Control of Mitochondrial Outer Membrane Permeabilization and Bcl-xL Levels by Thioredoxin 2 in DT40 Cells*

Mitochondria play a central role in the initiation of apoptosis, which is regulated by various factors such as ATP synthesis, reactive oxygen species, redox status, and outer membrane permeabilization. Disruption of chicken thioredoxin 2 (Trx2), a mitochondrial redox-regulating protein, results in apoptosis in DT40 cells. To investigate the mechanism of this apoptosis, we prepared transfectants expressing control (DT40-TRX2-/-), human thioredoxin 2 (TRX2) (DT40-hTRX2), or redox-inactive TRX2 (DT40-hTRX2CS) in conditional Trx2-deficient DT40 cells containing a tetracycline-repressible Trx2 gene. Production of ATP was not significantly changed by down-regulation of Trx2 expression. The generation of reactive oxygen species was enhanced by the down-regulation of Trx2 expression in DT40-TRX2-/-. Unexpectedly, the change was blocked in both DT40-hTRX2 and DT40-hTRX2CS cells. The down-regulation of Trx2 expression caused the release of cytochrome c and apoptosis-inducing factor on day 3, and apoptosis on day 5. These changes were also suppressed in both DT40-hTRX2 and DT40-hTRX2CS cells, suggesting that TRX2 regulates mitochondrial outer membrane permeabilization and apoptosis by redox-active site cysteine-independent mechanisms. The down-regulation of Trx2 expression caused a decrease in the protein level of Bcl-xL on day 3, whereas the protein level of Bcl-2 did not change until day 4, and the mRNA level of Bcl-xL was unchanged. The decrease in Bcl-xL was not blocked by a caspase 3 inhibitor but blocked in both DT40-hTRX2 and DT40-hTRX2CS. These findings indicate a link between the redox active site cysteine-independent action of TRX2 and the level of Bcl-xL in the regulation of apoptosis.

Mitochondria play a central role in the initiation of apoptosis, which is regulated by various factors such as ATP synthesis, reactive oxygen species, redox status, and outer membrane permeabilization. Disruption of chicken thioredoxin 2 (Trx2), a mitochondrial redox-regulating protein, results in apoptosis in DT40 cells. To investigate the mechanism of this apoptosis, we prepared transfectants expressing control (DT40-TRX2؊/؊), human thioredoxin 2 (TRX2) (DT40-hTRX2), or redox-inactive TRX2 (DT40-hTRX2CS) in conditional Trx2-deficient DT40 cells containing a tetracyclinerepressible Trx2 gene. Production of ATP was not significantly changed by down-regulation of Trx2 expression. The generation of reactive oxygen species was enhanced by the down-regulation of Trx2 expression in DT40-TRX2؊/؊. Unexpectedly, the change was blocked in both DT40-hTRX2 and DT40-hTRX2CS cells. The down-regulation of Trx2 expression caused the release of cytochrome c and apoptosis-inducing factor on day 3, and apoptosis on day 5. These changes were also suppressed in both DT40-hTRX2 and DT40-hTRX2CS cells, suggesting that TRX2 regulates mitochondrial outer membrane permeabilization and apoptosis by redox-active site cysteine-independent mechanisms. The downregulation of Trx2 expression caused a decrease in the protein level of Bcl-xL on day 3, whereas the protein level of Bcl-2 did not change until day 4, and the mRNA level of Bcl-xL was unchanged. The decrease in Bcl-xL was not blocked by a caspase 3 inhibitor but blocked in both DT40-hTRX2 and DT40-hTRX2CS. These findings indicate a link between the redox active site cysteine-independent action of TRX2 and the level of Bcl-xL in the regulation of apoptosis.
Mitochondria play an essential role in the control of apoptotic death. In the mitochondrial pathway of apoptosis, there is frequently a dissipation of mitochondrial outer membrane permeabilization (MOMP), 3 accompanied by the release of apoptogenic proteins (cytochrome c (1), AIF (2), Smac/DIABLO (3), and HtrA2/Omi (4)) from mitochondria, the generation of reactive oxygen species (ROS) (5), a change in redox state (6), a release of Ca 2ϩ (7), and morphological alterations of mitochondria (8). Meanwhile, MOMP is considered to be directly controlled by factors such as the Bcl-2 family. Among Bcl-2 family proteins, proapoptotic members such as Bax and Bak increase MOMP during apoptosis, whereas anti-apoptotic members such as Bcl-2 and Bcl-xL inhibit the change in membrane permeability (9). Bcl-xL is reported to be able to bind to and close the voltage-dependent anion channel (VDAC), one of the components of permeability transition (PT) pores on the outer membrane, while blocking Bax/VDAC interaction, VDAC opening, and subsequent cytochrome c release (10). Thus, accumulating evidence indicates a model of regulation through the protein-conduction pores in the outer mitochondrial membrane by Bcl-2 or Bcl-xL (9,11). However, the molecular mechanisms are still not completely understood.
