[Frontiers in Bioscience E3, 395-409, January 1, 2011]
The respiratory-dependent assembly of ANT1 regulates cytochrome c release
Faustin Benjamin 1, Rossignol Rodrigue 1, Deniaud Aurelien 2, Rocher Christophe 1, Claverol Stephane 3, Malgat Monique1, Brenner Catherine 4, Letellier Thierry 1
1
TABLE OF CONTENTS
1. ABSTRACT
The adenine nucleotide translocator (ANT) is a control point of several fundamental cell processes, as diverse as cell energy supply, mitochondrial DNA maintenance, and apoptosis. This paper describes six individual structures of the carrier, distinguished according to ANT1 oligomeric and conformational states, as well as associations with other proteins. Transitions between these structures depend on energy demand and thus contribute to a metabolic reserve of oxidative phosphorylation (OXPHOS) activity. Moreover, at low respiratory chain activity, we demonstrate that, unlike a mitochondrial Ca2+ upload, Bax, a pro-apoptotic Bcl-2-family protein, is able to trigger a massive release of cytochrome c from one of these ANT1 structures. These new insights emphasize the close relationship between structural rearrangements of ANT and molecular apoptotic events at distinct cell energy levels. OXPHOS functioning has to therefore be considered a crucial control point for the events leading to these contrasting pathways.
2. INTRODUCTION
The adenine nucleotide translocator (ANT) plays an essential role in the oxidative phosphorylation (OXPHOS) system developed in aerobic eukaryotic cells. One of the major functions of this nuclear encoded protein is to catalyze the exchange between cytosolic ADP and ATP synthesized within mitochondria. This exchange consequently provides the energy for several ATP-dependent functions occurring within the cell (1). The carrier can be specifically inhibited and stabilized into two distinct conformations, i.e. cytosolic c or matricial m, induced by binding carboxyatractyloside (CATR) or bongkrekic acid (BA), respectively (2).
Various ANT isoforms are expressed in mammalian cells, i.e. ANT1 is known as the specific heart/muscle isoform but is also expressed to a lesser extent in kidney and brain (3). It is now assumed that the minimal functional state of this carrier is the homo-dimer (4). However, several oligomerization states, such as tetramers or coupled dimers, have emerged from nucleotide binding data (5). In addition to this role in bioenergetics, ANT remains crucially involved in mtDNA maintenance (6). Besides these living functions, this carrier also plays a key role in programmed cell death (7; 8). ANT1 is described as a core component in the permeability transition pore complex (PTPC), as its large opening under diverse apoptotic stimuli, such as Ca2+ upload, leads to mitochondrial membrane potential (D Y m) collapse (9). PTPC is essential promoting the pro-apoptotic effect of Bax by forming pores in the external membrane and, thus, making it possible for apoptogenic factors, e.g. cytochrome c (cyt c), to diffuse from mitochondria to the cytosol (10; 11). Another release mechanism, demonstrated in cultured cells, is also mediated by Bax, but independently of PTPC (12; 13).
We recently reported that ANT might adopt various oligomeric and conformational states, changing according to the cell energy state (14). In this research, we demonstrated by analyzing various ANT1 structures using blue-native polyacrylamide gel electrophoresis (BN-PAGE)/ sodium dodecyl sulfate-PAGE (SDS-PAGE) that there could be up to six different forms of ANT1 in functional muscle mitochondria. We showed that these structures contribute to a metabolic reserve of oxidative phosphorylation (OXPHOS) activity through ANT1 oligomerization and conformational changes from a PTPC-like structure. These mechanisms enable OXPHOS to adapt to mitochondrial mutations or various cell energy states. Moreover, we suggest that one of the complexes formed by the interaction of ANT1 monomer with VDAC1 and endogenous Bax monomer forms a control point in the dual mitochondrial functions of energy production and apoptosis. Indeed, this structure is associated with 80% of the mitochondrial cyt c content at low OXPHOS activity. However, at higher energy states, ANT1 oligomerization leads to disassociate cyt c from this complex, thus responding to a respiratory flux increase.
In addition, we show that the recombinant proapoptotic protein Bax is able to target this structure, inducing a massive release from the bound pool of cyt c. This suggests this complex may be implicated in apoptosis, thus constituting an "apoptotic induction unit" for cyt c release, independent of the SMAC/Diablo release pathway. In contrast, the Ca2+-induced permeability transition triggers the release of free-respiratory active cyt c, thereby affecting mitochondrial energy functions.
As we have shown, ANT is a control point in the cell metabolism, enabling the respiratory chain to adapt to varying energy demands, as well as having a regulatory function in programmed cell death.
3. MATERIAL AND METHODS
3.1. Animals and mitochondrial preparations
All the methods are fully described in (15). Protein concentration was estimated by the Biuret method.
3.2. Human muscle biopsies
Three patients were selected at the Centre Hospitalier Regional de Bordeaux and diagnosed by clinical investigations, biological studies (especially blood lactate and pyruvate), and histo-enzymological studies. Controls were taken from patients who underwent muscle biopsy for neuromuscular symptoms, but were ultimately found to be free of any mitochondrial disease. Muscle biopsies were performed on the quadriceps muscle under general anesthesia (with the informed consent of the patients or their parents). Muscle biopsies were stored in 10 ml ice-cold buffer (EGTA 10 mM, MgCl2 5 mM, taurine 20 mM, dithiothreitol 0.5 mM, imidazol 20 mM, K+MES 0.1 M, ATP 5 mM, phosphocreatine 20 mM) and used in the following hour. The mitochondrial respiratory rate was measured on saponin- permeabilized muscle fibers, using the procedures developed by (16). Mitochondrial oxygen consumption on various respiratory substrates was measured as described in a medium containing 10 mM EGTA, 5 mM MgCl2, 20 mM taurine, 0.5 mM dithiothreitol, 20 mM imidazol, 0.1 M K+MES, 3 mM KH2PO4, 5mg/ml BSA, and expressed in natomO/min/mg of fibers.
