High vectorial proton stoichiometry by cytochrome c oxidase from the thermophilic bacterium PS3 reconstituted in liposomes

The stoichiometry of vectorial H+ ejection, coupled to ferrocytochrome c oxidation by a three-subunit bacterial cytochrome c oxidase (EC 1.9.3.1) from the thermophilic bacterium PS3, was measured. Three methods of measuring the H+/e- ratio were applied to proteoliposomes containing a relatively small amount of PS3 cytochrome oxidase, which showed a relatively low oxidation rate and a very low H+ leakage, as follows: (a) simultaneous measurements of H+ ejection and cytochrome c oxidation upon addition of a yeast ferrocytochrome c pulse, which enable us to calculate the H+/e- ratio as H+ ejected per cytochrome c oxidized; (b) computer simulations to find out the fit for the pH meter trace by changing the H+/e- ratio and the velocity constant of leakage; and (c) two successive measurements of initial rates of H+ movement in the absence and presence of carbonyl cyanide p-trifluoromethoxyphenylhydrazone. The H+/e- ratios obtained were 1.39, the 10-s value after ferrocytochrome c addition in (a), 1.35 in (b), and 1.33 in (c). This high H+/e- stoichiometry observed, exceeding 1 and as high as 1.4, is discussed with respect to the controversy of the H+/e- ratio at the cytochrome oxidase site.     

Characteristics of the redox-linked proton ejection in beef-heart cytochrome c oxidase reconstituted in liposomes

In this paper a study is presented of the characteristics of redox-linked proton ejection exhibited by isolated beef-heart cytochrome c oxidase incorporated in asolectin vesicles. The enzyme was 90% oriented 'right-side out' as in the mitochondrial membrane. The effects on the H+/e- stoichiometry of the modalities of activation of electron flow, the pH of the medium and its ionic composition were investigated. The results obtained show that, whilst ferrocytochrome c pulses of the aerobic oxidase vesicles at neutral pH and in the presence of saturating concentrations of valinomycin and K+ to ensure charge compensation produced H+/e- ratios around 1 (as has been shown previously), oxygen pulses of reduced anaerobic vesicles supplemented with cytochrome c, gave H+/e- ratios around 0.3. The H+/e- ratios exhibited, with both reductant and oxidant pulses, a marked pH dependence. Maximum values were observed at pH 7.0-7.7, which decreased to negligible values at acidic pH with apparent pKa of 6.7-6.3. Mg2+ and Ca2+ caused a marked depression of the H+/e- ratio, which in the presence of these cations and after a few ferrocytochrome pulses, became negligible. Analysis of cytochrome c oxidation showed that the modalities of activation of electron flow and divalent cations exerted profound effects on the kinetics of cytochrome c oxidation by oxidase vesicles. The observations presented seem to provide interesting clues for the nature and mechanism of redox-linked proton ejection in reconstituted cytochrome c oxidase.     

Effect of heat treatment on oxidase activity and proton-pumping capability of proteoliposome-incorporated beef heart cytochrome aa3

By incubating beef heart cytochrome c oxidase at 43-45 degrees C, selective inactivation of the H+-pumping function is possible without affecting cytochrome c oxidase activity; proteoliposomes reconstituted with heated enzyme (43.5 degrees C for 60 min at pH 7.0) showed an apparent H+/e- ratio of only 0.3 and a turnover with cytochrome c plus ferrocyanide as substrate of 20 s-1, while those with the intact enzyme showed an apparent H+/e- ratio somewhat greater than 1.0 and a turnover of 19 s-1. This decrease in the H+/e- ratio could not be attributed to a stimulation of H+ permeability upon heating, since the respiratory control ratio and the magnitude of membrane potential formation remained almost the same in the two cases. A pH-dependent Em (midpoint redox potential) change of cytochrome a in the presence of cyanide was still observed after the heat treatment. Heating induced a small spectral shift in the Soret region of the oxidized (resting) enzyme; the peak of the heated enzyme was at 421 nm, while that of the intact enzyme was at 419 nm. The spectral shift obtained by pulsing the enzyme with oxygen under turnover conditions is also altered.     

H+/e- stoichiometry of mitochondrial cytochrome complexes reconstituted in liposomes. Rate-dependent changes of the stoichiometry in the cytochrome c oxidase vesicles.

