On the oxidation pathways of the mitochondrial bcl complex from beef heart

Transcription

1 Eur. J. Biochem. 160, (1986) 0 FEBS 1986 On the oxidation pathways of the mitochondrial bcl complex from beef heart Effects of various inhibitors Mauro DEGLI ESPOSTI Ah-Lim TSA12, Graham PALMER2 and Giorgio LENAZ Institute of Botany, University of Bologna Department of Biochemistry, Rice University, Houston, Texas (Received June 1/July 22, 1986) - EJB We have investigated the oxidation of the reduced ubiquinol : cytochrome c reductase (bcl complex) isolated from beef heart mitochondria. The oxidation of cytochrome cl by both potassium ferricyanide and cytochrome c in the ascorbate-reduced bcl complex is not a first-order process. This is taken as evidence that cytochrome ci is in rapid equilibrium with the Rieske iron-sulphur center. Among the several inhibitors tested, only 5-n-undecyl- 6-hydroxy-4,7-dioxobenzothiazole and stigmatellin are seen to affect this redox equilibrium between the highpotential centers of the beef heart bcl complex. The oxidation of cytochrome b by cytochrome c in both the succinate-reduced and the fully reduced bcl complex is blocked by all the inhibitors tested. This inhibition occurs simultaneously with an acceleration in the oxidation of cytochrome cl, even after extraction of the endogenous ubiquinone which is present in the bcl preparation. Almost complete extraction of ubiquinone from the bcl complex has no effect upon the rapid phase of cytochrome b oxidation, nor does it alter the inhibition of cytochrome b oxidation by the various inhibitors. The oxidation of cytochrome b by exogenous ubiquinones is stimulated by myxothiazol and partially inhibited by antimycin. However, the addition of both these inhibitors together completely blocks the oxidation of cytochrome b by quinones. In contrast, the simultaneous addition of antimycin and myxothiazol has no such synergistic effect upon the oxidation of cytochrome b by cytochrome c. Our data show that intramolecular electron transfer from cytochrome(s) b to the Rieske iron-sulphur center can take place in the bcl complex without involvement of endogenous ubiquinone-10. This electron pathway is sensitive to all the inhibitors of the enzyme. The ubiquino1:cytochrome c oxidoreductase (EC , bcl complex or respiratory complex 111) spans the central portion of the mitochondrial respiratory chain [l, 21. This redox enzyme consists of about ten different subunits, only three of whch carry a prosthetic group 121. A very hydrophobic polypeptide of 42 kda coordinates two b-type haems (bl and bh [I]) which possess different spectroscopic and potentiometric characteristics [2]. An amphiphilic subunit of 28 kda covalently binds a c-type haem (cytochrome cl), while a second amphiphilic polypeptide of 25 kda contains a 2Fe-2S cluster, usually called the Rieske iron-sulphur protein [2]. In addition to these four intrinsic prosthetic groups, a complement of endogenous ubiquinone (mostly ubiquinone- 10 in mammals) is usually found in purified preparations of the bcl complex; this quinone is present in an amount which is approximately equimolar with respect to cytochrome c1 [l Correspondence to G. Lenaz, Istituto Botanico, Universita di Bologna, Via Irnerio 42, Bologna, Italy Abbreviations. BAL, 2,3-dimercaptopropanol (British anti- Lewisite); 2Fe-2S, Rieske iron-sulphur center; HQNO, 2-nonyl-4- hydroxy-quinoline-n-oxide; Q-10, ubiquinone-10; Q-1, ubiquinone- 1 ; 4-2, ubiquinone-2; UHDBT, 5-n-undecyl-6-hydroxy-4,7-dioxobenzothiazole ; HPLC, high-performance liquid chromatography. Enzymes. Ubiquinol :cytochrome c reductase (bcl complex) (EC ); succinate:ubiquinone reductase (complex 11) (EC ); pronase (EC ). The two b cytochromes and ubiquinone are deeply embedded in the membrane sector of the complex [2] and only react rapidly with hydrophobic electron donors and acceptors. Exogenous ubiquinols and ubiquinones, and their analogs, constitute the most effective redox partners of the b cytochromes [2-41. The high-potential cytochrome b (bh) can react slowly with ferricyanide [5], particularly in the presence of antimycin [6]. The physiological relevance of such a reaction is however questionable [2, 4, 61. Cytochrome c1 and the Rieske iron-sulphur center have their prosthetic groups located in a hydrophilic domain of the complex and are believed to protrude out of the cytoplasmic face of the membrane [I, 21. These redox centers react readily with membrane-impermeant redox mediators such as ferricyanide, ascorbate and the physiological electron acceptor cytochrome c [2]. The latter protein has been shown to react only with cytochrome c1 [2, 71. The 2Fe-2S cluster and c1 are in very rapid redox equilibrium [8,9] and possess a midpoint potential much higher than that of both ubiquinone and the cytochromes b [2,4,9, 101. The use of a variety of specific inhibitors of the bcl complex has been of great value in understanding the observed kinetics of electron transfer [4, (for a review see [12, 171). According to the concepts of the Q-cycle, antimycin, HQNO and funiculosin interact at center i, whereas a large group of inhibitors including myxothyazol, mucidin, stigmatellin and hydroxy analogs of ubiquinone such as UHDBT, interacts at center 0 [12, 14, 15, 17, 181. An in-

2 548 hibitor from either group hardly modifies the reduction of cytochrome b by ubiquinol, but the combined action of an inhibitor from each group (e. g. antimycin plus myxothiazol) is able to block completely the reduction of the b-cytochromes [12, 17, 191. Such a phenomenon has been called the 'doublekill' of cytochrome b [2, 12, 15, 171. This phenomenon is an important piece of evidence in favour of cyclic schemes of electron transfer in the bcl complex [12, 15, 17, 191 of which Mitchell's Q-cycle hypothesis [20] remains the most valuable, particularly in its most recent versions [12-15, 211. The reoxidation pathways of the prereduced bcl complex have not been extensively studied [9]. The oxidation of cytochrome b by exogenous quinones was first investigated in [22] and then partially in [23] and [24]. The oxidation of all the cytochromes has been studied in the isolated bcl complex from yeast mitochondria [8,9,25], but only the data in [6] and [26] are so far available for the beef heart system. However, the