Recruitment of NF-kappaB into mitochondria is involved in adenine nucleotide translocase 1 (ANT1)-induced apoptosis.

Overexpression of adenine nucleotide translocase-1 (ANT1) is known to induce apoptosis (Bauer, M. K., Schubert, A., Rocks, O., and Grimm, S. (1999) J. Cell Biol. 147, 1493-1501), but the mechanisms involved remain unclear. In this study we show that ANT1 overexpression results in a recruitment of the IkappaBalpha-NF-kappaB complex into mitochondria, with a coincident decrease in nuclear NF-kappaB DNA binding activity. In this situation, NF-kappaB transcriptionally regulated genes with antiapoptotic activity, such as Bcl-XL, MnSOD2, and c-IAP2, are down-regulated, and consequently, cells are sensitized to apoptosis. Accordingly, co-expression of p65 partially interferes with the proapoptotic effect of ANT1 overexpression. Despite the high identity of the two isoforms, overexpression of ANT2 does not exert an apoptotic effect; this lack of apoptotic activity is correlated with the absence of mitochondrial IkappaBalpha-NF-kappaB recruitment or changes in NF-kappaB activity. Thus, we propose that the mitochondrial recruitment of NF-kappaB observed following ANT1 overexpression has an important role in ANT1 proapoptotic activity.

Apoptosis is a form of cell death that plays a role in development, tissue homeostasis, and disease (1). The induction of apoptosis is governed by an elaborate array of checks and balances in the cell. Studies of apoptosis induction in "in vitro" systems have demonstrated that mitochondria are required for the apoptosis stimulated by a variety of different factors (2).
The ANT 1 protein is localized in the inner mitochondrial membrane and exchanges cytosolic ADP for mitochondrial ATP (3). Three isoforms (ANT1, ANT2, and ANT3) with tissuespecific expression patterns have been described in humans (4). ANT interacts with several proteins of the outer mitochondrial membrane (peripheral benzodiazepine receptor, porin/VDAC, and Bax) as well as the matrix (cyclophilin D) to form the permeability transition pore (PTP) (5). The PTP appears to be an important regulator of the apoptotic process. Opening of the pore leads to a loss of mitochondrial transmembrane potential, ⌬⌿ m , which can ultimately culminate in matrix swelling and outer membrane rupture, allowing the release of apoptogenic proteins such as cytochrome c, apoptosis-inducing factor, and procaspases (6,7). Proteins of the bcl-2 family essentially control the release of cytochrome c. Antiapoptotic members of the family (Bcl-2 and Bcl-XL) prevent cytochrome release, whereas the proapoptotic members Bax and Bak exert the opposite effect (8). Bax has been shown to interact with ANT to induce PTP opening and cytochrome c release (9). Several pharmacological compounds interfere with PTP. For instance, cyclosporin A, through its binding to cyclophilin D, prevents PTP opening, and bongkrekic acid and atractyloside are, respectively, a blocker and an inducer of apoptosis via binding of two different conformational states of ANT (10). In addition, alongside their modulation of pore formation by ANT, Bcl-2 and Bax also have been reported to influence ANT ADP/ATP antiporter activity (11). Although ANT is a pore component, recently it has been shown, using mitochondria lacking ANT, that this protein is not essential for MPTP activity (12).
In addition, it has been demonstrated that overexpression of ANT1 (but not ANT2) can induce apoptosis (13); these results demonstrate the specificity of ANT in this pathway. The lack of apoptosis induction by ANT2 contrasts with the high 80% amino acid sequence identity of ANT2 with ANT1 (14). Moreover, the apoptotic activity of ANT1 appears not to depend upon its known role in ADP/ATP exchange because cell death still occurs in several transport-inactive mutants. A critical region of ANT1 (amino acids 102-141) required for apoptosis has been identified (13), and intriguingly, this region overlaps with the Bax binding site of ANT (9) and contains a Vprbinding peptide motif (15). Moreover, these sites also overlap the region of ANT in which the three isoforms exhibit the highest degree of divergence, suggesting an isoform-specific regulation of ANT. Interestingly, hearts of ANT1-deficient mice exhibit a striking hypertrophic cardiomyopathy (16) that could be related to inhibition of apoptosis because of the lack of ANT1 expression in these animals.