The H+/e- stoichiometry of protonmotive cytochrome c oxidase, isolated from bovine heart mitochondria and reconstituted in liposomes, has been determined by making use of direct spectrophotometric measurements of the initial rates of e- flow and H+ translocation. It is shown that the ----H+/e- ratio for redox-linked proton ejection by the oxidase varies from around 0 to a maximum of 1 as a function of the rate of overall electron flow in the complex.     

The cytochrome chain of mitochondria exhibits variable H+/e- stoichiometry.

A study is presented of the ----H+/e- stoichiometry for H+ pumping by the cytochrome chain in isolated rat liver mitochondria under level-flow and steady-state conditions. It is shown that the ----H+/e- stoichiometry for the cytochrome chain varies under the influence of the flow rate and transmembrane delta microH+. The rate-dependence is shown to be associated with cytochrome c oxidase, whose ----H+/e- ratio varies from 0 to 1, whilst the ----H+/e- ratio for the span covered by cytochrome c reductase is invariably 2.     

New insights on the cytochrome c oxidase proton pump.

Cytochrome c oxidase vesicles were used to show that, under appropriate experimental conditions: (1) no net deprotonation of the vesicular membrane or of the incorporated enzyme occurs during the oxidation of ferrocytochrome c; (2) the pH equilibration kinetics of a respiration-induced pH gradient across the bilayer are a simple function of the ohmic proton-conductance properties of the membrane; (3) a fairly constant stoichiometry (0.8-0.7) of the numbers of protons pumped per molecule of ferrocytochrome c oxidized, i.e. the H+/e- ratio, over a wide range of dioxygen molecules reduced (1-12) is observed.     

Upper and lower limits of the proton stoichiometry of cytochrome c oxidation in rat liver mitoplasts

The stoichiometry of vectorial H+ translocation coupled to oxidation of added ferrocytochrome c by O2 via cytochrome-c oxidase of rat liver mitoplasts was determined employing a fast-responding O2 electrode. Electron flow was initiated by addition of either ferrocytochrome c or O2. When the rates were extrapolated to level flow, the H+/O ratios in both cases were less than but closely approached 4; the directly observed H+/O ratios significantly exceeded 3.0. The mechanistic H+/O ratio was then more closely fixed by a kinetic approach that eliminates the necessity for measuring energy leaks and is independent of any particular model of the mechanism of energy transduction. From two sets of kinetic measurements, an overestimate and an underestimate and thus the upper and lower limits of the mechanistic H+/O ratio could be obtained. In the first set, the utilization of respiratory energy was systematically varied through changes in the concentrations of valinomycin or K+. From the slope of a plot of the initial rates of H+ ejection (JH) and O2 uptake (JO) obtained in such experiments, the upper limit of the H+/O ratio was in the range 4.12-4.19. In the second set of measurements, the rate of respiratory energy production was varied by inhibiting electron transport. From the slope of a plot of JH versus JO, the lower limit of the H+/O ratio, equivalent to that at level flow, was in the range 3.83-3.96. These data fix the mechanistic H+/O ratio for the cytochrome oxidase reaction of mitoplasts at 4.0, thus confirming our earlier measurements (Reynafarje, B., Alexandre, A., Davies, P., and Lehninger, A. L. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 7218-7222). Possible reasons for discrepancies in published reports on the H+/O ratio of cytochrome oxidase in various mitochondrial and reconstituted systems are discussed.     

Kinetic studies on the reaction between cytochrome c oxidase and ferrocytochrome c.