results of the latter studies have been obtained mostly in mitochondria and appear to be still not complete. Herein we present a detailed study of the various oxidation reactions of the prereduced bcl complex isolated from beef heart mitochondria. MATERIALS AND METHODS Materials Beef heart mitochondria1 particles were prepared as described by Yu and Yu [27]. The bcl complex was isolated by the method of Rieske [28]. Some samples of bcl complex were kindly donated by Prof. C. A. Yu [27] and Prof. G. von Jagow [29]. The purified enzymes usually contained nmol cytochrome cl/mg protein (biuret) and exhibited a catalytic activity with 15 pm Q-2-H2 and 20 pm cytochrome c of pmol cytochrome c reduced min-' (mg protein)-' at 25 "C in M potassium phosphate, 1 mm EDTA, ph 7.4. Crude succinate cytochrome c reductase was obtained as fraction S1 [28] during the purification procedure of the bcl complex as previously described [16]. Treatment of fraction S1 with BAL plus oxygen was performed as described by De Vries [12]. Lipid and ubiquinone depletion of the isolated bcl complex was performed using six cycles of the cholate/ ammonium sulphate procedure [30]. The final preparation was dissolved in 50 mm potassium phosphate, 1 mh4 EDTA, 0,5% potassium cholate, ph 7.4, containing a mixture of sonicated phospholipids (Sigma, 25% cardiolipin, 28% phosphatidylethanolamine and 47% phosphatidylcholine) at a ratio of mg lipids/mg protein [30]. After min of incubation at O'C, the ubiquino1:cytochrome c reductase activity was restored within 80% of the initial value. The bcl complex depleted of iron-sulphur protein was prepared from the enzyme purified as in [29] by the method described by Engel et al. [31]. Such a depleted complex had less than 10% of the original activity. SDS-gel electrophoresis carried out as described in [31] indicated that more than 80% of the polypeptide band corresponding to the Rieske protein was missing in the depleted complex. Cytochrome c from either horse or beef heart (Sigma type VI and V respectively) was dissolved in 10 mm Tris/Cl, 1 mm EDTA, ph 7.4 at 1-2 mm. Ferricyanide solutions were prepared in the same buffer as the assays immediately before the experiment and stored in the dark. The absorption coefficient employed for the determination of the ferricyanide solutions was 1 mm-' cm-' at 420 nm [8]. Ubiquinones (a generous gift from Eisai Co., Tokyo, Japan), were dissolved in ethanol and ubiquinols were prepared as in [28]. Antimycin and HQNO were obtained from Sigma. Funiculosin was a generous gift from Dr. Bollinger (Sandoz, Basel, Switzerland), whereas mucidin was kindly donated by Dr. Musilek (Institute of Microbiology, Academy of Science, Praha, Czechoslovakia). Myxothiazol and stigmatellin were donated by Dr. Thierbach (Gesellschaft fur Biotechnologische Forschung, Braunschweig, FRG). UHDBT was purchased from Dr. B. L. Trumpower (Dartmouth College, Hanover, NH, USA). All the inhibitors were dissolved in ethanol or dimethylsulphoxide at 2 1 mm. Spectrophotometric determinations The concentration of the inhibitors and of the ubiquinols was determined spectrophotometrically in ethanol using published absorption coefficients [17, 241. The concentration of reduced cytochrome c was determined from the absorbance difference at nm with an absorption coefficient of 19mM-' cm-' [24] and the concentration of cytochrome cl was determined from the absorbance difference at nm after ascorbate reduction using an absorption coefficient of 17.5 mm-' cm-' [30]. Cytochrome b was measured at nm after dithionite reduction using an absorption coefficient of 25 mm-' cm-' [24]. Ubiquinones were determined at 275 nm with an absorption coefficient of 14.2mM-' cm-' [24]. The content of ubiquinone-10 of the preparations was measured on methanol/petroleum ether extractions, either by spectrophotometry at nm with an absorption coefficient of 8.8 mm-' cm-' [32], or by a more sensitive and specific HPLC technique [3]; the results obtained with the two methods were in good agreement [3]. No significant difference in the quantitative determination of Q-10 was found after complete proteolytic digestion of the bcl complex with pronase. The ubiquinone content of the Rieske preparation ranged over mol/mol cytochrome c1 and decreased by 90%, independently of the initial concentration, on extraction of ubiquinone-10 as described above. The preparation obtained by the method of Von Jagow [29] usually contained a markedly lower amount of intrinsic ubiquinone ( mol/mol cytochrome cl). Kinetic measurements The kinetic experiments were performed either in a dualwavelength Sigma Biochem spectrophotometer at room temperature or in a Gibson-Durrum single-wavelength spectrophotometer thermostated at 8 C [8, 91; the latter was connected to an Olis data system for analysis of the absorbance changes [S]. The Sigma dual-wavelength instrument was equipped with both a high-performance rapidmixing apparatus [24] and a stopped-flow apparatus which gave complete mixing at a 5: 1 ratio of enzyme/substrate in 5 ms [24]. In both systems the traces were stored and subsequently plotted in a appropriate format [16,24]. Cytochrome b oxidation was usually measured at nm when cytochrome c was employed in the oxidant mixture, or at nm in the presence of the inhibitors antimycin, HQNO and myxothiazol. At these wavelength pairs we found that the spectral interferences of cytochrome c with the absorbance changes of cytochrome b were negligible. Cytochrome c1 oxidation was measured at nm, where the spectral interference due to oxidized cytochrome c

