Active site rearrangement and structural divergence in prokaryotic respiratory oxidases

Cytochrome bd–type quinol oxidases catalyze the reduction of molecular oxygen to water in the respiratory chain of many human-pathogenic bacteria. They are structurally unrelated to mitochondrial cytochrome c oxidases and are therefore a prime target for the development of antimicrobial drugs. We determined the structure of theEscherichia colicytochrome bd-I oxidase by single-particle cryo–electron microscopy to a resolution of 2.7 angstroms. Our structure contains a previously unknown accessory subunit CydH, the L-subfamily–specific Q-loop domain, a structural ubiquinone-8 cofactor, an active-site density interpreted as dioxygen, distinct water-filled proton channels, and an oxygen-conducting pathway. Comparison with another cytochrome bd oxidase reveals structural divergence in the family, including rearrangement of high-spin hemes and conformational adaption of a transmembrane helix to generate a distinct oxygen-binding site.

Cytochromes bd-I and bo3 are essential for the bactericidal effect of microcin J25 on Escherichia coli cells

Microcin J25 has two targets in sensitive bacteria, the RNA polymerase, and the respiratory chain through inhibition of cellular respiration. In this work, the effect of microcin J25 in E. coli mutants that lack the terminal oxidases cytochrome bd-I and cytochrome bo₃ was analyzed. The mutant strains lacking cytochrome bo₃ or cytochrome bd-I were less sensitive to the peptide. In membranes obtained from the strain that only expresses cytochrome bd-I a great ROS overproduction was observed in the presence of microcin J25. Nevertheless, the oxygen consumption was less inhibited in this strain, probably because the oxygen is partially reduced to superoxide. There was no overproduction of ROS in membranes isolated from the mutant strain that only express cytochrome bo₃ and the inhibition of the cellular respiration was similar to the wild type. It is concluded that both cytochromes bd-I and bo₃ are affected by the peptide. The results establish for the first time a relationship between the terminal oxygen reductases and the mechanism of action of microcin J25.

Exposure of bovine cytochrome c oxidase to high triton X-100 or to alkaline conditions causes a dramatic change in the rate of reduction of compound F

The final step in the catalytic cycle of cytochrome oxidase, the reduction of oxyferryl heme a(3) in compound F, was investigated using a binuclear polypyridine ruthenium complex ([Ru(bipyridine)(2)](2)(1,4-bis[2-(4'-methyl-2, 2'-bipyrid-4-yl)ethenyl]benzene)(PF(6))(4)) as a photoactive reducing agent. In the untreated dimeric enzyme, the rate constant for reduction of compound F decreased from 700 s(-1) to 200 s(-1) as the pH was increased from 7.5 to 9.5. Incubation of dimeric enzyme at pH 10 led to an increase in the rate constant to 1650 s(-1), which was independent of pH between pH 7.4 and 10. This treatment resulted in a decrease in the sedimentation coefficient consistent with the irreversible conversion of the enzyme to a monomeric form. Similar results were obtained when the enzyme was incubated with Triton X-100 at pH 8.0. These treatments, which have traditionally been used to convert dimeric enzyme to monomeric form, have no effect on the steady-state activity. The data indicate that either the conversion of the bovine oxidase to a monomeric form or some structural change coincident with this conversion strongly influences the rate constant of this step in the catalytic cycle, perhaps by influencing the proton access to the heme-copper binuclear center.

Perfusion-induced redox differences in cytochrome c oxidase: ATR/FT-IR spectroscopy

Attenuated total reflection (ATR) spectroscopy brings an added dimension to studies of structural changes of cytochrome c oxidase (CcO) because it enables the recording of reaction-induced infrared difference spectra under a wide variety of controlled conditions (e.g. pH and chemical composition), without relying on light or potentiometric changes to trigger the reaction. We have used the ATR method to record vibrational difference spectra of CcO with reduction induced by flow-exchange of the aqueous buffer. Films of CcO prepared from Rhodobacter sphaeroides and beef heart mitochondria by reconstitution with lipid were adhered to the internal reflection element of the ATR device and retained their full functionality as evidenced by visible spectroscopy and time-resolved vibrational spectroscopy. These results demonstrate that the technique of perfusion-induced Fourier-transform infrared difference spectroscopy can be successfully applied to a large, complex enzyme, such as CcO, with sufficient signal/noise to probe vibrational changes in individual residues of the enzyme under various conditions.

Role of the K-channel in the pH-dependence of the reaction of cytochrome c oxidase with hydrogen peroxide

The reaction of cytochrome c oxidase (COX) from Rhodobacter sphaeroides with hydrogen peroxide has been studied at alkaline (pH 8.5) and acidic (pH 6.5) conditions with the aid of a stopped-flow apparatus. Absorption changes in the entire 350-800 nm spectral range were monitored and analyzed by a global fitting procedure. The reaction can be described by the sequential formation of two intermediates analogous to compounds I and II of peroxidases: oxidized COX + H2O2 --> intermediate I --> intermediate II. At pH as high as 8.5, intermediate I appears to be a mixture of at least two species characterized by absorption bands at approximately 607 nm (P607) and approximately 580 nm (F-I580) that rise synchronously. At acidic pH (6.5), intermediate I is represented mainly by a component with an alpha-peak around 575 nm (F-I575) that is probably equivalent to the so-called F* species observed with the bovine COX. The data are consistent with a pH-dependent reaction branching at the step of intermediate I formation. To get further insight into the mechanism of the pH-dependence, the peroxide reaction was studied using two mutants of the R. sphaeroides oxidase, K362M and D132N, that block, respectively, the proton-conducting K- and D-channels. The D132N mutation does not affect significantly the Ox --> intermediate I step of the peroxide reaction. In contrast, K362M replacement exerts a dramatic effect, eliminating the pH-dependence of intermediate I formation. The data obtained allow us to propose that formation of the acidic form of intermediate I (F-I575, F*) requires protonation of some group at/near the binuclear site that follows or is concerted with peroxide binding. The protonation involves specifically the K-channel. Presumably, a proton vacancy can be generated in the site as a consequence of the proton-assisted heterolytic scission of the O-O bond of the bound peroxide. The results are consistent with a proposal [Vygodina, T. V., Pecoraro, C., Mitchell, D., Gennis, R., and Konstantinov, A. A. (1998) Biochemistry 37, 3053-3061] that the K-channel may be involved in the delivery of the first four protons in the catalytic cycle (starting from reduction of the oxidized form) including proton uptake coupled to reduction of the binuclear site and transfer of protons driven by cleavage of the dioxygen O-O bond in the binculear site. Once peroxide intermediate I has been formed, generation of a strong oxene ligand at the heme a3 iron triggers a transition of the enzyme to the "peroxidase conformation" in which the K-channel is closed and the binuclear site becomes protonically disconnected from the bulk aqueous phase.

Site-directed mutation of the highly conserved region near the Q-loop of the cytochrome bd quinol oxidase from Escherichia coli specifically perturbs heme b595

Cytochrome bd is one of the two quinol oxidases in the respiratory chain of Escherichia coli. The enzyme contains three heme prosthetic groups. The dioxygen binding site is heme d, which is thought to be part of the heme-heme binuclear center along with heme b(595), which is a high-spin heme whose function is not known. Protein sequence alignments [Osborne, J. P., and Gennis, R. B. (1999) Biochim. Biophys Acta 1410, 32--50] of cytochrome bd quinol oxidase sequences from different microorganisms have revealed a highly conserved sequence (GWXXXEXGRQPW; bold letters indicate strictly conserved residues) predicted to be on the periplasmic side of the membrane between transmembrane helices 8 and 9 in subunit I. The functional importance of this region is investigated in the current work by site-directed mutagenesis. Several mutations in this region (W441A, E445A/Q, R448A, Q449A, and W451A) resulted in a catalytically inactive enzyme with abnormal UV--vis spectra. E445A was selected for detailed analysis because of the absence of the absorption bands from heme b(595). Detailed spectroscopic and chemical analyses, indeed, show that one of the three heme prosthetic groups in the enzyme, heme b(595), is specifically perturbed and mostly missing from this mutant. Surprisingly, heme d, while known to interact with heme b(595), appears relatively unperturbed, whereas the low-spin heme b(558) shows some modification. This is the first report of a mutation that specifically affects the binding site of heme b(595).