Thioredoxin 2 (TRX2) is a small redox protein containing the thioredoxin-active site Trp-Cys-Gly-Pro-Cys (-WCGPC-). TRX2 is encoded by a nuclear gene and localized to the mitochondria by a mitochondrial leader sequence (12). TRX2 has two cysteines in its active site. Whereas the thioredoxin (TRX) system in the cytosol is composed of TRX, thioredoxin reductase, and peroxiredoxin, TRX2, mitochondrial thioredoxin reductase (TRXR2), and mitochondrial peroxiredoxin 3 (Prdx-3) constitute a redox-reducing system in mitochondria (12)(13)(14)(15). TRX plays an important role in defense against oxidative stress and performs a wide variety of biological functions, such as regulation of gene expression, control of growth, and apoptosis (16). Studies show that the mitochondrial TRX2 system is essential for cell viability. We previously showed that chicken DT40 cells conditionally deficient in Trx2 undergo apoptosis in the absence of exogenous stress, accompanied by an accumulation of intracellular ROS, the activation of caspase 9 and caspase 3, and the release of cytochrome c into the cytosol (17). Overexpression of TRX2 confers resistance to tert-butyhydroperoxide-induced apoptosis in osteosarcoma cells (143B) (18). The overexpression induced an increase in mitochondrial membrane potential and enhanced resistance to etoposide in HEK-293 cells (19). An absence of TRX2 causes massive apoptosis, exencephaly, and early embryonic death in homozygous mice, suggesting a crucial role for TRX2 in cell survival in vivo (20). Cre-mediated inactivation of TRXR2 is also associated with embryonic death. TRXR2-deficient embryos are smaller and severely anemic and show increased apoptosis in the liver (21). These previous results col-lectively show the role of the mitochondrial TRX2 system in the regulation of apoptosis. TRX2 is an important example of how a deficiency in a single gene product causes the induction of mitochondrial apoptosis without any stimuli. Therefore, the investigation of how a deficiency in TRX2 results in apoptosis may provide an insight into the mechanism behind mitochondrial apoptotic events.
Our previous study shows that disruption of Trx2 results in apoptosis in chicken DT40 cells. We here took advantage of this model to investigate the mechanism of the mitochondrial apoptosis, analyzing ATP synthesis, ROS production, membrane permeability transition based on changes in flow cytometric staining with 3,3Ј-dihexyloxacarbocyanine iodide (DiOC6), and the level of Bcl-2 family proteins, Bcl-xL, and Bcl-2. Because TRX2 is considered to play an important part in regulating the redox status of mitochondria, we also analyzed the role of the redox reaction in the regulation of apoptosis by introducing human wild-type TRX2 (hTRX2) or mutant TRX2 (hTRX2CS) in which cysteines in the active site were replaced with serines (-WSGPS-), into conditional Trx2-deficient chicken DT40 cells.
Cell Culture and Transfection-Trx2-deficient chicken DT40 cells were cultured in RPMI 1640 medium (Sigma) supplemented with penicillin, streptomycin, 10% fetal calf serum (Clontech) and 1% chicken serum (JRH, Biosciences) at 39°C under 5% CO 2 . We used the conditional Trx2-knock-out DT40 system (tet-off) in which the expression of a transgenic chicken Trx2 is suppressed by treatment with doxycycline (Dox) (17). The expression of the transgenic chicken Trx2 is suppressed by adding Dox (2 g/ml) into the culture medium at every cell passage. Transfection of vectors (pCMV-hTRX2-FLAG, pCMV-hTRX2CS-FLAG, or control pCMV-FLAG) into Trx2-deficient chicken DT40 cells was performed by electroporation as described previously (22). Drug-resistant colonies were selected with the same medium containing 400 g/ml G418 (Nacalai).
