Rates of cyanide binding to the catalytic intermediates of mammalian cytochrome c oxidase, and the effects of cytochrome c and poly(L-lysine)

MpaR-driven expression of an orphan terminal oxidase subunit supports Pseudomonas aeruginosa biofilm respiration and development during cyanogenesis

IMPORTANCE

Cyanide is an inhibitor of heme-copper oxidases, which are required for aerobic respiration in all eukaryotes and many prokaryotes. This fast-acting poison can arise from diverse sources, but mechanisms by which bacteria sense it are poorly understood. We investigated the regulatory response to cyanide in the pathogenic bacterium Pseudomonas aeruginosa, which produces cyanide as a virulence factor. Although P. aeruginosa has the capacity to produce a cyanide-resistant oxidase, it relies primarily on heme-copper oxidases and even makes additional heme-copper oxidase proteins specifically under cyanide-producing conditions. We found that the protein MpaR controls expression of cyanide-inducible genes in P. aeruginosa and elucidated the molecular details of this regulation. MpaR contains a DNA-binding domain and a domain predicted to bind pyridoxal phosphate (vitamin B6), a compound that is known to react spontaneously with cyanide. These observations provide insight into the understudied phenomenon of cyanide-dependent regulation of gene expression in bacteria.

Cyanide-dependent control of terminal oxidase hybridization by Pseudomonas aeruginosa MpaR

Pseudomonas aeruginosa is a common, biofilm-forming pathogen that exhibits complex pathways of redox metabolism. It produces four different types of terminal oxidases for aerobic respiration, and for one of these-the cbb3-type terminal oxidases-it has the capacity to produce at least 16 isoforms encoded by partially redundant operons. It also produces small-molecule virulence factors that interact with the respiratory chain, including the poison cyanide. Previous studies had indicated a role for cyanide in activating expression of an "orphan" terminal oxidase subunit gene called ccoN4 and that the product contributes to P. aeruginosa cyanide resistance, fitness in biofilms, and virulence-but the mechanisms underlying this process had not been elucidated. Here, we show that the regulatory protein MpaR, which is predicted to be a pyridoxal phosphate-binding transcription factor and is encoded just upstream of ccoN4, controls ccoN4 expression in response to endogenous cyanide. Paradoxically, we find that cyanide production is required to support CcoN4's contribution to respiration in biofilms. We identify a palindromic motif required for cyanide- and MpaR-dependent expression of ccoN4 and co-expressed, adjacent loci. We also characterize the regulatory logic of this region of the chromosome. Finally, we identify residues in the putative cofactor-binding pocket of MpaR that are required for ccoN4 expression. Together, our findings illustrate a novel scenario in which the respiratory toxin cyanide acts as a signal to control gene expression in a bacterium that produces the compound endogenously.

Characterisation of the Cyanate Inhibited State of Cytochrome c Oxidase

Heme-copper oxygen reductases are terminal respiratory enzymes, catalyzing the reduction of dioxygen to water and the translocation of protons across the membrane. Oxygen consumption is inhibited by various substances. Here we tested the relatively unknown inhibition of cytochrome c oxidase (CcO) with isocyanate. In contrast to other more common inhibitors like cyanide, inhibition with cyanate was accompanied with the rise of a metal to ligand charge transfer (MLCT) band around 638 nm. Increasing the cyanate concentration furthermore caused selective reduction of heme a. The presence of the CT band allowed for the first time to directly monitor the nature of the ligand via surface-enhanced resonance Raman (SERR) spectroscopy. Analysis of isotope sensitive SERR spectra in comparison with Density Functional Theory (DFT) calculations identified not only the cyanate monomer as an inhibiting ligand but suggested also presence of an uretdion ligand formed upon dimerization of two cyanate ions. It is therefore proposed that under high cyanate concentrations the catalytic site of CcO promotes cyanate dimerization. The two excess electrons that are supplied from the uretdion ligand lead to the observed physiologically inverse electron transfer from heme a₃ to heme a.

