Mitochondrion 11 (2011) 334–341
Contents lists available at ScienceDirect
Mitochondrion j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / m i t o
Tight control of mitochondrial membrane potential by cytochrome c oxidase Consiglia Pacelli a, Dominga Latorre a, Tiziana Cocco a, Ferdinando Capuano a, Christian Kukat b, Peter Seibel b, Gaetano Villani a,⁎ a b
Department of Medical Biochemistry, Biology and Physics, University of Bari, 70124 Bari, Italy Molecular Cell Therapy, Center for Biotechnology and Biomedicine, University of Leipzig, 04103 Leipzig, Germany
a r t i c l e
i n f o
Article history: Received 26 July 2010 Received in revised form 24 November 2010 Accepted 3 December 2010 Available online 13 December 2010 Keywords: Mitochondria Cytochrome oxidase Respiration Membrane potential Flux control Cyanide
a b s t r a c t In the present work we have critically examined the use of the KCN-titration technique in the study of the control of the cellular respiration by cytochrome c oxidase (COX) in the presence of the mitochondrial membrane potential (Δψmito) in HepG2 cells. We clearly show that the apparent high inhibition threshold of COX in the presence of maximal Δψmito is due to the KCN-induced decrease of Δψmito and not to a low control of COX on the mitochondrial respiration. The tight control exerted by COX on the Δψmito provides further insights for understanding the pathogenetic mechanisms associated with mitochondrial defects in human neuromuscular degenerative disorders. © 2010 Elsevier B.V. and Mitochondria Research Society. All rights reserved.
1. Introduction Mitochondrial oxidative phosphorylation (OXPHOS) represents the major source of ATP in mammalian cells relying on aerobic energy metabolism. The mitochondrial respiratory complexes I (NADH:ubiquinone oxido-reductase), III (ubiquinol:ferricytochrome c oxido-reductase) and IV (cytochrome c oxidase, COX) build up a transmembrane proton electrochemical potential (ΔμH+) by coupling their electron transfer activities to H+-translocation from the matrix (negative, N) to the outer (positive, P) side of the inner mitochondrial membrane. The mitochondrial transmembrane ΔμH+ is composed of a proton gradient (ΔpHmito) and an electric membrane potential (Δψmito), with the latter being the major component under physiological conditions (Cohen et al., 1978; Ferguson and Sorgato, 1982). The ΔμH+ exerts a strong control on the respiratory ﬂux and is utilized backward for ATP synthesis by the OXPHOS complex V (ATP synthase). The three redox-driven proton pumps and the ATP synthase contain subunits encoded by the mitochondrial DNA that are essential for their biogenesis and function (Attardi and Schatz, 1988). The threshold effects, i.e., the capacity of an heteroplasmic mitochondrial Abbreviations: MCA, metabolic control analysis; COX, cytochrome c oxidase; ΔμH+, proton electrochemical potential; Δψmito, mitochondrial electric membrane potential; OXPHOS, oxidative phosphorylation; TMPD, N,N,N′,N′-tetramethyl-p-phenylenediamine; DNP, 2,4-dinitrophenol; CCCP, carbonyl cyanide 3-chlorophenylhydrazone; GM, glutamate + malate; AT, ascorbate + TMPD. ⁎ Corresponding author. Department of Medical Biochemistry, Biology and Physics, Piazza G. Cesare, 11, 70124, Bari, Italy. Tel.: +39 080 5448534; fax: +39 080 5448538. E-mail address: [email protected]
DNA mutation to produce an OXPHOS defect (Boulet et al., 1992; Chomyn et al., 1992; Hayashi et al., 1991; Yoneda et al., 1994), have called attention to the control that a given step exerts on the OXPHOS pathway. The most utilized method to analyze the control exerted by a single enzymatic step on the whole metabolic ﬂux is the inhibitor titration technique (Groen et al., 1982). In this approach, an enzymatic activity is titrated by a speciﬁc inhibitor both as isolated step and as a step integrated in the metabolic pathway. In this way, at each concentration of inhibitor used, the percentage of inhibition of the isolated enzymatic step and how this speciﬁc inhibition affects the overall pathway can be measured. The reliability of this method depends on how comparable the conditions are when titrating the enzyme as isolated or integrated step, respectively, and on whether or not a re-distribution of the control to steps not directly participating to the metabolic ﬂux occurs. In particular, the utilization of the KCN-titration technique in intact cells has revealed a tight in vivo control exerted by COX on the cellular respiration (Villani and Attardi, 1997; Villani et al., 1998). Most of the experiments on intact cells have been, however, carried out in the presence of OXPHOS uncouplers in order to work under maximal respiratory rates by endogenous substrates, as well as to avoid the interference of the variable ΔμH+ on the same ﬂuxes during the inhibitor titration. As an extension of the above in vivo studies, the control of respiration by COX has been recently analysed by the KCN-titration technique in HepG2 cells in the presence of speciﬁc ionophores and/or OXPHOS inhibitors known to modulate the mitochondrial membrane potential (Δψmito) (Dalmonte et al., 2009; Piccoli et al., 2006).
