Bioscience Reports, Vol. 21, No. 3, June 2001 ( 2001)
Real-time Measurement of Intracellular Reactive Oxygen Species using Mito tracker Orange (CMH2TMRos) Soo-Mi Kweon,1,2 Hyun-Jung Kim,1 Zee-Won Lee,1 Soo-Jung Kim,1 Seung-IL Kim,1 Sang-Gi Paik,2 and Kwon-Soo Ha1,3 Receiûed March 13, 2001 We have investigated a novel method to monitor real changes of intracellular ROS by the use of CMH2TMRos (a reduced form of MitoTracker orange) in Swiss 3T3 fibroblasts. Arachidonic acid induced a rapid increase of CMTMRos fluorescence with a maximal elevation at 120–150 sec, which was determined by scanning every 10 sec with a confocal microscope. The fluorescence increase by arachidonic acid was completely inhibited by 2-MPG but not by catalase, indicating a major contribution of superoxide to the oxidation of CMH2TMRos. Incubation with glucose oxidase, exogenous H2O2 , KO2 and lysophosphatidic acid also increased the CMTMRos fluorescence, which was blocked by 2-MPG. These results suggested that CMH2TMRos is a useful fluorophore for real-time monitoring of intracellular ROS and also indicated that CMH2TMRos detects primarily superoxide in cells even though the fluorophore can be oxidized by both superoxide and H2O2 . KEY WORDS: Reactive oxygen species; arachidonic acid; lysophosphatidic acid; MitoTracker orange; real-time measurement ABBREVIATIONS: H2DCFDA, 2,7-dichlorodihydrofluorescin diacetate; LPA, lysophosphatidic acid; 2-MPG, N-(2-mercaptopropionyl)-glycine; RFI, relative fluorescence intensity; ROS, reactive oxygen species
INTRODUCTION It has been suggested that reactive oxygen species (ROS) act as an important second messenger in various intracellular signaling [1]. ROS is required for the activation of various enzymes such as NF-κ B, phospholipase A2 and D, protein kinase C, and mitogen-activated protein kinase in response to agonists [2–6]. ROS is also involved in the expression of early growth factor-1 and vascular endothelial growth factors [7, 8], and the activation of Akt, p70S 6K and G-proteins such as Gi, Go [9–11]. In addition, ROS is known to mediate the formation of stress fibers [12, 13] and increase of intracellular Ca2+ [14–16]. 1
Biomolecule Research Team, Korea Basic Science Institute, Yeoeun-dong 52, Yusung-ku, Taejon 305333, Korea. 2 Department of Biology, Chungnam National University, Taejon 305-764, Korea. 3 To whom all correspondence should be addressed. Fax: C82-42-865-3419; E-mail:
[email protected] 341 0144-8463兾01兾0600-0341$19.50兾0 2001 Plenum Publishing Corporation
342
Measurement of intracellular ROS is critical in the investigation of regulation and function of ROS. 2,7-dichlorodihydrofluorescin diacetate (H2DCFDA) has been most widely used to determine intracellular ROS in various cells [4, 7, 16–18]. However, H2DCFDA is not ideal for the precise monitoring of intracellular ROS because the fluorophore is diffused back to extracellular medium after being trapped within cells and is light-sensitive [19, 20]. Dihydrorhodamine 123 has been used to determine the production of ROS in mitochondria [17, 21], but the fluorophore is also diffusive through plasma membrane after being trapped in the cells [19]. Recently, intracellular ROS has been measured by the use of dihydroethidium and MitoTracker red CM-H2XRos. It has been reported that dihydroethidium is oxidized mainly by superoxide and exhibits red fluorescent after binding to DNA [20, 22, 23]. MitoTracker red, originally developed to stain mitochondria [20], has been used to localize mitochondrial ROS in B-lymphoma cells [18]. In addition, intracellular ROS has been measured by electron spin resonance and ferricytochrome c reduction [24]. However, there is no report presenting real-time changes of intracellular ROS. In this report, we present a novel method to monitor the real-time changes of intracellular ROS by the use of CMH2TMRos (the reduced form of MitoTracker orange) in Swiss 3T3 fibroblasts. It is known that CMH2TMRos is oxidized by intracellular ROS into fluorescent CMTMRos, which is sequestered in mitochondria by thiol reactivity of its chloromethyl moiety [20]. Arachidonic acid induced realtime changes of CMTMRos fluorescence, with a maximal increase at 120–150 sec. The increase of CMTMRos fluorescence was completely inhibited by N-(2mercaptopropionyl)-glycine (2-MPG), but not by Aspergillus niger catalase. Catalase has been used to specifically inhibit the increase of intracellular H2O2 [4, 9]. 