In stopped-flow experiments in which oxidized cytochrome c oxidase was mixed with ferrocytochrome c in the presence of a range of oxygen concentrations and in the absence and presence of cyanide, a fast phase, reflecting a rapid approach to an equilibrium, was observed. Within this phase, one or two molecules of ferrocytochrome were oxidized per haem group of cytochrome a, depending on the concentration of ferrocytochrome c used. The reasons for this are discussed in terms of a mechanism in which all electrons enter through cytochrome a, which, in turn, is in rapid equilibrium with a second site, identified with 'visible' copper (830 nm-absorbing) Cud (Beinert et al., 1971). The value of the bimolecular rate constant for the reaction between cytochromes c2+ and a3+ was between 10(6) and 10(7) M(-1)-S(-1); some variability from preparation to preparation was observed. At high ferrocytochrome c concentrations, the initial reaction of cytochrome c2+ with cytochrome a3+ could be isolated from the reaction involving the 'visible' copper and the stoicheiometry was found to approach one molecule of cytochrome c2+ oxidized for each molecule of cytochrome a3+ reduced. At low ferrocytochrome c concentrations, however, both sites (i.e. cytochrome a and Cud) were reduced simultaneously and the stoicheiometry of the initial reaction was closer to two molecules of cytochrome c2+ oxidized per molecule of cytochrome a reduced. The bleaching of the 830 nm band lagged behind or was simultaneous with the formation of the 605 nm band and does not depend on the cytochrome c concentration, whereas the extinction at the steady-state does. The time-course of the return of the 830 nm-absorbing species is much faster than the bleaching of the 605 nm-absorbing component, and parallels that of the turnover phase of cytochrome c2+ oxidation. Additions of cyanide to the oxidase preparations had no effect on the observed stoicheiometry or kinetics of the reduction of cytochrome a and 'visible' copper, but inhibited electron transfer to the other two sites, cytochrome a3 and the undetectable copper, Cuu.     

Palmitate decreases proton pumping of liver-type cytochrome c oxidase.

The H+/e- stoichiometry of reconstituted cytochrome c oxidase from bovine kidney, containing subunit VIaL (liver type), is 0.5 under standard conditions but 1.0 on addition of 1% cardiolipin to the lipid mixture (asolectin). Low concentrations of palmitate (half-maximal effect at 0.5 microm), but not laurate, myristate, stearate, oleate, 1-hexadecanol, palmitoyl glycerol and palmitoyl CoA, decreased the H+/e- ratio in the presence of cardiolipin from 1.0 to 0.5, accompanied by an increase of coupled, but not of uncoupled respiration of proteoliposomes. Cardiolipin and palmitate did not influence the H+/e- stoichiometry and respiration of reconstituted cytochrome c oxidase from bovine heart, containing subunit VIaH (heart-type). The H+/e- stoichiometry of the heart enzyme, however, is decreased from 1.0 to 0.5 by 5 mm intraliposomal ATP (instead of 5 mm ADP). It is assumed that palmitate binds to subunit VIaL. The partial uncoupling of proton pumping in cytochrome c oxidase is suggested to participate in mammalian thermogenesis.     

Phospholipid vesicles containing bovine heart mitochondrial cytochrome c oxidase and subunit III-deficient enzyme: analysis of respiratory control and proton translocating activities

Phospholipid vesicles containing bovine heart mitochondrial cytochrome c oxidase (COV) or subunit III (Mr 29884)-deficient enzyme (COV-III) were characterized for electron transfer and proton translocating activities in order to investigate the relationship between the respiratory control ratio (RCR) and the apparent proton translocated to electron transferred stoichiometry (H+/e- ratio) in these preparations. We did not observe a quantitative correlation between the RCR value and the H+/e- ratio in the preparations. Significant deviation between these two parameters was observed in COV-III and also in COV. However, a new parameter, RCRval, did show a linear relationship with the H+/e- ratio of each preparation. Subunit III (SIII)-deficient cytochrome c oxidase isolated by either native gel electrophoresis or chymotrypsin treatment and incorporated into COV-III exhibited H+/e- ratios of 0.34 +/- 0.10, compared to 0.63 +/- 0.09 for COV, emphasizing that the 50% decrease of proton translocating activity is independent of the method of removal of SIII from the enzyme. COV and COV-III also showed similar rates of alkalinization of the extravesicular media after the initial proton translocation reaction (0.07-0.09 neq OH-/s), suggesting that these two preparations had similar endogenous proton permeabilities. In contrast, cytochrome c oxidase (COX) treated with Triton X-100 (3 mg/mg COX) and incorporated into phospholipid vesicles [COV (+TX)] exhibited slower rates of alkalinization (0.04 neq OH-/s), while having a H+/e- ratio similar to that of COV (0.66 +/- 0.10). The passive proton permeabilities of these preparations were tested by valinomycin-induced K+/H+ exchange activity. COV (+TX) and COV-III exhibited similar pseudo-first-order rate constants (10 peq OH-/s), while COV had a 20-fold higher rate constant. These results taken together suggest that the different preparations of COX-containing phospholipid vesicles have different biophysical properties. In addition, the decrease in proton-pumping activity observed in COV-III is due to removal of SIII from COX, suggesting that SIII may act either as a passive proton-conducting channel or as a regulator of COX conformation and/or functional activities.     