Western Blotting and Immunofluorescence Staining-Western blot analysis was performed as described previously (23). Immunofluorescence staining was performed according to the manufacturer's instructions. Briefly, cells were incubated with 200 nM MitoTracker Red CMXRos for 30 min at 39°C, washed with PBS, and attached to a microscope slide by centrifugation on Cytospin 3 (Shandon). After fixation and permeabilization, the cells were stained with anti-FLAG antibody, followed by FITC-conjugated anti-mouse IgG. Images were acquired with a confocal laser microscope (Leica).
Detection of Apoptosis by Annexin V-FITC Staining-Annexin V-FITC staining was performed with or without Dox treatment for 5 days. The cells were washed with PBS, resuspended in 500 l of binding buffer (10 mM HEPES NaOH, pH 7.4, 150 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , and 1.8 mM CaCl 2 ), incubated for 15 min with 5 l of Annexin V-FITC in the dark, and subjected to flow cytometric analysis (FACS-Calibur, BD Biosciences).
Caspase 3 Activation Assay-The activation of caspase 3 was measured as previously described (24). Briefly, cells were collected and washed twice with ice-cold PBS(Ϫ), and resuspended in lysis buffer (50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 10 mM EGTA, and 50 M digitonin). They were then incubated at 37°C for 10 min. Lysates were centrifuged at 15,000 rpm for 3 min, and supernatants were collected. A total of 10 g of protein was incubated with 50 M Ac-DEVD-MCA at 37°C, and the release of 7-amino-4-methyl-coumarin (AMC) was monitored with a spectrofluorometer (Spectra Fluor, TECAN) using an excitation wavelength of 360 nm and an emission wavelength of 465 nm. As a negative control, 10 l of ICE reaction buffer (50 mM Tris-HCl pH 7.4, 1 mM EDTA, and 10 mM EGTA) was incubated with 50 M Ac-DEVD-MCA at 37°C and monitored in the same way.
Preparation of Cytosolic Fraction-For the cytochrome c and AIF release assay, a soluble cytosolic fraction was prepared as follows. Cells were harvested and washed twice with ice-cold PBS(Ϫ). The cell pellets were suspended in ice-cold sucrose buffer (250 mM sucrose, 10 mM Tris-HCl (pH 7.4), 2 mM EDTA, and protease inhibitor mixture (Roche Applied Science)). The cells were homogenized using a Dounce homogenizer with 30 -50 strokes. Unbroken cells and nuclei were centrifuged at 1,000 ϫ g for 10 min at 4°C. The supernatant was further centrifuged at 12,000 ϫ g for 1 h at 4°C. The resulting supernatant and precipitates were stored as the soluble cytosolic and mitochondrial fraction, respectively.
ATP Production Assay-Cellular ATP levels were assayed using ATPlite (PerkinElmer Life Sciences) and LumiCount according to the manufacturer's instructions with minor modifications. Briefly, 1 ϫ 10 6 cells were washed with ice-cold PBS twice, then resuspended with cell lysis solution. The amount of protein was assayed and adjusted to 0.1 g/l with cell lysis solution. Next, 50 l of cell lysate was added to 100 l of PBS per well of a microplate, and the plate was shaken for 5 min in an orbital shaker at 700 rpm. The substrate solution (50 l) was added to the plate, and the microplate was shaken for 5 min. The plate was left in the dark for 10 min, and the luminescence was measured using an ARVO SX counter (PerkinElmer Life Sciences). The standard curve for ATP was made with a series of dilutions.
Flow Cytometric Estimation of Intracellular ROS-The level of intracellular ROS was monitored by flow cytometric analysis with DCFH-DA, as described previously (23). Cells were incubated in the medium with 5 M DCFH-DA for 15 min at 39°C, washed with PBS, and suspended in 1 ml of PBS. The intensity of fluorescence was analyzed using a flow cytometer.
Measurement of Mitochondrial Membrane Potential-Mitochondrial membrane potential was measured using DiOC6. Briefly, 1 ϫ 10 5 cells were resuspended in 0.5 ml of ice-cold PBS, then cultured with 40 nM DiOC6 for 15 min in 5% CO 2 at 39°C, and analyzed by flow cytometry. Data acquisition and analysis were performed using Cell Quest Software.