1567-7249/$ – see front matter © 2010 Elsevier B.V. and Mitochondria Research Society. All rights reserved. doi:10.1016/j.mito.2010.12.004
C. Pacelli et al. / Mitochondrion 11 (2011) 334–341
The aim of the present work was to carefully monitor the dynamic changes of Δψmito in response to the KCN inhibition of COX in HepG2 cells. The low control of COX on cellular respiration in the presence of maximal Δψmito is only apparent because the high inhibition thresholds are due to the KCN-induced decrease of Δψmito. The tight control exerted by COX on the mitochondrial respiration and on the redox-driven Δψmito has been also conﬁrmed by carrying out the same analysis in digitonin-permeabilized HepG2 cells where a high ΔμH+ (state IV) has been obtained by means of a physiological ATP-regenerating system. 2. Materials and methods 2.1. Cell and culture conditions The HepG2 human hepatoma cell line (ATCC HB-8065) was grown in Dulbecco's modiﬁed Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS), 1% (v/v) L-glutamine, 1% (v/v) penicillin/streptomycin, at 37 °C in a humidiﬁed atmosphere of 5% CO2. Cells were collected by trypsinization at 70–80% conﬂuence. 2.2. Measurement of oxygen consumption rates in HepG2 cells The respiratory activities of intact and digitonin-permeabilized HepG2 cells were measured polarographically with a Clark-type oxygen electrode in a water-jacketed chamber, (Hansatech Instruments, Norfolk, UK), magnetically stirred at 37 °C and 25 °C respectively, as previously described (Villani and Attardi, 1997, 2001). For the measurement of respiration rates in intact or permeabilized cells, exponentially growing cells were collected by trypsinization and centrifugation, and then transferred into the polarographic chamber at about 1 × 107 cells per ml in TD Buffer (0.137 M NaCl, 5 mM KCl, 0.7 mM Na2HPO4, 25 mM Tris–HCl, pH 7.4), or buffer A (75 mM sucrose, 5 mM KH2PO4, 40 KCl, 0.5 EDTA, 3 mM MgCl2, 30 mM Tris–HCl, pH 7.4), respectively. Cell permeabilization was obtained by adding digitonin at the optimal concentration of 10 μg/106 cells (Villani et al., 1998) directly into the oxygraphic chamber. For the analysis of COX activity as respiratory chain integrated step in permeabilized cells, 5 mM glutamate + 5 mM malate (GM) were utilized as exogenous NAD-dependent respiratory substrates. When studying COX activity as in situ isolated step, TMPD (0.4 mM) was used as membrane-permeant electron donor for the endogenous cytochrome c pool, in the presence of ascorbate (10 mM) as the primary reducing agent and antimycin A (40 nM) as upstream electron ﬂux inhibitor. The ATPase inhibitor oligomycin was used both on intact and permeabilized cells at the concentration of 1 μg/ml. The KCN-titration experiments of respiratory ﬂuxes in oligomycinblocked permeabilized HepG2 cells were essentially carried out as previously described (Villani and Attardi, 2001). After recording the oligomycin-blocked respiration rate, KCN was added at a given concentration and the oxygen consumption recorded until a constant slope was obtained. After blocking the respiratory ﬂux by the addition of antimycin A, AT (ascorbate+ TMPD) were added to measure the KCNinhibited COX activity as initial rate, corrected by the KCN-insensitive auto-oxidation rate of TMPD, as determined in a separate experiment (Villani and Attardi, 2001). Under our experimental conditions, full uncoupling of the endogenous respiration of intact cells could be equally obtained by the addition of 30 μM DNP (2,4-dinitrophenol), or of 0.25 μM CCCP (carbonyl cyanide 3-chlorophenylhydrazone) as well as of 1.4 ng/ 106 cells valinomycin. For the state IV respiration in permeabilized cells, Buffer A was supplemented with an ATP-regenerating system composed of ATP 0.5 mM, phosphoenolpyruvate (PEP) 1 mM, pyruvate kinase (PK) 20 U/ml. The state IV respiration rate measured in the presence of the ATP-regenerating system was not changed by further addition of oligomicyn.
2.3. Measurement of Δψmito in permeabilized HepG2 cells Mitochondrial Δψ was measured in digitonin-permeabilized cells by following the safranin ﬂuorescence quenching at 525 nm (excitation), 575 nm (emission) (Akerman and Saris, 1976; Di Paola et al., 2000). All determinations were performed with a Jasco FP6200 spectroﬂuorimeter utilizing a quartz cuvette (1 cm path length) magnetically stirred at 25 °C. Cells were collected by trypsinization and resuspended in Buffer A at 1 × 106 cells/ml. The cell suspension was transferred to the cuvette and supplied with digitonin (10 μg/106 cells) for cell permeabilization. After 2 min, 2 μM safranin was added and the formation of the Δψmitodependent ﬂuorescent signal was induced by the respiratory substrates as indicated. The maximal (100%) Δψmito-dependent ﬂuorescent signal of safranin, corresponding to the static-head membrane potential, was measured upon addition of oligomycin (1 μg/ml), or by supplementing the cell suspension with the ATP-regenerating system (ATP 0.5 mM, PEP 1 mM, PK 20 U/ml) as indicated, and was the same under both the experimental conditions. Full dissipation (0%) of the Δψmito-dependent ﬂuorescent signal of safranin could be equally obtained by addition of 30 μM DNP, or of 0.25 μM CCCP as well as of 1.4 ng/106 cells valinomycin. The ﬂuorescence variations measured in our experiments fell in the range of linear response of the probe, as calibrated with potassium diffusion potentials of known magnitude in deenergized rat liver mitochondria (Akerman and Wikstrom, 1976) (data not shown). 3. Results 3.1. Control of respiratory ﬂuxes by the mitochondrial membrane potential in intact cells The respiratory ﬂux control by COX has been previously analysed in HepG2 cells by carefully optimizing the experimental setup for the measurement of respiratory capacities in intact and digitonin-permeabilized cells (Villani et al., 1998). The respiration rate by endogenous substrates of HepG2 cells has been previously shown to increase by about 65% upon addition of the uncoupler DNP (Villani et al., 1998) (see also (Desquiret et al., 2006)), thus showing that a steady-state mitochondrial ΔμH+ controls the respiratory ﬂux of intact cells under basal conditions. An immediate estimate of the control exerted by the mitochondrial ΔμH+ on the respiratory ﬂuxes can be obtained by measuring the relative changes of the respiration rates (respiratory control ratios, RCRs) in response to cell-permeant OXPHOS inhibitors known to maximize (oligomycin) or collapse (uncouplers) the mitochondrial ΔμH+. As shown in Fig. 1, the endogenous respiration rate of intact HepG2 cells was 3.8-fold lower (RCRbas/oligo) than the basal one, when the mitochondrial ATP synthase was blocked by oligomycin, i.e., under conditions of maximal (static-head) mitochondrial ΔμH+. When the oligomycin-blocked ΔμH+ was removed by the addition of DNP, the respiration rate reached a value that is 6.5-fold (RCRunc/oligo) and 2.0-fold (RCRunc/bas) higher as compared with the oligomycin-blocked and the basal endogenous respiration rates, respectively. The RCRunc/bas did not change if the uncoupler was added directly to naive cells (cf. (Desquiret et al., 2006; Villani et al., 1998)). Under our experimental conditions, a similar increase of the basal, or of the oligomycin-blocked respiration rate by endogenous substrates could be obtained by the addition of saturating amounts of DNP, CCCP or valinomycin. The digitonin-permeabilization of HepG2 cells preserved the integrity of the mitochondrial inner membrane as indicated by the RCRunc/oligo of 6.2, value very close to the one measured in intact cells. Therefore, the range of control of the GM-elicited respiratory ﬂux by the mitochondrial ΔμH+ was comparable in intact and digitoninpermeabilized cells (Fig. 1). The addition of oligomycin to permeabilized cells respiring on NAD-dependent substrates under phosphorylating conditions (state III), resulted in a 4.9-fold decrease of the state III respiration rate (RCRstate III/oligo). The inhibitory effect of oligomycin on the state III respiration rate was lower as compared to that exerted on the
C. Pacelli et al. / Mitochondrion 11 (2011) 334–341
Fig. 1. Respiratory control ratios (RCRs) in intact and permeabilized HepG2 cells. Oxygen consumption rates by endogenous substrates in intact cells were measured directly (basal respiration), after addition of 1 μg/ml oligomycin (oligomycin-blocked respiration) and of 30 μM DNP (uncoupled respiration). bas/oligo, ratio of basal vs oligomycin-blocked endogenous respiration rate; unc/oligo, ratio of uncoupled vs oligomycin-blocked endogenous respiration rate; unc/bas, ratio of uncoupled vs basal endogenous respiration rate. The data shown represent the means ± EMS from 3 to 7 independent determinations. GM-dependent oxygen consumption rates in digitoninpermeabilized cells, were measured in the presence of 0.5 mM ADP (state III), after addition of 1 μg/ml oligomycin (oligomycin-blocked) and, ﬁnally, after addition of 30 μM DNP (uncoupled). state III/oligo, ratio of state III vs oligomycin-blocked respiration rate; unc/oligo, ratio of uncoupled vs oligomycin-blocked respiration rate; unc/state III, ratio of uncoupled vs state III respiration rate. The data shown represent the means ± EMS from 4 to 10 independent experiments.