The mechanism of proton translocation by the cytochrome system of mitochondria. Characterization of proton-transfer reactions associated with oxidoreductions of terminal respiratory carriers.

A direct kinetic analysis is presented of rapid proton-releasing reactions at the outer or C-side of the membrane, in ox heart and rat liver mitochondria, associated with aerobic oxidation of reduced terminal respiratory carriers in the presence of antimycin. Valinomycin plus K+ enhances the rate of cytochrome c oxidation and the rate and extent of H+ release. In the presence of valinomycin the leads to H+/e- ratio, computed on the basis of total electron flow from respiratory carriers to oxygen, varies with pH, remaining always lower than 1, and is unaffected by N-ethylmaleimide. 2-Heptyl-4-hydroxyquinoline N-oxide and 5-(n-undecyl)-6-hydroxy-4,7-dioxobenzothiazole, at concentrations which inhibit in the presence of antimycin the oxygen-induced reduction of b cytochromes, cause also a marked depression of the H+ release associated with aerobic oxidation of terminal respiratory carriers. Aerobic oxidation of the cytochrome system in mitochondria and of isolated b-c1 complex and cytochrome c oxidase results in scalar proton release from ionizable groups (redox Bohr effects). In mitochondria and submitochondrial particles, about 70% of the oxidoreductions of the components of the cytochrome system are linked to scalar proton transfer by ionizable groups. In isolated b-c1 complex scalar proton transfer, resulting from redox Bohr effect, amounts to 0.9H+ per Fe-S protein (190 muT). In isolated cytochrome c oxidase, Bohr protons amount to 0.8 per haem a + a3. The results presented indicate that the H+ release from mitochondria during oxidation of terminal respiratory carriers derives from residual antimycin-insensitive electron flow in the quinone-cytochrome c span and from redox Bohr effects in the b-c1 complex and cytochrome c oxidase. There is no sign of proton pumping by cytochrome oxidase during its transition from the reduced to the active 'pulsed' state and the first one or two turnovers.     

Implications of the integrated rate law for the reactions of Paracoccus denitrificans nitrite reductase.

The integrated rate law for the reaction of the nitrite reductase of Paracoccus denitrificans, a cytochrome cd, has been established for turnover assays using donor ferrocytochromes c and either nitrite or molecular oxygen as the ultimate acceptor. The time course for the concentration of ferrocytochrome follows the law: formula: (see text), where S is the concentration of donor ferrocytochrome c, So is the initial concentration, t is time, and u1, u2, and u3 are empirical parameters that are constant for a given experiment but depend upon the initial substrate concentration. In particular, all the u1 increase with decreasing initial ferrocytochrome concentration. Saturation of reaction rates at high donor ferrocytochrome concentrations was not observed. The parameter u1 was proportional to the enzyme concentration while u2 and u3 were not. The form of the integrated rate law and the behavior of the u1 impose severe restrictions on possible kinetic schemes for the activity of the enzyme. Contemporary mechanisms that have been proposed for mitochondrial oxidase aa3 are examined and found to be inadequate to explain the reactivity of cytochrome cd. The simplest interpretations of the cytochrome cd data suggest that the enzyme does not bind the ferri and ferro forms of donor cytochromes c with equal affinity and that the enzyme is subject to inhibition by a product of reaction. Eucaryotic horse cytochrome c reacts with the Paracoccus cytochrome cd with 77% of the activity when Paracoccus cytochrome c550 is used as the electron donor.     

Cobalt-cytochrome c. I. Preparation, properties, and enzymic activity.

An improved procedure for the preparation of cobalt-cytochrome c has been developed. Various factors influencing the cobalt insertion process are discussed. The optical spectra of cobalt-cytochrome c suggest a six-coordinated species. The spectral shifts occurring with oxidation-reduction are compared with those observed for deoxy-cobaltohemoglobin and ferrocytochrome c and attributed to the effect of d(z2) electron on stereoelectronic interactions between the axial ligands and the porphyrin pi systems. Cobalt-cytochrome c has Em,7 = -140 +/- 20 mV as compared to an Em,7 of +250mV for ferrocytochrome c. An explanation for this negative Em,7 is offered. Cobaltocytochrome c is oxidized by cytochrome oxidase at about 45% of the rate for native cytochrome c. On the other hand cobalticytochrome c was not reduced by microsomal NADH or NADPH cytochrome c reductase nor by mitochondrial NADH or succinate cytochrome c reductase. It appears that the integrity of the reductase binding site is destroyed and the oxidase binding site has been modified by cobalt substitution.     