Abortive assembly of succinate-ubiquinone reductase (complex II) in a ferrochelatase-deficient mutant of Escherichia coli

Heme molecules play important roles in electron transfer by redox proteins such as cytochromes. In addition, a structural role for heme in protein folding and the assembly of enzymes has been suggested. Previous results obtained using Escherichia coli hemA mutants, which are unable to synthesize 5-aminolevulinic acid, a precursor of porphyrins and hemes, have demonstrated a requirement for heme biosynthesis in the assembly of a functional succinate-ubiquinone reductase (SQR or complex II), which is a component of the aerobic respiratory chain. In the present study, in order to investigate the role of the heme in the assembly of E. coli SQR, we used a hemH (encodes ferrochelatase) mutant that lacks the ability to insert iron into the porphyrin ring. The hemH mutant failed to insert functional SQR into the cytoplasmic membrane, and the catalytic portion of SQR [the flavoprotein subunit (Fp) and the iron-sulfur protein subunit (Ip)] was localized in the cytoplasm of the cell. It is of interest to note that protoporphyrin IX accumulated in the mutant cells and inactivated the cytoplasmic succinate dehydrogenase (SDH) activity associated with the catalytic Fp-Ip complex. In contrast, SQR was assembled into the membrane of a heme-permeable hemH double mutant when hemin was present in the culture. Only a low level of SQR activity was found in the membrane when hemin was replaced by non-iron metalloporphyrins: Mn-, Co-, Ni-, Zn- and Cu-protoporphyrin IX, or protoporphyrin IX These results indicate that heme iron is indispensable for the functional assembly of SQR in the cytoplasmic membrane of E. coli, and provide a new insight into the biological role of heme in the molecular assembly of the multi-subunit enzyme complex.

pH-dependent structural changes at the Heme-Copper binuclear center of cytochrome c oxidase

The resonance Raman spectra of the aa3 cytochrome c oxidase from Rhodobacter sphaeroides reveal pH-dependent structural changes in the binuclear site at room temperature. The binuclear site, which is the catalytic center of the enzyme, possesses two conformations at neutral pH, assessed from their distinctly different Fe-CO stretching modes in the resonance Raman spectra of the CO complex of the fully reduced enzyme. The two conformations (alpha and beta) interconvert reversibly in the pH 6-9 range with a pKa of 7.4, consistent with Fourier transform infrared spectroscopy measurements done at cryogenic temperatures (D.M. Mitchell, J.P. Sapleigh, A.M.Archer, J.O. Alben, and R.B.Gennis, 1996, Biochemistry 35:9446-9450). It is postulated that the different structures result from a change in the position of the Cu(B) atom with respect to the CO due to the presence of one or more ionizable groups in the vicinity of the binuclear center. The conserved tyrosine residue (Tyr-288 in R. sphaeroides, Tyr-244 in the bovine enzyme) that is adjacent to the oxygen-binding pocket or one of the histidines that coordinate Cu(B) are possible candidates. The existence of an equilibrium between the two conformers at physiological pH and room temperature suggests that the conformers may be functionally involved in enzymatic activity.

On the role of the K-proton transfer pathway in cytochrome c oxidase

Cytochrome c oxidase is a membrane-bound enzyme that catalyzes the four-electron reduction of oxygen to water. This highly exergonic reaction drives proton pumping across the membrane. One of the key questions associated with the function of cytochrome c oxidase is how the transfer of electrons and protons is coupled and how proton transfer is controlled by the enzyme. In this study we focus on the function of one of the proton transfer pathways of the R. sphaeroides enzyme, the so-called K-proton transfer pathway (containing a highly conserved Lys(I-362) residue), leading from the protein surface to the catalytic site. We have investigated the kinetics of the reaction of the reduced enzyme with oxygen in mutants of the enzyme in which a residue [Ser(I-299)] near the entry point of the pathway was modified with the use of site-directed mutagenesis. The results show that during the initial steps of oxygen reduction, electron transfer to the catalytic site (to form the "peroxy" state, P(r)) requires charge compensation through the proton pathway, but no proton uptake from the bulk solution. The charge compensation is proposed to involve a movement of the K(I-362) side chain toward the binuclear center. Thus, in contrast to what has been assumed previously, the results indicate that the K-pathway is used during oxygen reduction and that K(I-362) is charged at pH approximately 7.5. The movement of the Lys is proposed to regulate proton transfer by "shutting off" the protonic connectivity through the K-pathway after initiation of the O(2) reduction chemistry. This "shutoff" prevents a short-circuit of the proton-pumping machinery of the enzyme during the subsequent reaction steps.

Expression and mutagenesis of the NqrC subunit of the NQR respiratory Na(+) pump from Vibrio cholerae with covalently attached FMN

The Na(+)-translocating NADH:quinone oxidoreductase (Na(+)-NQR) is present in the membranes of a number of marine bacteria and pathogenic bacteria. Two of the six subunits of the Na(+)-NQR, NqrB and NqrC, have been previously shown to contain covalently bound flavin adenine mononucleotide (FMN). In the current work, the cloning of nqrC from Vibrio cholerae is reported. The gene has been expressed in V. cholerae and shown to contain one equivalent of covalently bound FMN. In contrast, no covalent flavin was detected when threonine-225 was replaced by leucine. The data show that the FMN attachment does not require assembly of the enzyme and are consistent with the unusual threonine attachment site.

Direct evidence for the protonation of aspartate-75, proposed to be at a quinol binding site, upon reduction of cytochrome bo3 from Escherichia coli

Aspartate-75 (D75) was recently suggested to participate in a ubiquinone-binding site in subunit I of cytochrome bo(3) from Escherichia coli on the basis of a structural model [Abramson, J., Riistama, S., Larsson, G., Jasaitis, A., Svensson-Ek, M., Laakkonen, L., Puustinen, A., Iwata, S., and Wikström, M. (2000) Nat. Struct. Biol. 7 (10), 910-917]. We studied the protonation state of D75 for the reduced and oxidized forms of the enzyme, using a combined site-directed mutagenesis, electrochemical, and FTIR spectroscopic approach. The D75H mutant is catalytically inactive, whereas the more conservative D75E substitution has quinol oxidase activity equal to that of the wild-type enzyme. Electrochemically induced FTIR difference spectra of the inactive D75H mutant enzyme show a clear decrease in the spectroscopic region characteristic of protonated aspartates and glutamates. Strong variations in the amide I region of the FTIR difference spectrum, however, reflect a more general perturbation due to this mutation of both the protein and the bound quinone. Electrochemically induced FTIR difference spectra on the highly conservative D75E mutant enzyme show a shift from 1734 to 1750 cm(-1) in direct comparison to wild type. After H/D exchange, the mode at 1750 cm(-1) shifts to 1735 cm(-1). These modes, concomitant with the reduced state of the enzyme, can be assigned to the nu(C=O) vibrational mode of protonated D75 and E75, respectively. In the spectroscopic region where signals for deprotonated acidic groups are expected, band shifts for the nu(COO(-))(s/as) modes from 1563 to 1554-1539 cm(-1) and from 1315 to 1336 cm(-1), respectively, are found for the oxidized enzyme. These signals indicate that D75 (or E75 in the mutant) is deprotonated in the oxidized form of cytochrome bo(3) and is protonated upon full reduction of the enzyme. It is suggested that upon reduction of the bound ubiquinone at the high affinity site, D75 takes up a proton, possibly sharing it with ubiquinol.

Relocation of internal bound water in bacteriorhodopsin during the photoreaction of M at low temperatures: an FTIR study

Changes in the FTIR difference spectra upon photoconversion of the M intermediate to its photoproduct(s) M' were studied in wild-type bacteriorhodopsin and several mutants at low temperatures. The studies aimed at examining whether internally bound water molecules interact with the chromophore and the key residues Asp85 and Asp96 in M, and whether these water molecules participate in the reprotonation of the Schiff base. We have found that three water molecules are perturbed by the isomerization of the chromophore in the M --> M' transition at 80 K. The perturbation of one water molecule, detected as a bilobe at 3567(+)/3550(-) cm(-)(1), relaxed in parallel with the relaxation of an Asp85 perturbation upon increasing temperature from 80 to 100 and 133 K (before the reprotonation of the Schiff base). Two water bands of M at 3588 and 3570 cm(-)(1) shift to 3640 cm(-)(1) upon photoconversion at 173 K. These bands were attributed to water molecules which are located in the vicinity of the Schiff base and Asp85 (Wat85). In the M to M' transition at 80 and 100 K, where the Schiff base remained unprotonated, the Wat85 pair stayed in similar states to those in M. The reprotonation of the Schiff base at 133 K occurred without the restoration of the Wat85 band around 3640 cm(-)(1). This band was restored at higher temperatures. Two water molecules in the region surrounded by Thr46, Asp96, and Phe219 (Wat219) were perturbed in the M to M' transition at 80 K and relaxed in parallel with the relaxation of the perturbation of Asp96 upon increasing the temperature. Mutant studies show that upon photoisomerization of the chromophore at 80 K one of the Wat219 water molecules moves closer to Val49 (located near the lysine side chain attached to retinal, and close to the Schiff base). These data along with our previous results indicate that the water molecules in the cytoplasmic domain participate in the connection of Asp96 with the Schiff base and undergo displacement during photoconversions, presumably shuttling between the Schiff base and a site close to Asp96 in the L to M to N transitions.