Nuclear factor B (NF-B) is a dimeric transcription factor involved in the expression of proteins necessary for innate immunity (17), apoptosis, and cell proliferation (18). NF-B typically forms a heterodimer with the p50 and p65 (RelA) subunits and is mainly regulated by intracellular compartmentalization. The inactive form of NF-B is retained in the cytoplasm upon association with the inhibitory IB␣ protein (19). Exposure of cells to a variety of stimuli, including tumor necrosis factor-␣ (TNF␣), induces phosphorylation of IB␣, which allows subsequent ubiquitination and degradation of the inhibitor, thus leading to nuclear entry and DNA binding of NF-B * This work was supported by Grant PM1999-0171 from the Ministerio de Ciencia y Tecnologia (Spain) and by Grant 2001-SGR-00117 from the Generalitat de Catalunya (Spain). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ Both authors contributed equally to this work and share senior co-authorship.
Using the yeast two-hybrid system, Bottero et al. (21) demonstrated an interaction between IB␣ and mitochondrial ANT. Electron microscopy experiments and biochemical approaches demonstrate the presence of IB␣ associated with p65(NF-B) in the mitochondrial intermembrane space. Intriguingly, NF-B (p50) has been found in liver mitochondria, where it regulates mitochondria-specific gene expression (22). This mitochondrial gene regulation by NF-B may control key processes involved in cell growth and apoptosis.
NF-B plays a central role in the prevention of apoptosis. The induction of antiapoptotic genes, such as Bcl-XL and the caspase inhibitors c-IAP2 and c-IAP1, is associated with nuclear NF-B activity (23). The scavenger enzyme superoxide dismutase MnSOD2 is also regulated by nuclear NF-B and participates in the redox regulation that is essential for protecting cells from apoptosis (24). Inhibition of NF-B activity by E2F1 and c-Myc has been associated with the suppression of MnSOD2 expression; this mechanism is responsible for apoptosis induction (25). Recently, more complex aspects of NF-B regulation have been proposed. For example, studies indicate that NF-B shuttling into and out of the nucleus could play an important role in apoptosis induction (26).
The aim of this study is to determine the molecular mechanisms involved in the induction of apoptosis by ANT1 overexpression in transiently transfected cells. We report here that the overexpression of ANT1 (but not ANT2) potentiates the recruitment of p65(NF-B) into the mitochondria of overexpressing cells. Subsequently, nuclear NF-B activity diminishes the transcription of antiapoptotic genes such as Bcl-XL, MnSOD2, and IAP2. Consequently, overexpression of p65(NF-B) is able to interfere with the proapoptotic activity of ANT1 overexpression. In summary the increase in the mitochondrial NF-B pool induced by ANT1 overexpression plays an important role in the induction of apoptosis by ANT1.
Cell Culture and Transfection-HeLa cells (ATCC-CC-1) were cultured in Dulbecco's modified Eagle's medium containing 100 units/ml penicillin/streptomycin and supplemented with 10% fetal bovine serum (Invitrogen). The cells were incubated under an atmosphere of 95% air and 5% CO 2 at 37°C. Cells were transfected by standard calcium phosphate methods. For 6-well plates a total of 4 g of DNA/well was used. A co-transfected GFP expression plasmid was used in a 1:10 ratio to total DNA to assess the transfection efficiency (around 80% in all cases assayed). To isolate floating cells from attached cells, the medium in the plates was gently swirled to suspend floating cells, which were then harvested by centrifugation, washed, and resuspended with icecold PBS. Attached cells were trypsinized, washed, and resuspended with ice-cold PBS. Total cells resulted from the combination of floating and attached cells. Floating and attached cells were counted using a cytometer (Coulter Electronic Limited, Luton, UK), and the percentage of floating cells was calculated.
Cell Fractionation-HeLa cells were collected, washed with PBS, and centrifuged at 500 ϫ g for 2 min. The pellet was resuspended in 1 ml of homogenization buffer (250 mM sucrose, 1 mM EGTA, 10 mM Hepes, pH 7.4, 1 mM PMSF, Complete Mini protease inhibitor mixture tablets (Roche Applied Science)). Cells were disrupted by 50 strokes in a glass homogenizer and then centrifuged at 1,500 ϫ g for 10 min at 4°C to remove unbroken cells, plasma membranes, and nuclei. Supernatants were further centrifuged at 10,000 ϫ g for 10 min, and the resulting pellet contained the mitochondria-enriched fraction. The supernatant was designated as the cytosolic fraction after it was subjected to further ultracentrifugation at 100,000 ϫ g for 60 min at 4°C. Protein was quantified using the Bradford method (Bio-Rad).