basal endogenous respiration rate of intact cells (RCRbas/oligo). On the other hand, the addition of DNP resulted in a lower increase of the state III respiration rate (RCRunc/state III =1.3) in permeabilized cells as compared to the stimulatory effect of the uncoupler on the basal endogenous respiration rate of intact cells (RCRunc/bas =2). Taken together, these observations would suggest that, under basal conditions, intact HepG2 cells respire at a rate that is slightly below the phosphorylating state III value. 3.2. Measurement of Δψmito in digitonin-permeabilized cells The ﬂuorescent probe safranin has been utilized for the direct measurement of the Δψmito in digitonin-permeabilized HepG2 cells. Fig. 2 shows a time scan of the response of mitochondrial respiratory ﬂuxes and Δψmito to different OXPHOS substrates and inhibitors in digitonin-permeabilized HepG2 cells. The polarographic and spettroﬂuorimetric measurements have been conducted in parallel and the two tracings superimposed. As expected, the addition of NADdependent respiratory substrates to permeabilized cells resulted in a slow respiration rate (state IV) and in the parallel formation of a transmembrane Δψmito. It should be noted that the response time of the Δψmito ﬂuorescent signal is limited by the equilibration kinetics of the probe. The state III respiration elicited by the addition of ADP, was associated with a limited drop (− 20 ± 2.4 %, n = 9) of the Δψmito that was brought up to its maximal value (set as 100 %) by oligomycin. As shown in Fig. 2, the extent of the state IV Δψmito was comparable (95 ± 0.8 %, n = 9) to the maximal Δψmito. The block of the respiratory ﬂux by the complex III inhibitor antimycin A caused the almost complete dissipation of the Δψmito. At this point, the antimycin Ainduced block of the oxygen consumption could be bypassed by the artiﬁcial complex IV substrate composed of TMPD as membranepermeant one-equivalent electron donor plus ascorbate as primary reducing agent (AT). The AT-elicited oxygen consumption was associated with the production of a Δψmito that amounts to 64% ± 4.3% (n = 9) of the maximal one. The AT-supported potential is dependent on the proton motive activity of COX as demonstrated by the fact that it is dissipated by the subsequent addition of a saturating
Fig. 2. Parallel measurements of ΔΨmito and oxygen consumption rates in permeabilized HepG2 cells. Representative tracings of the parallel measurements of oxygen consumption rates (grey line) and ΔΨmito (black line) in digitonin-permeabilized cells. GM, glutamate (5 mM)+ malate (5 mM); ADP, 0.5 mM ADP; Oligo, 1 μg/ml oligomycin; Ant, 40 nM antimycin A; AT, ascorbate (10 mM) + TMPD (0.4 mM); KCN, 2 mM KCN; CCCP, 0.25 μM CCCP. For details see under Section 2.
amount of KCN, also resulting in the inhibition of the COX-dependent oxygen consumption. The loss of the AT-supported Δψmito was complete, since further addition of the uncoupler CCCP had no effect on the safranin signal. This also demonstrates that the chemical oxygen consumption due to the auto-oxidation of TMPD does not contribute to the observed AT-elicited Δψmito. 3.3. Control of oligomycin-blocked respiratory ﬂuxes by COX in digitonin-permeabilized cells The control exerted by COX on the mitochondrial respiration in the presence of maximal Δψmito has been analysed by the inhibitor titration technique. For this purpose, a KCN-titration of COX activity either as respiratory chain integrated step (GM-dependent respiration) or as isolated step (AT as electron donors and antimycin A as upstream inhibitor), has been carried out in oligomycin-blocked permeabilized HepG2 cells. The results are illustrated in Fig. 3 showing a threshold plot, i.e., the plot of the relative GM-dependent respiration rate against the percentage of inhibition of isolated COX activity at each KCN concentration. The dependence of the integrated respiratory chain ﬂux on COX activity was bi-phasic and a threshold value of 76% could be obtained as the intersection of the least-square regression lines of the two inhibition slopes. This result was in agreement with that obtained by the same analysis in intact HepG2 cells (Piccoli et al., 2006), thus conﬁrming that digitonin-permeabilized cells can represent a valid tool for the study of the metabolic control analysis of mitochondrial bioenergetics (Kunz et al., 2000; Villani et al., 1998). 3.4. Flux–force relationship In order to obtain a more detailed analysis of the control exerted by the Δψmito (force) on the mitochondrial respiration (ﬂux), we have carried out a valinomycin titration of both parameters in oligomycinblocked permeabilized cells. Valinomycin is an ionophore that dissipates the mitochondrial membrane potential by virtue of its K+-uniport properties. Fig. 4 shows representative tracings of the parallel measurements of the valinomycin-induced variations of the mitochondrial respiratory ﬂux and static-head Δψmito supported either by GM (Fig. 4A) or AT (Fig. 4B). As expected (cf. Fig. 2), the maximal ATdependent Δψmito was lower as compared with the GM-dependent one (see legend to Fig. 4C). The membrane potential decreased in response to increasing concentrations of valinomycin and the full collapse of Δψmito
C. Pacelli et al. / Mitochondrion 11 (2011) 334–341
Δψmito, when utilizing AT as electron donors for the KCN-titration of the antimycin A-isolated oxygen consumption ﬂux via COX. 3.5. KCN-titration of static-head Δψmito in oligomycin-blocked permeabilized cells Following the preliminary observation that saturating concentrations of KCN could fully collapse the AT-elicited Δψmito in oligomycin-blocked
Fig. 3. Control of respiratory ﬂuxes by COX in oligomycin-blocked permeabilized HepG2 cells. The threshold graphic of COX activity was obtained by plotting the relative GMdependent respiration rate against the percentage of inhibition of the isolated COX activity at the same KCN concentrations. The indicated threshold value (arrow) was obtained by the intersection of the least-square regression lines of the two inhibition slopes. Inset KCN-titration curves of the GM (open symbols) – and the AT (closed symbols) – dependent respiratory ﬂuxes in oligomycin-blocked permeabilized cells. For details see under Section 2. The data shown represent the means ± EMS from 3 to 8 independent experiments; the error bars that fall within the individual data symbols are not shown.