Proton transfer from glutamate 286 determines the transition rates between oxygen intermediates in cytochrome c oxidase

We have investigated the electron-proton coupling during the peroxy (P(R)) to oxo-ferryl (F) and F to oxidised (O) transitions in cytochrome c oxidase from Rhodobacter sphaeroides. The kinetics of these reactions were investigated in two different mutant enzymes: (1) ED(I-286), in which one of the key residues in the D-pathway, E(I-286), was replaced by an aspartate which has a shorter side chain than that of the glutamate and, (2) ML(II-263), in which the redox potential of Cu(A) is increased by approximately 100 mV, which slows electron transfer to the binuclear centre during the F-->O transition by a factor of approximately 200. In ED(I-286) proton uptake during P(R)-->F was slowed by a factor of approximately 5, which indicates that E(I-286) is the proton donor to P(R). In addition, in the mutant enzyme the F-->O transition rate displayed a deuterium isotope effect of approximately 2.5 as compared with approximately 7 in the wild-type enzyme. Since the entire deuterium isotope effect was shown to be associated with a single proton-transfer reaction in which the proton donor and acceptor must approach each other (M. Karpefors, P. Adelroth, P. Brzezinski, Biochemistry 39 (2000) 6850), the smaller deuterium isotope effect in ED(I-286) indicates that proton transfer from E(I-286) determines the rate also of the F-->O transition. In ML(II-263) the electron-transfer to the binuclear centre is slower than the intrinsic proton-transfer rate through the D-pathway. Nevertheless, both electron and proton transfer to the binuclear centre displayed a deuterium isotope effect of approximately 8, i.e., about the same as in the wild-type enzyme, which shows that these reactions are intimately coupled.

Proton pumping by cytochrome oxidase: progress, problems and postulates

The current status of our knowledge about the mechanism of proton pumping by cytochrome oxidase is discussed. Significant progress has resulted from the study of site-directed mutants within the proton-conducting pathways of the bacterial oxidases. There appear to be two channels to facilitate proton translocation within the enzyme and they are important at different parts of the catalytic cycle. The use of hydrogen peroxide as an alternative substrate provides a very useful experimental tool to explore the enzymology of this system, and insights gained from this approach are described. Proton transfer is coupled to and appears to regulate the rate of electron transfer steps during turnover. It is proposed that the initial step in the reaction involves a proton transfer to the active site that is important to convert metal-ligated hydroxide to water, which can more rapidly dissociate from the metals and allow the reaction with dioxygen which, we propose, can bind the one-electron reduced heme-copper center. Coordinated movement of protons and electrons over both short and long distances within the enzyme appear to be important at different parts of the catalytic cycle. During the initial reduction of dioxygen, direct hydrogen transfer to form a tyrosyl radical at the active site seems likely. Subsequent steps can be effectively blocked by mutation of a residue at the surface of the protein, apparently preventing the entry of protons.

Q-band ENDOR (electron nuclear double resonance) of the high-affinity ubisemiquinone center in cytochrome bo3 from Escherichia coli

Electron nuclear double resonance (ENDOR) was performed on the protein-bound, stabilized, high-affinity ubisemiquinone radical, QH*-, of bo3 quinol oxidase to determine its electronic spin distribution and to probe its interaction with its surroundings. Until this present work, such ENDOR studies of protein-stabilized ubisemiquinone centers have only been done on photosynthetic reaction centers whose function is to reduce a ubiquinol pool. In contrast, QH*- serves to oxidize a ubiquinol pool in the course of electron transfer from the ubiquinol pool to the oxygen-consuming center of terminal bo3 oxidase. As documented by large hyperfine couplings (>10 MHz) to nonexchangeable protons on the QH*- ubisemiquinone ring, we provide evidence for an electronic distribution on QH*- that is different from that of the semiquinones of reaction centers. Since the ubisemiquinone itself is physically nearly identical in both QH*- and the bacterial photosynthetic reaction centers, this electronic difference is evidently a function of the local protein environment. Interaction of QH*- with this local protein environment was explicitly shown by exchangeable deuteron ENDOR that implied hydrogen bonding to the quinone and by weak proton hyperfine couplings to the local protein matrix.

Mutations in the putative H-channel in the cytochrome c oxidase from Rhodobacter sphaeroides show that this channel is not important for proton conduction but reveal modulation of the properties of heme a

As the final electron acceptor in the respiratory chain of eukaryotic and many prokaryotic organisms, cytochrome c oxidase catalyzes the reduction of oxygen to water, concomitantly generating a proton gradient. X-ray structures of two cytochrome c oxidases have been reported, and in each structure three possible pathways for proton translocation are indicated: the D-, K-, and H-channels. The putative H-channel is most clearly delineated in the bovine heart oxidase and has been proposed to be functionally important for the translocation of pumped protons in the mammalian oxidase [Yoshikawa et al. (1998) Science 280, 1723-1729]. In the present work, the functional importance of residues lining the putative H-channel in the oxidase from Rhodobacter sphaeroides are examined by site-directed mutagenesis. Mutants were generated in eight different sites and the enzymes have been purified and characterized. The results suggest that the H-channel is not functionally important in the prokaryotic oxidase, in agreement with the conclusion from previous work with the oxidase from Paracoccus denitrificans [Pfitzner et al. (1998) J. Biomembr. Bioenerg. 30, 89-93]. Each of the mutants in R. sphaeroides, with an exception at only one position, is enzymatically active and pumps protons in reconstituted proteoliposomes. This includes H456A, where in the P. denitrificans oxidase a leucine residue substituted for the corresponding residue resulted in inactive enzyme. The only mutations that result in completely inactive enzyme in the set examined in the R. sphaeroides oxidase are in R52, a residue that, along with Q471, appears to be hydrogen-bonded to the formyl group of heme a in the X-ray structures. To characterize the interactions between this residue and the heme group, resonance Raman spectra of the R52 mutants were obtained. The frequency of the heme a formyl stretching mode in the R52A mutant is characteristic of that seen in non-hydrogen-bonded model heme a complexes. Thus the data confirm the presence of hydrogen bonding between the heme a formyl group and the R52 side chain, as suggested from crystallographic data. In the R52K mutant, this hydrogen bonding is maintained by the lysine residue, and this mutant enzyme retains near wild-type activity. The heme a formyl frequency is also affected by mutation of Q471, confirming the X-ray models that show this residue also has hydrogen-bonding interactions with the formyl group. Unlike R52, however, Q471 does not appear to be critical for the enzyme function.

Sequencing and preliminary characterization of the Na+-translocating NADH:ubiquinone oxidoreductase from Vibrio harveyi

The Na(+)-translocating NADH: ubiquinone oxidoreductase (Na(+)-NQR) generates an electrochemical Na(+) potential driven by aerobic respiration. Previous studies on the enzyme from Vibrio alginolyticus have shown that the Na(+)-NQR has six subunits, and it is known to contain FAD and an FeS center as redox cofactors. In the current work, the enzyme from the marine bacterium Vibrio harveyi has been purified and characterized. In addition to FAD, a second flavin, tentatively identified as FMN, was discovered to be covalently attached to the NqrC subunit. The purified V. harveyi Na(+)-NQR was reconstituted into proteoliposomes. The generation of a transmembrane electric potential by the enzyme upon NADH:Q(1) oxidoreduction was strictly dependent on Na(+), resistant to the protonophore CCCP, and sensitive to the sodium ionophore ETH-157, showing that the enzyme operates as a primary electrogenic sodium pump. Interior alkalinization of the inside-out proteoliposomes due to the operation of the Na(+)-NQR was accelerated by CCCP, inhibited by valinomycin, and completely arrested by ETH-157. Hence, the protons required for ubiquinol formation must be taken up from the outside of the liposomes, which corresponds to the bacterial cytoplasm. The Na(+)-NQR operon from this bacterium was sequenced, and the sequence shows strong homology to the previously reported Na(+)-NQR operons from V. alginolyticus and Haemophilus influenzae. Homology studies show that a number of other bacteria, including a number of pathogenic species, also have an Na(+)-NQR operon.