3.3. Oxygraphic and swelling measurements
Mitochondrial oxygen consumption was monitored at 30�C in a 1 ml thermostatically-controlled chamber equipped with a Clark oxygen electrode connected to a computer, which gave an on-line display of the respiratory rate value. The respiration buffers were used as described in (15), with the addition of respiratory substrates (10 mM pyruvate with 10 mM malate, or 25 mM succinate with 0.4 �g/ml rotenone, or 3 mM ascorbate with 0.5 mM TMPD, or 2 �M CCCP. However, in the case of Ca2+ addition, EGTA in the respiration buffer was replaced by 5 �M and the respiratory rate recorded at state 4. The mitochondrial concentration used was 1 mg/ml and state 3 according to (17) was obtained by adding 2 mM ADP. Mitochondrial oxygen consumption was expressed in natomO/min/mg proteins. MPT was measured by the convenient swelling assay, carried out in an Uvikon 940 spectrophotometer. Swelling curves were recorded at 540 nm. Cuvettes containing the mitochondrial suspension were maintained at 25oC. Incubation conditions and MPT induction are indicated in the Figure legends.
3.4. Titration curves, threshold curves, and threshold value determination
Titration curves of the step catalyzed by ANT were performed using CATR (Sigma), as described by (14). Threshold curves come from the titration curves and represent the state 3 respiratory rate (%) as a function of ANT inhibition (%) The threshold values were determined as described (15).
3.5. Production and purification of recombinant BaxD C
Bax D 21 was produced as a GST-fusion recombinant protein in E. coli. The GST N-terminal tag was cleaved by thrombin and Bax D 21 was purified as described by (18). To induce oligomerization, Bax D 21 was incubated with TX-100 (TX-100/protein ratio of 1.4 g/g) at 4oC for 30 min. Mitochondria respiring with different respiratory substrates were then incubated with oligomerized Bax D 21 at 30oC for 30 min.
3.6. Electrophoresis techniques
Isolated rat mitochondria from muscle (1mg/ml) were subjected to oxygraphic measurements at state 3 on various respiratory substrates. Mitochondrial proteins (300 �g) were taken from the oxygraph, solubilized by digitonin, and then subjected to BN-PAGE/SDS-PAGE procedures, as described (14). However, for the BaxD C experiments, the ultracentrifugation previously used to remove solubilized complexes was reduced to 20 000 rpm, at 4oC, for 30 min.
3.7. Western-blot
Mitochondrial membrane proteins solubilized by digitonin (digitonin/protein ratio of 3g/g) were subjected to BN-PAGE/SDS-PAGE, electrotransferred to PVDF membranes (Bio-Rad), and subjected to primary ab in 5% milk. Polyclonal abs specific for human ANT1 isoform (Ab1, amino acids 2-11 and Ab2, amino acids 161-171) were purchased from Oncogene Research Products. Abs for Porin (31HL) and cyclophilin D were purchased from Calbiochem; and cytochrome c from Zymed. All abs were used at dilutions recommended by the manufacturer. After washing steps, membranes were incubated with horseradish peroxydase-conjugated purified secondary abs (Bio-Rad) Proteins were detected by enhanced chemiluminiscence (ECL; Amersham Pharmacia) and revealed by autoradiography. For cytochrome c release measurements, isolated muscle mitochondria (1mg/ml) respiring with the various substrates at state 4 in the presence of Ca2+ were centrifuged (20 min, 8000 rpm, 4oC) The supernatant was then removed with 1�g/ml protease inhibitors (PMSF, leupeptin and pepstatin), subjected to SDS-PAGE (20% SDS, 1% b -mercaptoehanol), and immunoblotted, as described above.
3.8. Protein detection and analysis
The proteins separated by second (m) SDS-PAGE dimension were subjected either to Western-blotting or Sypro Rubyâ staining (Molecular Probes), according to the manufacturer's instructions. Fluorescent proteins were detected with an UV-transilluminator system and recorded using a Herada 429K DV camera. The protein spots were cut and subjected to tandem mass spectrometry analysis. Densitometry analysis on Western-blots was carried out using Image J (NIH) software.
4. RESULTS
4.1. Distribution of ANT1 oligomeric and conformational states in functional mitochondria
In a recent study, we showed that ANT1 might exist in several oligomeric states within the inner membrane of muscle mitochondria respiring with ADP on pyruvate-malate (14). Indeed, when membrane proteins were solubilized with digitonin and BN-PAGE/SDS-PAGE coupled with Western-blots using a polyclonal antibody specific for ANT1 (abs, Ab1), we observed the presence of ANT1 monomers, dimer and coupled-dimer functional forms in the ADP/ATP exchange (M1, D1, 2D1; Figure 1. A) When Ab2 that recognizes the second matricial loop of ANT1, was used, two novel forms were observed, i.e. D2 dimers (50 kDa) associated with porin in a complex exceeding 300 kDa and free M2 monomers (31 kDa) (Figure 1A) Thus, when the N-terminal domain (N-ter) was exposed, the second matricial loop was not recognized, and vice versa. These results indicate that the two ANT1 conformations, i.e. cytosolic c and matricial m, are revealed by the two distinct antibodies used.
The proteins potentially bound to ANT1 monomers in the c conformation (M1) within the 75 kDa complex were studied. Sypro rubyâ staining of mitochondria respiring on pyruvate-malate with ADP revealed the proteomic pattern (Figure 1Ba) The Western-blot immunoblotted with anti-ANT1 (Ab1), anti-porin, anti CyD and anti-cyt c (Figure 1Bb) showed the three ANT1 c forms associated with porin, therefore D1 was also interacting with CyD within the 300 kDa complex. Surprisingly, the 75 kDa complex was composed of M1, monomeric porin and cyt c. In addition, cyt c was also observed in a 16 kDa minority free form in the first dimension. Spots indicated in Figure 1Bd was subjected to a tandem mass spectrometry analysis. Peptide sequences compared with the NCBI databank suggest a good correlation for the presence of the ANT isoform 1 in M1 and M2 and the voltage-dependent anion-channel isoform 1 (VDAC1) in the 75 kDa complex, and VDAC2 in free state (Xcorr and gi indicated) Unfortunately, the thin fluorescent staining of cyt c could not be analyzed by mass spectrometry. Another 100 kDa spot in a complex of over 300 kDa in the first dimension was an association of ANT (isoform not specified) with mitochondrial creatine kinase (mtCK). Figure 1. Bc shows the presence of D2 (matricial conformation), porin, and CyD within a complex exceeding 300 kDa, as well as free 31 kDa M2 monomers. As previously shown, Cyt c was co-localized with porin in the 75 kDa complex but a minority was also found in free form.