3 were not significant. The buffer used contained M potassium phosphate, 1 mm EDTA, ph 7.4. The endogenous cholate present in the bel complex [12] was sufficient to maintain the solubility of the enzyme at the concentrations of 1-4 pm employed in the kinetic experiments. No significant change of the reaction rates was found in the presence of 0.1 '30 potassium deoxycholate or Triton QS-30, detergents which have been used previously in the study of the yeast enzyme [8, 9, 251. The oxidants used were either potassium ferricyanide alone or a mixture of equivalents of ferricyanide plus equivalents cytochrome clequivalent of cytochrome c1 [33]. In the cytochrome c/ferricyanide mixture the excess of ferricyanide maintains cytochrome c essentially in the oxidized state during the course of the reoxidation of the bcl complex. Under the conditions used, ferricyanide reacts more than two orders-of-magnitude faster with ferrocytochrome c than with ferrocytochrome c1 (see also [34, 351). Any significant interference in the observed absorbance changes of cytochrome b and c1 due to the rapid reduction of cytochrome c by the bcl complex is therefore minimized by the oxidation of cytochrome c by ferricyanide. The presence of cytochrome c in the ferricyanide solution produces an order-of-magnitude increase in the observed rates of cytochrome b or c1 oxidation, due to specific reaction between the reduced enzyme and its natural electron acceptor [36]. Most of the experiments have been performed under conditions in which the rate of cytochrome c1 oxidation was at least twofold faster than that of cytochrome b oxidation, thus excluding the possibility that the interaction of the oxidants with cytochrome c1 could be significantly rate-limiting in the overall process [9, 331. The bcl complex was reduced by one of three procedures. (a) Addition of a small excess of solid potassium ascorbate followed by gel filtration on a short Bio-Gel P6 column yields enzyme complex in which only cytochrome cl and the 2Fe-2S cluster are reduced [8]. (b) Because the bcl preparation contains a small amount of complex 11, incubation of the enzyme with mm succinate followed after 5 min with 2.5 mm malonate leads to reduction of 50-70% of the dithionite-reducible cytochrome b. About 30% of the endogenous Q-10 is reduced by this protocol. (c) Complete reduction of all the components is obtained by photochemical reduction using deazaflavin radical as reductant [9, 371. RESULTS Oxidation of the uscorbate-reduced bcl complex The ascorbate-reduced bel complex contains about two electrons/molecule cytochrome c1 ; these are located in cytochrome c1 and the Rieske iron-sulphur center when care is taken to use a minimal amount of ascorbate, thus avoiding any possible reduction of cytochrome b. We initially performed a large set of experiments with potassium ferricyanide to ensure a complete and irreversible oxidation of the enzyme; complete oxidation is expected since the midpoint potential of ferricyanide is much hlgher than that of the high-potential electron carriers of the complex [Z, 81. The rate of the reaction of ferricyanide with cytochrome c1 was found to be insensitive to temperature, to be unmodified by the presence of detergents and to decrease upon lowering the ionic strength of the medium; these observations are consistent with published date obtained on the oxidation of the purified cytochrome c1 by ferricyanide [35]. However, the progress curve for the oxida- 549 tion of cytochrome c1 in the bcl complex does not follow a first-order process even at very high ferricyanide concentrations, contrary to observations on the oxidation of purified cytochrome c1 with this reagent [8, 351. This deviation from first-order behaviour can be ascribed to the rapid redox equilibrium between cytochrome c1 and the Rieske center [4, 8, 91. Thus, the reaction between ferricyanide and the ascorbate-reduced beef heart bcl complex can be reproduced by the following minimal Scheme 1 [8, 91: where fox 4' + Fe (CN)Z- + c:' + Fe(CN):- k2/kl = K. The equilibrium constant K is given by the difference in midpoint potential between the Rieske iron-sulphur center and cytochrome cl, according to AE, (mv) = 59 log K. We have evaluated K for different preparations of the bcl complex by static titrations of the extent of cytochrome c1 oxidation upon addition of stoichiometric concentrations of ferricyanide (Fig.1A); K can be evaluated from the extent of cytochrome c1 reduction when only one electron is left in the complex [8]. A value of 2.5 has been found routinely for both the Rieske [27] and the Yu [26] preparations of the bcl complex, indicating that the midpoint potential of the ironsulphur center is about 24mV more positive than that of cytochrome el, in good agreement with the published potentiometric data [2, 4, 12, 151. The preparation of Von Jagow [28] yields a significantly higher value of K, also in accordance with potentiometric data showing that the midpoint potential of the Rieske center is somewhat more positive in this enzyme preparation [38]. We find that antimycin (cf. Fig. 1 A), funiculosin, myxothiazol and mucidin have little effect on the value of K deduced from the ferncyanide titrations, as expected from published potentiometric data [2, 4, 10, 381. UHDBT, however, does produce a dramatic increase, raising K to 65 (Fig. 1 A); this effect is expected from the striking increase in the midpoint potential of the Rieske cluster induced by this inhibitor [lo, 151. Scheme 1 has been utilized to simulate the time course of the oxidation of cytochrome c1 by an excess of ferricyanide as in [S] (Fig. 1 B). The dashed line shows the calculated time course for the oxidation of the iron-sulphur center; the rate of this reaction is seen to be slower that that of the oxidation of cytochrome c1 because the midpoint potential of the Rieske center is higher than that of cl. A minimal value of 600 s-l for k2 was required to produce a reasonable fit to the kinetic data, while values of kz greater than 600 s-' did not change the simulated time course. Thus, in the beef heart bcl complex there is a very rapid electron transfer between cytochrome c1 and the iron-sulphur protein, as already demonstrated for the yeast enzyme [8]. Addition of UHDBT (data not shown) increases the observed rate of cytochrome c1 oxidation by approximately twofold, and renders the process essentially first-order. This indicates that: (a) the large increase in K induced by UHDBT eliminates the redox buffering of cytochrome c1 by the Rieske center; (b) ferricyanide reacts predominantly with cytochrome cl, otherwise no clear effect of UHDBT on the observed