Proteinase K Treatment-Mitochondrial extracts were incubated in homogenization buffer without protease inhibitors in the presence of 15 ng/ml proteinase K (Sigma) for 10 min on ice. Protease inhibitors were added to stop the reaction, and proteins were analyzed by Western blotting as described below.
Assessment of Apoptosis by Flow Cytometry-The viability of transiently transfected cells was analyzed at different times after transfection of the vectors used. Apoptotic cells were detected by flow cytometry after staining with fluorescein isothiocyanate-conjugated annexin V and propidium iodide (PI) using a commercially available kit (Annexin V-FLUOS kit, Roche Applied Science). Cells were considered apoptotic when they were annexin V-positive and PI-negative. Staining of cells by PI was an indicator of the loss of plasma membrane integrity. Flow cytometry was performed using an EPICS-XL-MCL (Beckman Coulter, Inc., Fullerton, CA) cytometer.
Mitochondrial Membrane Potential-The mitochondrial membrane potential (⌬⌿ m ) was measured by flow cytometry using hexamethylindodicarbocyanine iodide (DilC 1 (5)) (Molecular Probes). Transiently transfected cells were loaded with 250 ng/ml DilC 1 (5) for 20 min in Dulbecco's modified Eagle's medium without fetal bovine serum. At the end of the incubation the cells were washed twice in PBS and resuspended in a total volume of 0.5 ml of PBS, and the ⌬⌿ m was analyzed by flow cytometry in a Coulter EPICS-XL-MCL (Beckman Coulter, Inc.).
Nuclear Extract Preparation-Extracts were prepared using a modified method from Dignam et al. (27). For the isolation of nuclear protein extracts HeLa cells were washed with PBS, scraped off the plates in PBS, and briefly centrifuged (16,000 ϫ g, 4°C, 30 s). Cells were washed again with ice-cold PBS, pelleted, and resuspended at 4°C in 400 l of buffer A (10 mM Hepes, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, 0.5 mM PMSF, and protease inhibitors (5 g each of aprotinin, leupeptin, and pepstatin/ml)). Cells were allowed to swell on ice for 15 min; then 25 l of Nonidet P-40 (0.5%) was added, and the suspension was thoroughly mixed for 10 s. The homogenate was centrifuged (16,000 ϫ g, 4°C, 60 s), and the nuclear pellet was resuspended in 50 l of ice-cold buffer C (20 mM Hepes, pH 7.9, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 1 mM PMSF, and protease inhibitors (5 g each of aprotinin, leupeptin, and pepstatin/ ml)). Nuclear lysates were maintained on ice for 15 min with occasional mixing. The nuclear extract was cleared (16,000 ϫ g, 4°C, 5 min), and the supernatant containing the proteins from the nuclear extract was transferred to a fresh tube and stored at Ϫ80°C. Protein concentrations were determined by the Bradford assay (Bio-Rad).
Electrophoretic Mobility Shift Assays-HeLa cells transiently transfected with ANT1pcDNA3, ANT2pcDNA3, or empty pcDNA3 vector were used at various time points following transfection. In separate experiments HeLa cells were stimulated with 20 ng/ml TNF␣ for 60 min before the cells were collected. For gel retardation assays, a doublestranded oligonucleotide containing a consensus NF-B DNA binding site (5Ј-TCT AGA GTT GAG GGG ACT TTC CCA G-3Ј, obtained from Roche Applied Science) was end-labeled using [␣-32 P]dCTP and Klenow enzyme. Nuclear protein extracts (10 g) were incubated for 10 min on ice with binding buffer (25 mM Hepes, pH 7.6, 0.5 mM dithiothreitol, 12.5 mM ZnSO 4 , 50 mM KCl, 1 mg/ml bovine serum albumin, 5% glycerol, 0.1% Nonidet P-40, and 2.5 g of poly(dI-dC) (deoxyinosinic-deoxycytidylic acid)). The DNA probe (30,000 cpm) was added and incubated for 20 min at room temperature in a final volume of 25 l. Samples were run on 5% non-denaturing polyacrylamide gels in 0.5ϫ Tris-borate-EDTA at 350 V and 4°C for 60 min. Supershift experiments were performed by adding anti-p65 antibody (Santa Cruz Biotechnology) to the sample after the initial incubation and keeping the complex on ice for 10 min, followed by addition of radiolabeled oligonucleotide and incubation for 20 min at room temperature. In the competition experiments, 100-fold molar excess of unlabeled double-stranded oligonucleotides was included.