Mechanism of ubiquinol oxidation by the bc(1) complex: different domains of the quinol binding pocket and their role in the mechanism and binding of inhibitors

Structures of mitochondrial ubihydroquinone:cytochrome c oxidoreductase (bc(1) complex) from several animal sources have provided a basis for understanding the functional mechanism at the molecular level. Using structures of the chicken complex with and without inhibitors, we analyze the effects of mutation on quinol oxidation at the Q(o) site of the complex. We suggest a mechanism for the reaction that incorporates two features revealed by the structures, a movement of the iron sulfur protein between two separate reaction domains on cytochrome c(1) and cytochrome b and a bifurcated volume for the Q(o) site. The volume identified by inhibitor binding as the Q(o) site has two domains in which inhibitors of different classes bind differentially; a domain proximal to heme b(L), where myxothiazole and beta-methoxyacrylate- (MOA-) type inhibitors bind (class II), and a distal domain close to the iron sulfur protein docking interface, where stigmatellin and 5-n-undecyl-6-hydroxy-4,7-dioxobenzothiaole (UHDBT) bind (class I). Displacement of one class of inhibitor by another is accounted for by the overlap of their volumes, since the exit tunnel to the lipid phase forces the hydrophobic "tails" to occupy common space. We conclude that the site can contain only one "tailed" occupant, either an inhibitor or a quinol or one of their reaction products. The differential sensitivity of strains with mutations in the different domains is explained by the proximity of the affected residues to the binding domains of the inhibitors. New insights into mechanism are provided by analysis of mutations that affect changes in the electron paramagnetic resonance (EPR) spectrum of the iron sulfur protein, associated with its interactions with the Q(o)-site occupant. The structures show that all interactions with the iron sulfur protein must occur at the distal position. These include interactions between quinone, or class I inhibitors, and the reduced iron sulfur protein and formation of a reaction complex between quinol and oxidized iron sulfur protein. The step with high activation energy is after formation of the reaction complex, likely in formation of the semiquinone and subsequent dissociation of the complex into products. We suggest that further progress of the reaction requires a movement of semiquinone to the proximal position, thus mapping the bifurcated reaction to the bifurcated volume. We suggest that such a movement, together with a change in conformation of the site, would remove any semiquinone formed from further interaction with the oxidized [2Fe-2S] center and also from reaction with O(2) to form superoxide anion. We also identify two separate reaction paths for exit of the two protons released in quinol oxidation.

Examination of the reaction of fully reduced cytochrome oxidase with hydrogen peroxide by flow-flash spectroscopy

The reaction of cytochrome c oxidase with hydrogen peroxide has been of great value in generating and characterizing oxygenated species of the enzyme that are identical or similar to those formed during turnover of the enzyme with dioxygen. Most previous studies have utilized relatively low peroxide concentrations (millimolar range). In the current work, these studies have been extended to the examination of the kinetics of the single turnover of the fully reduced enzyme using much higher concentrations of peroxide to avoid limitations by the bimolecular reaction. The flow-flash method is used, in which laser photolysis of the CO adduct of the fully reduced enzyme initiates the reaction following rapid mixing of the enzyme with peroxide, and the reaction is monitored by observing the absorbance changes due to the heme components of the enzyme. The following reaction sequence is deduced from the data. (1) The initial product of the reaction appears to be heme a(3) oxoferryl (Fe(4+)=O(2)(-) + H(2)O). Since the conversion of ferrous to ferryl heme a(3) (Fe(2+) to Fe(4+)) is sufficient for this reaction, presumably Cu(B) remains reduced in the product, along with Cu(A) and heme a. (2) The second phase of the reaction is an internal rearrangement of electrons and protons in which the heme a(3) oxoferryl is reduced to ferric hydroxide (Fe(3+)OH(-)). In about 40% of the population, the electron comes from heme a, and in the remaining 60% of the population, Cu(B) is oxidized. This step has a time constant of about 65 micros. (3) The third apparent phase of the reaction includes two parallel reactions. The population of the enzyme with an electron in the binuclear center reacts with a second molecule of peroxide, forming compound F. The population of the enzyme with the two electrons on heme a and Cu(A) must first transfer an electron to the binuclear center, followed by reaction with a second molecule of peroxide, also yielding compound F. In each of these reaction pathways, the reaction time is 100-200 micros, i.e., much faster than the rate of reaction of peroxide with the fully oxidized enzyme. Thus, hydrogen peroxide is an efficient trap for a single electron in the binuclear center. (4) Compound F is then reduced by the final available electron, again from heme a, at the same rate as observed for the reduction of compound F formed during the reaction of the fully reduced oxidase with dioxygen. The product is the fully oxidized enzyme (heme a(3) Fe(3+)OH(-)), which reacts with a third molecule of hydrogen peroxide, forming compound P. The rate of this final reaction step saturates at high concentrations of peroxide (V(max) = 250 s(-)(1), K(m) = 350 mM). The data indicate a reaction mechanism for the steady-state peroxidase activity of the enzyme which, at pH 7.5, proceeds via the single-electron reduction of the binuclear center followed by reaction with peroxide to form compound F directly, without forming compound P. Peroxide is an efficient trap for the one-electron-reduced state of the binuclear center. The results also suggest that the reaction of hydrogen peroxide to the fully oxidized enzyme may be limited by the presence of hydroxide associated with the heme a(3) ferric species. The reaction of hydrogen peroxide with heme a(3) is very substantially accelerated by the availability of an electron on heme a, which is presumably transferred to the binuclear center concomitant with a proton that can convert the hydroxide to water, which is readily displaced.

Glutamate-89 in subunit II of cytochrome bo3 from Escherichia coli is required for the function of the heme-copper oxidase

Recent electrostatics calculations on the cytochrome c oxidase from Paracoccus denitrificans revealed an unexpected coupling between the redox state of the heme-copper center and the state of protonation of a glutamic acid (E78II) that is 25 A away in subunit II of the oxidase. Examination of more than 300 sequences of the homologous subunit in other heme-copper oxidases shows that this residue is virtually totally conserved and is in a cluster of very highly conserved residues at the "negative" end (bacterial cytoplasm or mitochondrial matrix) of the second transmembrane helix. The functional importance of several residues in this cluster (E89II, W93II, T94II, and P96II) was examined by site-directed mutagenesis of the corresponding region of the cytochrome bo(3) quinol oxidase from Escherichia coli (where E89II is the equivalent of residue E78II of the P. denitrificans oxidase). Substitution of E89II with either alanine or glutamine resulted in reducing the rate of turnover to about 43 or 10% of the wild-type value, respectively, whereas E89D has only about 60% of the activity of the control oxidase. The quinol oxidase activity of the W93V mutant is also reduced to about 30% of that of the wild-type oxidase. Spectroscopic studies with the purified E89A and E89Q mutants indicate no perturbation of the heme-copper center. The data suggest that E89II (E. coli numbering) is critical for the function of the heme copper oxidases. The proximity to K362 suggests that this glutamic acid residue may regulate proton entry or transit through the K-channel. This hypothesis is supported by the finding that the degree of oxidation of the low-spin heme b is greater in the steady state using hydrogen peroxide as an oxidant in place of dioxygen for the E89Q mutant. Thus, it appears that the inhibition resulting from the E89II mutation is due to a block in the reduction of the heme-copper binuclear center, expected for K-channel mutants.

Mass spectrometric determination of dioxygen bond splitting in the "peroxy" intermediate of cytochrome c oxidase

The "peroxy" intermediate (P form) of bovine cytochrome c oxidase was prepared by reaction of the two-electron reduced mixed-valence CO complex with (18)O(2) after photolytic removal of CO. The water present in the reaction mixture was recovered and analyzed for (18)O enrichment by mass spectrometry. It was found that approximately one oxygen atom ((18)O) per one equivalent of the P form was present in the bulk water. The data show that the oxygen-oxygen dioxygen bond is already broken in the P intermediate and that one oxygen atom can be readily released or exchanged with the oxygen of the solvent water.

Vibrational modes of ubiquinone in cytochrome bo(3) from Escherichia coli identified by Fourier transform infrared difference spectroscopy and specific (13)C labeling

In this study we present the infrared spectroscopic characterization of the bound ubiquinone in cytochrome bo(3) from Escherichia coli. Electrochemically induced Fourier transform infrared (FTIR) difference spectra of DeltaUbiA (an oxidase devoid of bound ubiquinone) and DeltaUbiA reconstituted with ubiquinone 2 and with isotopically labeled ubiquinone 2, where (13)C was introduced either at the 1- or at the 4-position of the ring (C=O groups), have been obtained. The vibrational modes of the quinone bound to the discussed high-affinity binding site (Q(H)) are compared to those from the synthetic quinones in solution, leading to the assignment of the C=O modes to a split signal at 1658/1668 cm(-)(1), with both carbonyls similarly contributing. The FTIR spectra of DeltaUbiA reconstituted with the labeled quinones indicate an essentially symmetrical and weak hydrogen bonding of the two C=O groups from the neutral quinone with the protein and distinct conformations of the 2- and 3-methoxy groups. Perturbations of the vibrational modes of the 5-methyl side groups are discussed for a signal at 1452 cm(-)(1). Only negligible shifts of the aromatic ring modes can be reported for the reduced and the protonated form of the quinone. Alterations of the protein upon quinone binding are reflected in the electrochemically induced FTIR difference spectra. In particular, difference signals at 1640-1633 cm(-)(1) and 1700-1670 cm(-)(1) indicate variations of beta-sheet secondary structure elements and loops, bands at 1706 and 1678 cm(-)(1) are tentatively attributed to individual amino acids, and a difference signal a 1540 cm(-)(1) is discussed to reflect an influence on C=C modes of the porphyrin ring or on deprotonated propionate groups of the hemes. Further tentative assignments are presented and discussed. The (13)C labeling experiments allow the assignment of the vibrational modes of a bound ubiquinone 8 in the electrochemically induced FTIR difference spectra of wild-type bo(3).

