The effect of ΔμH+ on the interaction of rotenone with Complex I of submitochondrial particles

Biochimica et Biophysica Acta, 1140 (1992) 169-174

169

© 1992 Elsevier Science Publishers B.V. All rights reserved 0005-2728/92/$05.00

BBABIO 43725

The effect of A H+ on the interaction of rotenone with Complex I of submitochondrial particles Alexander B. Kotlyar and Menachem Gutman Laser Laboratory for Fast Reactions in Biology, Department of Biochemistry, Tel Aviv University, Tel Aviv (Israel)

(Received 13 April 1992)

Key words: NADH-ubiquinone reductase; Respiratory chain; Electron transfer; Rotenone inhibition; (Bovine heart mitochondrion) The inhibition by rotenone of the forward (NADH-oxidase) and reverse (AtzH+-dependent succinate-NAD ÷ reductasc activities of submitochondrial vesicles was measured. The inhibition of NADH-oxidase, measured in the presence of uncoupler, followed a monophasic inhibition curve with K i ~<2 nM. The reverse electron flow was only partially (40%) inhibited at these rotenone concentrations. The rest of the activity was less sensitive to the inhibitor (K i = 30 nM). The lower affinity for the inhibitor of the reverse electron flow is a consequence of enhanced rate of rotenone dissociation caused by the high A/~H+ value required for this reaction. The analysis of the results indicates that the AS-SMP preparation consists of two subpopulations: one with a relatively low degree of coupling, which exhibits high sensitivity to rotenone and the other which is highly coupled with lower affinity to the inhibitor.

Introduction The mitochondrial NADH-ubiquinone oxidoreductase (EC 1.6.99.3) commonly termed Complex I carries a A~ia+-dependent electron transfer between the water-soluble dinucleotide - N A D H and lipid-soluble ubiquinone. One flavin (FMN [1]) and at least four Fe-S [2-4] centers each with its own midpoint potential [5] are involved in this A~u+-dependent redox reaction. The enzyme, either m e m b r a n e - b o u n d or when purified, can reduce several artificial electron acceptors. More then one reducing site is involved in these reactions. The reduction of the artificial oxidants like ferricyanide, Wurster's Blue or 2,6-dichlorophenolindophenol is not coupled with energy transduction steps. On the other hand, reaction of m e m b r a n a l Complex I (in submitochondrial vesicles or incorporated in liposomes) with ubiquinone and its water soluble homologs (Q0, Q1, Q2 or duroquinone [6-10]) is coupled with energy accumulation. The reverse electron flow in m e m b r a n e - b o u n d mito-

Correspondence to: A.B. Kotlyar, Laser Laboratory for Fast Reactions in Biology, Department of Biochemistry, George S. Wise Faculty of Life Sciences, Ramat Aviv, 69978, Tel Aviv, Israel. Abbreviations: Qn, homologs of ubiquinone having n isoprenoid units in position 6 of the quinone ring; BSA, bovine serum albumin; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; SMP, submitochondrial particles.

chondrial Complex I (Ap~H+-dependent q u i n o l - N A D reductase) is regarded as a reversal of the events associated with the native sequence of electron-transfer steps during oxidation of N A D H . The free energy needed for transfer of electrons from high midpoint potential donor (quinol) through Complex I to N A D + is supplied either by A T P hydrolysis (ATP-dependent reaction) or oxidation of succinate (succinate energy supported reaction). The first, ATP-dependent, reaction was originally discovered in intact mitochondria [11,12] and later demonstrated in submitochondrial particles [13,14]. This reaction is sensitive to uncouplers, inhibitors of ATPase and inhibitors of Complex I (rotenone and piericidine). The reverse electron transfer supported by succinate oxidation is sensitive to respiratory inhibitors like antimycin, cyanide, malonate, and is insensitive to excess of oligomycin [15,16]. The reverse reaction as well as the forward one is inhibited by rotenone and piericidine - classical inhibitors of the N A D H - Q reductase segment of mitochondrial respiratory chain. Both inhibitors compete for common sites on Complex I [17-19] and suppress N A D H - Q reductase activity of the enzyme at stoichiometrical concentrations [17,19]. The binding of these inhibitors, especially in the presence of N A D H [21,22], is very tight, but not covalent [17,23,24]. The affinity of rotenone is smaller than for piericidine and can be partially reversed by successive washing with BSA [19,23].