BN-PAGE on gradient gel was used to characterize the over-300 kDa complexes, under the same conditions as above (Figure 1Be) and OXPHOS complexes were revealed by Blue Coomassie staining. Subunits of each complex were then separated in modified second dimension, stained with Sypro Rubyâ , and subsequently immunoblotted with either anti-ANT1 (Ab1) (Figure 1. Bf) or anti-ANT1 (Ab2) (Figure 1Bg), together with anti-porin and anti-CyD. Western-blots showed D1, porin and CyD associated in a 550-600 kDa complex, on the one hand, and D2 with porin but not CyD in a 400 kDa complex, on the other hand. The D2 form of ANT1 also interacted with the two respiratory chain supercomplexes of 1400 and 1600 kDa described by (19). The membrane used with anti-ANT1 (Ab2) was then reprobed with anti-Cyt c (Figure 1Bh) and showed that cyt c was associated in complexes under 100 kDa in the first dimension.
4.2. Oligomeric and conformational states of ANT1 depend on respiratory chain activity
As previously described (14), the increase in respiration from pyruvate-malate to ascorbate-TMPD via succinate led to a rise in 2D1 (contrarily to D1), accompanied by a considerable decrease in ANT1 M1 monomers, which were no longer observed on ascorbate-TMPD (Figure 2A) Interestingly, adding CCCP uncoupler disassociated the ANT1 oligomers and ANT1-porin associations, resulting in an increase in M1. As for Ab1, M2 forms were expressed more than D2, but no coupled dimers in the m conformation were detected (Figure 2A) By replacing pyruvate-malate with succinate, the increase in D2 content was followed by synthesis of coupled dimers in the m conformation (2D2, 110 kDa) Unlike 2D1, all 2D2 forms were associated with complexes over 300 kDa. When respiration was further increased on ascorbate-TMPD, 2D2 forms increased considerably compared to D2 whereas M2 was no longer observed. Adding CCCP induced the same disassociation effect on the m structures. Consequently, these two types of ANT1 monomers are able to constitute preferentially coupled dimeric structures, irrespective of their c or m conformation. However, each oligomer conformation interacts with specific macrocomplexes.
In order to analyze these respiratory-induced transitions for complexes exceeding 300 kDa, we carried out the same procedure using 5-13% gradient BN-PAGE. The second modified dimension was immunoblotted with either anti-ANT1 (Ab1) or anti-ANT1 (Ab2) and reprobed with anti-CyD and anti-porin (Figure 2B) Porin signals were used to quantify the various ANT1 forms. As in Figure 1. Bf and Bg, D1 was associated with porin and CyD with a 550-600 kDa complex, unlike D2, which was associated with porin but not CyD in a 400 kDa complex. Small amounts of the latter structure were also associated with two respiratory chain supercomplexes. As previously shown, the respiratory increase from pyruvate-malate to ascorbate-TMPD induced an increase in both 2D1 and 2D2 content. However, while 2D1 interacted with porin in a 220 kDa complex, 2D2 was mainly associated with the 1600 kDa supercomplex on ascorbate-TMPD. Nevertheless, the transition from pyruvate-malate to succinate led to a larger decrease in D1 than CyD (2 fold and 1.4 fold, respectively), while D2 increased in both cases. These observations may indicate that an additional structural mechanism occurs at lower fluxes besides oligomerization.
4.3. ANT may be mobilized from a structure responsible for the mitochondrial permeability transition
In Figure 2B, the amount of D1 followed the decrease in M1 when respiration is raised from pyruvate to succinate, unlike D2, which increased under the same respiratory conditions. This suggests that there are two distinct mechanisms, depending on the respiratory chain activity level, i.e. (i) a mobilization of monomers as described above and (ii) probably conformational changes in the carrier for lower transitions of respiratory steady-states. On permeabilized human muscle fibers, we titrated the mitochondrial respiration by CATR and determined the amount of functional ANT involved in the flux (Figure 3 A) The amount of inhibitor necessary to inhibit all the carriers involved in the flux (Imax), as well as the inhibition constant (Ki) are given in Table 1. As previously observed, the amount of ANT participating in the respiratory steady-state increased when pyruvate-malate was replaced with succinate (less than 2-fold), but the Ki values remained constant. Likewise, adding 10 mM creatine in the presence of pyruvate-malate with ADP increased both the respiratory steady-state and the Imax value, almost to the same extent as succinate (Table 1) Moreover, the Ki value also remained constant. On the contrary, adding Ca2+ in the presence of pyruvate-malate with ADP induced an uncoupling of respiration and decreased the Imax value by half. Threshold curves shown in Figure 3, B represent the effect of ANT inhibition by CATR on the respiratory rate. Our results show that ANT activity can be inhibited by up to 80% on pyruvate-malate with or without creatine, with only a slight effect on respiration. However, this excess enzyme activity was not observed with Ca2+, as there was only half the functional ANT in the ADP/ATP exchange. Creatine was recently found to inhibit the permeability transition by enhancing the microcompartmentation between ANT and mtCK (20). Moreover, we show here that its addition in the cell context makes it possible to increase the functional amount of ANT. This effect probably occurs through conformational changes in the carriers from the c state within a PTP-like structure with CyD to the m state associated with mtCK.