Detergent-solubilized Escherichia coli cytochrome bo3 ubiquinol oxidase: a monomeric, not a dimeric complex

The protein molecular weight, M(r), and hydrodynamic radius, R(s), of Triton X-100-solubilized Escherichia coli cytochrome bo3 were evaluated by computer fitting of sedimentation velocity data with finite element solutions to the Lamm equation. Detergent-solubilized cytochrome bo3 sediments as a homogeneous species with an S20,w of 6.75 s and a D20,w of 3.71 x 10(-7) cm2/s, corresponding to a R(s) of 5.8 nm and a M(r) of 144,000 +/- 3500. The protein molecular weight agrees very well with the value of 143,929 calculated from the four known subunit sequences and the value of 143,025 measured by MALDI mass spectrometry for the histidine-tagged enzyme. We conclude that detergent-solubilized E. coli ubiquinol oxidase is a monomeric complex of the four known subunits.

Chromophore-protein-water interactions in the L intermediate of bacteriorhodopsin: FTIR study of the photoreaction of L at 80 K

Using FTIR spectroscopy, perturbations of several residues and internal water molecules have been detected when light transforms all-trans bacteriorhodopsin (BR) to its L intermediate having a 13-cis chromophore. Illumination of L at 80 K results in an intermediate L' absorbing around 550 nm. L' thermally converts to the original BR only at >130 K. In this study, we used the light-induced transformation of L to L' at 80 K to identify some amino acid residues and water molecules that closely interact with the chromophore and distinguish them from those residues not affected by the photoreaction. The L minus L' FTIR difference spectrum shows that the chromophore in L' is in the all-trans configuration. The perturbed states of Asp96 and Val49 and of the environment along the aliphatic part of the retinal and Lys216 seen in L are not affected by the L --> L' photoreaction. On the other hand, the environments of the Schiff base of the chromophore, of Asp115, and of water molecules close to Asp85 returned in L' to their state in which they originally had existed in BR. The water molecules that are affected by the mutations of Thr46 and Asp96 also change to a different state in the L --> L' transition, as indicated by transformation of a water O-H vibrational band at 3497 cm-1 in L into an intense peak at 3549 cm-1 in L'. Notably, this change of water bands in the L --> L' transition at 80 K is entirely different from the changes observed in the BR --> K photoreaction at the same temperature, which does not show such intense bands. These results suggest that these water molecules move closer to the Schiff base as a hydrogen bonding cluster in L and L', presumably to stabilize its protonated state during the BR to L transition. They may contribute to the structural constraints that prevent L from returning to the initial BR upon illumination at 80 K.

Aspartate-132 in cytochrome c oxidase from Rhodobacter sphaeroides is involved in a two-step proton transfer during oxo-ferryl formation

The aspartate-132 in subunit I (D(I-132)) of cytochrome c oxidase from Rhodobacter sphaeroides is located on the cytoplasmic surface of the protein at the entry point of a proton-transfer pathway used for both substrate and pumped protons (D-pathway). Replacement of D(I-132) by its nonprotonatable analogue asparagine (DN(I-132)) has been shown to result in a reduced overall activity of the enzyme and impaired proton pumping. The results from this study show that during oxidation of the fully reduced enzyme the reaction was inhibited after formation of the oxo-ferryl (F) intermediate (tau congruent with 120 microseconds). In contrast to the wild-type enzyme, in the mutant enzyme formation of this intermediate was not associated with proton uptake from solution, which is the reason the DN(I-132) enzyme does not pump protons. The proton needed to form F was presumably taken from a protonatable group in the D-pathway (e.g., E(I-286)), which indicates that in the wild-type enzyme the proton transfer during F formation takes place in two steps: proton transfer from the group in the pathway is followed by faster reprotonation from the bulk solution, through D(I-132). Unlike the wild-type enzyme, in which F formation is coupled to internal electron transfer from CuA to heme a, in the DN(I-132) enzyme this electron transfer was uncoupled from formation of the F intermediate, which presumably is due to the impaired charge-compensating proton uptake from solution. In the presence of arachidonic acid which has been shown to stimulate the turnover activity of the DN(I-132) enzyme (Fetter et al. (1996) FEBS Lett. 393, 155), proton uptake with a time constant of approximately 2 ms was observed. However, no proton uptake associated with formation of F (tau congruent with 120 micros) was observed, which indicates that arachidonic acid can replace the role of D(I-132), but it cannot transfer protons as fast as the Asp. The results from this study show that D(I-132) is crucial for efficient transfer of protons into the enzyme and that in the DN(I-132) mutant enzyme there is a "kinetic barrier" for proton transfer into the D-pathway.

Cu XAS shows a change in the ligation of CuB upon reduction of cytochrome bo3 from Escherichia coli

Copper X-ray absorption spectroscopy (XAS) has been used to examine the structures of the Cu(II) and Cu(I) forms of the cytochrome bo3 quinol oxidase from Escherichia coli. Cytochrome bo3 is a member of the superfamily of heme-copper respiratory oxidases. Of particular interest is the fact that these enzymes function as redox-linked proton pumps, resulting in the net translocation of one H+ per electron across the membrane. The molecular mechanism of how this pump operates and the manner by which it is linked to the oxygen chemistry at the active site of the enzyme are unknown. Several proposals have featured changes in the coordination of CuB during enzyme turnover that would result in sequential protonation or deprotonation events that are key to the functioning proton pump. This would imply lability of the ligands to CuB. In this work, the structure of the protein in the immediate vicinity of CuB, in both the fully oxidized and fully reduced forms of the enzyme, has been examined by Cu XAS, a technique that is particularly sensitive to changes in metal coordination. The results show that in the oxidized enzyme, CuB(II) is four-coordinate, consistent with three imidazoles and one hydroxyl (or water). Upon reduction of the enzyme, the coordination of CuB(I) is significantly altered, consistent with the loss of one of the histidine imidazole ligands in at least a substantial fraction of the population. These data add to the credibility that changes in the ligation of CuB might occur during catalytic turnover of the enzyme and, therefore, could, in principle, be part of the mechanism of proton pumping.

Expression and one-step purification of a fully active polyhistidine-tagged cytochrome bc1 complex from Rhodobacter sphaeroides

The fbcB and fbcC genes encoding cytochromes b and c1 of the bc1 complex were extended with a segment to encode a polyhistidine tag linked to their C-terminal sequence allowing a one-step affinity purification of the complex. Constructions were made in vitro in a pUC-derived background using PCR amplification. The modified fbc operons were transferred to a pRK derivative plasmid, and this was used to transform the fbc- strain of Rhodobacter sphaeroides, BC17. The transformants showed normal rates of growth. Chromatophores prepared from these cells showed kinetics of turnover of the bc1 complex on flash activation which were essentially the same as those from wild-type strains, and analysis of the cytochrome complement and spectral and thermodynamic properties by redox potentiometry showed no marked difference from the wild type. Chromatophores were solubilized and mixed with Ni-NTA-Sepharose resin. A modification of the standard elution protocol in which histidine replaced imidazole increased the activity 20-fold. Imidazole modified the redox properties of heme c1, suggesting ligand displacement and inactivation when this reagent is used at high concentration. The purified enzyme contained all four subunits in an active dimeric complex. This construction provides a facile method for preparation of wild-type or mutant bc1 complex, for spectroscopy and structural studies.

Observation of a novel transient ferryl complex with reduced CuB in cytochrome c oxidase

The reaction between mixed-valence (MV) cytochrome c oxidase from beef heart with H2O2 was investigated using the flow-flash technique with a high concentration of H2O2 (1 M) to ensure a fast bimolecular interaction with the enzyme. Under anaerobic conditions the reaction exhibits 3 apparent phases. The first phase (tau congruent with 25 micros) results from the binding of one molecule of H2O2 to reduced heme a3 and the formation of an intermediate which is heme a3 oxoferryl (Fe4+=O2-) with reduced CuB (plus water). During the second phase (tau congruent with 90 micros), the electron transfer from CuB+ to the heme oxoferryl takes place, yielding the oxidized form of cytochrome oxidase (heme a3 Fe3+ and CuB2+, plus hydroxide). During the third phase (tau congruent with 4 ms), an additional molecule of H2O2 binds to the oxidized form of the enzyme and forms compound P, similar to the product observed upon the reaction of the mixed-valence (i.e., two-electron reduced) form of the enzyme with dioxygen. Thus, within about 30 ms the reaction of the mixed-valence form of the enzyme with H2O2 yields the same compound P as does the reaction with dioxygen, as indicated by the final absorbance at 436 nm, which is the same in both cases. This experimental approach allows the investigation of the form of cytochrome c oxidase which has the heme a3 oxoferryl intermediate but with reduced CuB. This state of the enzyme cannot be obtained from the reaction with dioxygen and is potentially useful to address questions concerning the role of the redox state in CuB in the proton pumping mechanism.