170 Both rotenone and piericidine inhibit the catalytic functions of the enzyme which span the A/zH+-dependent step [6,25,26]. Thus, we may conclude that the inhibitory site may be directly involved in the energy conservation. If so, then the inhibitor-binding properties of the enzyme should depend on atzH+. Until now, such an observation has not been reported. In this work we describe the effect of rotenone on the forward and reverse reactions carried out by m e m brane-bound mitochondrial Complex I. The data demonstrate a large change in the rotenone affinity caused by electrochemical potential.

7

6

54 E]~FECrOR

NADH

3

NAD*+SUCCINAI~~ ] A34o=0,01

Materials and Methods

Submitochondrial particles were p r e p a r e d from bovine heart mitochondria. The coupling of the vesicles was increased by addition of small amounts of oligomycin as described before [16]. In all assays the vesicles were preexposed to N A D H to potentiate the activity of Complex I [16]. The energy-linked reduction of N A D ÷ was followed spectrophotometrically at 340 nm at 30 o C. The reaction mixture contained 0.25 M sucrose, 1 m g / m l BSA, 4 m M N A D +, 20 m M succinate, 0.2 m M E D T A and 0.02 M H e p e s (potassium salts; p H 8.0). NADH-oxidase was followed spectrophotometrically at 340 nm at 3 0 ° C in solution containing 0.25 M sucrose, 1 m g / m l BSA, 2 ~ g / m l gramicidin-D, 50 ~tM N A D H , 0.2 m M E D T A and 0.02 M H e p e s (potassium salts; p H 8.0). N A D H - Q o reductase activity was recorded spectrophotometrically in the same assay mixture in the presence of 0.1 tzM myxothiazol and 400 IzM Qo. The reaction was started by the addition of QoAll experiments were carried out in the presence of low enzyme concentrations (25-50 i z g / m l in the assay). The combination of low amount of enzyme and high level of coupling allowed us to monitor the reaction for more than 10 min before anaerobiosis was established. Protein was determined with a biuret reagent [27]. All chemicals were obtained from Sigma. Results

The effect of rotenone on succinoxidase-driven reverse electron flow The time-course of N A D ÷ reduction supported by oxidation of succinate is shown in Fig. 1. After rapid oxidation of added N A D H (the procedure needed for activation of Complex I [16]), succinate and N A D ÷ were added simultaneously. U n d e r these conditons the oxidation of succinate generates sufficient A/Zn+ to drive the electron flow from the quinol pool back to N A D ÷, as seen by the rise in the curve. The rate of N A D H accumulation progressively slows until a steady

Fig. 1. The effect of rotenone on the succinate-supported reverse electron transfer. AS-particles were treated with oligomycin (0.4 /zg/mg) and assayed as described in Materials and Methods. The absorbance at 340 nm is drawn vs. time with indication for the addition of particles (50 ~g/ml), NADH (2 izM), NAD + (1 mM) plus succinate (20 mM) and effectors. The effeetors used for the various lines are: (1) 20 mM malonate; (2) 10 tzl of ethanol; (3) 5.10 -7 M rotenone; (4) 2.10 -7 M rotenone; (5) 10-7 M rotenone; (6) 10-8 M rotenone; (7) 5.10 -8 M rotenone.

state is established. The steady-state level is characterized by the equal rates of the reverse reaction (A/~n~dependent succinate-NAD + reductase) and forward oxidation of the generated N A D H ; the level is stable as long as the system remains aerobic. The steady state is susceptible to modulation by inhibitors of succinate and N A D H dehydrogenases. Addition of malonate (see curve 1 on Fig. 1) causes selective inhibition of the A~u~-generating system with subsequent decrease of N A D H concentration. On the other hand, rotenone, which inhibits the N A D H - Q sector of the electron path, shifts the steady state to the higher level of N A D + reduction. The effect of rotenone is concentration-dependent; within the 1 0 - s - 1 0 -v M range it markedly favours N A D H accumulation. The effect of rotenone on the accumulation of N A D H may be explained by assuming that at low concentration it inhibits the N A D H oxidation more than the reverse AtzH~-driven N A D + reduction. This proposal is supported by the data represented on Fig. 2. In this experiment rotenone was added to SMP, either catalyzing succinate-driven N A D + reduction (curve 1, Fig. 2) or respiring on N A D H (curve 2, Fig. 2). The inhibitor caused a syncrhonous time-dependent inhibition of both activities, yet the final rate differed in both cases; the rate of NADH-oxidase was inhibited by more than 95%, while the rate of N A D ÷ reduction was slowed only by 60%. The dependence of the steady rate of direct and reverse reactions measured as described in Fig. 2 on the rotenone concentration is