4.4. Cytochrome c is mobilized through the respiratory-induced oligomerization of ANT1
As described in Figure 1, cyt c is mainly associated with porin and M1 in the 75 kDa complex. Furthermore, small amounts have also observed in free-state in mitochondria respiring on pyruvate-malate with ADP. Likewise, Figure 4. A shows that, under these conditions, 81% of cyt c is associated. When respiration increased from pyruvate-malate to succinate, the decrease in bound cyt c was concomitant with a proportional increase in the free-state (Bar graphs in Figure 4A) This phenomenon is more marked between succinate and ascorbate-TMPD as the bound state completely disappears in favor of the free form (80% of total cyt c increase). We reproduced the same experiments in brain and liver mitochondria, known to express ANT1 (to a lesser extent than muscle) and, mainly, ANT2 (Figure 4B) We observed the presence of bound cyt c in brain (less than in muscle) but not in liver. Thus, this cyt c mobilization, mainly in response to high fluxes, depends on the amount of ANT1 and is linked to M1 oligomerization.
Cyt c in the 75 kDa complex may be more tightly bound to the inner membrane than in its free state. In Figure 5, we added various concentrations of Ca2+ and analyzed the cyt c release capacities at the different respiratory chain activity levels induced. In isolated mitochondria, a low matricial Ca2+ uptake was sufficient to induce a permeability transition leading to the release of several apoptogenic factors, such as cyt c (21). Figure 5A shows that cyt c was mainly released by Ca2+ as the respiratory rate increased. This effect was inhibited by 10 �M CsA, whereas adding CCCP induced a strong release compared to pyruvate-malate only. As control, we showed that the mitochondrial swelling induced by adding 100 �M Ca2+ was strictly the same, irrespective of the energy substrate used (Figure 5A). To determine whether the cyt c pool was affected by Ca2+, we analyzed 2D gels of mitochondria respiring on various substrates in the presence of 100 �M Ca2+ (Figure 5B) As shown in the densitometry analysis, this addition induced a decrease in bound cyt c (Cb) and an increase in free cyt c (Cf) This mobilization of cyt c resulted from the uncoupling effect of Ca2+ responsible for increasing respiration. However, this effect was attenuated as respiration continued to increase. Consequently, the released cyt c seems to follow ANT mobilization when respiration increases. This indicates that interactions within the 75 kDa complex may confer a tight binding of cyt c stored within the inner membrane, as compared to the loosely-bound free state specifically released by a Ca2+-induced PT.
To illustrate this point, we determined the respiratory rate of mitochondria at state 4 with various energetic substrates and after adding Ca2+ (Figure 5C) The uncoupling effect induced by 50 �M Ca2+ was observed when the flux increased on succinate, and accentuated on ascorbate-TMPD, but not on pyruvate-malate. However, when the Ca2+ concentration increased, the uncoupling effect disappeared even on succinate and respiration was inhibited 2-fold on pyruvate-malate with 100 �M Ca2+. This inhibition was considerably more marked in the presence of CCCP. Moreover, the uncoupling effect on succinate was abolished by adding 10 �M CsA, which specifically inhibits the permeability transition by closing the PTPC. Adding 2mM Ca2+, which was associated with the strong release of cyt c from mitochondria observed in Figure 5A, strongly inhibited mitochondrial respiration. These data indicate that the mobilization of cyt c by disassociation of the 75 kDa complex may maintain respiration despite a slight release by 50 �M Ca2+. Thus, the mobilized cyt c may become involved in electron transfer as well as being releasable by external membrane disruptions. Consequently, functional cyt c (electron donor in the respiratory chain) is loosely bound to the mitochondrial inner membrane in free form, unlike the tightly bound form in the 75 kDa complex.
4.5. Respiratory-induced structural changes in ANT1 regulate Bax-triggered cytochrome c release from mitochondria
Previous studies have shown that Bax interacts with ANT to induce a pore-forming structure, leading to the release of cyt c and other apoptotic factors from mitochondria (10; 11; 22; 23; 24; 25). By modifying the experimental procedures (see Material and Methods), on 2D gels of mitochondria respiring on pyruvate-malate and ADP, Bax monomers were associated in distinct complexes exposing the N-ter domain (Figure 6A) Indeed, B1, B2, and B3 Bax monomers were observed to interact in complexes exceeding 300 kDa, 300 kDa, and 100 kDa, respectively, whereas B4 was mainly found in free monomeric form. Several studies have shown that, unlike monomers, the oligomeric form is active in permeabilizing the outer membrane. Thus, inactive endogenous Bax monomers are weakly bound to muscle mitochondria and may interact in specific complexes. Immunoblots with anti-cyt c revealed two pools, as previously described. The tightly bound cyt c was exactly colocalized with B3 monomers of Bax associated in a 100 kDa complex. Moreover, 2D gels immunoblotted with anti-smac/Diablo showed 30 kDa monomers in the second dimension associated in 57 kDa dimers from the first dimension. The results obtained on various respiratory substrates show that, like cyt c, the Bax monomer in the 100 kDa complex disassociated when respiration increased, resulting in a larger amount of B1, but not B4. As in the case of ANT1 oligomerization accompanied by mobilization of bound cyt c, this effect is rather marked between succinate and ascorbate-TMPD (B3 remained undetectable at high respiratory rates) Similarly to ANT1 monomers and cyt c, adding CCCP disassociated Bax from the various complexes, increasing B4. Unlike Bax, SMAC/Diablo was insensitive to respiratory modulations.
To determine the molecular events that occur during the mitochondrial relocation of Bax during apoptosis, we used the same 2D gels as before with purified Bax with the C-terminal domain deleted (BaxD C) to induce oligomerization of BaxD C, we previously incubated the protein with various concentrations of TX-100 (Figure 6A) The results of adding BaxD C (solubilized with TX-100/protein ratio of 1.4 g/g) to isolated mitochondria revealed two new forms i.e. (i) Bax dimers (Bd) maintained in the second dimension (45 kDa), mainly associated in a 150 kDa complex but also in a complex exceeding 300 kDa, and (ii) Bax oligomers (Bo, 65 kDa) in a complex exceeding 300 kDa. B1 was also observed to increase under these conditions, probably due to disassociation of Bo. However, when BaxD C was added, B3 disappeared. When respiration increased, BaxD C mainly induced modifications in B3, Bo, and Bd, but not B1, B2, and B4. Indeed, the decrease in Bd in the 150 kDa complex (mainly during the transition from succinate to ascorbate-TMPD) correlated quantitatively with the increase in Bo in a complex exceeding 300 kDa. Thus, respiratory-induced variations in Bd accompanied the oligomerization of ANT1 monomers. Immunoblots with anti-cyt c showed that adding BaxD C induced an increase in Cf by disassociating the cyt c bound in the 100 kDa structure. This bound pool was no longer associated with Bd in the 150 kDa complex, irrespective of the substrate used, indicating that Bd was linked to ANT1 c monomers and porin, but not cyt c.