RESULTS
In an effort to understand more fully the mechanism of ANT apoptosis induction in mammalian cells, we transiently transfected cells with individual ANT1-and ANT2-encoding plasmids. First of all, we characterized the apoptotic model induced specifically by ANT1 overexpression.
Comparison of the effects of overexpression of ANT1 and ANT2 in HeLa cells revealed an increase in the percent of detached and floating cells in the culture medium 24 h after transfection with an ANT1 expression construct (Fig. 1A). Overall transfection efficiency, measured by co-transfection with GFP, was about 70 -80% in all transfection assays analyzed, and Western analysis, using a non-isoform-specific antibody that detects both ANT isoforms, demonstrated ANT increases in homogenate and also in mitochondria of HeLa transfected cells, indicating a correct localization of overexpressed protein in the mitochondrial membrane (Fig. 1B).
To characterize the floating and attached cells we analyzed apoptotic parameters such as procaspase-9 cleavage and annexin V-FITC staining at different time points following ANT1 or ANT2 transfection. Analysis was performed in total ANT1transfected cells or in floating cells separated from attached cells. Lysates of both populations were prepared separately as described under "Experimental Procedures." The number of floating cells in the medium after ANT2 or vector transfection was low, and it was not possible to collect sufficient amounts to process for subsequent analysis.
A significant increase in cleaved procaspase-9 was observed in total ANT1-transfected cells and also in floating and attached cells obtained after ANT1 overexpression, whereas the cleavage was negligible after ANT2 or empty vector transfection in HeLa cells (Fig. 1C). These data demonstrate that caspase-9 is involved in apoptosis induced by ANT1. Flow cytometric determinations of DilC 1 (5) were used to analyze changes in mitochondrial membrane potential (⌬⌿ m ) in ANT1-and ANT2-transfected cells. Untransfected control cells treated with the chemical uncoupler CCCP (carbonyl cyanide m-chlorophenylhydrazone) were used to quantify the maximum decrease of ⌬⌿ m in HeLa control cells. We co-transfected ANT1 or ANT2 together with an expression vector for GFP into HeLa cells and analyzed changes in the number of cells with low ⌬⌿ m after 4 h of co-expression. Fig. 3A shows that ANT1-transfected, GFP-positive cells had an increased number of cells with a decreased ⌬⌿ m compared with vectortransfected, GFP-positive cells or ANT2 GFP-positive cells (Fig. 3B). Interestingly, the increase in the number of cells with a decreased ⌬⌿ m was observed in attached ANT1 cells, as well as in ANT1 floating cells. The increase in the number of cells with annexin V labeling and a low ⌬⌿ m , especially in attached ANT1-transfected cells after 4 h of transfection, suggests that this increase is an early and isoform-specific effect of ANT1 overexpression.
Recently it has been described that ANT binds to the IB␣-NF-B complex in mitochondria (21). We hypothesized that as a result of ANT overexpression an increase of mitochondrial IB␣-NF-B complex could be induced. We studied the mitochondrial localization of p65(NF-B) in ANT1 or ANT2 tran-siently transfected cells (Fig. 4). Mitochondria were first isolated from cells and subjected to a proteinase K treatment, whose function was to hydrolyze proteins, hence allowing only NF-B protected by mitochondrial membranes to be detected. Western blot analysis of ANT revealed an increase in expressed protein in homogenates and also in mitochondria from floating and attached HeLa cells. Expression of VDAC and ␤-actin was examined as a control for protein loading. dria, homogenate, or cytosol of HeLa transfected cells. The increase of NF-B localization in mitochondria from HeLa ANT1-transfected cells thus seems to be specific to ANT1 overexpression and could lead to a decrease in nuclear NF-B activity.
To verify the NF-B nuclear binding activity following transient transfection of ANT1 or ANT2, we analyzed nuclear extracts obtained from HeLa cells 24 h after transfection by EMSA (electrophoretic mobility shift assay). Two bands corresponding to NF-B DNA binding were observed in nuclear extracts from cells transfected with empty vector (Fig. 5A).
Specificity of the two DNA binding complexes was assessed in competition experiments by adding an excess of unlabeled NF-B oligonucleotide. The use of a p65-specific antibody led to a reduction in DNA binding without the appearance of a supershifted band. Nuclear extracts from ANT1-and ANT2-overexpressing cells revealed differences in NF-B DNA binding activity; these differences were an important decrease in ANT1-and no changes in ANT2-transfected cells (Fig. 5A).