171 SMP

SNIP

A MOa0.00s

/

AMO ,0.025

t2)

T

NAD~ SUCCINATE SMP

~

i

CIDIN

ROTENONE

1 rain

Fig. 2. Time-dependent evolution of inhibition of NADH-oxidase and NAD+-reductase activities by rotenone. The trace depicts the change of absorbance at 340 nm associated with succinate-supported NAD+-reductase reaction (curve 1) and NADH-oxidase reaction (curve 2). The arrows indicate the sequential addition of SMP (25 ~g/ml), NADH (1 or 50 /zM), NAD ÷ (4 mM) plus succinate (20 mM) and rotenone (5' 10 -8 M).

T

A Noi0.02

4 min

summarized in Fig. 3. The NADH-oxidase reaction exhibits an extremely high sensitivity to rotenone (curve 2). The value of K i, calculated for the inhibition of oxidase reaction is approximately equal to 10 -9 M. In contrast to NADH-oxidase, the inhibition of the energy-linked s u c c i n a t e : N A D + - r e d u c t a s e exhibits a biphasic curve. Approx. 40% of the activity is suppressed by low concentrations (less than 10 -8 M) of

1°°I ~ e 8oii1 >_" I£, 60" <

40

20

o

IO

20

30

ROTENdNE, I O - 8 I"I

Fig. 3. Dependence of NADH-oxidase and NAD+-reductase activity of submitochondrial particles on the concentration of rotenone. NAD+-reductase (curve 1) and NADH-oxidase (curve 2) reactions were assayed as shown in Fig. 2. The steady rates of both reactions were measured 4 min after the addition of rotenone to the reaction mixture.

Fig. 4. Rotenone-induced inhibition of NADH-oxidase in the coupled vesicles. NADH-oxidase activity was monitored as described in Materials and Methods section. The addition of SMP (25 /zg/ml), rotenone ((5"10 _8 M) and gramicidin (2 p,g/ml) is indicated by arrows. BSA (10 mg/ml), where indicated, was added together with gramicidin.

rotenone, while the residual fraction of the activity (approx. 60%) is inhibited by much higher concentrations. The dual response to rotenone of the two activities carried out by the same enzyme may stem from nonidentical conditions in the two assay systems. N A D H oxidase reaction was measured in the presence of uncoupler, while the reverse reaction was assayed under tightly coupled conditions. To check whether the A/zi_i+ determines the observed differences of the two activities we examined the rotenone-induced inhibition of NADH-oxidase reaction under conditions identical to those used for measuring the reverse reaction. Submitochondrial vesicles, respiring on N A D H , were treated with rotenone. The inhibition evolved over time until after about 4 min and a steady residual rate was attained (see Fig. 4). At that point the vesicles still maintain A/z~+, which limits the rate of respiration as shown by the effect of gramicidin; the addition of uncoupler stimulates respiration 20-fold. That burst of respiration is slowed in time to a new steady rate which is practically equal to that measured in the absence of gramicidin. This delayed inhibition was attributed to the reaction of the enzyme with the free rotenone in

172 100

8O

>_I'--

40

20 ? 0

10

20

30

ROTENONE, I0 - 8 M F i g • 5 • ] ~h lbi~ ion by r o t e n o n e o f the N ~ H I

o x i d ase activity m e as I

ured with coupled and uncoupled submitochondrial vesicles. Curve 1: respiring particles were preincubated with rotenone for 4-8 min (until steady respiration was established) and the enzymatic activity was measured immediately after addition of gramicidin plus BSA to the reaction mixture (see Fig. 4). Curve 2: the steady rate of NADHoxidase was measured with uncoupled (by gramicidin) vesicles in presence of rotenone. 100% activity corresponds to 0.9 ~mol of NADH oxidized per min per mg protein.