was obtained at an ionophore concentration of 1.4 ng/106 cells with both respiratory substrates. No further variation of the ﬂuorescent signal was observed upon addition of DNP (or CCCP). On the other hand, valinomycin showed very different effects on the two respiratory ﬂuxes. In fact, while the GM-dependent respiratory ﬂux was highly stimulated by the addition of valinomycin, the AT-supported oxygen consumption rate, with a starting rate 5.6-fold higher as compared with the GM-dependent one, was poorly enhanced at similar ionophore concentrations. The ﬂux (JO2)– force (Δψmito) relationship is better illustrated by the graphic in Fig. 4C, obtained by plotting the relative respiration rates against the percent inhibition of the Δψmito at each valinomycin concentration. A linear dependence of the respiratory ﬂuxes on the Δψmito could be observed with both respiratory substrates. However, upon full dissipation of the Δψmito, while the GM-dependent respiration rate could be increased up to about 6-fold as compared with the oligomycin-blocked rate, the AT-dependent oxygen consumption rate showed only a relatively slight increase of 1.7fold. Therefore, the Δψmito exerts a much stronger control on the respiratory chain activity elicited by physiological NAD-dependent substrates as compared with the artiﬁcial AT-supported oxygen consumption via COX. This difference would lead per se to overestimate the COX inhibition threshold in the case of any KCN-induced decrease of the
Fig. 4. Control of respiratory ﬂuxes by Δψmito in oligomycin-blocked permeabilized HepG2 cells. Representative tracings of parallel valinomycin-titrations of GM- (A) or AT- (B) dependent oxygen consumption rates (grey trace) and ΔΨmito (black trace) in digitonin-permeabilized cells. Cells resuspended in buffer A supplemented with ADP (0.5 mM) and oligomycin (1 μg/ml) were incubated for two minutes with digitonin (10 μg/106 cells). Oxygen consumption rates and ΔΨmito were measured after addition of the respiratory substrates. GM, glutamate (5 mM) + malate (5 mM); AT, ascorbate (10 mM) + TMPD (0.4 mM), in the presence of 40 nM antimycin A; Val, 0.4 ng/106 cells valinomycin; DNP, 30 μM DNP; KCN, 2 mM KCN. (C) Flux–force plot. The relationship between respiratory ﬂuxes (JO2) and ΔΨmito by GM (open symbols) and AT (close symbols) is obtained by plotting the respiration rates expressed as percent of the oligomycin-blocked rates (100% GM = 0.9 ± 0.06 nmoles O2/min/106 cells; 100% AT = 4.9 ± 0.2 nmoles O2/min/106 cells) against the ΔΨmito expressed as percent of the maximal uncoupler-sensitive ΔΨmito at each valinomycin concentration (100% GMdependent ΔΨmito = 56.9 ± 3.9 ﬂuorescence A.U.; 100% AT-dependent ΔΨmito = 37.5 ± 1.5 ﬂuorescence Α.U.). Data represent the means ± EMS from 3 independent experiments (the error bars that fall within the individual data symbols are not shown).
C. Pacelli et al. / Mitochondrion 11 (2011) 334–341
permeabilized cells (Fig. 2), we wanted to analyse the dynamic response of Δψmito to the range of the inhibitor concentrations utilized for the KCN-titration of COX activity illustrated in Fig. 3. As shown in the representative tracings in Fig. 5A, KCN produced a step-wise decrease of both the GM- and AT-dependent Δψmito very similar to that obtained upon addition of valinomycin. Similar concentrations of KCl did not produce any change on the Δψmito-dependent safranin signal with both respiratory substrates (data not shown). The KCN-titration curve of the AT-dependent Δψmito in oligomycin-blocked permeabilized cells was quasilinear at subsaturating concentrations of the inhibitor (Fig. 5B). On the other hand, the variation of the GM-dependent Δψmito over the same range of KCN concentrations produced a curve that was clearly sigmoidal in shape (Fig. 5B). Interestingly, the GM-dependent Δψmito was completely dissipated at KCN concentrations of 200–250 μM, at which the respiratory rates sustained by the same substrates and by AT were inhibited by about 50% and 85%, respectively (see Fig. 3). Any aspeciﬁc inﬂuence of KCN on the safranin signal was also excluded, since the ATP-
elicited Δψmito, due to the ATP hydrolase activity of complex V, was completely insensitive to the COX inhibitor (data not shown). 3.6. KCN-dependent loss of the in vivo control by Δψmito on respiratory ﬂuxes in oligomycin-blocked intact cells The gradual decrease of Δψmito due to the KCN inhibition of COX as respiratory chain integrated step in oligomycin-blocked cells should correspond to a lower control exerted by the residual Δψmito on the respiratory ﬂux. This could be proved in intact cells by measuring, at each concentration of KCN utilized for the titration experiments, the increase of the residual KCN-inhibited respiration rate obtained by the addition of a saturating amount of valinomycin (Fig. 6). Indeed, the stimulatory effect of valinomycin on the endogenous respiration rate of oligomycin-blocked intact cells declined at increasing concentrations of KCN as shown in Fig. 6, and was completely lost at 300 μM KCN, i.e. at a concentration shown to cause full dissipation of Δψmito (Fig. 5). Very similar results were obtained by carrying out the same analysis in permeabilized oligomycin-blocked cells respiring on GM (data not shown). 