Terpene-containing PEGylated liposomes as transdermal carriers of a hydrophilic compound

We investigated the effect of PEGylated liposomes (PLs) containing a terpene on the penetration of a hydrophilic compound through porcine skin. PLs composed of N-(carbonyl-methoxypolyethyleneglycol-2000)-1,2-distearoyl-sn-glycero-3-phosphoethanolamine (PEG2000-DSPE), the sodium salt of PEG2000-DSPE, phosphatidylcholine (PC), cholesterol (Chol), Tween 20, and d-limonene were prepared as carriers for fluorescein sodium (NaFI). The physicochemical characteristics of PLs and their effects on in vitro skin penetration were evaluated. Tape stripping was used to evaluate NaFI deposition in skin layers, and confocal laser scanning microscopy (CLSM) was used to investigate the depth of skin penetration and the pathways used by NaFI-loaded vesicles. PLs containing d-limonene were smaller and conferred higher entrapment efficiency and skin penetration on NaFI than did PLs and conventional liposomes (CLs). The deposition of NaFI from PLs with d-limonene was greater in epidermis and dermis (6.10±1.74 µg) than stratum corneum (2.06±0.47 µg). CLSM images revealed that NaFI penetrated into the deepest skin layer with maximum fluorescence intensity. NaFI penetrated deeper (180 µm) in follicular than nonfollicular regions (145 µm), suggesting a transfollicular pathway predominates in skin penetration by NaFI-loaded PLs. In conclusion, grafting PEG onto ultra-deformable liposomes may enhance transdermal NaFI delivery and may be used as a carrier to prolong liposome circulation time.

The inhibition of mitochondrial cytochrome oxidase by the gases carbon monoxide, nitric oxide, hydrogen cyanide and hydrogen sulfide: chemical mechanism and physiological significance

The four gases, nitric oxide (NO), carbon monoxide (CO), hydrogen sulfide (H(2)S) and hydrogen cyanide (HCN) all readily inhibit oxygen consumption by mitochondrial cytochrome oxidase. This inhibition is responsible for much of their toxicity when they are applied externally to the body. However, recently these gases have all been implicated, to greater or lesser extents, in normal cellular signalling events. In this review we analyse the chemistry of this inhibition, comparing and contrasting mechanism and discussing physiological consequences. The inhibition by NO and CO is dependent on oxygen concentration, but that of HCN and H(2)S is not. NO and H(2)S are readily metabolised by oxidative processes within cytochrome oxidase. In these cases the enzyme may act as a physiological detoxifier of these gases. CO oxidation is much slower and unlikely to be as physiologically important. The evidence for normal physiological levels of these gases interacting with cytochrome oxidase is equivocal, in part because there is little robust data about their steady state concentrations. A reasonable case can be made for NO, and perhaps CO and H(2)S, inhibiting cytochrome oxidase in vivo, but endogenous levels of HCN seem unlikely to be high enough.

Characterization of Novel Mixed Polyethyleneglycol Modified Liposomes

In our previous paper, mixed polyetheleneglycol (PEG) modification of liposomes by a mixture of 1-monomethoxypolyethyleneglycol-2,3-distearoylglycerol (PEG-DSG) with short polyoxyethylene chain and PEG-DSG with long one was shown to increase fixed aqueous layer thickness (FALT) around the liposomal membrane, and this was useful in vivo. In this study, we investigated the characterization of mixed PEG modification of liposomes with different anchors (PEG2000-DSG and PEG2000-cholesterol (CHO)). When the liposomes was modified by a mixture of PEG2000-DSG and PEG2000-CHO, FALT was increased compared to that of each single PEG-lipids modification and the most suitable mix modification (PEG2000-DSG:PEG2000-CHO = 3:1) showed a maximum FALT. This phenomenon was speculated to be based on the difference in the insertion state of the PEG anchor unit in the liposomal membrane. PEG-CHO-modified liposomes (single or mixed PEG-modified liposomes) were easily incorporated into the liposomal membranes compared with that of single PEG-DSG-modified liposomes. Namely, it was considered that the cholesterol anchor as a single chain was able to be easily introduced, compared with the DSG anchor as two chains, and induced some interaction with both PEG modification. In conclusion, it is expected that novel PEG-modified liposomes with PEG2000-DSG and PEG2000-CHO (3:1) had superior physicochemical properties.