the reaction mixture. Thus the uncoupler, which abolished zl~Ia., increased the sensitivity of the enzyme to rotenone. If the experiment is conducted in the presence of BSA, which sequesters all free rotenone, the rate measured in the presence of uncoupler is linear with time. Albumin added alone to the coupled rotenone-inhibited particles has no effect on the initial rate of respiration. This type of experiment was repeated with various rotenone concentrations and the results are summarized in Fig. 5. As seen in curve 1, in the coupled conditions about 15% of NADH-oxidase activity is insensitive to low rotenone concentrations and inhibition requires much higher concentrations. Addition of uncoupler to reaction medium leads to the transformation of the biphasic pattern of inhibition to a monophasic, highly sensitive one (see Fig. 5, curve 2), similar to that shown in Fig. 2, curve 2.

The effect of AIZH + on the affinity for rotenone The effect of A~.+ on affinity for rotenone may be due to a slower rate of rotenone binding or to the increased rate of dissociation. To evaluate the mechanism of this process we monitored the rate of rotenone dissociation from the enzyme. A concentrated suspension of particles (2 mg/ml) was preincubated in the presence of NADH with stoichiometric concentrations of rotenone. After equilibration, the samples were diluted in a reaction mixtures containing BSA and substrates. The catalysis was monitored over time, looking for the incremental rate due to the dissociation of

rotenone and its trapping by BSA. The results are summarized in Fig. 6. The release of NADH-oxidase (in presence of uncoupler) is a simple first-order reaction. On the other hand, the activation of the reverse reaction is biphasic. Approx. 70% of enzymatic activity is being liberated during a short time period (within 1 min) while the residual activity is being recovered much more slowly, the rate of both reactions reaching the level of a control sample after 15-20 min. Rapid activation of NAD +-reductase reaction is attributed to the rapid dissociation of rotenone from the enzyme in the presence of high electrochemical potential. The diminished affinity for rotenone detected in the coupled system may be due to the reduction of the quinone pool under a state of respiratory control. This possibility seems quite probable because the target of rotenone binding is located at or near the Q-binding site of the enzyme [17,19,28]. To check whether the redox state of ubiquinone has an effect on the rotenone binding we measured the sensitivity of uncoupled vesicles to rotenone when the ubiquinone pool is completely reduced. The particles were inhibited by KCN (1 mM) and mixothiazol (0.1 /xg/ml) to impose a total block of quinol oxidation, and the quinone pool was reduced by NADH (100 /xM)+ succinate (20 mM). The reduced vesicles were incubated for 4 min with rotenone; the reaction was initiated by addition of Qo and initial rate of NADH oxidation was measured

80 1 60

:2 ~

40

.<

20

.

.

.

.

i

.

.

.

.

i

.

.

.

.

0 50 100 t, sec Fig. 6. The effect of rotenone dissociation on the activities of NADH-oxidase and succinate-supported NAD+-reductase. SMP (2 mg/ml) were pretreated for 5 min at 20 ° C with 300/xM NADH. 2 min later, 1 /zM rotenone was added to the preincubation mixture. Curve 1: Enzyme sample was withdrawn 10 min after the addition of rotenone and aerobic succinate-supported NAD+-reductase (see Materials and Methods) was followed continuously with time. Curve 2: The activity of uncoupled (2 /~g/ml gramicidin) NADH-oxidase. Activities are normalized with respect to those of uninhibited particles and are plotted against time passed since addition of SMP to the reaction mixture. 100% corresponds to 0.12/~mol of NAD + reduced and 1.2/xmol of NADH oxidized per min per mg of protein.