3.7. KCN-titration of Δψmito and respiratory ﬂuxes in permeabilized cells under state IV conditions A more physiological model of static-head Δψmito was obtained by exposing permeabilized cells respiring on NAD-dependent substrates to the phosphoenolpyruvate (PEP)/pyruvate kinase (PK) ATP-regenerating system. As shown in Fig. 7A, the addition of ATP resulted in a small decrease of the GM-dependent Δψmito since the ADP, mostly formed by the cellular ATP-consuming processes, could be cycled to ATP by OXPHOS, with consequent utilization of the Δψmito. As expected, the further addition of PEP and PK brought back the Δψmito to its maximal value. However, the overall Δψmito was only in part sensitive to KCN due to the variable fraction contributed by the ATP hydrolase activity of complex V that could be monitored through its sensitivity to oligomicyn. During the KCN inhibition of the respiratory chain, the oligomycinsensitive Δψmito somewhat compensated for the loss of the redox-driven
Fig. 5. KCN-titration of Δψmito in permeabilized oligomycin-blocked HepG2 cells. A. Representative tracings of spectroﬂuorimetric measurement of KCN-titration of Δψmito elicited by GM (grey trace) or AT (black trace) in permeabilized oligomycin-blocked HepG2 cells. After establishment of static-head membrane potential, discrete concentrations (μM) of KCN were added as indicated. GM, glutamate (5 mM)+malate (5 mM); AT, ascorbate (10 mM)+TMPD (0.4 mM), in the presence of 40 nM antimycin A; CCCP, 0.25 μM CCCP; KCN, KCN. For experimental details, see under Section 2. B. KCN-titration curves for GM—(open symbols; 100%=56.2±3.5 ﬂuorescence A.U.) and AT—(close symbols; 100%=37.3±2.3 ﬂuorescence A.U.) supported Δψmito. Values are expressed as percentage of the maximal uncoupler-sensitive Δψmito. The data represent the mean values±EMS from 3 independent experiments (the error bars that fall within the individual data symbols are not shown).
Fig. 6. KCN-induced loss of the control by Δψmito on endogenous respiration in oligomycin-blocked HepG2 intact cells. Oxygen consumption measurements were carried out in intact cells as described in Section 2. The RCRVal values represent the ratios of the endogenous respiration rates of oligomycin-blocked cells measured after the addition of valinomycin (1.4 ng/106 cells) vs the endogenous respiration rate measured before the addition of valinomycin, at different KCN concentrations. The data represent the means ± EMS from 3 independent experiments. The inset shows a representative tracing of the effect of valinomycin on oligomycin-blocked endogenous respiration rates in the absence (black line) and in the presence of 150 μM KCN (grey line). Cells, 1 × 107/ml cells; Oligo, 1 μg/ml oligomycin; Val, 1.4 ng/106 cells valinomycin; Ant, 40 nM antimycin A.
C. Pacelli et al. / Mitochondrion 11 (2011) 334–341
Fig. 7. KCN-titration of state IV respiration and ΔΨmito in permeabilized HepG2 cells. The state IV was obtained by an ATP-regenerating system as explained in Section 2. A. Representative tracing of spectroﬂuorimetric measurement of ΔΨmito. GM, glutamate (5 mM)+ malate (5 mM); ATP, 0.5 mM adenosine-5′-triphosphate; PEP, 1 mM Phosphoenolpyruvate; PK, 20 U/ml pyruvate kinase; KCN 100, 100 μM KCN; oligo, 1 μg/ml oligomicyn; CCCP, 0.25 μM CCCP. B. KCN-titration curves of GM-supported ΔΨmito. The values of the ΔΨmito (close symbols; see the representative a-value in panel A for 100 μM KCN) and of the fraction of the oligomycin-sensitive ΔΨmito contributed by the ATP hydrolase activity (open symbols; see the representative b-value in panel A for 100 μM KCN) after each KCN concentration, are expressed as percentage of the maximal uncoupler-sensitive ΔΨmito (see 100% in panel A). The data represent the mean values ± EMS from 4 to 6 independent experiments. C. KCN-induced loss of the control by Δψmito on oligomycin-blocked (close symbols) and state IV (open symbols) respiratory ﬂuxes in permeabilized cells. The RCRVal values have been calculated as above described (see Fig. 6 and Section 2) and expressed as percentage values of the maximal-fold increase obtained by the addition of valinomycin in the absence of KCN. The data represent the mean values± EMS from 4 to 6 independent experiments. *pb 0.05. D. KCN-titration curves of GM-dependent respiratory ﬂuxes in oligomycin-blocked (close symbols; 100%= 0.86 ± 0.02 nmoles O2/min/106 cells) and state IV (open symbols; 100% = 0.80± 0.03 nmoles O2/min/106 cells) permeabilized cells. The inset shows the threshold graph of COX activity for state IV respiration. The data represent the mean values ± EMS from 3 to 8 independent experiments; the error bars that fall within the individual data symbols are not shown. *pb 0.05.