4 m Ol C - 0 C L : LL FelCNi:-/cyt c, Time (5) Fig. 1. Oxidation of the ascorbate-reduced bcl complex by potassium ferricyanide. (A) Titration of the extent of cytochrome CI reduction in ascorbate-reduced bcl complex (2 pm) as a function of ferricyanide concentration. The buffer used was 0.1 M potassium phosphate and the experiment was carried out at 25 C. (0-0) Control sample; (0-0) in the presence of 5 mol antimycin/mol cytochrome cl; (A---- A) in the presence of 5 mol UHDBT/mol cytochrome cl. (B) The time course of the oxidation of cytochrome c1 in ascorbatereduced bcl complex. The data represent the measured absorbance values, whereas the solid line was produced by a computer simulation of Scheme 1 (see the text and [8]). The constants used in the fit were: K = 2.5; k,. = 6 s-'; k2 = 600 s- '. The broken line is the computed time course for the oxidation of the Rieske iron-sulphur cluster. The concentration of bcl and ferricyanide were 4 pm and 40 pm respectively; the buffer was the same as that used in A and the temperature was 8 C. The Durrum instrument was used in this experiment Table 1. Apparent rate constants of cytochrome c1 oxidation by ferricyanide The hcl complex previously reduced by ascorbate was mixed with a 20-fold excess of potassium ferricyanide. The experiments have been carried out in M phosphate buffer at 25 C. The rate constants were evaluated from the initial rates of cytochrome c1 oxidation in the rapid-mixing apparatus. The inhibitors were added at concentrations 5-10-fold higher than that of the enzyme, which was pm Enzyme Inhibitor x Apparent Relative rate constant rate M-1 s-l Native none o UHDBT antimycin o funiculosin o myxothiazol Iron-sulphurdepleted none o UHDBT antimycin funiculosin o myxo thiazol reaction pattern would be seen [9]; (c) when cytochrome c1 is 'disconnected' from the iron-sulphur center, its rate of reaction with ferricyanide increases and becomes first-order, approaching the values obtained with the purified protein [35]. Similar results have been obtained: (a) upon adding funiculosin to the yeast bcl complex [9]; (b) adding stigmatellin to the beef heart bcl complex or removing the iron-sulphur protein from the beef heart enzyme (data not shown). In these cases the oxidation of cytochrome c1 is restored to a first-order process. The other inhibitors tested showed no significant effect on the oxidation reaction of the ascorbate-reduced bcl complex by ferricyanide (Table 1). Essentially similar results to those found with ferricyanide alone were obtained when a stoichiometric amount of cytochrome c (1 : 1 cytochrome cl) was added together with a 20-fold excess of ferricyanide. Oxidation of the succinate-reduced bcl complex Upon the addition of a low concentration of succinate to isolated bcl complex, both cytochrome b and c1 are reduced relatively slowly. Cytochrome b is reduced by more than 60%, whereas cytochrome c1 is reduced completely. The addition of malonate does not change the reduction level of cytochrome cl, but does induce a slight reoxidation of cytochrome b, yielding a final stable value of about 55% reduction (Fig. 2). Addition of a cytochrome clferricyanide mixture produces a rapid and complete reoxidation of both cytochrome c1 and cytochrome b. The rate of cytochrome c1 oxidation exceeds that of cytochrome b, indicating that the interaction of the oxidant with cytochrome c1 is not rate-limiting in the overall oxidation process [9, 331. Even at the lowest ratio of inhibitor to enzyme which is required for maximal inhibition of the steady-state activity of the bcl complex, all the inhibitors tested suppress cytochrome b oxidation and simultaneously stimulate cytochrome c1 oxidation (Table 2). Among all the compounds tested, funiculosin is the most effective inhibitor of the oxidation of cytochrome b (Fig. 2 and Table 2, see also [9]). However, with the ascorbate-reduced complex, the stimulation of cytochrome c1 oxidation is only produced by UHDBT and stigmatellin (cf. Table l), as expected because these inhibitors raise the Em of the Rieske iron-sulphur protein and kinetically decouple cytochrome c1 from 2Fe-2S (cf. Fig. 1 A). None of the other inhibitors appear to modify the redox equilibrium between these high-potential electron carriers of the enzyme (cf. Table 1, see also [38]). Thus the acceleration of cytochrome c1 oxidation observed in the succinate-reduced complex has to be due to other factors. One possibility considered was that the significant amount of ubiquinol-10 present in the enzyme before the addition of the oxidant (about 0.4 mol/mol cytochrome c1 in the experiment of Table 2) could promote re-reduction of the highpotential centers of the bcl complex during the oxidation process. The inhibitors would block this re-reduction and consequently the rate of cytochrome c1 oxidation would be

5 ~ ~~~ 551 seen to be faster in the presence of the inhibitors than in their absence. To test this possibility, we repeated the experiments of Table 2 using ubiquinone-depleted bcl complex; in this preparation the content of Q-10-H2 before the reaction with the oxidant was one order-of-magnitude lower than in the succinate rnalonote oxidant <I 20s 6) (1 0.2s ') Fig. 2. Oxidation 05 cytochrome b and cl in the succinate-reduced bcl complex. The reduction of cytochrome c1 (lower traces) and of cytochrome b (upper traces) by 0.6 mm succinate in the native bcl complex (1.1 pm, containing 1.4 mol ubiquinone-lo/mol cytochrome c,) was carried out in M potassium phosphate buffer at 25 C. The addition of malonate (2.5 mm) decreased the rate of b reduction by succinate over 80% (data not shown). On the right, the oxidation of the two cytochromes by 1 equivalent of beef heart cytochrome c plus 25 equivalents of ferricyanide per equivalent of cytochrome c1 are shown on a much faster time scale. The traces were obtained in the dual-wavelength mode using the rapid-mixing apparatus. The dashed lines represent the data obtained in the presence of 8 mol funiculosin/mol cytochrome cl. The higher absorbance changes at nm in this case reflect an increased contribution by cytochrome b, because it is reduced to a higher extent in the presence of the inhibitor [9] (see the upper line) native enzyme. Theoretically this should lead to a dramatic decrease in the rate of reaction of ubiquinol-10 with the oxidized complex [13, 141. However, in the ubiquinonedepleted bcl the various inhibitors stimulated the oxidation of cytochrome c1 in much the same way as in the native enzyme (Table 3). These data imply that the acceleration of cytochrome c1 oxidation induced by all the inhibitors is not a consequence of the inhibition of an important