I

173

>_-

100¢

Discussion

80

The recognition that the affinity of Complex I to rotenone varies with the magnitude of A/.~H+ is instrumental to understanding of the biphasic titration curves as shown in Figs. 3 and 5. To explain this phenomenon we assume that in uncoupled vesicles the sensitivity of Complex I is much higher (K i ~<2 nM) than in presence of A/~H+ (K i ~ 30 nM). On the basis of these affinities, associated with the state of coupling, we can explain the biphasic inhibition curves measured with submitochondrial vesicles. The submitochondrial vesicles, as prepared, consist of two subpopulations. One is inherently well coupled and exhibits low affinity to rotenone and a high rate (or efficiency) of NAD ÷ reduction. The catalytic properties of this highly coupled population can be estimated from the rates measured at rotenone concentration which inhibits the high-affinity population yet only slightly affects the low-affinity one. Indeed, at about 20 nM rotenone (see Fig. 3) we select a population representing 12% of total NADH-oxidase and 60% of total NAD +-reductase activity. The high-affinity population, that quantitated by the loss of activity at about 20 nM rotenone, is contributing about 90% of the total NADH oxidase of the vesicles but only 40% of the NAD +-reductase capacity (see Table I). The low-affinity population exhibits extremely high respiratory control. As seen in Fig. 4 addition of gramicidin to this population stimulates the respiration 20-fold. Well-known subpopulations contaminating submitochondrial vesicles are unsealed fragments or mixed-oriented vesicles [29,30], which accelerate their NADHor succinate-oxidase activity upon addition of cytochrome c. The NADH-oxidase activity of these membrane fragments isolated by centrifugation of extensively sonicated mitochondria was accelerated almost 20-fold when supplemented by cytochrome c (Kotlyar, unpublished results). As our mesaurements were car-

I->

60 o

1-40

2O

0

2

4

ROTENONE,

6

8

I0

I 0 -8 M

Fig. 7. Inhibition by rotenone of the NADH:Q-reductase measured with reduced and oxidized submitochondrial particles. (o) SMP (25 /~g/ml) were preincubated with rotenone in a reaction mixture containing: 0.25 M sucrose, 1 mg/ml BSA, 0.2 mM EDTA, 100 ,~M NADH, 20 mM succinate, 2 /~g/ml gramicidin, 1 mM CN-, 0.1 ~ g / m l mixopthiazol, 20 mM Hepes (potassium salts; pH 8.0) for 4 min at 30°C. NADH-Q reductase reaction was started by the addition of 400 ~M Qo and the initial rate of NADH oxidation was monitored. (e) the vesicles were treated with rotenone as above except that only Qo (400 /~M) was present during the incubation period.

(open symbols). In a parallel experiment, the incubation of the enzyme with rotenone was in presence of Qo (400 /xM) which lowered the reduction level of ubiquinone pool. The sensitivity to rotenone of the two systems is given in Fig. 7. In both cases the inhibition is established in the nanomolar range, independent on the redox state of the quinone, following a curve typical for uncoupled preparation (see Fig. 3, curve 2). Thus we conclude that reduction of the endogenous quinone in the membrane does not induce the state of low affinity, which is established only under conditions of high A/xr~+. TABLE I

Catalytic properties of the submitochondrial particle populations Sample

Particles a as prepared Population 1 Population 2

Sensitivity to rotenone

NADH-oxidase b activity (%)

Succinate_NAD + c reductase activity (%)

Respiratory d control

Low High

100 12 88

100 60 40

5.5 20 < 5.5

a Particles were pretreated with oligomycin (0.4/.~gfml), b NADH-oxidase reaction was assayed in the presence of gramicidin as described in Materials and Methods; 100% corresponds to the specific activity of 1.2/.tmole of NADH oxidized per min per mg of protein, c The energy-linked reduction of NAD + was measured as described in Materials and Methods; 100% corresponds to the specific activity of 0.14 /zmol of NAD + reduced per min per mg of protein, d Respiratory control is calculated as a ratio between the rates of NADH oxidation after and before the addition of gramicidin (see Fig. 4 curve 2).

174 ried out in absence of added cytochrome c, the contribution of these populations to the overall rate is negligible and cannot be identified with the less coupled population. The fact that rotenone binding is controlled by Atzn+ is an indication for a major change in the enzyme structure which follows the build-up or collapse of A/.~H+.The energy-rich state, having low affinity to rotenone, is effective in binding of the quinol molecule and its subsequent oxidation by Fe-S clusters. In contrast, the highly rotenone-sensitive non-energized state of enzyme is functioning during oxidation of NADH. In our opinion, stabilization of Complex I by rotenone in one of its conformation states may be the mechanism of inhibition.