Δψmito until reaching a steady-state value of 44% of the maximal Δψmito at saturating concentrations of KCN (Fig. 7B). Accordingly, a signiﬁcantly higher residual control of the GM-dependent respiratory ﬂux by Δψmito was maintained even at saturating KCN concentrations, as compared with oligomycin-blocked cells (Fig. 7C). Furthermore, under these conditions, the GM-dependent respiratory ﬂux became signiﬁcantly more sensitive to KCN (Fig. 7D), as compared with oligomycin-blocked cells, thus leading to a lower apparent inhibition threshold of COX (68%; see inset in Fig. 7D). 4. Discussion In the present paper we show that: 1) the AT-dependent isolated COX activity is not suitable to study the control of cellular respiration by COX in the presence of the Δψmito by the inhibitor titration technique; 2) the apparent COX thresholds obtained by the inhibitor titration technique in oligomycin-blocked or state IV respiring HepG2 cells are inﬂuenced by the KCN-induced variations of the Δψmito; 3) COX exerts a tight control on the respiration-driven Δψmito; 4) the limiting role of COX on mitochondrial respiration does not depend on the mitochondrial energy state. The mitochondrial respiratory chain builds up and is in its turn inhibited by the transmembrane ΔμH+. This represents a strong limitation for the use of the MCA to the coupled mitochondrial respiration by the inhibitor titration technique. In fact, under coupled
conditions, i.e. when the permeability of the inner mitochondrial membrane is preserved, any variation of the respiration rate in consequence of the addition of a speciﬁc inhibitor, would result from a new dynamic equilibrium between the electron transfer activities of the respiratory complexes and the mitochondrial ΔμH+. Therefore, MCA in intact cells has been carried out in the presence of uncouplers, both to work under maximal respiratory ﬂuxes and to avoid the strong control exerted by the membrane potential on the respiratory ﬂuxes (Villani and Attardi, 1997; Villani et al., 1998). Recently, the in vivo analysis of the respiratory ﬂux control by COX has been extended by applying the inhibitor titration technique to HepG2 cells in the presence of speciﬁc ionophores or OXPHOS inhibitors known to modulate the mitochondrial electric membrane potential (Δψmito) (Dalmonte et al., 2009; Piccoli et al., 2006). Unfortunately, the direct analysis of the response of the Δψmito to the different drugs and/or inhibitors was not detailed, or even missing. In particular the so-called state IV (or state II) respiration, highly controlled by a maximal Δψmito, has been mimicked by blocking the mitochondrial OXPHOS complex V with its speciﬁc inhibitor oligomycin that prevents the utilization of the membrane potential for ATP synthesis (Piccoli et al., 2006). The Δψmito has been monitored in intact cells by laser-scanning confocal microscopy using the probe MitoCaptureTM (Piccoli et al., 2006). In this way, the Δψmito could be barely detected under basal
C. Pacelli et al. / Mitochondrion 11 (2011) 334–341
conditions, with this result being in contrast with the evidence that most of naive HepG2 cells have a high Δψmito as measured, in vivo, by ﬂow cytometric analysis after JC-1 staining (Ferraresi et al., 2004; Pinti et al., 2003). Using the same imaging technique, the ﬂuorescent signal of oligomycin-blocked HepG2 cells did not change up to 300 μM KCN, while only a slight decrease could be detected at 1 mM KCN (Piccoli et al., 2006). The above discrepancies could be explained by the common limitations relative to the use of imaging techniques for the quantitative measurements of the Δψmito with ﬂuorescent potentiometric probes in intact cells (Bernardi et al., 1999). In the particular case of dual-emitting potential probes such as JC-1 or Mitocapture, attention should be paid when quantifying green-emission signals (as in the case of (Piccoli et al., 2006)), since the green ﬂuorescence derives from the passive binding of the dye monomer to any cellular membrane. Furthermore, the need of measuring respiratory ﬂuxes by polarography and Δψmito by image analysis on cell suspensions and attached cells, respectively, poses important restrictions on the possibility of monitoring in the same experimental conditions the parallel changes of the two parameters in response to OXPHOS inhibitors. To bypass these problems, we have utilized the ﬂuorescent probe safranin to measure dynamic changes of Δψmito in digitoninpermeabilized HepG2 cells. The degree of control of the respiration by Δψmito is comparable when analysed in intact and permeabilized HepG2 cells respiring on endogenous and NAD-dependent substrates, respectively (Fig. 1). Similarly, the control exerted by Δψmito on the AT-sustained oxygen consumption is very poor, as compared with that exerted on the respiration rate elicited by the physiological NADdependent substrates and by endogenous substrates in permeabilized (Fig. 4) and intact cells (Piccoli et al., 2006), respectively. This could be due to the lower Δψmito elicited by AT as compared with the one measured with NAD-dependent substrates, as well as to the peculiar chemistry of the electron transfer from the primary reductant ascorbate to cytochrome c / COX, via TMPD (Crinson and Nicholls, 1992; Ferguson-Miller et al., 1978). At any event, this observation pinpoints a fundamental limitation to the utilization of the standard KCN-titration technique in the presence of Δψmito, since this would differentially affect the respiratory ﬂuxes depending on whether COX is analysed as an integrated respiratory chain complex or as an isolated step. Obviously, such a limitation becomes more important if Δψmito changes during the KCN-titration experiments, as clearly shown in Fig. 5. The difference between the KCN-titration curves of the GM- and the AT-elicited Δψmito adds a further restriction to the MCA analysis of COX by the inhibitor titration technique in oligomycin-blocked cells. Importantly, the complete suppression of the GM-dependent Δψmito is obtained at KCN concentrations close to the COX inhibition threshold. Therefore, the Δψmito decreases during the apparent lowcontrolling phase (ﬁrst linear slope) of the KCN-titration curve of the respiratory ﬂux shown in Fig. 3. This is functionally conﬁrmed in oligomycin-blocked intact cells by the fall of the control of the respiratory ﬂux by Δψmito during the COX inhibitor titration, as evidentiated by the KCN-dependent decline of the stimulatory effect of valinomycin on the endogenous respiration rate (Fig. 6). It is worth noting that, when comparing the independent results obtained by KCN-titration of the GM-dependent Δψmito (Fig. 5) and of the stimulatory effect of valinomycin (Fig. 6), the complete loss of Δψmito and, therefore, of its functional control on respiration, is practically lost at the same KCN concentration (≈250 μM). The presence of oligomycin is far from mimicking a high cellular energetic (or resting) state, especially in cells with an efﬁcient glycolytic metabolism such as HepG2 cells (Akerman and Jarvisalo, 1980). In fact, oligomycin also blocks the ATP hydrolase activity of complex V that would be activated by the high ATP/ADP ratio (Nicholls, 2006) and by the inhibition of the respiratory chain as recently shown in cells containing pathogenic mtDNA mutations