2-MPG also inhibited real-time changes of CMTMRos fluorescence in response to lysophosphatidic acid (LPA), but catalase did not. Incubation of the cells with glucose oxidase, H2O2 and KO2 elevated the CMTMRos fluorescence, which was inhibited by 2-MPG. These results support that CMH2TMRos is a useful fluorophore to monitor real-time changes of intracellular ROS. MATERIALS AND METHODS Chemicals and Reagents Fetal bovine serum, bovine serum albumin, penicillin兾streptomycin solution, and Dulbecco’s modified Eagle’s medium were obtained from Gibco-BRL (Gaithersburg, MD). Reduced MitoTracker orange (CMH2TMRos) and H2DCFDA were purchased from Molecular Probes (Eugene, OR). Arachidonic acid, LPA, Aspergillus niger catalase, 2-MPG, KO2 , H2O2 and glucose oxidase were from Sigma (St. Louis, MO). Cell Culture Swiss 3T3 fibroblasts, obtained from American Type Culture Collection (ATCC CCL 92), were maintained at 37°C in Dulbecco’s modified Eagle’s medium supplemented with 25 mM HEPES, pH 7.4, 10% (v兾v) fetal bovine serum, 100 unit兾ml
Real-time Measurement of ROS
343
penicillin, and 100 µg兾ml streptomycin (culture medium) [13]. For experiments, cells were cultured on round coverslips in 12-well plates for 2 days and then serumstarved for 1 day with Dulbecco’s modified Eagle’s medium supplemented with 5 µg兾ml apotransferrin, 1 mg兾ml bovine serum albumin, 25 mM HEPES, pH 7.4, 2 mM glutamine 100 units兾ml penicillin, and 100 µg兾ml streptomycin (serum-free medium).
Measurement of Intracellular ROS with H2DCFDA The amount of intracellular H2O2 using H2DCFDA was measured by the procedures of Koo et al. [13]. Briefly, serum-starved cells on round coverslips were stabilized in serum-free medium without phenol red for at least 30 min and stimulated with arachidonic acid. Sometimes, cells were pre-incubated with 5 mM 2-MPG or 500 unit兾ml catalase for 30 min. For the last 5 min of stimulation, 5 µM H2DCFDA was added to measure intracellular H2O2 . The cells were then immediately observed by a laser scanning confocal microscope (Carl Zeiss LSM410). The samples were excited by a 488 nm Ar laser and images were filtered by a longpass 515 nm filter. About thirty cells were randomly selected from three separate experiments and DCF fluorescence intensities of treated cells were compared with those of unstimulated control cells (fold stimulation).
Real-time Measurement of Intracellular ROS with CMH2TMRos For the real-time measurement of intracellular ROS, serum-starved cells were incubated with 150 nM CMH2TMRos for 20 min in serum-free medium without phenol red and washed three times with the same medium. Each coverslip containing stained cells was mounted on a perfusion chamber (supplied by SEC, Seoul, Korea), subjected to a confocal microscope (Carl Zeiss LSM410 or LSM510), and then scanned every 10 sec with a 543 nm He兾Ne laser and a 570 nm longpass emission filter. Agonists were applied to the cells by using a 1-ml syringe during scanning. Serial images (about 30 images) obtained from the continuous scanning were processed by the time series program in the confocal microscope to determine the changes of relative CMTMRos fluorescence (RFI) in mitochondria. The changes of intra-mitochondrial CMTMRos fluorescence were defined as real-time changes of intracellular ROS in these experiments, since CMTMRos fluorescence represents the level of intracellular ROS. Fold stimulation of intracellular ROS was also determined by comparing intensities of the CMTMRos fluorescence. Cells were labeled with CMH2TMRos and scanned every 10 sec by a confocal microscope as explained above. Then, about 20– 30 cells were randomly selected from three separate experiments and the CMTMRos fluorescence intensities at 180 sec wre compared with those at 0 sec (fold stimulation).