Influence of Reduction of Hemeaand CuAon the Oxidized Catalytic Center of CytochromecOxidase: Insight from Organic Solvents†

Purified bovine heart cytochrome c oxidase (CcO) has been extracted from aqueous solution into hexane in the presence of phospholipids and calcium ions. In extracts, CcO is in the so-called "slow" form and probably situated in reverse micelles. At low water:phospholipid molar ratios, electron transfer from reduced heme a and Cu(A) to the catalytic center is inhibited and both heme a3 and Cu(B) remain in the oxidized state. The rate of binding of cyanide to heme a3 in this oxidized catalytic center is, however, dependent on the redox state of heme a and Cu(A). When heme a and Cu(A) are reduced, the rate is increased 20-fold compared to the rate when these two centers are oxidized. The enhanced rate of binding of cyanide to heme a3 is explained by the destabilization of an intrinsic ligand, located at the catalytic site, that is triggered by the reduction of heme a and Cu(A).

Carboxyfluorescein Leakage from Poly(ethylene glycol)-Grafted Liposomes Induced by the Interaction with Serum

The effects of fetal bovine serum (FBS) on carboxyfluorescein (CF) leakage from poly(ethylene glycol)-grafted liposomes (PEG-liposomes) were investigated. PEG-liposomes were prepared from dipalmitoylphosphatidylcholine (DPPC) and distearoyl-N-monomethoxy poly(ethylene glycol)-succinyl-phosphatidylethanolamines (DSPE-PEG) having PEG molecular weights of 1000, 2000, 3000 and 5000. The presence of FBS dramatically increased CF leakage from liposomes near the gel-liquid crystalline phase transition temperature, but had little effect at lower and higher temperatures. The CF leakage from PEG-liposomes whose molecular weight in PEG units was above 2000 was suppressed compared with that of liposomes without PEG. And, there was hardly any difference in the effect of the PEG molecular weight of the PEG-lipids on CF leakage from PEG-liposomes with FBS when PEG-lipids with a molecular weight in PEG units above 2000 were used. On the other hand, the leakage of CF from liposomes containing 0.145 mol fractions of DSPE-PEG1000 was larger than that of liposomes without PEG. Furthermore, the effects of FBS on the cooperative units of lipid molecules during the gel-liquid crystalline phase transition of liposomes were examined. However, the cooperative units of liposomes with FBS had little change compared with that of liposomes without FBS.

Cyanide stimulated dissociation of chloride from the catalytic center of oxidized cytochrome c oxidase

A comparison of bovine cytochrome c oxidase isolated in the presence and the absence of chloride salts reveals that only enzyme isolated in the presence of chloride salts is a mixture of a complex of oxidized enzyme with chloride (CcO.Cl) and chloride-free enzyme (CcO). Using a spectrophotometric method for chloride determination, it was shown that CcO.Cl contains one chloride ion that is released into the medium by a single turnover or by cyanide binding. Chloride is bound slowly within the heme a(3)-Cu(B) binuclear center of oxidized enzyme in a manner similar to the binding of azide. The pH dependence of the dissociation constant for the formation of the CcO.Cl complex reveals that chloride binding proceeds with the uptake of one proton. With both forms of the enzyme the dependence of the rate of reaction for cyanide binding upon cyanide concentration asymptotes a limiting value indicating the existence of an intermediate. With CcO.Cl this limiting rate is 10(3) higher than the rate of the spontaneous dissociation of chloride from the binuclear center and we propose that the initial step is the coordination of cyanide to Cu(B) and in this intermediate state the rate of dissociation of chloride is substantially enhanced.

The binding of cyanide to cytochrome d in intact cells, spheroplasts, membrane fragments and solubilized enzyme from Salmonella typhimurium