Semiquantitative Reverse Transcription-PCR-Total RNA was isolated from cells with a RNeasy Mini Kit (Qiagen). Reverse transcription was performed with a SuperScript III First-Strand Synthesis System kit (Invitrogen). PCRs were carried out using the following oligonucleotide primers: chicken Bcl-xL sense (5Ј-CGTACCAGAGCTTT-GAGCAGGT-3Ј) and antisense (5Ј-GACCAAGCACAAGCACAAT-CAC-3Ј) and chicken glyceraldehyde-3-phosphate dehydrogenase (GAPDH) sense (5Ј-ATGGTGAAAGTCGGAGTCAACGG-3Ј and antisense (5Ј-ACAGTGCCCTTGAAGTGTCC-3Ј). PCR was performed under the following conditions: 96°C for 3 min; 30 cycles of 96°C for 30 s; 63°C for 30 s; 72°C for 30 s; and a final elongation at 72°C for 8 min to amplify the chicken Bcl-xL gene; 96°C for 3 min, 20 cycles of 96°C for 30 s; 55°C for 30 s; 72°C for 30 s; and a final elongation at 72°C for 8 min to amplify the GAPDH gene. The amplification of the product is not saturated at this number of cycles.

Cell Death Induced by Trx2 Deficiency Is Rescued by Reconstitution
with Human TRX2 or TRX2CS-To investigate the mechanism of death in Trx2-deficient cells, we used the conditional Trx2-knock-out DT40 system (tet-off) in which the expression of a transgenic chicken Trx2 is suppressed by treatment with Dox (17). We transfected the control vector or the expression vector for FLAG-tagged human TRX2 (hTRX2) or double mutant human TRX2 (hTRX2CS), which encodes mutated cysteines (-WSGPS-) in the active site (Fig. 1A) into chicken Trx2-deficient DT40 cells to produce stable transfectants: DT40-TRX2Ϫ/Ϫ, DT40-hTRX2, or DT40-hTRX2CS, respectively (Fig. 1B). With Dox treatment, the expression of transgenic chicken Trx2 was hardly detectable on day 3, whereas that of hTRX2-FLAG or hTRX2CS-FLAG remained unchanged (Fig. 1C). Confocal microscopic images showed that both hTRX2-FLAG and hTRX2CS-FLAG proteins are expressed in mitochondria (Fig. 1D). The proliferative nature of the transfectants was then monitored as shown in Fig. 1E. The Trx2knock-out cells (DT40-TRX2Ϫ/Ϫ) transfected with the control vector showed a reduction in viability after Dox treatment. In contrast, the hTRX2-transfected cells (DT40-hTRX2) did not show any change in viability after Dox treatment. The hTRX2CS-transfected cells (DT40-hTRX2CS) did not show a change upon Dox treatment either. These results suggested that human TRX2 can substitute for chicken Trx2 and that not only hTRX2 but also hTRX2CS is capable of supporting the survival of conditional Trx2-deficient DT40 cells.
Suppression of Apoptosis in hTRX2-or hTRX2CS-transfected Cells-We previously reported that a deficiency of Trx2 promotes apoptosis. Therefore, we analyzed the mechanism of the apoptosis, comparing DT40-TRX2Ϫ/Ϫ, DT40-hTRX2, and DT40-hTRX2CS. We first analyzed the staining of Annexin V-FITC in cells cultured with Dox for 5 days using a flow cytometer. Although the staining was enhanced in DT40-TRX2Ϫ/Ϫ after Dox treatment, it was unchanged in DT40-hTRX2 or DT40-hTRX2CS ( Fig. 2A). Downregulation of Trx2 expression on Dox treatment clearly induced activation of caspase-3 in DT40-TRX2Ϫ/Ϫ. In contrast, caspase-3 was not activated in either DT40-hTRX2 or DT40-hTRX2CS (Fig. 2B). All these results showed that either hTRX2 or hTRX2CS protects cells from apoptosis induced by the knock-out of chicken Trx2.
ATP Production Is Unchanged in Trx2-deficient Cells-One of the most important functions of mitochondria is the synthesis of ATP via oxidative phosphorylation during respiration. A previous report suggested that TRX2 interferes with the activity of ATP synthase (19). We then investigated the effect of Trx2 deficiency on ATP synthesis by assessing ATP production in the transfectants. As shown in Fig. 3, treatment with oligomycin, an inhibitor of ATP synthase, suppressed ATP production rapidly in control cells (DT40-TRX2Ϫ/Ϫ without Dox treatment). The down-regulation of Trx2 expression in DT40-TRX2Ϫ/Ϫ caused by Dox treatment did not result in a significant decrease in the production of ATP. In addition, the level of production did not differ significantly between DT40-hTRX2 and DT40-hTRX2CS. These results suggest that the death caused by Trx2 deficiency is not due to a decrease in the production of ATP.