Protonmotive stoichiometry of rat liver cytochrome c oxidase: determination by a new rate/pulse method

The stoichoimetry of vectorial H+ ejection coupled to electron flow through the cytochrome c oxidase (EC 1.9.3.1) of rat liver mitochondria was determined by a new rate/pulse method. This is a modification of the oxygen-pulse method. Electron flow through the oxidase is initiated by adding oxygen to suspensions of anaerobic mitochondria at a known and constant rate. Cytochrome c oxidase was examined directly or in combination with cytochrome c reductase (ubiquinol:ferricytochrome c oxidoreductase). In both cases the----H0+/2e- ratio was found to be constant during the time-course of oxygen reduction, and thus independent of delta pH. The stoichiometries observed were consistent with mechanistic stoichiometries of 2 and 6 for cytochrome c oxidase alone and cytochrome c oxidase together with cytochrome c reductase, respectively. The stoichiometry of cytochrome c reductase alone was also examined, by using ferricyanide in place of oxygen. The results obtained were consistent with the accepted mechanistic stoichiometry of 4 for this enzyme.     

Complex-formation between cytochrome c and cytochrome c peroxidase. Kinetic studies

1. The kinetics of ferrocytochrome c peroxidation by yeast peroxidase are described. Kinetic differences between the older and more recent preparations of the enzyme most probably arise from differences in intrinsic turnover rates. 2. The time-courses of cytochrome c peroxidation by the enzyme follow essentially first-order kinetics in phosphate buffer. Deviations from first-order kinetics occur in acetate buffer, and are due to a higher enzymic turnover rate in this medium accompanied by a greater tendency to autocatalytic peroxidation of cytochrome c. 3. The kinetics of ferrocytochrome c peroxidation by yeast peroxidase are interpreted in terms of a mechanism postulating formation of reversible complexes between the peroxidase and both reduced and oxidized cytochrome c. Formation of these complexes is inhibited at high ionic strengths and by polycations. 4. Oxidized cytochrome c can act as a competitive inhibitor of ferrocytochrome c peroxidation by peroxidase. The K(i) for ferricytochrome c is approximately equal to the K(m) for ferrocytochrome c and thus probably accounts for the observed apparent first-order kinetics even at saturating concentrations of ferrocytochrome c. 5. The results are discussed in terms of a possible analogy between the oxidations of cytochrome c catalysed by yeast peroxidase and by mammalian cytochrome oxidase.     

The mechanism of transmembrane delta muH+ generation in mitochondria by cytochrome c oxidase.

In rat liver mitochondria treated with rotenone, N-ethylmaleimide or oligomycin the expected alkalinization caused by proton consumption for aerobic oxidation of ferrocyanide was delayed with respect to ferrocyanide oxidation, unless carbonyl cyanide p-trifluoromethoxyphenylhydrazone was present. 2. When valinomycin or valinomycin plus antimycin were also present, ferricyanide, produced by oxidation of ferrocyanide, was re-reduced by hydrogenated endogenous reductants. Under these circumstances the expected net proton consumption caused by ferrocyanide oxidation was preceded by transient acidification. It is shown that re-reduction of formed ferricyanide and proton release derive from rotenone- and antimycin-resistant oxidation of endogenous reductants through the proton-translocating segments of the respiratory chain on the substrate side of cytochrome c. The number of protons released per electron flowing to ferricyanide varied, depending on the experimental conditions, from 3.6 to 1.5. 3. The antimycin-insensitive re-reduction of ferricyanide and proton release from mitochondria were strongly depressed by 2-n-heptyl-4-hydroxyquinoline N-oxide. This shows that the ferricyanide formed accepts electrons passing through the protonmotive segments of the respiratory chain at the level of cytochrome c and/or redox components of the cytochrome b-c1 complex situated on the oxygen side of the antimycin-inhibition site. Dibromothymoquinone depressed and duroquinol enhanced, in the presence of antimycin, the proton-release process induced by ferrocyanide respiration. Both quinones enhanced the rate of scalar proton production associated with ferrocyanide respiration, but lowered the number of protons released per electron flowing to the ferricyanide formed. 4. Net proton consumption caused by aerobic oxidation of exogenous ferrocytochrome c by antimycin-supplemented bovine heart mitochondria was preceded by scalar proton release, which was included in the stoicheiometry of 1 proton consumed per mol of ferrocytochrome c oxidized. This scalar proton production was associated with transition of cytochrome c from the reduced to the oxidized form and not to electron flow along cytochrome c oxidase. 5. It is concluded that cytochrome c oxidase only mediates vectorial electron flow from cytochrome c at the outer side to protons that enter the oxidase from the matrix side of the membrane. In addition to this consumption of protons the oxidase does not mediate vectorial proton translocation.     