Sequence analysis of cytochrome bd oxidase suggests a revised topology for subunit I

Numerous sequences of the cytochrome bd quinol oxidase (cytochrome bd) have recently become available for analysis. The analysis has revealed a small number of conserved residues, a new topology for subunit I and a phylogenetic tree involving extensive horizontal gene transfer. There are 20 conserved residues in subunit I and two in subunit II. Algorithms utilizing multiple sequence alignments predicted a revised topology for cytochrome bd, adding two transmembrane helices to subunit I to the seven that were previously indicated by the analysis of the sequence of the oxidase from E. coli. This revised topology has the effect of relocating the N-terminus and C-terminus to the periplasmic and cytoplasmic sides of the membrane, respectively. The new topology repositions I-H19, the putative ligand for heme b595, close to the periplasmic edge of the membrane, which suggests that the heme b595/heme d active site of the oxidase is located near the outer (periplasmic) surface of the membrane. The most highly conserved region of the sequence of subunit I contains the sequence GRQPW and is located in a predicted periplasmic loop connecting the eighth and ninth transmembrane helices. The potential importance of this region of the protein was previously unsuspected, and it may participate in the binding of either quinol or heme d. There are two very highly conserved glutamates in subunit I, E99 and E107, within the third transmembrane helix (E. coli cytochrome bd-I numbering). It is speculated that these glutamates may be part of a proton channel leading from the cytoplasmic side of the membrane to the heme d oxygen-reactive site, now placed near the periplasmic surface. The revised topology and newly revealed conserved residues provide a clear basis for further experimental tests of these hypotheses. Phylogenetic analysis of the new sequences of cytochrome bd reveals considerable deviation from the 16sRNA tree, suggesting that a large amount of horizontal gene transfer has occurred in the evolution of cytochrome bd.

Magnetic circular dichroism used to examine the interaction of Escherichia coli cytochrome bd with ligands

The interactions of the fully reduced and fully oxidized cytochrome bd from E. coli with ligands CO, NO, and CN- have been studied by a combination of absorption and magnetic circular dichroism (MCD) spectroscopy. In the reduced cytochrome bd, MCD resolves individual bands due to the high-spin heme b595 and the low-spin heme b558 components of the enzyme, allowing one to separately monitor their interactions along with ligand binding to the heme d component. The data show that at low concentrations, the ligands bind almost exclusively to heme d. At high concentrations, the ligands begin to interact with the low-spin heme b558. At the same time, no evidence for significant binding of the ligands to the high-spin heme b595 is revealed in either the reduced or the fully oxidized cytochrome bd complex. The data support the model [Borisov, V. B., Gennis, R. B., and Konstantinov, A. A. (1995) Biochemistry (Moscow) 60, 231-239] according to which the two high-spin hemes d and b595 share a high-affinity ligand binding site with a capacity for only a single molecule of the ligand; i.e., there is a strong negative cooperativity with respect to ligand binding to these two hemes with cytochrome d having an intrinsic ligand affinity much higher than that of heme b595.

Single electron reduction of cytochrome c oxidase compound F: resolution of partial steps by transient spectroscopy

The final step of the catalytic cycle of cytochrome oxidase, the reduction of oxyferryl heme a3 in compound F, was investigated using a binuclear polypyridine ruthenium complex (Ru2C) as a photoactive reducing agent. The net charge of +4 on Ru2C allows it to bind electrostatically near CuA in subunit II of cytochrome oxidase. Photoexcitation of Ru2C with a laser flash results in formation of a metal-to-ligand charge-transfer excited state, Ru2C, which rapidly transfers an electron to CuA of cytochrome oxidase from either beef heart or Rhodobacter sphaeroides. This is followed by reversible electron transfer from CuA to heme a with forward and reverse rate constants of k1 = 9.3 x 10(4) s-1 and k-1 = 1.7 x 10(4) s-1 for R. sphaeroides cytochrome oxidase in the resting state. Compound F was prepared by treating the resting enzyme with excess hydrogen peroxide. The value of the rate constant k1 is the same in compound F where heme a3 is in the oxyferryl form as in the resting enzyme where heme a3 is ferric. Reduction of heme a in compound F is followed by electron transfer from heme a to oxyferryl heme a3 with a rate constant of 700 s-1, as indicated by transients at 605 and 580 nm. No delay between heme a reoxidation and oxyferryl heme a3 reduction is observed, showing that no electron-transfer intermediates, such as reduced CuB, accumulate in this process. The rate constant for electron transfer from heme a to oxyferryl heme a3 was measured in beef cytochrome oxidase from pH 7.0 to pH 9.5, and found to decrease upon titration of a group with a pKa of 9.0. The rate constant is slower in D2O than in H2O by a factor of 4.3, indicating that the electron-transfer reaction is rate-limited by a proton-transfer step. The pH dependence and deuterium isotope effect for reduction of isolated compound F are comparable to that observed during reaction of the reduced, CO-inhibited CcO with oxygen by the flow-flash technique. This result indicates that electron transfer from heme a to oxyferryl heme a3 is not controlled by conformational effects imposed by the initial redox state of the enzyme. The rate constant for electron transfer from heme a to oxyferryl heme a3 is the same in the R. sphaeroides K362M CcO mutant as in wild-type CcO, indicating that the K-channel is not involved in proton uptake during reduction of compound F.

The post-translational modification in cytochrome c oxidase is required to establish a functional environment of the catalytic site

Mutation of tyrosine-288 to a phenylalanine in cytochrome c oxidase from Rhodobacter sphaeroides drastically alters its properties. Tyr-288 lies in the CuB-cytochrome a3 binuclear catalytic site and forms a hydrogen bond with the hydroxy group on the farnesyl side chain of the heme. In addition, through a post-translational modification, Y288 is covalently linked to one of the histidine ligands that is coordinated to CuB. In the Y288F mutant enzyme, the "as-isolated" preparation is a mixture of reduced cytochrome a and oxidized cytochrome a3. The cytochrome a3 heme, which is largely six-coordinate low-spin in both oxidation states of the mutant, cannot be reduced by cytochrome c, but only by dithionite, possibly due to a large decrease in its reduction potential. It is postulated that the Y288F mutation prevents the post-translational modification from occurring. As a consequence, the catalytic site becomes disrupted. Thus, one role of the post-translational modification is to stabilize the functional catalytic site by maintaining the correct ligands on CuB, thereby preventing nonfunctional ligands from coordinating to the heme.

Tryptophan-136 in subunit II of cytochrome bo3 from Escherichia coli may participate in the binding of ubiquinol

In the cytochrome c oxidases, the role of subunit II is to provide the electron entry site into the enzyme. This subunit contains both the binding site for the substrate, cytochrome c, and the CuA redox center, which is initially reduced by cytochrome c. Cytochrome bo3 and other quinol oxidases that are members of the heme-copper oxidase superfamily have a homologous subunit II, but the CuA site is absent, as is the docking site for cytochrome c. Speculation that subunit II in the quinol oxidases may also be important as an electron entry site is supported by the demonstration several years ago that a photoreactive substrate analogue, azido-Q, covalently labeled subunit II in cytochrome bo3. In the current work, a sequence alignment of subunit II of heme-copper quinol oxidases is used as a guide to select conserved residues that might be important for the binding of ubiquinol to cytochrome bo3. Results are presented for point mutants in 24 different residue positions in subunit II. The membrane-bound enzymes were examined by optical spectroscopy and by determining the activity of ubiquinol-1 oxidase. In each case, the Km for ubiquinol-1 was determined as a measure of possible perturbation to a quinol binding site. The only mutant that had a noticeably altered Km for ubiquinol-1 was W136A, in which the Km was about sixfold increased. Thus, W136 may be at or close to a substrate (ubiquinol)-binding site in cytochrome bo3. In the cytochrome c oxidases, the equivalent tryptophan (W121 in Paracoccus denitrificans) has been identified as the "electron entry site".

The cbb3-type cytochrome c oxidase from Rhodobacter sphaeroides, a proton-pumping heme-copper oxidase

Rhodobacter sphaeroides expresses a bb3-type quinol oxidase, and two cytochrome c oxidases: cytochrome aa3 and cytochrome cbb3. We report here the characterization of the genes encoding this latter oxidase. The ccoNOQP gene cluster of R. sphaeroides contains four open reading frames with high similarity to all ccoNOQP/fixNOQP gene clusters reported so far. CcoN has the six highly conserved histidines proposed to be involved in binding the low spin heme, and the binuclear center metals. ccoO and ccoP code for membrane bound mono- and diheme cytochromes c. ccoQ codes for a small hydrophobic protein of unknown function. Upstream from the cluster there is a conserved Fnr/FixK-like box which may regulate its expression. Analysis of a R. sphaeroides mutant in which the ccoNOQP gene cluster was inactivated confirms that this cluster encodes the cbb3-type oxidase previously purified. Analysis of proton translocation in several strains shows that cytochrome cbb3 is a proton pump. We also conclude that cytochromes cbb3 and aa3 are the only cytochrome c oxidases in the respiratory chain of R. sphaeroides.