Increased Intracellular Levels of ROS on Loss of Trx2-Because TRX2 is important for reducing intramitochondrial levels of ROS, we next examined the basal intracellular level of ROS by conducting flow cytometric analyses using DCFH-DA. Down-regulation of Trx2 expression augmented the level of ROS in DT40-TRX2Ϫ/Ϫ, whereas no significant increase in ROS was observed in DT40-hTRX2 after 5 days of Dox treatment. We showed the cessation of ROS production also in DT40-hTRX2CS (Fig. 4).
Trx2 Deficiency Caused Disruption of Mitochondrial Membrane Potential (⌬ m )-A previous report showed that overexpression of TRX2 increases mitochondrial membrane potential (⌬ m ) in mammalian cells (19). We then analyzed the ⌬ m in Trx2-deficient cells on day 5 after Dox treatment using DiOC6. We showed that ⌬ m decreases in DT40-TRX2Ϫ/Ϫ. In contrast, no significant change in ⌬ m was detected in either DT40-hTRX2 or DT40-hTRX2CS (Fig.  5A). Moreover, down-regulation of Trx2 expression also induced the release of cytochrome c and AIF into the cytosol in DT40-TRX2Ϫ/Ϫ, whereas the release was blocked in DT40-hTRX2 and DT40-hTRX2CS (Fig. 5B).  MARCH 17, 2006 • VOLUME 281 • NUMBER 11

JOURNAL OF BIOLOGICAL CHEMISTRY 7387
Trx2 Deficiency Caused the Bcl-xL Level to Decrease-Because members of the Bcl-2 family play a crucial role in the regulation of MOMP, we examined the level of Bcl-xL, an important member of this family, using antibodies available for the detection of chicken proteins. As shown in Fig. 6A, the expression of Bcl-xL protein in total cellular lysates decreased on day 3 after Dox treatment in DT40-TRX2Ϫ/Ϫ, whereas the level of tubulin did not change. In contrast, the Bcl-xL level was unchanged in DT40-hTRX2 and DT40-hTRX2CS (Fig. 6A). The level of Bcl-xL did not change until day 2 after Dox treatment and decreased on day 3, although the expression of chicken transgenic Trx2 was hardly detectable at 24 h after Dox treatment (Fig. 6B). The mRNA level of Bcl-xL did not decrease in DT40-TRX2Ϫ/Ϫ, DT40-hTRX2, or DT40-hTRX2CS (Fig. 6C). Treatment with Z-VAD-FMK, a caspase inhibitor, did not influence the decrease in the level of Bcl-xL. The level of Bcl-2 did not change until day 4 after Dox treatment and decreased on day 5 (Fig. 6D). These results indicate that Trx2-deficiency causes a decrease in the level of Bcl-xL protein in a manner independent of the reducing activity of TRX2. We further examined whether Bcl-xL plays an important anti-apoptotic role in DT40-hTRX2 and DT40-hTRX2CS cells by conducting siRNA experiments. Transfection of siRNA targeting Bcl-xL in DT40-hTRX2 and DT40-hTRX2CS cells caused a down-regulation of Bcl-xL expression (Fig. 6E) and resulted in decreased cell numbers, compared with the cells transfected with control siRNA, when these cells were treated with Dox for more than 5 days (Fig. 6F).   A, decrease of mitochondrial membrane potential in Trx2-deficient cells. Mitochondrial membrane potential was analyzed by FACS using DiOC6 on day 0 (dash) and day 5 (line) after Dox treatment. B, Western blot analysis of the release of cytochrome c and AIF into the cytosol. The cells were cultured in the absence or presence of Dox (2 g/ml) for 3 or 5 days. The cytosolic fraction was obtained as described under "Experimental Procedures." The experiment was repeated three times.

DISCUSSION
The chicken Trx2 gene encodes a protein consisting of 151 amino acids, whereas the human TRX2 gene encodes a protein of 166 amino acids. Mature chicken Trx2 is 91.6% identical to its human counterpart (12). Our results showed that human TRX2 could substitute for chicken Trx2 to rescue DT40 cells from death (Fig. 1E). Considering the importance of maintaining the redox environment in mitochondria, the structure and function of TRX2 should be highly conserved.