re-reduction process. The time course of cytochrome b oxidation is biphasic; the first phase usually accounts for more than 65% of the overall absorbance change and is much faster than the second phase. Removal of ubiquinone from the complex does not significantly affect the rate of the faster phase of cytochrome b oxidation, nor the responses to the several inhibitors we have tested (Table4). Extensive depletion of the quinone (up to 0.1 mol/mol cytochrome cl) only slightly increases the extent of the second, slower phase of cytochrome b oxidation, even at the maximal rates we measured (about 12 s- ', Fig.3A). Addition of exogenous 4-2 at equimolar concentrations to the Q-depleted complex reduces the extent of the slower kinetic phase, without any significant stimulation of the rapid phase of b oxidation [33]. The Von Jagow preparation of the bcl complex [29] intrinsically contains much less ubiquinone than the Rieske preparation (see also [38]). However, the oxidation of cytochrome b is equally fast in the two preparations; in both cases, addition of exogenous Q-2 only decreases the extent of the slow phase of the time course (data not shown). These results agree well with those previously found in the yeast bcl complex, where the rapid oxidation of cytochrome b is also independent of the content of endogenous 4-6 [9]. From the data in Tables 2 and 4 it appears that the myxothiazol-like inhibitors tend to block oxidation of cytochrome b to a lesser extent than the antimycin-like inhibitors, particularly after depletion of ubiquinone. Similar data have been obtained in intact beef heart mitochondria using ferricyanide as the oxidant (L. Landi and M. Degli Esposti, unpublished observations). We have therefore investigated whether the addition of a center 'i' inhibitor to the bcl complex Table 2. Effect of various inhibitors on the oxidation of cytochrome b and c1 in succinate-reduced bcl complex The experimental conditions were identical to those in Fig.2. The kinetic values express only the apparent first-order rate constant for the rapid phase of the oxidation process; this usually accounted for 70-85% of the overall absorbance change. The oxidation rates with the first two inhibitors have been measured after the extra reduction of cytochrome b was complete. Due to the clear red shift induced by these inhibitors on the spectrum of cytochrome b [lo], the traces were monitored at nm. The concentration of the oxidants was nonsaturating, in order to able to resolve the kinetic traces with the rapid-mixing apparatus (cf. Fig.2). Employing higher concentrations of cytochrome c and ferricyanide cytochrome b oxidation reached a maximal rate of 12 s-l (cf. Fig.3A), whereas the turnover rate of the bcl complex using saturating concentrations of ubiquinol-1 and cytochrome c was around 13 s- under the same experimental conditions (data not shown) Inhibitor Molar ratio to c1 Cytochrome b oxidation Cytochrome c1 oxidation kobs cf. control kdbs cf. control None s % S % Antimycin HQNO Funiculosin Myxothiazol Stigmatellin Mucidin UHDBT

6 552 Table 3. Comparison of the effect of inhibitors on the oxidation of cytochrome c1 The experimental conditions were the same as in Table 2; ferricyanide was in a 20-fold excess over cytochrome ci. Ubiquinone-depeleted bcl complex, retaining 0.2 mol Q/mol enzyme, was reduced by 0.2 mm succinate. The amount of ubiquinol present in this preparation before the oxidant pulse was determined to be less than 0.04 mol/mol cytochrome c1 by the HPLC technique. The final concentration of the enzyme was 1.2 FM Inhibitor Molar ratio to c, Cytochrome c plus ferricyanide Ferricyanide kobs cf. control kt.3 cf. control S-1 % S-1 % None Antimycin Funiculosin M yxothiazol UHDBT oo 364 Table 4. Effect of various inhibitors on the oxidation of cytochrome b in a ubiquinone-depleted bcl complex Inhibitors as indicated were incubated at the molar ratios given in the table with bcl complex which had been depleted of the endogenous Q to a level of 0.2 mol/mol cytochrome cl. Other conditions as in Fig. 2 and Table 2. The kinetic values express the apparent first-order rate constants of the rapid phase of cytochrome b oxidation, which accounted for about 60-75% of the overall absorbance changes. With the first two inhibitors, no extra reduction of cytochrome b could be detected, due to the extensive depletion of ubiquinone; the presence of ubiquinone is necessary to demonstrate a significant onset of such a phenomenon [6] Inhibitor Molar ratio to c1 kobs cf. Control S-1 Yo None Antimycin HQNO Funiculosin Myxothiazol Stigmatellin Mucidin UHDBT already treated with a center 0 inhibitor could suppress completely the oxidation of cytochrome 6, similarly to the double-kill of the reduction of cytochrome b As shown in Fig. 3 B, funiculosin clearly inhibits cytochrome b reduction in the presence of stigmatellin, with a classical double-kill effect [19, but it does not significantly affect the subsequent reoxidation pattern. The same type of results has been obtained with antimycin plus myxothiazol. Experiments performed with the Durrum stopped-flow system using bcl complex which had been reduced photochemically with deazariboflavin gave results which are essentially similar to those reported above, in agreement with the data found in the yeast enzyme [9]. Oxidation of cytochrome b by exogenous ubiquinones The succinate-reduced bc, complex can selectively be depleted of the electrons from the b cytochromes and the endogenous ubiquinone using exogenous ubiquinones [22], particularly ubiquinone-1 and -2. These homologs are also the best substrates for the reductase activity of the enzyme [24]. Although this oxidation reaction may be simply described by the equilibrium [4, 141: 6 + Q-1 b3+ + Q-1 - we find that the rates of cytochrome b oxidation vary hyperbolically with the ubiquinone concentration (see also [24]). The overall mechanisms of the reaction should therefore include a binding step of the quinone to the reduced enzyme, viz, Scheme 2: where k- = Kd and can be determined from the doublereciprocal plot of the observed rate constant as a function of ubiquinone concentration [39]. In the isolated bcl complex the values of Kd and kz for the oxidation of cytochrome b by Q-1 are 2 pm and 3.9 s-l respectively. The semiquinone radical