To investigate whether adding BaxD C released cyt c and whether this effect was sensitive to respiratory changes, we incubated isolated mitochondria respiring with the previous substrates with ADP in the presence of BaxD C (with 1,4g/g TX-100) Mitochondria were isolated and the pellet immunoblotted with anti-cyt c and anti-porin (Figure 6B) The densitometry analysis showed that 50 nM BaxD C, as well as detergent only, had no effect on the total cyt c content of mitochondria respiring on pyruvate-malate, while 1�M BaxD C released 40% of the total. Contrary to control, this effect was gradually abolished when respiration was increased by succinate and, finally, ascorbate-TMPD (29% and 6%, respectively) Interestingly, adding CCCP strongly enhanced this release (75%) Likewise, adding 1 �M BaxD C also induced SMAC/Diablo release from mitochondria (48%), but this effect was insensitive to respiratory states. Nevertheless, Figure 6C shows that BaxD C-induced cyt c release at low respiration did not alter the mitochondrial functions. Indeed, despite a strong cyt c and smac/Diablo release on pyruvate-malate, the respiratory rate and ATP synthesis (data not shown) remained at control values, emphasizing the absence of D y m and D pH collapse, even when 40% of the cyt c was released into the cytosol from the mitochondria.
5. DISCUSSION
Previous studies have demonstrated that ANT1 monomers may be mobilized and assembled into active oligomers sensitive to OXPHOS functioning (14). This paper describes six distinct forms of the carrier present in muscle mitochondria under phosphorylation conditions. These forms are distinguished according to their oligomeric (monomers, dimers, or coupled dimers) and conformational states (cytosolic or matricial), as well as associations with other proteins (Figure 7).
These ANT1 structures may feature distinct mitochondrial functions, associated with the respiratory steady-state. M2 monomers are found in free state, unlike M1, present in a 95 kDa complex. Moreover, we observed that 80% of the mitochondrial cyt c content interacted in a complex composed of M1/VDAC1/endogenous Bax monomer, with reinforced binding to the inner membrane. According to the apparent mass restriction, only one or two molecules of cyt c, together with a Bax monomer, may interact per complex. These results emphasize the structural specificity of ANT1 c monomers to interact with cyt c and endogenous Bax. We show that M1 and M2 monomers are able to oligomerize and create new active c- and m-coupled dimers (2D1, 2D2) during a strong increase in respiration, resulting in a metabolic reserve of oxidative phosphorylation (OXPHOS) activity. The tightly bound cardiolipin in the crystallographic structure (26) may be sensitive to D y m or D pH changes and induce this assembly.
We evidenced another metabolic reserve mechanism produced by conformational changes in the carrier. 2D gradient gel analysis showed a PTPC-like structure combining ANT1 c dimers (D1) with porin and CyD in a 550-600 kDa complex. These data correlate with those showing that adding CATR or Atr stabilized ANT1 in the c conformation required for PTP induction (27), and that ANT1 but not ANT2 is the dominant apoptosis-inducing factor in mammalian cells (28, 29). Four hexokinase isoform 2 (HK 2) molecules and octameric mitochondrial creatine kinase (mtCK) can be added to these structures to form, respectively, the 600 and 400 kDa complexes described by (30). Our experiments demonstrated that an increase in respiration (with succinate or mtCK substrates) induced a conformational change in ANT1 from c to m (in the 600 and 400 kDa complexes, respectively) (Figure 7) As for the assembly process, the conformational change from c to m may also result in an increase in ADP/ATP exchange that could be reinforced by channeling with mtCK, as characterized by the 400 kDa complex. Indeed, we observed that the amount of functional ANT in human muscle fibers increased with respiration. Creatine, reported to inhibit the permeability transition (20), induced the same increase as succinate. On the contrary, Ca 2+, known to induce a CsA-sensitive permeability transition at low concentrations (He and Lemasters, 2002), reduced the functional ANT involved in nucleotide exchange. Thus, as described for oligomerization (14), this D1/D2 transition resulted in the excess of enzyme activity shown in the threshold curves. This excess maintained a normal respiratory rate until ANT activity had been 80% inhibited. This property can be enhanced by channeling between the mtCK and dimeric ANT m. Consequently, this dynamic equilibrium may be regulated by respiratory chain activity. In addition, CsA displaces this equilibrium via the structural effect of closing PTPC (14), whereas creatine induces this transition physiologically via a metabolic effect. These data may explain why different cell energy functioning states (various OXPHOS substrates or other metabolic processes) do not display the same sensitivity towards permeability transition inducers.
Similar mechanisms to ANT involving oligomerization and conformational changes have already been demonstrated for other OXPHOS complexes. Complex I exists in activated and deactivated forms with distinct kinetic and structural properties (30), dimeric Complex III consists of active and "silent" monomers for binding cyt c (31), and, finally, dimerization of ATP synthase may be regulated by the matricial pH-dependent binding of the IF1 inhibitor (32) ATP oligomerization is, therefore, involved in the morphology of mitochondrial cristae (33). As for ANT, these molecular mechanisms may be crucial for regulating OXPHOS activity and, thus, the cell energy supply.