Proton translocation by cytochrome oxidase vesicles catalyzing the peroxidatic oxidation of ferrocytochrome c

Coupled with the peroxidatic oxidation of ferrocytochrome c under anaerobic conditions, proteoliposomes reconstituted with a purified preparation of bovine heart cytochrome oxidase ejected protons into the external medium with an apparent H+/e- ratio of 0.9. At the same time, protons in the intravesicular space were consumed. Dicyclohexylcarbodiimide significantly inhibited the proton translocation. Cyanide (0.14 mM) completely inhibited both the peroxidase and proton translocating activities. On the contrary, in the presence of 1 mM CO the proton ejection was abolished almost completely, but 50% of the peroxidase activity persisted. This result suggests the operation of multiple mechanisms in the peroxidase reaction and that the CO-sensitive one is coupled to the proton translocation.     

Chemiluminescence of Acanthamoeba castellanii.

We have investigated ferrocytochrome c-induced proton ejection from reconstituted cytochrome c oxidase-containing vesicles using careful control of the number of enzyme turnovers. Ferrocytochrome c caused the appearance of protons at the vesicle exterior, and this could be abolished by using a protonophore. In addition, its decay was dependent on the permeability of the vesicle membranes to protons and the number of turnovers of the oxidase. These observations indicate that the ejection of protons was the result of genuine translocation. The possibility of this translocation occurring via a Mitchellian loop as a result of the presence of a reduced hydrogen carrier contaminating the enzyme was considered and excluded. Proton-translocating activity in this reconstituted system depended critically on the ratio of enzyme to lipid used in the reconstitution process and we propose a rationale to account for this. We conclude that our data provide strong support for the proposal that cytochrome c oxidase acts as a proton pump and that approx. 0.9 H+ is excluded per ferrocytochrome c molecule oxidized.     

Evidence against proton pump activity by cytochrome c oxidase of Pseudomonas AM1

Proteoliposomes reconstituted from purified cytochrome c oxidase of Pseudomonas AM1 and from a heptyl beta-D-thioglucoside-extract of its membranes showed respiratory control but did not show H+ pumping upon a pulse with reduced cytochrome c. The stoichiometries of respiration-dependent H+ translocation in the resting cells respiring ascorbate via N,N,N',N'-tetramethyl-p-phenylenediamine were measured by the oxygen-pulse and initial rate methods. The apparent H+/O ratio of about 2 was due to 2H+ release from the hydrogen-donating substrate. These results strongly suggested that Pseudomonas AM1 does not pump H+ intrinsically, although the enzyme catalyzes electron transfer across the membranes.     

Energetics of ATP-driven reverse electron transfer from cytochrome c to fumarate and from succinate to NAD in submitochondrial particles

The maximum Gibbs free energies of reverse electron transfer from succinate to NAD+ and from cytochrome c to fumarate driven by ATP hydrolysis in submitochondrial particles from beef heart were measured as a function of the Gibbs free energy of ATP hydrolysis. The ratio of the energies delta G'redox/delta G'ATP was 1.40 from succinate to NAD+ and 0.89 from cytochrome c to succinate. The ratio, equivalent to a thermodynamic P/2e-ratio, was dependent on whether the electrochemical proton gradient was primarily a membrane potential or a pH gradient for the cytochrome c to fumarate reaction. The results are consistent with H+/ATP = 3 for F1 ATPase, H+/2e- = 4 for NADH-CoQ reductase, and H+(matrix)/2e- = 2 for succinate-cytochrome c reductase.