may either rapidly dismute [21] or be stabilized through the specific binding to the complex [12]. It should be noted that cytochrome c1 and the 2Fe-2S cluster stay reduced during the oxidation of cytochrome b by ubiquinones, as these centers have a much higher redox potential than the quinones [41. Fig. 4 shows the typical behaviour of the oxidation of cytochrome b by Q-1 in the succinate-reduced bcl complex, and the effects induced on the reaction by either antimycin of myxothiazol. The latter antibiotic stimulates the oxidation process, whereas antimycin significantly decreases the rate and the extent of cytochrome b oxidation. The inhibitory effect of antimycin is less pronounced in coupled mitochondria than in the isolated bcl complex. Subjecting the complex to treatments known to labilize the b cytochromes, for instance by ageing or by treatment with detergents [40], causes the oxidation of b by quinones to become more sensitive to antimycin (M. Degli Esposti, unpublished results). The addition of myxothiazol to the antimycin-treated bcl complex markedly enhances the degree of inhibition of the rate of b oxidation by ubiquinone-1 (Fig.4A). Elimination of the iron-sulphur cluster, either by extraction or by treatment of the enzyme with BAL, also increases the rate of oxidation in much the same way as myxothiazol. Again, addition of antimycin leads to complete inhibition of cytochrome b oxidation b) ubiquinones (erg. Fig.4B). Table 5 summarizes the dependence of the rates of b oxidation on the concentration of Q-1 in the same succinatecytochrome c reductase preparation as that used for the data

7 A 0.8 D o L v? LL 0.2 oxidant U L Time (5) succinate Time (5) Fig. 3. (A) Time course of the oxidation of cytochrome b by the c-ferricyanide mixture. (B) Effect of inhibitors on the reduction and the oxidation qf cytochrorne b. (A) The reaction with horse-heart cytochrome c and ferricyanide (2 and 20 equivalents/cytochrome c1 respectively) was measured in the stopped-flow system at nm at 25 C. (--- --) Ubiquinone-depleted bc, complex (residual Q-10 was 0.1 mol/mol cytochrome el); (-)native bcl complex (containing 1.1 mol Q-lO/mol cytochrome cl). Other experimental conditions as in Fig.2. except that the enzyme concentration was 3 pm for the native and 2.2 pm for the ubiquinone-depleted sample respectively. The reaction was about 96% sensitive to antimycin in both cases. (B) The experimental conditions were the same as those employed in the experiment of Fig.2, including the addition of 2.5 mm malonate (which is not shown in the figure), but note the different time scales used for cytochrome b oxidation. (- ----) The trace obtained in the presence of 3 mol stigmatellin/mol cytochrome c1 ;(-) the trace obtained in the presence of stigmatellin plus the further addition of 4 mol funiculosin/mol cytochrome cl. In the latter case the reduction rate of cytochrome b by succinate was 6% of that measured in the presence of stigmatellin alone. The rate of b oxidation in the presence of funiculosin alone was about 12-16% of that in the presence of stigmatellin A Q1 U._... 1 '. I '.....! 0.01A Q1 c t c J... - t 5u= -3 L,e Time (5) Fig.4. The effect of (A) antimycin and myxothiazol and (B) BAL and antimycin on ihe oxidation of cytochrome b by Q-I. (A) Oxidation of isolated bc, complex reduced by 1 mm succinate under the same conditions as those described in the legend to Fig.2, except that the enzyme concentration was 0.9 pm and the Q-1 Concentration was 20 pm. (-) Control trace; (- --) after addition of 13 mol myxothiazol/mol cytochrome c1 ;(....) after addition of 12 mol antimycin/mol cytochrome cl; ( ) after addition of both antimycin and myxothiazol, with the latter added after completion of cytochrome b reduction, to avoid the large inhibition by the combined action of the two inhibitors [19]. The lower and upper separated lines represent the extent of the fully oxidized and of the dithionite-reduced cytochrome b respectively. (B) Crude succinate:cytochrome c reductase suspended at 0.9 pm cytochrome c1 in 0.1 M potassium phosphate buffer was reduced by 4 mm succinate (SUC) and then treated with 10 mm malonate (mal). The ubiquinone-1 concentration was 9 pm. ( ) Trace obtained in the preparation which was treated with 10 mm BAL during rapid stirring for 30 min at room temperature [12] and then centrifuged at x g for 20 min to remove the excess of the reagent; this procedure could not be applied to the isolated hcl complex. (-) Trace obtained in a control sample of the succinate:cytochrome c reductase preparation which was subjected to the same treatment at room temperature as that used for the BAL modification; no significant decrease of the enzymatic activity was found in this case. (....) Trace obtained after addition of 3 mol antimycin/mol cytochrome c1 in the control sample. (- ----, left) Reduction trace measured in the BAL-treated sample in the presence of the same concentration of antimycin; (- ----, right) oxidation trace of the BAL-treated preparation which was first reduced by succinate plus malonate, and then treated with antimycin, to avoid the striking inhibition of cytochrome h reduction [12]. The rates of reduction of cytochrome b by succinate in the presence of malonate were about one order-of-magnitude slower than the rates of oxidation by Q-1 shown in Fig.4B. Among the several inhibitors tested, HQNO (see Fig. 1 A and [15]) and is assumed to interact at center '0' and UHDBT (together with funiculosin, data not shown) ap- [15, 171. pear to be the most effective in inhibiting the rate of the The oxidation reaction promoted by ubiquinone-2 is oxidation of cytochrome b, inducing a striking increase in the qualitatively similar to that found with Q-1, with a dissociaapparent Kd for the reaction. This action of UHDBT is quite tion constant (Kd) of 2.3 pm and an apparent maximal rate surprising for this inhibitor also modifies the redox (kz) of 18 s-' under the same experimental conditions of equilibrium of the high-potential centers of the bcl complex those in the experiments of Table 4. The depletion of the

