Type 1 center and Asp129 in modulating activation energy shows that electron transfer from the Type 1 center to a nitrite-ligated Type 2 center is rate-limiting for nitrite reduction. Mutation of Ile289 to Ala and Val caused minor perturbation to enzyme activity, but as detected by ENDOR, allowed formate binding. Thus, bulky Ile289 may exclude non-nitrite ligands from the Type 2 active site.

The ability of Agrobacetrium tumefaciens to perform balanced transitions from aerobic to anaerobic respiration was studied by monitoring oxygen depletion, transcription of nirK and norB, and the concentrations of nitrite, nitric oxide (NO) and nitrous oxide in stirred batch cultures with different initial oxygen, nitrate or nitrite concentrations. Nitrate concentrations (0.2-2 mM) did not affect oxygen depletion, nor the oxygen concentration at which denitrification was initiated (1-2 microM). Nitrite (0.2-2 mM), on the other hand, retarded the oxygen depletion as it reached approximately 20 microM, and caused initiation of active denitrification as oxygen concentrations reached 10-17 microM. Unbalanced transitions occurred in treatments with high cell densities (i.e. with rapid transition from oxic to anoxic conditions), seen as NO accumulation to muM concentrations and impeded nitrous oxide production. This phenomenon was most severe in nitrite treatments, and reduced the cells' ability to respire oxygen during subsequent oxic conditions. Transcripts of norB were only detectable during the period with active denitrification. In contrast, nirK transcripts were detected at low levels both before and after this period. The results demonstrate that the transition from aerobic to anaerobic metabolism is a regulatory challenge, with implications for survival and emission of trace gases from denitrifying bacteria.