These data support a recruitment of mitochondrial NF-B because of increased ANT1 protein in transient transfection assays, which then affects nuclear NF-B activity. Moreover, this effect seems to be ANT1 isoform-specific because NF-B activity was neither decreased in nuclear extracts nor increased in mitochondria from ANT2-overexpressing cells.
To analyze the time course of NF-B DNA binding activity, we transiently transfected HeLa cells with ANT1 vector and isolated nuclear extracts at different time points following transfection. As shown in Fig. 5B, ANT1 overexpression at 8, 14, and 24 h caused a dramatic and progressive reduction in NF-B DNA binding activity. EMSAs were also used to compare the levels of nuclear NF-B activity in floating and attached ANT1-overexpressing cells 16 h after transfection. There was a marked decrease in NF-B DNA binding activity in floating cells according to their previously described apoptotic phenotype. Attached cells also showed a decrease in NF-B DNA binding activity that can be associated with the gradual acquisition of apoptotic character that cells exhibit at this time point following ANT1 expression.
To determine whether an inductor of the NF-B pathway, such as TNF␣, is capable of regulating activation signals in cells with a low nuclear NF-B activity, control cells and cells transfected with either vector alone or ANT1 were stimulated with 2 nM TNF␣ for 1 h, and nuclear extracts were obtained. As shown in Fig. 6, TNF␣ causes an increase in NF-B DNA binding activity in control cells or in HeLa cells transfected with empty vector. Moreover, TNF␣ treatment causes an activation of NF-B DNA binding activity in ANT1-transfected cells. However, the NF-B nuclear activity was lower in ANT1 than in vector-transfected cells. These results suggest a diminished nuclear NF-B availability in ANT1-transfected cells.
To examine whether mitochondrial NF-B recruitment in ANT1-overexpressing cells leads to changes in the regulation of antiapoptotic genes, we analyzed changes in the expression of Bcl-XL, MnSOD2, and c-IAP2 in ANT1-transfected cells by RT-PCR. All of these proteins are involved in the antiapoptotic effect of NF-B. The levels of MnSOD2 and c-IAP2 were sig-nificantly reduced in cells transfected with ANT1 after 8 h of overexpression. In contrast, no changes were observed in Bcl-XL expression in this situation (Fig. 7A). After 14 h of ANT1 overexpression we observed a significant decrease in the expression of the Bcl-XL, MnSOD2, and c-IAP2 genes in floating as well as attached ANT1-transfected cells (Fig. 7B). These results suggest that the biological function of NF-B in the regulation of antiapoptotic genes was decreased in ANT1transfected cells.
Finally, we examined whether p65(NF-B) overexpression could compensate the recruitment of NF-B into mitochondria induced by ANT1 expression. For this purpose, we co-transfected HeLa cells with p65(NF-B) vector together with ANT1 or empty vector and analyzed the effects on the number of floating cells, mortality measured by propidium iodide staining, and apoptosis measured as annexin-positive cells. As shown in Fig. 8A, Western blot analysis demonstrates that the expression of p65(NF-B) increases in p65-overexpressing cells. We obtained two bands most probably corresponding to endogenous human p65 and exogenous overexpressed mouse p65. In the same blot ANT overexpression can be observed after 24 h of transfection. As shown in Fig. 8B, the increase in the number of floating cells induced by ANT1 overexpression is reduced by about 50% upon co-transfection with NF-B. When we analyzed annexin-positive cells as a percentage of vectortransfected cells (Fig. 8C), we observed that the increase induced by ANT overexpression could be significantly reduced by co-transfection with NF-B. Moreover quantification of mortality measured by PI staining (Fig. 8D) also demonstrates that co-transfection of NF-B together with ANT1 significantly reduces ANT1 mortality induced by transfection with ANT1 alone. Taken together the results demonstrate that overexpression of NF-B is able to counteract the apoptotic effect of ANT1.

DISCUSSION
The results presented here demonstrate that the ANT1 isoform specifically induces programmed cell death when overexpressed in HeLa cells. These results are essentially in agreement with those of Bauer et al. (13). Furthermore, our results demonstrate that the recruitment of NF-B into mitochondria, specifically by ANT1 and not by ANT2 overexpression, plays an important role in this apoptosis pathway. Thus, the diminished nuclear availability of NF-B observed in ANT1-overexpressing cells results in a down-regulation of antiapoptotic genes, causing cells to become sensitized to induction of apoptosis. Importantly, the overexpression of NF-B reduces the ANT1 apoptotic effect.