344
RESULTS Changes of Intracellular ROS Induced by Arachidonic Acid In order to study the changes of intracellular ROS in response to arachidonic acid, Swiss 3T3 fibroblasts were labeled with a cell-permeable ROS-sensitive fluorophore H2DCFDA and the level of intracellular ROS was determined using a confocal microscope. As shown in Fig. 1A, arachidonic acid strongly increased the DCF fluorescence compared to unstimulated cells, indicating an enhancement of intracellular ROS. However, incubation with A. niger catalase, which has been used to specifically inhibit the increase of intracellular H2O2 [4, 9], blocked the increase of DCF fluorescence by arachidonic acid. These results indicated that H2O2 mainly contributed to the oxidation of H2DCF in response to arachidonic acid. The increase of DCF fluorescence by arachidonic acid was also inhibited by a thiol-reducing agent, 2-MPG. 2-MPG is known to scavenge intracellular ROS by maintaining the cytosolic pool of reduced glutathione [22]. The elevation of intracellular ROS by arachidonic acid was dose-dependent (Fig. 1B). Arachidonic acid induced a significant increase of intracellular ROS by 50 µM and a sub-maximal increase by 100 µM. Swiss 3T3 cells were also incubated with 100 µM arachidonic acid for various times and the level of intracellular ROS from 30 randomly selected cells were measured using a confocal microscope. As shown in Fig. 1C, arachidonic acid significantly increased intracellular ROS at 1 min and induced a maximal increase at 5 min. Then, the increased level slowly decreased over 30 min. Real-time Changes of Intracellular ROS using CMH2TMRos Even though H2DCFDA has been most widely used to detect intracellular ROS in various cells [7, 4, 13, 16, 17, 22], the fluorophore was not ideal for real-time measurement of intracellular ROS because H2DCFDA was easily photo-oxidized into DCF by scanning with a confocal microscope (data not shown). There have been reports suggesting the possible use of MitoTrackers (reduced form), such as MitoTracker red and CMH2TMRos, in the determination of intracellular ROS and redox state, respectively. [18, 25]. Thus, we have applied CMH2TMRos (the reduced form of MitoTracker orange) to the real-time monitoring of intracellular ROS in response to arachidonic acid. To test the possibility whether CMH2TMRos can be used to detect intracellular ROS produced by arachidonic acid, Swiss 3T3 cells were labeled with CMH2TMRos and CMTMRos images were obtained by scanning every 10 sec with a confocal microscope. As shown in Fig. 2A, arachidonic acid induced an apparent increase of CMTMRos fluorescence at 30 sec and a maximal increase at 150 sec. The CMTMRos fluorescence increased in most cells, without any changes of shape and distribution of mitochondria during scanning. However, incubation with control buffer did not cause any changes in the CMTMRos fluorescence (Fig. 2B), showing that the fluorophore was light-resistant in our scanning condition. These results suggested that CMH2TMRos could be used to monitor the changes of intracellular ROS. Thus, about 30 serial images of CMTMRos fluorescence were obtained by scanning every 10 sec and the CMTMRos fluorescence in mitochondria was processed
Real-time Measurement of ROS
Fig. 1. Changes of intracellular ROS by arachidonic acid in Swiss 3T3 cells. (A) Serum-starved cells were incubated with 500 unit兾 ml catalase (CAT) or 5 mM 2-MPG for 30 min, and then treated with 100 µM arachidonic acid (AA) and 5 µM H2DCFDA for 5 min. DCF images were obtained using a confocal microscope as described under Materials and Methods. (B) Cells were incubated with various concentrations of arachidonic acid and 5 µM H2DCFDA for 5 min. (C) Cells were incubated with 100 µM arachidonic acid for the indicated times and stained with 5 µM H2DCFDA for the last 5 min. The level of intracellular ROS (fold stimulation) was determined using a confocal microscope as described under Materials and Methods. Data are the meansJSD from three separate experiments.
345
346
Fig. 2. Real-time changes of CMTMRos fluorescence by arachidonic acid. Serum-starved Swiss 3T3 cells were labeled with 150 nM CMH2TMRos for 20 min and stimulated with 100 µM arachidonic acid (AA). (A) CMTMRos images were obtained by scanning the cells every 10 sec by a confocal microscope as described under Materials and Methods. (B) Serial images (about 30 images) obtained from the continuous scanning were processed to determine the real-time changes of CMTMRos fluorescence by control (open circle) and arachidonic acid (closed circle). Results are expressed as the relative fluorescence intensity (RFI). Each trace is a single cell representative from at least three independent experiments.