Using matrix-assisted laser desorption ionization mass spectrometry to map the quinol binding site of cytochrome bo3 from Escherichia coli

The cytochrome bo3 ubiquinol oxidase contains at least one and possibly two binding sites for ubiquinol/ubiquinone. Previous studies used the photoreactive affinity label 3-[3H]azido-2-methyl-5-methoxy-6-geranyl-1,4-benzoquinone (azido-Q), a substrate analogue, to demonstrate that subunit II contributes to at least one of the quinol binding sites. In the current work, mass spectroscopy is used to identify a peptide within subunit II that is photolabeled by the azido-Q. Purified cytochrome bo3 was photolabeled as previously described using azido-Q that was not tritiated (i.e., not radiolabeled). Subunit II was then isolated from an SDS-PAGE gel and proteolyzed in situ with trypsin. The resulting peptides were eluted from the gel and then identified using matrix-assisted laser desorption ionization mass spectrometry. The resulting mass spectrum was compared to that obtained by analysis of subunit II that had not been exposed to the photolabel. Using the amino acid sequence, each peak in the mass spectrum of the unlabeled subunit II could be assigned to an expected trypsin fragment. Two additional peaks were observed in the mass spectrum of the photolabeled subunit with m/z 1931.9 and 2287.7. Subtraction of the mass of azido-Q from the peak at m/z 1931.9 results in a mass equivalent to that of a peptide consisting of amino acids 165-178. The assignment of the peak at m/z 2287.7 cannot be made unequivocally and may correspond either to the covalent attachment of azido-Q to peptide 254-270 or to a peptide resulting from incomplete proteolysis. The labeled peptide, 165-178, is within the water-soluble domain of subunit II, whose X-ray structure is known. This peptide is located near the site where CuA is located in the homologous cytochrome c oxidases and can be placed near the interface between subunits I and II.

Identification of a stable semiquinone intermediate in the purified and membrane bound ubiquinol oxidase-cytochrome bd from Escherichia coli

The quinol oxidase, cytochrome bd, functions as a terminal oxidase in the Escherichia coli respiratory chain, reducing O2 to water and using ubiquinol-8 or menaquinol-8 as its immediate reductant. The oxidation of quinol is by the low-spin ferri-haem, cytochrome b558. This occurs at a quinol-binding site by sequential one electron steps, requiring the stabilisation of the semiquinone intermediate. We have observed, by EPR spectroscopy, the properties of this semiquinone radical in appropriately poised samples of purified enzyme reconstituted with excess of ubiquinone-8 and menaquinone-8 analogues. The line width of the EPR spectrum is approximately 0.9 mT, which is consistent with a semiquinone anion of this type. The line shape is Gaussian. The semiquinone is highly stabilised with respect to free semiquinone; significant free radical can be observed at pH 7.0 and above. The pH dependence of the redox reactions indicate that the anionic form of the semiquinone and the neutral form of the quinol predominate in the pH range studied. The pH dependence of the mid-point potentials of the one electron reactions from pH 7.0-9.0 is 120 mV/pH change for the semiquinone anion to quinol (E2) and none for the quinone to semiquinone (E1). The semiquinone radical is attenuated on titration with putative inhibitors of this quinone-binding site. We have similarly studied the semiquinone in membrane preparations from a strain with overexpression of cytochrome bd oxidase. The data can be fitted with the assumption of a single quinone-binding site.

Localization of histidine residues responsible for heme axial ligation in cytochrome b556 of complex II (succinate:ubiquinone oxidoreductase) in Escherichia coli

Complex II (succinate:ubiquinone oxidoreductase) from Escherichia coli contains four different subunits. Two of the subunits (SDHC and SDHD) are hydrophobic and anchor the two more hydrophilic (flavin and iron-sulfur) subunits (SDHA and SDHB) to the cytoplasmic membrane. Previous studies have shown that the complex of SDHC/SDHD is required to maintain the heme B component of the enzyme and that the heme B is ligated to the protein by two histidine ligands. In the current work, the histidines within SDHC and SDHD have been systematically mutated. SDHC-His91 and SDHD-His14 were eliminated as potential ligands by these studies. SDHC-His84 and SDHD-His71 have been identified as the most likely heme axial ligands in the E. coli enzyme, suggesting that the heme bridges these two subunits in the membrane. Furthermore, the results show that the four-subunit Complex II assembles and retains function despite the absence of the heme B prosthetic group in the membrane. The results do not rule out completely SDHC-His30 as a candidate for heme ligation, but do show that mutation at this position prevents assembly of Complex II in the membrane.

Substitution of lysine-362 in a putative proton-conducting channel in the cytochrome c oxidase from Rhodobacter sphaeroides blocks turnover with O2 but not with H2O2

The recently reported X-ray structures of cytochrome oxidase reveal structures that are likely proton-conducting channels. One of these channels, leading from the negative aqueous surface to the heme a3/CuB bimetallic center, contains a lysine as a central element. Previous work has shown that this lysine (K362 in the oxidase from Rhodobacter sphaeroides) is essential for cytochrome c oxidase activity. The data presented demonstrate that the K362M mutant is impeded in the reduction of the heme a3/CuB bimetallic center, probably by interfering with the intramolecular movement of protons. The reduction of the heme-copper center is required prior to the reaction with dioxygen to form the so-called peroxy intermediate (compound P). This block can be by-passed to some extent by the addition of H2O2, which can react with the enzyme without prereduction of the heme-copper center and can then be reduced to water using electrons from cytochrome c. Hence, the K362M mutant, though lacking oxidase activity, exhibits cytochrome c peroxidase activity. Rapid mixing techniques have been used to determine the kinetics of this peroxidase activity at concentrations of H2O2 up to 0.5 M. The Km for peroxide is about 50 mM and the Vmax is 50 electrons s-1, which is considerably slower than the turnover that can be obtained for the oxidase activity of the wild-type enzyme (1200 s-1). The turnover of the mutant oxidase with H2O2 appears to be limited by the rate of reaction of the enzyme with peroxide to form compound P, rather than the rate of reduction of compound P to water by cytochrome c. The data require a reexamination of the proposed roles of the putative proton-conducting channels.

Role of the pathway through K(I-362) in proton transfer in cytochrome c oxidase from R. sphaeroides

In this study we have combined the use of site-directed mutants with time-resolved optical absorption spectroscopy to investigate the role of the protonatable subunit-I residues lysine-362 (K(I-362)) and threonine-359 (T(I-359)) in cytochrome c oxidase from Rhodobacter sphaeroides in electron and proton transfer. These residues have been proposed to be part of a proton-transfer pathway in cytochrome oxidases from Paracoccus denitrificans and bovine heart. Mutation of K(I-362) and T(I-359) to methionine and alanine, respectively, results in reduction of the overall turnover activities to <2% and approximately 35%, respectively, of those in the wild-type enzyme. The results show that in the absence of dioxygen, electron transfer between hemes a3 and a with a time constant of approximately 3 micros, not coupled to protonation reactions, is not affected in the mutant enzymes. However, the slower electron transfer between hemes a3 and a, coupled to proton release with a time constant of approximately 3 ms (at pH 9.0) is impaired in the KM(I-362) and TA(I-359) mutant enzymes. This is consistent with the slow reduction rate of heme a3 in the oxidized KM(I-362) enzyme because in the wild-type enzyme reduction of heme a3 is coupled to proton uptake. On the other hand, when reacting with O2, both the wild-type and mutant fully reduced enzymes become oxidized in approximately 5 ms, and proton uptake on this time scale is not affected. Hence, the results indicate that the KM(I-362) mutant enzyme is inactive because the proton-transfer pathway through K(I-362) and T(I-359) is involved in proton uptake during reduction of the oxidized binuclear center. Proton uptake during oxidation of the fully reduced enzyme takes place through a different pathway [through E(I-286) (Adelroth, P., et al. (1997) Biochemistry 36, 13824-13829)].

Matrix-assisted laser desorption ionization mass spectrometry of membrane proteins: demonstration of a simple method to determine subunit molecular weights of hydrophobic subunits

Matrix-assisted laser desorption ionization (MALDI) mass spectrometry has been used to obtain accurate molecular weight information for each subunit of several hydrophobic integral membrane proteins: cytochrome bo3 (4 subunits) and cytochrome bd (2 subunits) from E. coli, and the bc1 complex (3 subunits) and the cytochrome c oxidase (3 subunits) from Rhodobacter sphaeroides. The results demonstrate that the MALDI method is a convenient, quick, sensitive and reliable means for obtaining the molecular masses of the subunits of purified multisubunit membrane proteins.