Like ANT, cyt c displays contrasting functions in mitochondria. Indeed, on the one hand, it is necessary to the cell energy supply as an electron donor within the respiratory chain and, on the other hand, its release activates the apoptosome, thus contributing to the cell death process. It is not yet known how these contrasting pathways are activated according to cell energy status. A recent study shows that cyt c may be either loosely or tightly bound to the inner membrane by cardiolipin in liver mitochondria (composed mainly of ANT2). The loosely-bound state is required for cyt c release from mitochondria (34). We demonstrated that cyt c was distributed between two distinct pools at low respiratory chain activity, i.e. a large amount associated in a complex composed of ANT1 c monomers, a single Bax monomer, and porin, and a minority in free state. Interactions with this complex bind cyt c tightly to the inner membrane. These interactions may be hydrophobic or electrostatic between the positively charged Lys of cyt c and the negative facing of the external ring of crystallized ANT/CATR complex, rather than the non-polar bonds observed between cyt c and the CYT1 subunit of complex III (31). In addition, only the free state of cyt c is involved in electron transfer through the respiratory chain, rather than the tightly bound majority stored with the ANT1 c monomers. Ca2+-induced PT mobilized the bound cyt c, thus increasing the free active pool. At the molecular level, this mechanism is correlated with the CsA-sensitive tBid-induced mitochondrial cristae remodeling required to mobilize cyt c stores (35). We observed that only the loosely-bound, active cyt c was released under PT with Ca2+, as observed by (13). This probably explains the ∆Y m collapse observed under these conditions. Indeed, adding Ca2+ inhibits respiration for low fluxes and this effect is strongly enhanced with CCCP, which probably disrupts weak interactions between cardiolipin and free monomeric cyt c.
On the other hand, we observed that endogenous monomeric inactive Bax associates with ANT1 c monomers to constitute an "apoptotic induction unit" (AIU), a specific cyt c release pathway capable of inducing a massive release of this component from mitochondria. Indeed, added Bax∆C homodimerized with endogenous monomeric Bax already associated in this structure to mediate a pore-forming structure, releasing large amounts of stored cyt c (40% of total). This interaction disassociated the rest of the bound cyt c, thus increasing the monomeric respiratory active pool. Therefore, we can postulate that the mitochondrial relocation of the full-length native Bax occurring within cells may form a more efficient structure for releasing the bound cyt c completely (without affecting the active pool) These events would make it possible to activate the death pathway while maintaining the respiratory active free cyt c in the inner membrane and, thus, the normal rate of respiration and ATP synthesis required for caspase activation. This insight is supported by several studies showing that ∆Y m collapse is not due to a reduction in the cyt c pool, even after a large release into the cytosol (36). Moreover, the effect of Bax on such structure would explain that, once initiated, cyt c is massively and rapidly released from mitochondria without any membrane damage or protein import (37). The time collapse of ∆Y m downstream from the release step depends on the tissue type and may be due to activated caspases targeting specific sub-units of OXPHOS complexes, such as p75 of Complex I (38). The excess M1 structure at low fluxes constitutes a fine component for this fast release. Indeed, this structure represents 80% of the cyt c stored with inactive monomers. These mechanisms may impact the cell process as, several cell lines investigated used glucose as a glycolytic rather than a respiratory substrate. Consequently, as Bax-induced cyt c release clearly decreased with high respiratory chain activity (unlike Ca2+ upload), OXPHOS functioning may regulate the apoptotic induction triggered by Bax. Unlike cyt c, dimeric smac/Diablo is free in the intermembrane space and its release is insensitive to respiratory variations. A complex in excess of 300 kDa that is also recognized by exogenous Bax∆C probably mediates its release pathway.
We observed that the oligomerization of ANT1 c monomers made it possible for the mobilized cyt c to be taken up again in respiration, thus responding to a flux increase while increasing the amount releasable by Ca2+-induced PT. Hence, we may distinguish different cases in terms of the cell's oxidative energy status and sensitivity to different cell-death inducers. Consistent with previous data concerning Bax and Bak regulation of endoplasmic reticulum Ca2+ (39), we underline that combined mechanisms, including the permeabilization of OM by "multidomain " Bcl2 family members along with transient PT (triggering tightly- versus loosely-bound cyt c) are necessary to fully induce the mitochondrial apoptotic process. Hence, as illustrated in Figure 7, these new insights emphasize the close relationships between structural rearrangements of ANT and molecular apoptotic events at various cell energy levels. OXPHOS functioning has to be, therefore, considered a crucial control point for the events leading to these contrasting pathways.
6. ACKNOWLEDGMENT
This work was supported by INSERM, Université Victor Ségalen-Bordeaux 2, Région Aquitaine, A.F.M. (Association Française contre les Myopathies), F.R.M. (Fondation pour la Recherche Médicale), INCa (Insitut national du Cancer), ANR (Agence Nationale de la Recherche), A.R.C. (Association pour la Recherche contre le Cancer), Philippe Foundation, and the functional genomic facility of the Université Victor Ségalen Bordeaux 2 for mass spectrometry analysis. CB is supported by grants funded by Association pour la Recherche sur le Cancer (ARC), EGIDE, l'Agence Nationale pour la Recherche (ANR, ANR-08PCVI-0008-01) and l'Institut National pour le Cancer (INCa, 2008-1-PL BIO-04-CNRS ON1). The authors thank O. Sharaf el dein for his help in preparation of the manuscript.