Interestingly, after ANT1 transfection cells become detached from the plate, and the number of floating cells increases. Floating cells have typical features of apoptotic cells, such as annexin labeling and cleavage of procaspase-9. Thus, the increase in floating cells could be considered an apoptotic index, as it occurs in response to other apoptotic stimuli (28). When we analyzed attached cells, apoptotic parameters were observed in those that overexpressed ANT1 at 4 h, almost disappearing later by 24 h.
To investigate the molecular mechanism(s) by which ANT1 induces apoptosis, changes in cell number with low ⌬⌿ m after ANT1 or ANT2 transfection were studied. Results demonstrated that only the ANT1 isoform is capable of increasing the number of cells with a low potential. This effect is observed within 4 h of ANT1 overexpression. Thus, the early increase in annexin labeling and in cell number with low ⌬⌿ m observed in attached cells suggests that these cells acquired the phenotype of apoptotic cells early and then were progressively detached from the plate, becoming apoptotic, floating cells.
ANT is a protein that belongs to the mitochondrial mem-

FIG. 4. p65(NF-B) is increased in mitochondria from ANT1overexpressing HeLa cells.
A, mitochondria and cytosol extracts were isolated from HeLa cells 24 h after transfection with ANT1 or vector alone. Floating (F) and attached (A) cells were obtained separately as described under "Experimental Procedures" and were subjected to Western blot analysis using anti-p65, anti-ANT, and anti-VDAC antibodies. Extracts from mitochondrial fractions were incubated with proteinase K for 10 min at 4°C as described under "Experimental Procedures." B, homogenates (H), mitochondrial fractions (M), and cytosol extracts (C) were obtained from HeLa cells 24 h after transfection with ANT2 or vector alone. These fractions were subjected to Western blot analysis using the indicated antibodies. brane carrier family, and it has been shown that overexpression of other members of this family, like uncoupling proteins, could disorganize the inner mitochondrial membrane nonspecifically (29). This disruption could be involved in the death induction observed after protein overexpression. It is important to keep in mind that ANT2 also belongs to the same mitochondrial carrier family and that overexpression of ANT2 is not able to induce apoptosis; therefore, a nonspecific apoptotic effect of ANT1 overexpression can be discounted.
Intriguingly, the ANT2 isoform, which is over 90% identical to ANT1, was inactive for apoptosis and for changing ⌬⌿ m , despite the fact that transfected cells overexpressed ANT2 and that the protein was localized in mitochondria. Together these observations suggest functional differences between ANT isoforms. Moreover, it has been suggested that the ANT2 isoform, unlike ANT1, could import glycolytic ATP into mitochondria, and a differential regulation of ANT2 gene expression has also been demonstrated (30). However, it has also been reported that there are no important differences in the transport activity or kinetic parameters of ANT isoforms (31). In addition, deletion studies of ANT1 demonstrated that apoptosis induction by ANT1 does not depend upon its function as a mitochondrial carrier (13).
Some evidence has been reported indicating different affinities between certain proteins and ANT1 or ANT2 isoforms. For instance, ARL2 and its binding partner (BART) were found to be predominantly associated with ANT1 in mitochondria, and the structurally homologous ANT2 does not bind this complex (32). On the other hand, using specific antibodies it was found that ANT1 might have a higher affinity for cyclophilin D, an MPTP component, which suggests a greater involvement of ANT1 than ANT2 in MPTP activity (33). These observations predict the isoform specificity of ANT-protein interactions and could explain the different activity of specific isoforms in apoptosis induction.
It has been reported that the apoptotic effect of ANT1 is dependent upon MPTP activity because it is inhibited by cyclophilin D co-transfection or by treatment with bongkrekic acid, a known MPTP inhibitor (13). However, overexpression of ANT1 did not produce cell death in yeast, indicating that it is not ANT1 per se that has the ability to directly disturb the cell (13). Thus, the interaction between ANT1 and another protein must be responsible for inducing apoptosis. This putative protein probably inhibits MPTP; thus, ANT1 overexpression would be predicted to displace this protein, leading to MPTP opening. Recently, cyclophilin D has been described as being able to inhibit apoptosis (34). Thus, we can speculate that when ANT1 is overexpressed, cyclophilin D would be displaced from MPTP resulting in MPTP opening.