8 554 Table 5. Eflect of inhibitors on the oxidation of cytochrome b by ubiquinone-1 Crude succinate:cytochrome c reductase [16] was reduced by succinate under the same experimental conditions as in Fig.4B. The values of Kd and k2 (Scheme 2) have been obtained from the double-reciprocal plots of the rates of oxidation as a function of the quinone concentration [39] Inhibitor Inhibitor/cl Kd k2 mol/mol PM S-1 None Antimycin HQNO UHDBT M yxothiazol BAL (10 mm) Antimycin plus myxothiazol Antimycin plus BAL see footnote a see footnote a a The very slow reaction does not saturate as a function of Q-1 concentration, exhibiting an apparent second-order rate constant of 2 x 103 M- s-1. endogenous ubiquinone of the enzyme has little effect on the mode of cytochrome b oxidation by exogenous ubiquinones. DISCUSSION Some of the data of this work appear to be in conflict with the basic postulates common to both the Q-cycle and the b- cycle schemes for electron transport in the bcl complex [l, 4, 20, 211. a) Ubiquinone is clearly not required for the rapid antimycin-sensitive oxidation of cytochrome b by cytochrome c (cf. Fig.3A, see also [9, 331). A stoichiometric amount of ubiquinone should be required for the oxidation of both b cytochromes in all the cyclic schemes so far proposed [l, 4, 21, 261. It might be argued that the diffusion of ubiquinone between individual bcl complexes is sufficiently fast to support the oxidation we observe, even at low ratios of Q/ cytochrome cl. We find this to be most unlikely for the following reasons. First, experiments in which fully oxidized and fully reduced native yeast bcl were mixed in a stoppedflow instrument clearly showed that the reduction of cytochrome c1 present in the fully oxidized complex required several minutes for completion. In view of the large difference in redox potential between cytochrome c1 and the b cytochromes plus Q, ths experiment is persuasive evidence that the intercomplex transfer of reducing equivalents is extremely inefficient, even when the full complement of ubiquinone is present (A. L. Tsai, R. Kauten and G. Palmer, upublished results). Second, such rapid equilibration of the quinone implies that Q exhibits Q-pool behaviour [4, 331. However, we have demonstrated that the oxidation of cytochrome b by cytochrome c does not exhibit Q-pool behaviour, since it is unaffected by lipid dilution of (the residual) ubiquinone [33], nor is influenced by the content of Q [33] (and this work). b) The double-kill effect of antimycin plus either myxothiazol or BAL can be demonstrated with the oxidation of cytochrome b by exogenous ubiquinone (cf. Fig. 4), but not in the oxidation of cytochrome b by cytochrome c (cf. Fig. 3 B). The reverse reaction at center o, although thermodynami- cally allowed [9], usually is considered to be kinetically forbidden [4, 12, 14, 211. Moreover, the rate of the reverse electron transfer from the high-potential to the low-potential cytochrome b should be severely attenuated by the substantial gap in their midpoint potentials [4, lo]. c) The oxidation of cytochrome c1 is stimulated by all the inhibitors of the bcl complex when either the high-potential cytochrome b (cf. Tables 2 and 3) or both b cytochromes [9] are reduced, but not when they are oxidized, as in the ascorbatereduced complex (cf. Table 1, see also [8]). One would expect that the rate of c1 oxidation would be the same both in the absence and in the presence of inhibitors, unless these inhibitors prevent the normal electron flow to cytochrome cl. A redox buffering on cytochrome c1 is normally exerted by the Rieske center [8, 91, but this should be the same in the ascorbate-reduced as in the succinate-reduced or the fullyreduced bcl complex. Another electron carrier in the complex has to contribute to the redox buffering of cytochrome c1 in the succinate-reduced enzyme, in order to explain the stimulation of c1 oxidation in the presence of inhibitors which do not affect the redox equilibrium with the 2Fe-2S center, such as antimycin and myxothiazol (cf. Table 1). It is unlikely that this component is the endogenous ubiquinol which is present before the addition of the oxidant, since a very similar acceleration of cytochrome c1 oxidation is found in the ubiquinonedepleted complex (cf. Table 3 and [9]). According to the current cyclic schemes, no alteration of the mode of cytochrome c1 oxidation is to be expected with a specific inhibitor unless, as in the case of UHDBT [15] or stigmatellin [38], it clearly modifies the redox equilibrium with the Rieske iron-sulphur center. The second criticism noted above may be overcome by assuming that the reverse reaction at center 0 can occur [9]. Thus, the high-potential cytochrome b could also be oxidized in a myxothiazol-sensitive way through the reversal sequence bh + bl -+ ubiquinone. The resulting fully-reversible cyclic scheme, however, would still be unable clearly to account for points (a) and (c). The modified b-cycle scheme recently proposed by hch and Wirkstrom to explain the oxidation of fully-reduced cytochrome b is also based on an absolute requirement for ubiquinone in its semiquinone state [26] and can not explain point (a). The simplest explanation for our results postulates that upon reduction of the enzyme, and particularly after extraction of ubiquinone, direct intramolecular electron-transfer can take place from the b cytochromes to the Rieske center [4, 9,25,33]. We postulate that such an intramolecular reaction is highly sensitive to the conformational state of the enzyme. Since most of the specific inhibitors of the bcl complex clearly modify its conformation [2, 18,411, the binding of the inhibitors to the enzyme could produce, either directly or indirectly through structural transitions, an inhibition of the electron pathway inside the complex. The reaction scheme for b oxidation in the beef heart bcl complex would therefore follow the linear sequence : bb (antimycin, funiculosin, HQNO) I! \- I (myxothia!ol mucidin) >patellin, UHDBT) 2Fe-2S 7 c1 -+ oxidant The distance between the b haems and the 2Fe-2S cluster inside the complex (2-3 nm [21]) may be close enough to allow an electron transfer b + 2Fe-2S at a rate of s- according to Marcus theory [42]. The rates of cytochrome b oxidation we have measured are in this range (cf. Fig.3A, see also [9]) and are comparable to the rate of reduction of