7. REFERENCES
1. Klingenberg, M. Structure-function of the ADP/ATP carrier. Biochem Soc Trans, 20, 547-550 (1992)
2. Boulay, F., Lauquin, G.J. & Vignais, P.V. Localization of immunoreactive regions in the beef heart adenine nucleotide carrier using rabbit antisera against the carboxyatractyloside-liganded and the sodium dodecyl sulfate denaturated carrier forms. Biochemistry, 25, 7567-7571 (1986)
doi:10.1021/bi00371a046
PMid:2432931
3. Levy, S.E., Chen, Y.S., Graham, B.H. & Wallace, D.C. Expression and sequence analysis of the mouse adenine nucleotide translocase 1 and 2 genes. Gene, 254, 57-66 (2000)
doi:10.1016/S0378-1119(00)00252-3
PMid:10974536
4. Huang, S.G., Odoy, S. & Klingenberg, M. Chimers of two fused ADP/ATP carrier monomers indicate a single channel for ADP/ATP transport. Arch Biochem Biophys, 394, 67-75 (2001)
doi:10.1006/abbi.2001.2520
PMid:11566029
5. Block, M.R. & Vignais, P.V. Substrate-site interactions in the membrane-bound adenine-nucleotide carrier as disclosed by ADP and ATP analogs. Biochim Biophys Acta, 767, 369-376 (1984)
doi:10.1016/0005-2728(84)90207-X
6. Kaukonen, J., Juselius, J.K., Tiranti, V., Kyttala, A., Zeviani, M., Comi, G.P., Keranen, S., Peltonen, L. & Suomalainen, A. Role of adenine nucleotide translocator 1 in mtDNA maintenance. Science, 289, 782-785 (2000)
doi:10.1126/science.289.5480.782
PMid:10926541
7. Halestrap, A. Biochemistry: a pore way to die. Nature, 434, 578-579 (2005)
doi:10.1038/434578a
PMid:15800609
8. Kroemer G, Galluzzi L, & Brenner C. Mitochondrial membrane permeabilization in cell death. Physiol Rev. 2007;87:99-163 (2007)
doi:10.1152/physrev.00013.2006
PMid:17237344
9. Zamzami, N. & Kroemer, G. The mitochondrion in apoptosis: how Pandora's box opens. Nat Rev Mol Cell Biol, 2, 67-71 (2001)
doi:10.1038/35048073
PMid:11413468
10. Marzo, I., Brenner, C., Zamzami, N., Jurgensmeier, J.M., Susin, S.A., Vieira, H.L., Prevost, M.C., Xie, Z., Matsuyama, S., Reed, J.C. & Kroemer, G. Bax and adenine nucleotide translocator cooperate in the mitochondrial control of apoptosis. Science, 281, 2027-2031 (1998)
doi:10.1126/science.281.5385.2027
PMid:9748162
11. Pastorino JG, Tafani M, Rothman RJ, Marcinkeviciute A, Hoek JB, & Farber JL. Functional consequences of the sustained or transient activation by Bax of the mitochondrial permeability transition pore. J Biol Chem. 274:31734-9 (1999)
doi:10.1074/jbc.274.44.31734
PMid:10531385
12. Eskes R, Antonsson B, Osen-Sand A, Montessuit S, Richter C, Sadoul R, Mazzei G, Nichols A, & Martinou JC. Bax-induced cytochrome C release from mitochondria is independent of the permeability transition pore but highly dependent on Mg2+ ions. J Cell Biol. 143:217-24 (1998) doi:10.1083/jcb.143.1.217
doi:10.1083/jcb.143.1.217
PMid:9763433 PMCid:2132823
13. von Ahsen, O., Renken, C., Perkins, G., Kluck, R.M., Bossy-Wetzel, E. & Newmeyer, D.D. Preservation of mitochondrial structure and function after Bid- or Bax-mediated cytochrome c release. J Cell Biol, 150, 1027-1036 (2000) doi:10.1083/jcb.150.5.1027
doi:10.1083/jcb.150.5.1027
PMid:10973993 PMCid:2175243
14. Faustin, B., Rossignol, R., Rocher, C., Benard, G., Malgat, M. & Letellier, T. Mobilization of adenine nucleotide translocators as molecular bases of the biochemical threshold effect observed in mitochondrial diseases. J Biol Chem, 279, 20411-20421 (2004)
doi:10.1074/jbc.M314259200
PMid:14976187
15. Rossignol, R., Malgat, M., Mazat, J.-P. & Letellier, T. Threshold Effect and Tissue Specificity. J Biol Chem 274, 33426-33432 (1999)
doi:10.1074/jbc.274.47.33426
PMid:10559224
16. Letellier, T., Malgat, M., Coquet, M., Moretto, B., Parrot-Roulaud, F. & Mazat, J.P. Mitochondrial myopathy studies on permeabilized muscle fibers. Pediatr Res, 32, 17-22 (1992).
doi:10.1203/00006450-199207000-00004
PMid:1635840
17. Chance, B. & Williams, G.R. Adv. Enzymol., 17, 65-134 (1956)
18. Jurgensmeier, J.M., Xie, Z., Deveraux, Q., Ellerby, L., Bredesen, D. and Reed, J.C. Bax directly induces release of cytochrome c from isolated mitochondria. Proc Natl Acad Sci U S A, 95, 4997-5002 (1998)
doi:10.1073/pnas.95.9.4997
19. Schègger, H. & Pfeiffer, K. Supercomplexes in the respiratory chains of yeast and mammalian mitochondria. EMBO J, 19, 1777-1783 (2000)
doi:10.1093/emboj/19.8.1777
PMid:10775262 PMCid:302020
20. Dolder, M., Walzel, B., Speer, O., Schlattner, U. & Wallimann, T. Inhibition of the mitochondrial permeability transition by creatine kinase substrates. Requirement for microcompartmentation. J Biol Chem, 278, 17760-17766 (2003)
doi:10.1074/jbc.M208705200
PMid:12621025
21. Kroemer, G. & Reed, J.C. Mitochondrial control of cell death. Nat Med, 6, 513-519 (2000)
doi:10.1038/74994
PMid:10802706
21. Kroemer, G. & Reed, J.C. Mitochondrial control of cell death. Nat Med, 6, 513-519 (2000)
doi:10.1038/74994
PMid:10802706
22. Brenner C, Cadiou H, Vieira HL, Zamzami N, Marzo I, Xie Z, Leber B, Andrews D, Duclohier H, Reed JC & Kroemer G. Bcl-2 and Bax regulate the channel activity of the mitochondrial adenine nucleotide translocator. Oncogene. 19:329-36 (2000)
doi:10.1038/sj.onc.1203298
PMid:10656679
23. Cao G, Minami M, Pei W, Yan C, Chen D, O'Horo C, Graham SH, & Chen J. Intracellular Bax translocation after transient cerebral ischemia: implications for a role of the mitochondrial apoptotic signaling pathway in ischemic neuronal death. J Cereb Blood Flow Metab. 2, 321-33 (2001)