Effects of mutation of the conserved lysine-362 in cytochrome c oxidase from Rhodobacter sphaeroides

We describe the effects of a mutation, K362M, of the conserved lysine in cytochrome c oxidase from Rhodobacter sphaeroides, a residue located in a putative proton channel that may convey substrate protons to the binuclear center. Spectra of the "as prepared", ferricyanide-oxidized, and dithionite-reduced forms of the mutant protein confirm that the redox centers remain intact. Ligand binding kinetics of the ferricyanide-oxidized enzyme and of the dithionite-reducible fraction are similar to those of the wild type, indicating that the K channel is not the major route for CO, cyanide, formate, or peroxide entry into the structure. Protonation of the lysine residue is not redox-linked to heme a or CuB as judged from the essentially unaltered midpoint potentials of these centers in the cyanide-ligated enzyme. A difficulty in electron transfer from heme a to the binuclear center is indicated by the slow and only partial reduction of heme a3 by dithionite or ferrocytochrome c and by the presence of some reduced heme a in the as prepared mutant enzyme and under steady-state conditions. Further characterization of the K362M enzyme in the steady state shows that up to one electron, but not two, can enter the binuclear center easily. It is this inability to form the two-electron-reduced, oxygen-reactive R state that prevents activity. A model is proposed where the K channel serves as a dielectric well of high dielectric strength and low proton conductivity, rather than as a pathway for proton entry to the binuclear center. The function of this structure would be to decrease the cost of introducing a transiently uncompensated charge into the binuclear center prior to formation of a stable, charge-compensated R-state.

Glutamate 286 in cytochrome aa3 from Rhodobacter sphaeroides is involved in proton uptake during the reaction of the fully-reduced enzyme with dioxygen

The reaction with dioxygen of solubilized fully-reduced wild-type and EQ(I-286) (exchange of glutamate 286 of subunit I for glutamine) mutant cytochrome c oxidase from Rhodobacter sphaeroides has been studied using the flow-flash technique in combination with optical absorption spectroscopy. Proton uptake was measured using a pH-indicator dye. In addition, internal electron-transfer reactions were studied in the absence of oxygen. Glutamate 286 is found in a proton pathway proposed to be used for pumped protons from the crystal structure of cytochrome c oxidase from Paracoccus denitrificans [Iwata et al. (1995) Nature 376, 660-669; E278 in P.d. numbering]. It is the residue closest to the oxygen-binding binuclear center that is clearly a part of the pathway. The results show that the wild-type enzyme becomes fully oxidized in a few milliseconds at pH 7.4 and displays a biphasic proton uptake from the medium. In the EQ(I-286) mutant enzyme, electron transfer after formation of the peroxy intermediate is impaired, CuA remains reduced, and no protons are taken up from the medium. Thus, the results suggest that E(I-286) is necessary for proton uptake after formation of the peroxy intermediate and transfer of the fourth electron to the binuclear center. The results also indicate that the proton uptake associated with formation of the ferryl intermediate controls the electron transfer from CuA to heme a.

A conserved glutamic acid in helix VI of cytochrome bo3 influences a key step in oxygen reduction

We have compared the reactions with dioxygen of wild-type cytochrome bo3 and a mutant in which a conserved glutamic acid at position-286 of subunit I has been changed to an alanine. Flow-flash experiments reveal that oxygen binding and the rate of heme-heme electron transfer are unaffected by the mutation. Reaction of the fully (3-electron) reduced mutant cytochrome bo3 with dioxygen yields a binuclear center which is substantially in the P (peroxy) state, not the well-characterized F (oxyferryl) state which is the product of the reaction of the fully reduced wild-type enzyme with dioxygen [Puustinen, A., et al. (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 1545-1548]. These results confirm that proton uptake is important in controlling the later stages of dioxygen reduction in heme-copper oxidases and show that E286 is an important component of the channel that delivers these protons to the active site.

Factors determining electron-transfer rates in cytochrome c oxidase: studies of the FQ(I-391) mutant of the Rhodobacter sphaeroides enzyme

The mechanisms of internal electron transfer and oxygen reduction were investigated in cytochrome c oxidase from Rhodobacter sphaeroides (cytochrome aa3) using site-directed mutagenesis in combination with time-resolved optical absorption spectroscopy. Electron-transfer reactions in the absence of O2 were studied after flash photolysis of CO from the partly-reduced enzyme and the reaction of the fully-reduced enzyme with O2 was studied using the so-called flow-flash technique. Results from studies of the wild-type and mutant enzyme in which phenylalanine-391 of subunit I was replaced by glutamine (FQ(I-391)) were compared. The turnover activity of the mutant enzyme was approximately 2% ( approximately 30 s-1) of that of the wild-type enzyme. After flash photolysis of CO from the partly-reduced mutant enzyme approximately 80% of CuA was reduced, which is a much larger fraction than in the wild-type enzyme, and the rate of this electron transfer was 3.2 x 10(3) s-1, which is significantly slower than in the wild-type enzyme. The redox potentials of hemes a and a3 in the mutant enzyme were found to be shifted by about +30 and -70 mV, respectively, as compared to the wild-type enzyme. During the reaction of the fully-reduced FQ(I-391) mutant enzyme with O2 a rapid kinetic phase with a rate constant of 1.2 x 10(5) s-1, presumably associated with O2 binding, was followed by formation of the P intermediate with electrons from heme a3 and CuB with a rate of approximately 4 x 10(3) s-1, and oxidation of the enzyme with a rate of approximately 30 s-1. The dramatically slower electron transfer between the hemes during O2 reduction in the mutant enzyme is not only due to the slower intrinsic electron transfer, but also due to the altered redox potentials. In addition, the results show that the reduced overall activity of the mutant enzyme is due to the slower electron transfer from heme a to the binuclear center during O2 reduction. The relation between the intrinsic heme a/heme a3 electron-transfer rate and equilibrium constant, and the electron-transfer rate from heme a to the binuclear center during O2 reduction is discussed.

Subunit II of the cytochrome bo3 ubiquinol oxidase from Escherichia coli is a lipoprotein

The purified Escherichia coli cytochrome bo3 ubiquinol oxidase contains four subunits that are each integral components of the cytoplasmic membrane. The molecular weight of each of the subunits has been determined by matrix-assisted laser desorption ionization mass spectrometry (MALDI). The observed molecular weight of subunit II (CyoA) is considerably less than the calculated value from the deduced amino acid sequence, indicating possible posttranslational processing. The similarity of a portion of the sequence near the N-terminus of CyoA with the sequences of known prokaryotic membrane-bound lipoproteins suggested that CyoA is proteolytically processed to generate an N-terminus at Cys25, and that Cys25 is covalently modified by the addition of lipids. This would be consistent with the observed molecular mass, and was confirmed by demonstrating the incorporation of radioactive palmitic acid into subunit II of the cytochrome bo3 oxidase. Site-directed mutagenesis replacing Cys25 by alanine prevents the processing, generating a precursor form of CyoA with a higher molecular mass. The C25A mutant of CyoA still assembles as an active quinol oxidase capable of supporting growth of the cells by aerobic respiration. Hence, this unusual processing of a cytoplasmic membrane protein, which is already anchored to the membrane by two transmembrane helices, is not essential for either assembly or function.

The roles of the two proton input channels in cytochrome c oxidase from Rhodobacter sphaeroides probed by the effects of site-directed mutations on time-resolved electrogenic intraprotein proton transfer

The crystal structures of cytochrome c oxidase from both bovine and Paracoccus denitrificans reveal two putative proton input channels that connect the heme-copper center, where dioxygen is reduced, to the internal aqueous phase. In this work we have examined the role of these two channels, looking at the effects of site-directed mutations of residues observed in each of the channels of the cytochrome c oxidase from Rhodobacter sphaeroides. A photoelectric technique was used to monitor the time-resolved electrogenic proton transfer steps associated with the photo-induced reduction of the ferryl-oxo form of heme a3 (Fe4+ = O2-) to the oxidized form (Fe3+OH-). This redox step requires the delivery of a "chemical" H+ to protonate the reduced oxygen atom and is also coupled to proton pumping. It is found that mutations in the K channel (K362M and T359A) have virtually no effect on the ferryl-oxo-to-oxidized (F-to-Ox) transition, although steady-state turnover is severely limited. In contrast, electrogenic proton transfer at this step is strongly suppressed by mutations in the D channel. The results strongly suggest that the functional roles of the two channels are not the separate delivery of chemical or pumped protons, as proposed recently [Iwata, S., Ostermeier, C., Ludwig, B. & Michel, H. (1995) Nature (London) 376, 660-669]. The D channel is likely to be involved in the uptake of both "chemical" and "pumped" protons in the F-to-Ox transition, whereas the K channel is probably idle at this partial reaction and is likely to be used for loading the enzyme with protons at some earlier steps of the catalytic cycle. This conclusion agrees with different redox states of heme a3 in the K362M and E286Q mutants under aerobic steady-state turnover conditions.