This investigation focused on the kinetics of cyanide binding to oxidized and reduced cytochrome d in Salmonella typhimurium intact cells, spheroplasts, membrane fragments and solubilized enzyme, and on the effect of pH on this binding. Cyanide bound to the oxidized form of cytochrome d under all experimental conditions, inducing a trough at 649 nm in the oxidized-cyanide-minus-oxidized difference absorption spectra. V(max) of cyanide binding to oxidized cytochrome d at pH 7.0 was 14.0+/-2.0 pmol/min/mg protein (prot.) in intact cells, 37.0+/-3.5 pmol/min/mg prot. in spheroplasts, 125.0+/-6.0 pmol/min/mg prot. in membrane fragments, and 538.0+/-8.5 pmol/min/mg prot. in solubilized cytochrome d. The pseudo-first order rate constants were 0.004 s(-1) for intact cells, 0.005 s(-1) for spheroplasts, 0.007 s(-1) for membrane fragments and 0.025 s(-1) for the solubilized enzyme. The V(max) value was highest at pH 7.0 for intact cells and solubilized cytochrome d and at pH 8.0 for both spheroplasts and membrane fragments. The K(s) of binding at pH 7.0 was around 4 mM in intact cells, spheroplasts and membrane fragments, but was 10.5 mM in solubilized cytochrome d. This difference between the K(s) values suggested a change in conformation, upon solubilization, leading to a decrease in the affinity of cyanide for the solubilized enzyme. The K(s) value was nearly the same at all pH investigated (pH 5-10). Cyanide was found to also bind to the reduced form of cytochrome d in membrane fragments (K(s)=18+/-3 mM, V(max)=377+/-28 pmol/min/mg prot. at pH 7) and the solubilized enzyme (K(s)=18+/-1.2 mM, V(max)=649+/-45 pmol/min/mg prot. at pH 7) with a lower affinity of cyanide for the reduced cytochrome d than for the oxidized enzyme. Pseudo-first order rate constants were 0.025 s(-1) and 0.042 s(-1) respectively for membrane fragments and solubilized enzyme. The value of V(max) for cyanide binding to the reduced cytochrome d, whether membrane-bound or solubilized, increased slightly with pH (for pH 6-10) while the K(s) value dropped significantly with increasing pH. The pH dependence observed here might be interpretable as a possible role for conformational transition associated with energy transduction. Finally, this investigation pointed to the influence of the microenvironment of a protein within the cell on its reactivity.

Spectral and cyanide binding properties of the cytochrome aa3 (600 nm) complex from Bacillus subtilis

The cytochrome aa3 (600 nm) complex, or menaquinol oxidase, from Bacillus subtilis is a member of the cytochrome oxidase superfamily of respiratory membrane protein complexes. We have characterized some spectral properties of this enzyme and its reaction with cyanide. The magnetic circular dichroism (MCD) spectrum of the oxidized enzyme has a single band at 1560 nm in the near-infrared region assigned to bis-histidine-ligated, low-spin ferricytochrome a. The other heme, cytochrome a3, is presumably high-spin in the oxidized enzyme, as isolated. The absence of a trough in the MCD spectrum at 790 nm, observed previously with mammalian cytochrome c oxidase and assigned to CuA (Greenwood et al., Biochem. J. 215, 303-316, 1983), is consistent with the absence of this center from the menaquinol oxidase. When the heme ligand cyanide is added to oxidized menaquinol oxidase, a new MCD band appears at 2010 nm, while the band at 1560 nm is unperturbed. The new band is assigned to low-spin ferricytochrome a3 bound with cyanide. The long-wavelength position of this cyanide-induced band is proposed to arise from the close interaction of cytochrome a3 with the copper atom, CuB. The kinetics of cyanide binding to oxidized cytochrome aa3(600 nm) reveal a spectrally simple, yet kinetically complex process. The reaction is biphasic with second-order rate constants of 45 and 0.61 M-1s-1 at 1 mM KCN, with each phase constituting about 50% of the overall reaction. When the enzyme is subjected to a cycle of anaerobic reduction and air oxidation, the subsequent reaction with cyanide occurs in a single phase at the faster rate. This behavior is ascribed to different conformations of the binuclear center exhibiting different reactivities with cyanide, and is in keeping with that previously established for the structurally more complex mitochondrial cytochrome c oxidase. However, the electronic spectral characteristics of some of the species involved in these reactions are different in the present bacterial case from those of reported eukaryotic systems.