doi:10.1097/00004647-200104000-00001
24. Capano M & Crompton M. Biphasic translocation of Bax to mitochondria. Biochem J. 367:169-78 (2002).
doi:10.1042/BJ20020805
PMid:12097139 PMCid:1222873
25. Belzacq AS, Vieira HL, Verrier F, Vandecasteele G, Cohen I, PrÄvost MC, Larquet E, Pariselli F, Petit PX, Kahn A, Rizzuto R, Brenner C & Kroemer G. Bcl-2 and Bax modulate adenine nucleotide translocase activity. Cancer Res. 63:541-6 (2003)
26. Pebay-Peyroula, E., Dahout-Gonzalez, C., Kahn, R., Trezeguet, V., Lauquin, G.J. & Brandolin, G. Structure of mitochondrial ADP/ATP carrier in complex with carboxyatractyloside. Nature, 426, 39-44 (2003)
doi:10.1038/nature02056
PMid:14603310
27. Davidson, A.M. &Halestrap, A.P. Liver mitochondrial pyrophosphate concentration is increased by Ca2+ and regulates the intramitochondrial volume and adenine nucleotide content. Biochem J, 246, 715-723 (1987)
28. Bauer, M.K., Schubert, A., Rocks, O. & Grimm, S. Adenine nucleotide translocase-1, a component of the permeability transition pore, can dominantly induce apoptosis. J Cell Biol, 147, 1493-1502 (1999)
doi:10.1083/jcb.147.7.1493
PMid:10613907 PMCid:2174250
29. Le Bras M, Borgne-Sanchez A, Touat Z, El Dein OS, Deniaud A, Maillier E, Lecellier G, Rebouillat D, Lemaire C, Kroemer G, Jacotot E, & Brenner C. Chemosensitization by knockdown of adenine nucleotide translocase-2. Cancer Res. 66:9143-52 (2006)
doi:10.1158/0008-5472.CAN-05-4407
PMid:16982757
30. Beutner, G., Ruck, A., Riede, B. & Brdiczka, D. Complexes between porin, hexokinase, mitochondrial creatine kinase and adenylate translocator display properties of the permeability transition pore. Implication for regulation of permeability transition by the kinases. Biochim Biophys Acta, 1368, 7-18 (1998)
doi:10.1016/S0005-2736(97)00175-2
<
<
<31. Lange, C. & Hunte, C. Crystal structure of the yeast cytochrome bc1 complex with its bound substrate cytochrome c. Proc Natl Acad Sci U S A, 99, 2800-2805 (2002)
doi:10.1073/pnas.052704699
32. Cabezon, E., Montgomery, M.G., Leslie, A.G. &Walker, J.E. The structure of bovine F1-ATPase in complex with its regulatory protein IF1. Nat Struct Biol, 10, 744-750 (2003)
doi:10.1038/nsb966
PMid:12923572
33. Paumard, P., Vaillier, J., Coulary, B., Schaeffer, J., Soubannier, V., Mueller, D.M., Brethes, D., di Rago, J.P. & Velours, J. (2002) The ATP synthase is involved in generating mitochondrial cristae morphology. Embo J, 21, 221-230.
doi:10.1093/emboj/21.3.221
PMid:11823415 PMCid:125827
34. Ott, M., Robertson, J.D., Gogvadze, V., Zhivotovsky, B. & Orrenius, S. Cytochrome c release from mitochondria proceeds by a two-step process. Proc Natl Acad Sci U S A, 99, 1259-1263 (2002)
doi:10.1073/pnas.241655498
35. Scorrano, L., Ashiya, M., Buttle, K., Weiler, S., Oakes, S.A., Mannella, C.A. & Korsmeyer, S.J. A distinct pathway remodels mitochondrial cristae and mobilizes cytochrome c during apoptosis. Dev Cell, 2, 55-67 (2002)
doi:10.1016/S1534-5807(01)00116-2
PMid:11782314
36. Waterhouse, N.J., Goldstein, J.C., von Ahsen, O., Schuler, M., Newmeyer, D.D. & Green, D.R. Cytochrome c maintains mitochondrial transmembrane potential and ATP generation after outer mitochondrial membrane permeabilization during the apoptotic process. J Cell Biol, 153, 319-328 (2001)
doi:10.1083/jcb.153.2.319
PMid:11309413 PMCid:2169468
37. Goldstein, J.C., Waterhouse, N.J., Juin, P., Evan, G.I. & Green, D.R. The coordinate release of cytochrome c during apoptosis is rapid, complete and kinetically invariant. Nat Cell Biol, 2, 156-162 (2000)
doi:10.1038/35004029
PMid:10707086
38. Ricci, J.E., Munoz-Pinedo, C., Fitzgerald, P., Bailly-Maitre, B., Perkins, G.A., Yadava, N., Scheffler, I.E., Ellisman, M.H. & Green, D.R. Disruption of mitochondrial function during apoptosis is mediated by caspase cleavage of the p75 subunit of complex I of the electron transport chain. Cell, 117, 773-786 (2004)
doi:10.1016/j.cell.2004.05.008
PMid:15186778
39. Scorrano, L., Oakes, S.A., Opferman, J.T., Cheng, E.H., Sorcinelli, M.D., Pozzan, T. & Korsmeyer, S.J. BAX and BAK regulation of endoplasmic reticulum Ca2+: a control point for apoptosis. Science, 300, 135-139 (2003)
doi:10.1126/science.1081208
PMid:12624178
Abbreviations: Abs, antibodies; ADP, adenosine diphosphate; ATP, adenosine triphosphate; ANT, adenine nucleotide translocator; BN-PAGE, blue-native polyacrylamide gel electrophoresis; CATR, carboxyatractyloside; cyt c, cytochrome c; CsA, cyclosporin A; CyD, cyclophiline D; JO2, respiratory rate; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; PTPC, permeability transition pore complex; TMPD, N,N,N',N'-tetramethyl-p-phenylenediamine; OXPHOS, oxidative phosphorylation
Key Words: Apoptosis, Bax, Mitochondria, Metabolism, Respiration
Send correspondence to: Thierry Letellier, 1INSERM U688, Universite Victor Segalen-Bordeaux 2, 146 rue Leo-Saignat, F-33076 Bordeaux Cedex, France, Tel: 33-557571657, Fax: 33-557571431, E-mail:thierry.letellier@u-bordeaux2.fr