9 555 cytochrome c1 by ubiquinol-1 and to the turnover rate of cytochrome c reduction under the same experimental conditions (cf. Table 2). These values are clearly lower than those found when the enzymatic activity is measured under optimal conditions ( s- [24]), probably due to aggregation phenomena of the bel complex occurring at the relatively high concentrations needed for our kinetic measurements which may prevent the appropriate binding of the substrates [12]. If the above hypothesis is valid, one must ask if there might be any physiological relevance for the intramolecular b -+ iron-sulphur protein pathway. In fact, if such a reaction could be followed during the normal catalysis, it would lead to a short-circuiting of cyclic electron transfer mechanisms such as both the b-cycle and the Q-cycle [4,9,26]. The recent finding that the partially reduced bcl complex exhibits a lower catalytic turnover than that of the oxidized complex [43] may suggest that alternative electron routes could be followed, depending on the conformation associated with the different redox states. The large conformational changes which seem to occur in the complex upon reduction [2], in fact, might favour the intramolecular pathway b --f 2Fe-2S, for instance by reducing the distance between one b haem and the Rieske cluster. We would like to consider also the extreme possibility raised by the present findings that the double-kill of cytochrome b may not reflect the main path of electron flow in the enzyme, since it is found during reduction by ubiquinol [19] (cf. Figs 3B and 4B), but not upon oxidation by cytochrome c (cf. Fig.3B). Actually the double-kill is found when ubiquinone oxidizes cytochrome b in the presence of antimycin and myxothazol, or antimycin plus BAL treatment (cf. Fig.4). All those factors affecting center 0 might also change the molecular reactivity of cytochromes b with their hydrophobic redox partners Q and Q-H2, leading to a situation whch is completely different from that which exists in the native enzyme [44]. In support of this view, we note that all the inhibitors tested modify the Kd of Q-1 for cytochrome b oxidation (cf. Table 5) and that the relationship between the concentration of ubiquinol-1 and the rate of reduction of cytochrome c1 is clearly modified by many inhibitors such as antimycin, funiculosin, HQNO and BAL [ A detailed study on the chemical reactivity of isolated cytochrome b in the absence and in the presence of isolated iron-sulphur protein will clarify whether the latter speculation is valid. We thank Dr. R. Kauten for his kind help in performing some HPLC experiments. The work was supported by grants from Minister0 della Pubblica Istruzione and Centro Nazionale delle Ricercbe (Rome, Italy), the National Institute of Health (GM 21332) and the Welch Foundation (C636). REFERENCES 1. Wikstrom, M. K. F. & Saraste, M. (1985) in Bioenergetics (Ernster, E., ed.) pp.49-84, Elsevier, Amsterdam. 2. Rieske, J. S. & Ho, S. H. K. (1985) in Coenzyme Q (Lenaz, G., ed.) pp , Wiley, Chichester. 3. Tsai, A, L., Kauten, R. & Palmer, G. (1985) Anal. Biochem. 151, Rich, P. R. (1984) Biochim. Biophys. Acta 768, Kunz, W. S., Konstantinov, A,, Tsofina, L. & Liberman, E. A. (1984) FEBS Lett. 172, Pasquali, P., Degli Esposti, M., Landi, L., Cabrini, L. & Lenaz, G. (1985) J. Bioenerg. Biomembr. 17, Broger, C., Nalecz, M. J. & Azzi, A. (1980) Biochim. Biophys. Acta Tsai, A. L., Olson, J. S. & Palmer, G. (1983) J. Biol. Chem. 258, Palmer, G., Tsai, A. L.. Kauten, R., Degli Esposti, M., & Lenaz, G. (1985) in Achievements andperspectives ofthe mitochondria1 research (Quagliarello, E., et al., eds) pp , Elsevier, Amsterdam. 10. Tsai, A. L., Kauten, R. & Palmer, G. (1985) Biochim. Biophys. Acta 806, De Vries, S., Albracht, S. P. J., Berden, J. A. & Slater, E. C. (1982) Biochim. Biophys. Acta 681, De Vries, S. (1983) Ph. D. Thesis, University of Amsterdam. 13. Rich, P. R. (1983) Biochim. Biophys. Acta 722, Crofts, A. R., Meinhardt, S. W., Jones, K. R. & Snozzi, M. (1983) Biochim. Biophys. Acra 723, Bowyer, J. R.. Edwards, C. A,, Ohnishi, T. & Trumpower, B. L. (1982) J. Biol. Chem. 257, Degli Esposti, M. & Lenaz, G. (1982) FEBS Lett. 142, Von Jagow, G. &Link, T. A. (1986) Methods Enzymol. 126, Rieske, J. S. (1980) Pharmacol. Ther. 11, Von Jagow, G. & Engel, W. D. (1981) FEBSLett. 136, Mitchell, P. (1975) FEBS Lett. 56, Mitchell, P. & Moyle, J. (1985) in Coenzyme Q (Lenaz, G., ed.) pp , Wiley, Chichester. 22. Von Jagow, G. & Bohrer, C. (1975) Biochim. Biophys. Acta 387, Malvyia, A. N., Nicholls, P. & Elliott, W. B. (1980) Biochim. Biophys. Acta 589, Degli Esposti, M., Rotilio, G. & Lenaz, G. (1984) Biochim. Biophys. Acta 767, Tsai, A. L. (1983) Ph. D. Thesis, Rice University, Houston. 26. Rich, P. R. & Wikstrom, M. (1986) FEBS Lett. 194, Yu, C. A. & Yu, L. (1982) J. Biol. Chem. 257, Rieske, J. S. (1967) Methods Enzymol. 10, Engel, W. D., Shagger, H. & Von Jagow, G. (1980) Biochim. Biophys. Acta 592, Yu, C. A. & Yu, L. (1980) Biochemistry 19, Engel, W. D., Michalski, C. & Von Jagow, G. (1983) Eur. J. Biochem. 132, Kroger, A. (1978) Methods Enzymol. 53, Lenaz, G., Battino, M., Degli Esposti, M., Fato, R. & Parenti- Castelli, G. (1986) in Biomedical and clinical aspects ofcoenzyme Q (Folkers, K. & Yamamura, Y., eds) vol. 5, pp , Elsevier, Amsterdam. 34. Morton, R. A., Overnell, J. & Harbury, H. A. (1970) J. Biol. Chem. 245, Konig, B. W., Veerman, E. C. &Van Gelder, B. F. (1982) Biochim. Biophys. Acta 681, Kim, C. H., Balny, C. & King, T. E. (1984) Proc. Natl Acad. Sci. USA Massey, V. & Hemmerich, P. (1977) J. Biol. Chem. 252, Von Jagow, G. & Ohnishi, T. (1985) FEBS Lett. 185, Hiromi, K. (1979) Kinetics offast enzyme reactions, Kodansha, Tokyo. 40. Rieske, J. S. (1977) Arch. Biochem. Biophys. 145, Rieske, J. S., Ramesh, V. & Tripathy, C. (1984) in Biomedical and clinical of coenzyme Q (Folkers, K. & Yamamura, Y., eds) vo1.4, pp , Elsevier, Amsterdam. 42. Marcus, R. A. & Sutin, N. (1985) Biochim. Bioph-vs. Acta 811, Degli Esposti, M. & Lenaz, G. (1985) Proceedings of International Symposium in Mitochondria1 Research, p. 58, University of Bari, Bari. 44. Lenaz, G., Degli Esposti, M., Crimi, M., Saw, J., Svoboda, P. & Timoneda, J. (1985) in Water and ions in biological systems (Pulman, B., Packer, L. & Vasilescu, V., eds) pp , University of Bucarest, Bucarest. 45. Degli Esposti, M. & Lenaz, G. (1985) Third EBEC Short Reports, pp, , lcsv Press, Cambridge.