We showed here that a deficiency of Trx2 causes apoptosis and the activation of caspase 3, accompanied by a decrease in ⌬ m , and the release of AIF and cytochrome c into the cytosol. Intriguingly, these effects were independent of its conserved active site cysteines. As reported in our previous study, Trx2 may interact with cytochrome c to block its release (17). In a glutathione S-transferase pull-down assay using recombinant TRX2 protein, cytochrome c bound to the wild-type TRX2 but not to the mutant TRX2 (supplemental Fig. S1). Because both DT40-hTRX2 and DT40-hTRX2CS displayed resistance against apoptosis, TRX2 may have another regulatory mechanism to prevent apoptosis than the interaction with cytochrome c.
The synthesis of ATP is a requisite function of mitochondria and key to cell survival. However, the role of ATP in the control of apoptosis is debatable. In one report, apoptosis was triggered by inhibitors through a reduction in the level of ATP (25), whereas in other reports, a reduction of ATP blocked apoptosis and switched the mechanism to necrosis (26,27). It was proposed that TRX2 interacts with specific components of the mitochondrial respiratory chain, based on experiments using various inhibitors (19). In our study, we showed that production of ATP was maintained despite a disrupted membrane potential and outer membrane permeabilization (Fig. 3). The timing of the dissipation of mitochondrial inner membrane transmembrane potential and mitochondrial outer membrane permeabilization varied under different circumstances (9). In addition, the drop in production of ATP was reported to occur relatively late in the apoptotic process (28). Another report showed that ATP is required for downstream events in apoptosis (26). Therefore, our results showing that levels of ATP were maintained seem to be consistent with the previous results. We consider that the disruption of mitochondrial membrane potential in Trx2 deficiency is not complete enough to affect the mitochondrial bioenergetic machinery. In addition, an increase in glycolysis may compensate a decrease in mitochondrial ATP synthesis. Such a compensatory mechanism is supported by the report that the addition of glucose to glucose-free medium restored ATP levels in cells cultured with an inhibitor of oxidative phosphorylation (26). It is also possible that other redox regulating systems such as glutaredoxin 2 (29,30) compensate for the loss of ATP caused by the deficiency in Trx2.
ROS are generated in aerobic organisms during metabolism and in response to both internal and external stimuli. Imbalances in the production and removal of ROS have been hypothesized to exert a causative role in cellular toxicity. Mitochondria are an important site for the production of cellular ROS, which is regulated by mitochondrial anti-  MARCH 17, 2006 • VOLUME 281 • NUMBER 11 JOURNAL OF BIOLOGICAL CHEMISTRY 7389 oxidative enzymes, including Mn-superoxide dismutase, the mitochondrial glutathione (GSH)/glutathione peroxidase system, and the mitochondrial-specific TRX2/Prdx-3 system (31).

TRX2 Controls Mitochondrial Membrane Permeabilization
Our previous study showed that a deficiency of Trx2 causes a decrease in the level of GSH in mitochondria and an increase of ROS (17). The study also showed increased production of ROS on the downregulation of Trx2 expression and a cessation of the production on the substitution of TRX2, suggesting that TRX2 is required to suppress the production of ROS. The enhanced production of ROS when there is a deficiency of Trx2 may further aggravate the apoptotic cascade. Considering that the release of cytochrome c into the cytosol was detectable on day 3 (Fig. 5B), the production of ROS may be secondary to the disruption of MOMP, subsequent apoptotic cascade, and down-regulation of antioxidant systems. Consistent with this idea, gene chip data showed that the mRNA expression of various redox enzymes, including superoxide dismutase 2 and glutathione peroxidase 4 is down-regulated on day 5 but not day 3 after Dox treatment in DT40-TRX2Ϫ/Ϫ cells (data not shown). The suppression of ROS production by overexpressed TRX2 may be due to interaction with TRXR2 or Prdx-3, which compose the thiol-reducing system in mitochondria. Paradoxically, the redoxinactive TRX2CS mutant also suppressed the production of ROS, raising the possibility that the suppression of ROS production is caused by a mechanism independent of the redox active site of TRX2. The overexpression of hTRX2 or hTRX2CS may suppress the disruption of MOMP, maintaining the antioxidant systems, including cytosolic TRX and blocking subsequent ROS production.