The dinuclear center of cytochrome bo3 from Escherichia coli

For the study of the dinuclear center of heme-copper oxidases cytochrome bo3 from Escherichia coli offers several advantages over the extensively characterized bovine cytochrome c oxidase. The availability of strains with enhanced levels of expression allows purification of the significant amounts of enzyme required for detailed spectroscopic studies. Cytochrome bo3 is readily prepared as the fast form, with a homogeneous dinuclear center which gives rise to characteristic broad EPR signals not seen in CcO. The absence of CuA and the incorporation of protohemes allows for a detailed interpretation of the MCD spectra arising from the dinuclear center heme o3. Careful analysis allows us to distinguish between small molecules that bind to heme o3, those which are ligands of CuB, and those which react to yield higher oxidation states of heme o3. Here we review results from our studies of the reactions of fast cytochrome bo3 with formate, fluoride, chloride, azide, cyanide, NO, and H2O2.

Hydrogen peroxide is not released following reaction of cyanide with several catalytically important derivatives of cytochrome c oxidase

We have looked for the production of hydrogen peroxide following reaction of oxidized cytochrome c oxidase and two oxy derivatives (compounds P and F) with cyanide. In each case the final product was the cyanide adduct of cytochrome c oxidase. In no case release of hydrogen peroxide was detected, as gauged by the scopoletin plus horse radish peroxidase assay. The simplest conclusion is that none of these forms of the enzyme contains intact hydrogen peroxide.

EPR spectroscopy of Escherichia coli cytochrome bo which lacks CuB

The spectroscopic and ligand-binding properties of a copper-deficient cytochrome bo3, a member of the haem-copper superfamily of terminal oxidases, are reported and contrasted with those of the native enzyme. The enzyme lacks the copper atom (CuB) which is normally an integral part of the catalytic site. The consequences of loss of the CuB are the loss of antiferromagnetic coupling to the high-spin haem and an inability to form any of the integer-spin derivatives of the enzyme. Low-spin compounds of the normally high-spin haem are still formed with appropriate ligands, although these are modified.

Kinetics and mechanism for the binding of HCN to cytochrome c oxidase

The kinetics of cyanide binding to cytochrome c oxidase were systematically studied as a function of [HCN], [oxidase], pH, ionic strength, temperature, type and concentration of solubilizing detergent, and monomer-dimer content of oxidase. On the basis of these results a minimum reaction mechanism is proposed in which the spectrally visible rapid and slow cyanide binding reactions are two consecutive first-order reactions, not parallel reactions with different conformers of cytochrome c oxidase. The fast reaction (k'obs) follows saturation type kinetics to form an HCN complex that subsequently undergoes a slow reaction (k'obs). The fast k'obs reaction is independent of ionic strength but is strongly dependent upon pH. Two pK values were evaluated from the bell-shaped rate versus pH profile; one is due to an ionizable group on the protein (pKa = 7.45), while the other is that of HCN (pKHCN = 9.15). Therefore, oxidase is reactive toward HCN only when the group on the protein is unprotonated. The slow k'obs reaction is not a reaction of oxidase with either CN- or HCN; in fact, the product formed by the fast k'obs reaction, the oxidase-HCN complex, still undergoes the slow k" process even if all of the excess KCN is removed. The apparent rate constant of the slower phase (k"obs) is independent of all the variations done in this study, and it probably corresponds to either a slow conformational change in the protein or a change in ligand coordination at one of the metal centers after HCN binds to the bimetallic center of oxidase. Based upon the bell-shaped pH dependence of the fast phase and the pH independence of the slow phase, the mechanism also predicts that a single conformer of cytochrome c oxidase can exhibit either monophasic or biphasic cyanide binding kinetics depending upon the pH. At either very low or very high pH, the two rates become comparable in magnitude, which makes the reaction appear to be monophasic even though both reactions still occur. The amount of monomeric or dimeric oxidase only slightly affects the magnitude of k'obs and k"obs values, and both processes are clearly present in both types of oxidase.

Electron transfer and proton pumping in cytochrome oxidase

This article presents an outlook on the structure and function of terminal oxidases, the respiratory enzymes which catalyze the reduction of dioxygen to water in aerobic organisms. The structure of the redox active metals, their interactions with the protein matrix, and their role in electron transfer ligand binding and proton pumping are briefly reviewed.