Acknowledgements This research is supported by the US Navy N0001489-J 1622, the Israeli Ministry for Science and Technology and Koret Foundation. A.B.K. is a post-doctoral fellow of the Koret Foundation.

References 1 Hatefi, Y. and Rieske, J.S. (1967) Methods Enzymol. 10, 235-239. 2 Orme-Jonson, N.R., Hansen, R.E. and Beinert, H. (1974) J. Biol. Chem. 249, 1922-11927. 3 Albracht, S.P.J. and Subramaniam, J. (1977) Biochim. Biophys. Acta 462, 36-48. 4 Ohnishi, T. (1979) in Membrane Proteins in Energy Transduction (Capaldi, R.A., ed.), pp. 1-87, Marcel Dekker, New York. 5 Ingledew, W.J. and Ohnishi, T. (1980) Biochem. J. 186, 111-117. 6 Schatz, G. and Racker, E. (1966) J. Biol. Chem. 241, 1429-1438.

7 Lawford, H.G. and Garland, P.B. (1971) Biochem. J. 130, 10291044. 8 Ragan, C.I. and Racker, E. (1973) J. Biol. Chem. 248, 2563-2569. 9 Ragan, C.I. and Hinkle, P.C. (1975)J. Biol. Chem. 250, 8472-8480. 10 Ruzicka, F.J. and Crane, F.I. (1971) Biochim. Biophys. Acta 226, 221-223. 11 Klingenberg, M. and Slenczka, W. (1959) Biochem. Z. 331, 486517. 12 Chance, B. and Hollunger, G. (1960) Nature 185, 666-672. 13 Low, H. and Vallin, I. (1963) Biochim. Biophys. Acta 69, 361-374. 14 Hommes, F.A. (1963) Biochim. Biophys. Acta 77, 183-190. 15 Chance, B. and Hollunger, G. (1961) J. Biol. Chem. 236, 15341543. 16 Kotlyar, A.B. and Vinogradov, A.D. (1990) Biochim. Biophys. Acta 1019, 151-158. 17 Horgan, D.J., Ohno, H., Singer, T.P. and Casida, J.E. (1968) J. Biol. Chem. 243, 5967-5976. 18 Horgan, D.J. and Singer, T.P. (1967) Bicohem. J. 104, 50c-52c. 19 Horgan, D.J., Singer, T.P. and Casida, ,I.E. (1968) J. Biol. Chem. 243, 834-843. 20 Singer, T.P. and Gutman, M. (1971) Adv. Enzymol. 34, 79-153. 21 Bois, R. and Estabrook, R.W. (1969) Arch. Biochem. Biophys. 129, 362-369. 22 Van Belzen, R., Van Gaalen, M.C.M., Cuypers, P.A. and AIbracht, S.P.J. (1990) Biochim. Biophys. Acta 1017, 152-159. 23 Singer, T.P., Horgan, D.,I. and Casida, J.E. (1968) in Flavins and Flavoproteins (Yagi, K., ed.), pp. 192-215, University of Tokyo Press, Tokyo. 24 Gutman, M., Singer, T.P., Beinert, H. and Casida, J.E. (1970) Proc. Natl. Acad. Sci. USA 65, 763-770. 25 Ernster, L., Dallner, G. and Azzone, G.F. (1963) J. Biol. Chem. 238, 1124-1131. 26 Ohnishi, T. (1973) Biochim. Biophys. Acta 301, 105-128. 27 Gornall, A.G., Bardawill, C.S. and David, M.M. (1949) J. Biol. Chem. 177, 751-766. 28 Gutman, M., Coles, C.J., Singer, T.P. and Casida, J.E. (1971) Biochemistry 10, 2036-2043. 29 Huang, C.H., Keyhani, E. and Lee, C.P. (1973) Biochim. Biophys. Acta 305, 455-473. 30 Huang, C.H. and Lee, C.P. (1975) Biochim. Biophys. Acta 376, 398-414.