During the apoptotic process, MOMP occurs (11), liberating multiple death-promoting factors in the mitochondrial intermembrane space into the cytosol. Although the mechanism involved is still unclear, the most current model suggests that MOMP is regulated by a mitochondrial polyprotein channel called the mitochondrial permeability transition (mPT) pore, composed of VDAC on the outer membrane, the inner membrane adenine nucleotide (ADP/ATP) translocator, and the matrix protein cyclophilin D (9). However, a recent study showed that mitochondria from livers of adenine nucleotide (ADP/ATP) translocator knock-out mice still possess mPT pore activity (32). Studies on cyclophilin D-deficient mice indicate that the cyclophilin D-dependent mPT regulates some forms of necrotic death but not apoptotic death (33,34). Therefore, the importance of VDAC in the regulation of MOMP is indicated. Anti-apoptotic Bcl-2 family proteins such as Bcl-2 and Bcl-xL control the release of mitochondrial proteins by preventing permeabilization of the outer mitochondrial membrane (10). Our study showed that Trx2-deficiency causes a decrease of ⌬ m (Fig. 5A), together with the release of cytochrome c and AIF into the cytosol (Fig. 5B). Down-regulation of Trx2 expression also results in a decrease in the level of Bcl-xL. The expression of Trx2 decreased on day 2, whereas that of Bcl-xL decreased on day 3 (Fig. 6, A and B). The release of cytochrome c into the cytosol occurred on day 3, whereas caspase 3 was activated on day 4 after Dox treatment (data not shown). Treatment with a caspase inhibitor, Z-VAD-FMK, blocked activation of caspase 3, but did not change the decrease in Bcl-xL (Fig. 6D). Thus, the decrease in Bcl-xL protein preceded the activation of caspase 3. Bcl-2, another anti-apoptotic protein of the Bcl-2 family, did not change until 4 days after Dox treatment and decreased after 5 days of treatment (Fig. 6D). A schematic model showed the temporal behavior of the apoptotic changes (Fig. 7). Although the expression of many genes changed on day 3 after Dox treatment at the mRNA level (data not shown), the mRNA level of Bcl-xL did not significantly decrease (Fig. 6C). There was also no significant change in the mRNA levels of Prdx-3 and TRXR2 with or without Trx2 expression at the mRNA level (data not shown). We further performed a siRNA experiment and showed that the down-regulation of Bcl-xL expression resulted in a decrease of cell number in both DT40-hTRX2 and DT40-hTRX2CS (Fig. 6, E and F). These results collectively suggest that Bcl-xL plays an important protective role against apoptosis induced by a deficiency of Trx2 in DT40 cells. The mechanism behind the decrease in Bcl-xL protein in total cell lysate is currently unclear. The Bcl-xL protein level is regulated at several steps, including protein synthesis, transport into mitochondria, and protein degradation. Protein stability is a key regulatory mechanism in the control of cell development, the cell cycle, cell growth, and apoptosis. Proteolysis of mitochondrial proteins is regulated by multiple mechanisms (35,36). The ubiquitin/proteasome system regulates molecules such as Mcl-1, a member of the Bcl-2 family in mitochondria (37)(38)(39)(40), indicating the importance of the protease system in the regulation of levels of Bcl-2 family proteins. However, in our preliminary experiments, treatment with proteasome inhibitors did not change the decrease in Bcl-xL (data not shown). The interaction of TRX2 with the protein degradation machinery is speculative. To further elucidate the mechanism, proteomic analyses of binding partners and substrates of Trx2 are underway in our laboratory.
We here showed that TRX2 is a key molecule for maintaining MOMP and the Bcl-xL protein level in a redox active site cysteine-independent manner. Inhibitors of anti-apoptotic protein are considered to have clinical applications. The most recent study shows that ABT-737, a small molecule inhibitor of the anti-apoptotic proteins Bcl-2, Bcl-xL, and Bcl-w, killed cells from lymphoma and small-cell lung carcinoma lines, as well as primary cells from patients (41). In addition, the importance of the control of apoptosis and protein degradation has been

TRX2 Controls Mitochondrial Membrane Permeabilization
emphasized in the study of neuronal diseases such as Parkinson's disease. Elucidation of the mechanism by which apoptosis is induced by a deficiency of Trx2 may provide a novel therapeutic approach against cancer and neuronal diseases.