Electron transfer and ligand binding in terminal oxidases. Impact of recent structural information

A consensus structure for the active site of terminal oxidases has been recently proposed by Hosler et al. [(1993) J. Bioenerg. Biomem. 25, 121-135]. We exploit the novel structural information to propose a hypothesis for the large difference in the rate of internal electron transfer found when experiments are started either with the reduced or with the oxidized enzyme. This rationale also allows us to discuss the oxidation state of the prevailing oxygen reacting species with reference to the concentration of the two substrates (oxygen and cytochrome c) and to the structural state of the oxidase.

A comparison of three preparations of cytochrome c oxidase. Optical absorbance spectra, EPR spectra and reaction towards ligands

Three preparations of cytochrome c oxidase, the preparation as traditionally prepared in our laboratory as described by Van Buuren (1992; PhD Thesis, University of Amsterdam), a preparation according to Volpe and Caughey (Biochem. Biophys. Res. Commun. 61 (1974) 502-509) and a preparation of 'fast' cytochrome c oxidase (Brandt, U., Schägger, H. and Von Jagow, G. (1989) Eur. J. Biochem. 182, 705-711), are compared in their reaction with cyanide and carbon monoxide. The reaction with cyanide is nearly as fast for the Van Buuren preparation as for the 'fast' preparation, but much slower for the Volpe-Caughey preparation. Mixed-valence cytochrome c oxidase (cytochrome a3 and CuB reduced with carbon monoxide bound and cytochrome a and CuA oxidized) is prepared by anaerobic incubation with carbon monoxide. With the Van Buuren preparation complete formation of the species takes 4 h, whereas with the Volpe-Caughey preparation it takes 20 h. Longer incubation under CO results in partial reduction of cytochrome a and CuA. With the 'fast' preparation mixed-valence cytochrome c oxidase is formed after more than one day of incubation with CO, but it is stable for at least 3 days. The presence of oxidized cytochrome c did enhance the reactivity towards cyanide and towards carbon monoxide in cytochrome c oxidase of all three preparations. Furthermore, optical and EPR spectra of the preparations of cytochrome c oxidase are compared. The Volpe-Caughey preparation has an intense g' = 12 EPR-signal, the Van Buuren preparation has hardly any g' = 12 signal and the 'fast' preparation has no g' = 12 signal. In the 'fast' preparation the low-spin heme signal is shifted (from g = 3.00 to g = 2.97). The absorbance spectra of the three preparations in the Soret region are similar with a maximum at 424 nm. Only the 'fast' preparation as isolated was completely oxidized, whereas the other preparations were partially reduced. It was concluded that differences in the reaction of cytochrome c oxidase with ligands are determined by the internal or external ligand bound to the cytochrome a3-CuB couple.

Light-induced structural changes in cytochrome c oxidase: implication for the mechanism of electron and proton gating

We have investigated electrogenic events and absorbance changes following pulsed illumination of partly reduced cytochrome c oxidase in the absence of dioxygen and carbon monoxide (Hallén et al. (1993) FEBS Lett. 318, 134-138). In both types of experiment similar kinetics were observed; a rapid (tau < 0.5 micros) change was followed by relaxations with time constants of approx. 7 micros and 80 micros. Both the time constant and the activation energy of the 80 micros component were, within the experimental error, the same as those of one of the steps in the reduction of dioxygen by reduced cytochrome c oxidase. The absorbance changes showed a rapid haem reduction, followed by reoxidation. They were affected by CN(-) and N(-)3, ligands which bind in the binuclear centre of cytochrome c oxidase; the absorbance changes were quenched by CN(-) and in the presence of N(-)3, the amplitude of the 7 micros component increased whereas that of the 80 micros decreased. Based on these findings, a model is proposed which involves electron transfer from Cu(+)B to Fe(3+)A3, as a response to structural changes upon pulsed illumination. The same structural changes are also suggested to take place in the oxygen reduction. These changes may play an important role in the gating of electrons as well as protons, an obligatory feature of a redox-linked proton pump.