(Abramov et al., 2010). We have therefore conﬁrmed our results by carrying out the same analysis in a physiological state IV obtained by an ATP-regenerating system. Under these conditions, the ATP hydrolase activity partially compensates the KCN-induced loss of Δψmito and the GM-dependent respiratory ﬂux becomes more controlled by COX (Fig. 7). Indeed, the possible contribution of the allosteric inhibition of COX by ATP in the control of cellular respiration could also be envisaged (Ramzan et al., 2010). When applying the inhibitor titration technique to the MCA of the mitochondrial respiratory chain under oligomycin-blocked or coupled conditions, the constancy of the membrane potential should be monitored to make sure that a re-distribution of the control between the respiratory chain and the membrane potential does not take place. This would, in fact, mask the real effect of the speciﬁc inhibitor on the respiratory ﬂux. Actually, even an unchanged membrane potential, upon addition of a respiratory chain inhibitor, could be explained by a compensatory effect of the partially inhibited respiratory ﬂux. In the case of oligomycin-blocked or state IV respiration, condition under which the respiratory ﬂux is low and the respiratory capacity is at its maximal value, the partial inhibition of a speciﬁc respiratory complex (which is not yet rate-limiting), if associated with a temporary decrease of the membrane potential, would immediately lead to a compensatory increase of the respiratory ﬂux. An unchanged Δψmito and a low/null inhibition of the respiratory ﬂux would be, in this case, only apparent and would, as a matter of fact, represent a decrease of the available respiratory capacity, as well as a re-distribution of the overall control of the metabolic ﬂux. The ﬁnal outcome of these observations is that the measurement of the control by COX on the cellular respiration by the inhibitor titration technique can be carried out only in the presence of uncouplers or, eventually, under phosphorylating conditions (state III) if the residual Δψmito exerts a negligible and comparable control on both the integrated and isolated COX respiratory ﬂuxes. The results of the inhibitor titration technique and of the MCA applied to mitochondrial OXPHOS have been often referred to as a mimic of respiratory defects associated with mutations of mitochondrial DNA. In particular, the biochemical thresholds have been emphasized to play an important role, together with the genetic thresholds determined by the heteroplasmic level of the mitochondrial DNA mutations, in the various pathogenetic mechanisms. Most of the studies have been, however, focused on the impact of speciﬁc mutations/inhibitors on the mitochondrial respiration and/or ATP production. On the other hand, the physiological relevance of our ﬁndings is directly related to the functional roles played by the mitochondrial membrane potential in cell metabolism. Thus, even a partial deﬁciency of a respiratory complex (ahead of the respiratory inhibition threshold), although causing only a relatively small decrease in OXPHOS, could lead to a signiﬁcant decrease in Δψmito and its functionally related organellar processes, other than OXPHOS, such as protein and substrate import and calcium homeostasis. Therefore, the negative impact of a respiratory chain defect would predominantly decrease the cellular ATP production by OXPHOS under active phosphorylating conditions (state III), while differentially affecting Δψmito-dependent functions even in resting (state IV) or intermediate energy states. Indeed, impairment of mitochondrial Fe–S cluster formation by a reduction of the Δψmito-dependent protein import would also affect the nuclear genome stability (Veatch et al., 2009). Furthermore, Δψmito could be signiﬁcantly inﬂuenced under conditions during which COX becomes even more limiting due to its inhibition by NO (Brown, 2001; Brunori et al., 2004; Clementi et al., 1999) and/or hypoxia (Fukuda et al., 2007; Wiedemann and Kunz, 1998). Finally, the increased susceptibility to apoptosis of cells with impaired mitochondrial functions, as shown in cellular and animal models of mitochondrial diseases (Aure et al., 2006; Ghelli et al., 2003; Mirabella et al., 2000; Perier et al., 2005), could also be explained by their higher tendency to drops in Δψmito.
C. Pacelli et al. / Mitochondrion 11 (2011) 334–341
5. Conclusions In the present paper we demonstrate that COX exerts a tight control on the cellular respiration as well as on the mitochondrial membrane potential, thus strengthening the role of the terminal step of the respiratory chain in the regulation of the cellular energy metabolism in human physiopathology. Acknowledgments This work was supported by grants from National Research Project (PRIN 2003 n° 2003064310 and PRIN 2006 n° 2006069034_004) of the Italian Ministry for the University (MIUR). References Abramov, A.Y., Smulders-Srinivasan, T.K., Kirby, D.M., Acin-Perez, R., Enriquez, J.A., Lightowlers, R.N., Duchen, M.R., Turnbull, D.M., 2010. Mechanism of neurodegeneration of neurons with mitochondrial DNA mutations. Brain 133, 797–807. Akerman, K.E., Jarvisalo, J.O., 1980. Effects of ionophores and metabolic inhibitors on the mitochondrial membrane potential within isolated hepatocytes as measured with the safranine method. Biochem. J. 192, 183–190. Akerman, K.E., Saris, N.E., 1976. Stacking of safranine in liposomes during valinomycininduced efﬂux of potassium ions. Biochim. Biophys. Acta 426, 624–629. Akerman, K.E., Wikstrom, M.K., 1976. Safranine as a probe of the mitochondrial membrane potential. FEBS Lett. 68, 191–197. Attardi, G., Schatz, G., 1988. Biogenesis of mitochondria. Annu. Rev. Cell Biol. 4, 289–333. Aure, K., Fayet, G., Leroy, J.P., Lacene, E., Romero, N.B., Lombes, A., 2006. Apoptosis in mitochondrial myopathies is linked to mitochondrial proliferation. Brain 129, 1249–1259. Bernardi, P., Scorrano, L., Colonna, R., Petronilli, V., Di Lisa, F., 1999. Mitochondria and cell death. Mechanistic aspects and methodological issues. Eur. J. Biochem. 264, 687–701. Boulet, L., Karpati, G., Shoubridge, E.A., 1992. Distribution and threshold expression of the tRNA(Lys) mutation in skeletal muscle of patients with myoclonic epilepsy and ragged-red ﬁbers (MERRF). Am. J. Hum. Genet. 51, 1187–1200. Brown, G.C., 2001. Regulation of mitochondrial respiration by nitric oxide inhibition of cytochrome c oxidase. Biochim. Biophys. Acta 1504, 46–57. Brunori, M., Giuffre, A., Forte, E., Mastronicola, D., Barone, M.C., Sarti, P., 2004. Control of cytochrome c oxidase activity by nitric oxide. Biochim. Biophys. Acta 1655, 365–371. Chomyn, A., Martinuzzi, A., Yoneda, M., Daga, A., Hurko, O., Johns, D., Lai, S.T., Nonaka, I., Angelini, C., Attardi, G., 1992. MELAS mutation in mtDNA binding site for transcription termination factor causes defects in protein synthesis and in respiration but no change in levels of upstream and downstream mature transcripts. Proc. Natl Acad. Sci. USA 89, 4221–4225. Clementi, E., Brown, G.C., Foxwell, N., Moncada, S., 1999. On the mechanism by which vascular endothelial cells regulate their oxygen consumption. Proc. Natl Acad. Sci. USA 96, 1559–1562. Cohen, S.M., Ogawa, S., Rottenberg, H., Glynn, P., Yamane, T., Brown, T.R., Shulman, R.G., 1978. P nuclear magnetic resonance studies of isolated rat liver cells. Nature 273, 554–556. Crinson, M., Nicholls, P., 1992. Routes of electron transfer in beef heart cytochrome c oxidase: is there a unique pathway used by all reductants? Biochem. Cell Biol. 70, 301–308. Dalmonte, M.E., Forte, E., Genova, M.L., Giuffre, A., Sarti, P., Lenaz, G., 2009. Control of respiration by cytochrome c oxidase in intact cells: role of the membrane potential. J. Biol. Chem. 284, 32331–32335. Desquiret, V., Loiseau, D., Jacques, C., Douay, O., Malthiery, Y., Ritz, P., Roussel, D., 2006. Dinitrophenol-induced mitochondrial uncoupling in vivo triggers respiratory adaptation in HepG2 cells. Biochim. Biophys. Acta 1757, 21–30.