Recently an intramitochondrial pool of IB␣-NF-B has been demonstrated to be associated with ANT. The interaction was originally detected in a yeast two-hybrid system and confirmed in an in vitro system, although differences in the affinities of ANT isoforms to the complex were not analyzed (21). However, little is known about the function of the mitochondrial NF-B pathway.
We hypothesized that the mitochondrial pool of IB␣-NF-B would increase in response to ANT overexpression. Interestingly, our results demonstrate an increase in the reservoir of mitochondrial p65(NF-B), especially in apoptotic, floating A, nuclear extracts were isolated from HeLa cells transiently transfected with vector, ANT1, or ANT2 after 24 h of overexpression as described under "Experimental Procedures." They were subjected to EMSA using an NF-B consensus oligonucleotide. In competition assays, a 100-fold molar excess of unlabeled competitor oligonucleotide was added to the binding mixture. In supershift assays, nuclear extracts were preincubated with 1 g of an antibody (Ab) recognizing the NF-B subunit p65 prior to the binding reaction. B, nuclear extracts from vector-and ANT1transfected cells after 8, 14, and 24 h of ANT1 overexpression and nuclear extracts from attached (A) and floating (F) cells previously isolated after 16 h of ANT1 overexpression were subjected to EMSA. n.s, nonspecific.
ANT1-overexpressing cells, whereas no such increase was observed when ANT2 was expressed. The increase in mitochondrial IB␣-NF-B results in a decrease of this complex in the cytosol, and that decrease, in turn, has an important repercussion for nuclear NF-B activity. It has been assumed that inactivation of NF-B increases the sensitivity of cells to apoptotic stimuli (25,35,36). In agreement with this suggestion, antiapoptotic genes such as the Bcl-XL, MnSOD2, and c-IAP genes are down-regulated in ANT1-overexpressing cells indicating sensitization to an apoptotic stimulus. Together, these results indicate that the affinity of the IB␣-NF-B complex for ANT protein varies depending upon the ANT isoform.
In addition, the TNF␣ response of NF-B activity is lower in ANT1-overexpressing cells than in vector-transfected cells, suggesting a diminished cytosolic NF-B TNF␣-sensitive pool. We demonstrated that the co-transfection of p65(NF-B) together with ANT1 significantly reduces the apoptosis assessed as changes in the number of floating cells, cellular viability, or annexin V labeling, indicating that the cytosolic IB␣-NF-B pool is limiting in ANT1-transfected cells.
The results presented here highlighting the importance of mitochondrial recruitment of IB␣-NF-B by ANT1 overexpression could also explain the protective effect of cyclophilin D observed when it was co-expressed with ANT1 (34). In our model the overexpression of cyclophilin D would lead to IB␣-NF-B removal from ANT1 binding and consequently a recovery of nuclear NF-B. Thus, antiapoptotic genes up-regulated by NF-B can control apoptosis induction.
On the other hand, ANT1 is highly expressed in skeletal muscle and heart (16), where this protein is one of the most abundant in mitochondria and makes up around 10% of all proteins in the inner mitochondrial membrane (37). In this biological context ANT does not induce apoptosis, and it has an important bioenergetic role. A similar phenomenon occurs in late (24 h) attached transfected cells, as shown in this report, where despite ANT1 overexpression neither mitochondrial recruitment of NF-B nor induction of apoptotic parameters was observed. This suggests that other proteins must regulate ANT-protein interactions and consequently the apoptotic activity of ANT in these biological and experimental situations.
Lastly, three ANT isoforms have been described in humans, ANT1, -2, and -3. We have demonstrated that ANT3, like ANT1, induces apoptosis when overexpressed in HeLa cells (38). In this situation, a decrease in NF-B DNA binding and an increase in NF-B in mitochondria were observed (data not shown). Thus, we propose that as for ANT1 overexpression, recruitment of NF-B into mitochondria could be involved in the induction of apoptosis by ANT3 overexpression.
In summary, our results demonstrate that recruitment of mitochondrial IB␣-NF-B associated with ANT1 overexpression plays a critical role in the induction of apoptosis by ANT1. Our results also indicate the importance of ANT-protein interactions in the biological activity of this mitochondrial protein and suggest that the different isoform affinity for proteins like NF-B could be very important in explaining different ANT isoform activity.