The effect of antibodies to subunit V of cytochrome oxidase on cyanide inhibition of electron transfer

Binding of antibodies raised against subunit V of mammalian cytochrome oxidase to the intact membranous enzyme is redox-sensitive, suggesting the existence of 'open' and 'closed' protein conformers (Freedman, J.A., Cooper, C.E., Leece, B., Nicholls, P. and Chan, S.H.P. (1988) Biochem. Cell Biol. 66, 1210-1217). Similar open and closed states for the oxygen-reacting site have been proposed to explain cyanide binding kinetics (Jensen, P., Wilson, M.T., Aasa, R. and Malmström, B.G. (1984) Biochem. J. 224, 829-837). We therefore examined cyanide inhibition of oxidase activity polarographically and spectrophotometrically using soluble oxidase preincubated with and without anti-subunit V or non-immune rabbit gamma-globulin. The subunit-specific antibody decreased the cyanide 'on' rate and essentially eliminated the rapid phase of cyanide binding. We conclude that (i), bound antibody blocks HCN binding; (ii), antibody and HCN probably bind to the same conformation of the oxidase and (iii), the 'open'-'closed' conformation change that modulates binding of HCN may be similar to that which modulates antibody binding. The results are consistent with some reciprocating models of electron transfer and energy transduction by the oxidase (cf., Wikström, M.K.F., Krab, K. and Saraste, M. (1981) Cytochrome Oxidase: A Synthesis).

Ligand-binding properties and heterogeneity of cytochrome bo from Escherichia coli

Cyanide and formate induce spectral changes in E. coli cytochrome bo which are similar to those induced in bovine heart cytochrome-c oxidase (cytochrome aa3). Cyanide induces a red shift of 6 nm in the Soret band, whereas formate induces a blue shift of 2 nm. Cytochrome bo as purified shows multiphasic cyanide-binding kinetics. At least three phases can be seen with rate constants of 16, 1 and 0.1 M-1 s-1, respectively, at pH 7 and 20 degrees C. The enzyme after redox cycling ('pulsing') or in situ in E. coli membranes shows essentially monophasic binding with a rate constant of 15 M-1 s-1. Further evidence of heterogeneity in the enzyme as prepared comes from formate binding, which also shows at least three phases (rate constants of 1.4, 0.2 and 0.01 M-1 s-1, respectively, at pH 5 and 20 degrees C). The fast phase of cyanide binding is eliminated in less than 2 min by incubation with 40 mM formate, but the intermediate phase is unaffected by incubation for 3.5 h with 40 mM formate. Thus, the subpopulation that causes the fast phase of cyanide binding also causes the fast phase of formate binding. Formate-ligated cytochrome bo has similar cyanide-binding kinetics to the subpopulation that causes the slow phase of cyanide binding in cytochrome bo as prepared. It appears, from all this, that the subpopulations responsible for the fast and slow phase of cyanide binding are analogous to the 'fast' and 'slow' forms, respectively, of cytochrome aa3.(ABSTRACT TRUNCATED AT 250 WORDS)

A new spectral intermediate in cyanide binding with the oxidized cytochrome c oxidase

Reaction of cyanide with oxidized cytochrome c oxidase at a low concentration of the ligand and pH > 8 reveals an initial phase, not reported earlier, associated with a small blue shift of the absorption spectrum, which is followed by a conventional red shift of the heme alpha(3+)3. The initial blue shift resembles the spectral changes induced under the same conditions by low concentrations of azide and it is not observed in the presence of 0.3 mM azide. It is suggested that, similarly to NO, cyanide and HN3 cannot only bind to heme alpha 3 but to Cu(2+)B as well, perturbing the spectrum of alpha(3+)3 indirectly. A rapid binding to Cu(2+)B could provide the long-sought intermediate in the cyanide reaction with heme alpha(3+)3, the existence of which is implied by the Michaelis-Menten type kinetics of the latter process.

The osmochemistry of electron-transfer complexes

Detailed molecular mechanisms of electron transfer-driven translocation of ions and of the generation of electric fields across biological membranes are beginning to emerge. The ideas inherent in the early formulations of the chemiosmotic hypothesis have provided the framework for this understanding and have also been seminal in promoting many of the experimental approaches which have been successfully used. This article is an attempt to review present understanding of the structures and mechanisms of several osmoenzymes of central importance and to identify and define the underlying features which might be of general relevance to the study of chemiosmotic devices.