Di Paola, M., Cocco, T., Lorusso, M., 2000. Ceramide interaction with the respiratory chain of heart mitochondria. Biochemistry 39, 6660–6668. Ferguson, S.J., Sorgato, M.C., 1982. Proton electrochemical gradients and energytransduction processes. Annu. Rev. Biochem. 51, 185–217. Ferguson-Miller, S., Brautigan, D.L., Margoliash, E., 1978. Deﬁnition of cytochrome c binding domains by chemical modiﬁcation. III. Kinetics of reaction of carboxydinitrophenyl cytochromes c with cytochrome c oxidase. J. Biol. Chem. 253, 149–159. Ferraresi, R., Troiano, L., Rossi, D., Gualdi, E., Lugli, E., Mussini, C., Cossarizza, A., 2004. Mitochondrial membrane potential and nucleosidic inhibitors of HIV reverse transcriptase: a cytometric approach. Mitochondrion 4, 271–278. Fukuda, R., Zhang, H., Kim, J.W., Shimoda, L., Dang, C.V., Semenza, G.L., 2007. HIF-1 regulates cytochrome oxidase subunits to optimize efﬁciency of respiration in hypoxic cells. Cell 129, 111–122. Ghelli, A., Zanna, C., Porcelli, A.M., Schapira, A.H., Martinuzzi, A., Carelli, V., Rugolo, M., 2003. Leber's hereditary optic neuropathy (LHON) pathogenic mutations induce mitochondrial-dependent apoptotic death in transmitochondrial cells incubated with galactose medium. J. Biol. Chem. 278, 4145–4150. Groen, A.K., Wanders, R.J., Westerhoff, H.V., van der Meer, R., Tager, J.M., 1982. Quantiﬁcation of the contribution of various steps to the control of mitochondrial respiration. J. Biol. Chem. 257, 2754–2757. Hayashi, J., Ohta, S., Kikuchi, A., Takemitsu, M., Goto, Y., Nonaka, I., 1991. Introduction of disease-related mitochondrial DNA deletions into HeLa cells lacking mitochondrial DNA results in mitochondrial dysfunction. Proc. Natl Acad. Sci. USA 88, 10614–10618. Kunz, W.S., Kudin, A., Vielhaber, S., Elger, C.E., Attardi, G., Villani, G., 2000. Flux control of cytochrome c oxidase in human skeletal muscle. J. Biol. Chem. 275, 27741–27745. Mirabella, M., Di Giovanni, S., Silvestri, G., Tonali, P., Servidei, S., 2000. Apoptosis in mitochondrial encephalomyopathies with mitochondrial DNA mutations: a potential pathogenic mechanism. Brain 123 (Pt 1), 93–104. Nicholls, D.G., 2006. Simultaneous monitoring of ionophore- and inhibitor-mediated plasma and mitochondrial membrane potential changes in cultured neurons. J. Biol. Chem. 281, 14864–14874. Perier, C., Tieu, K., Guegan, C., Caspersen, C., Jackson-Lewis, V., Carelli, V., Martinuzzi, A., Hirano, M., Przedborski, S., Vila, M., 2005. Complex I deﬁciency primes Baxdependent neuronal apoptosis through mitochondrial oxidative damage. Proc. Natl Acad. Sci. USA 102, 19126–19131. Piccoli, C., Scrima, R., Boffoli, D., Capitanio, N., 2006. Control by cytochrome c oxidase of the cellular oxidative phosphorylation system depends on the mitochondrial energy state. Biochem. J. 396, 573–583. Pinti, M., Troiano, L., Nasi, M., Ferraresi, R., Dobrucki, J., Cossarizza, A., 2003. Hepatoma HepG2 cells as a model for in vitro studies on mitochondrial toxicity of antiviral drugs: which correlation with the patient? J. Biol. Regul. Homeost. Agents 17, 166–171. Ramzan, R., Staniek, K., Kadenbach, B., Vogt, S., 2010. Mitochondrial respiration and membrane potential are regulated by the allosteric ATP-inhibition of cytochrome c oxidase. Biochim. Biophys. Acta 1797, 1672–1680. Veatch, J.R., McMurray, M.A., Nelson, Z.W., Gottschling, D.E., 2009. Mitochondrial dysfunction leads to nuclear genome instability via an iron–sulfur cluster defect. Cell 137, 1247–1258. Villani, G., Attardi, G., 1997. In vivo control of respiration by cytochrome c oxidase in wild-type and mitochondrial DNA mutation-carrying human cells. Proc. Natl Acad. Sci. USA 94, 1166–1171. Villani, G., Attardi, G., 2001. In vivo measurements of respiration control by cytochrome c oxidase and in situ analysis of oxidative phosphorylation. Meth. Cell Biol. 65, 119–131. Villani, G., Greco, M., Papa, S., Attardi, G., 1998. Low reserve of cytochrome c oxidase capacity in vivo in the respiratory chain of a variety of human cell types. J. Biol. Chem. 273, 31829–31836. Wiedemann, F.R., Kunz, W.S., 1998. Oxygen dependence of ﬂux control of cytochrome c oxidase—implications for mitochondrial diseases. FEBS Lett. 422, 33–35. Yoneda, M., Miyatake, T., Attardi, G., 1994. Complementation of mutant and wild-type human mitochondrial DNAs coexisting since the mutation event and lack of complementation of DNAs introduced separately into a cell within distinct organelles. Mol. Cell. Biol. 14, 2699–2712.