to monitor real-time changes of intracellular ROS (Fig. 2B). Arachidonic acid induced a significant increase at 20–30 sec and a maximal increase at 120–150 sec. And then the increased level was continuously maintained. Considering that DCF fluorescence decreased after 5 min (Fig. 1C), these results indicated that oxidation of CMH2TMRos was not reversible after being sequestered in mitochondria. Thus,
Real-time Measurement of ROS
347
it is likely that CMH2TMRos is a useful fluorophore to monitor the real-time production of intracellular ROS. ROS Species Detected by CMH2TMRos In order to investigate ROS species detected by CMH2TMRos, Swiss 3T3 cells were preincubated with ROS scavengers, catalase or 2-MPG, and then investigated their effects on the response of CMTMRos fluorescence to arachidonic acid. As shown in Fig. 3A, incubation with 2-MPG completely inhibited the increase of CMTMRos fluorescence by arachidonic acid. However, contrary to the results obtained by H2DCFDA (Fig. 1A), catalase had no effect on the CMTMRos response to arachidonic acid at any concentration (Fig. 3B). These results suggested two possibilities, since arachidonic acid is known to produce superoxide and H2O2 [26, 27]. First, CMH2TMRos may detect superoxide, but not H2O2 . Secondly, CMH2TMRos may be able to detect both superoxide and H2O2 , but catalase had no effect because CMH2TMRos was primarily oxidized by superoxide produced by arachidonic acid in cells. To test the possibilities, Swiss 3T3 cells were labeled with CMH2TMRos and scanned every 10 sec by a confocal microscope after incubation with glucose oxidase, H2O2 and KO2 . Glucose oxidase increased DCF fluorescence in Swiss 3T3 fibroblasts (data not shown), consistent with the previous reports [9, 24, 28]. Incubation with glucose oxidase, H2O2 , and KO2 largely increased the CMTMRos fluorescence (Fig. 4A) and the increase was inhibited by incubation with 2-MPG (data not shown). Thus, 20–30 cells were randomly selected and relative fluorescence intensity of CMTMRos (fold stimulation) was determined by comparing the fluorescence intensities at 180 sec with those at 0 sec (Fig. 4B). Glucose oxidase increased the level of CMTMRos fluorescence by about 3.5 fold over the control level and the increase was inhibited by 2-MPG. Similar results were also obtained with the incubation of exogenous H2O2 . KO2 , a source of superoxide, also induced a significant increase of intracellular ROS, which as blocked by 2-MPG. Thus, these results supported the second possibility that CMH2TMRos is able to detect both superoxide and H2O2 , but it is mainly oxidized by superoxide in living cells. Real-time Changes of Intracellular ROS by LPA Since the previous results showed that CMH2TMRos was a useful fluorophore to study the changes of intracellular ROS by arachidonic acid, glucose oxidase, H2O2 , and KO2 , we have applied the fluorophore to the real-time changes of intracellular ROS by LPA in Swiss 3T3 cells (Fig. 5). LPA induced a maximal increase of intracellular ROS at 5 min and the increase was inhibited by ROS scavengers, catalase and 2-MPG (data not shown). LPA induced a fast increase of CMTMRos fluorescence with a maximal increase at 30–50 sec, and the increased level was continuously maintained during scanning. We have also studied the effects of ROS scavengers, 2-MPG and catalase, on the CMTMRos response to LPA. As shown in Fig. 5, 2-MPG completely inhibited the elevation of CMTMRos fluorescence by LPA, but catalase had only a minor inhibitory effect. These results suggested that
348
Fig. 3. Effects of 2-MPG (A) and catalase (B) on the real-time changes of CMTMRos fluorescence by arachidonic acid. (A) Serum-starved cells were preincubated with 5 mM 2-MPG for 30 min and labeled with 150 nM CMH2TMRos for 20 min. The stained cells were scanned every 10 sec using a confocal microscope after treating with 100 µM arachidonic acid (AA) and the serial images were processed to determine the real-time changes of CMTMRos fluorescence (RFI). Each trace is a single cell representative from at least three independent experiments. (B) Cells were incubated with various concentrations of A. niger catalase for 30 min and labeled with 150 nM CMH2TMRos for 20 min. The cells were treated with 100 µM arachidonic acid and scanned for 300 sec using a confocal microscope as described under Materials and Methods. Fold stimulation was determined by comparing CMTMRos fluorescence intensities at 180 sec with those at 0 sec as described under Materials and Methods. Results are meansJSD from three independent experiments.
Real-time Measurement of ROS
Fig. 4. Changes of CMTMRos fluorescence by glucose oxidase, H2O2 and KO2 . (A) Cells were labeled with 150 nM CMH2TMRos for 20 min, and treated with 5 unit兾ml glucose oxidase (GO), 1 mM H2O2 , or 2 mM KO2 , and then scanned every 10 sec for 300 sec using a confocal microscope. CMTMRos images shown were scanned at 0 sec (a, b, c) and 180 sec (d, e, f) with glucose oxidase, H2O2 and KO2 , respectively. (B) Cells were incubated with 5 mM 2MPG, labeled with 150 nM CMH2TMRos, and then scanned with 5 unit兾ml glucose oxidase (GO), 1 mM H2O2 , or 2 mM KO2 . Fold stimulation of CMTMRos fluorescence was determined as explained in the legend of Fig. 3B. Results are meanJSD from three independent experiments.
349
350
Fig. 5. Effects of 2-MPG and catalase on the real-time changes of CMTMRos fluorescence by LPA. Cells were incubated with 500 unit兾ml of A. niger catalase (CAT, open circle) or 5 mM 2-MPG (open triangle) for 30 min, and labeled with 150 nM CMH2TMRos for 20 min. The cells were treated with 50 µg兾ml LPA and scanned every 10 sec using a confocal microscope. Real-time changes of CMTMRos fluorescence (RFI) were determined by processing the serial images as described under Materials and Methods. Each trace is a single cell representative from at least three independent experiments.
CMH2TMRos could monitor the real-time changes of intracellular ROS in response to LPA and mainly detected superoxide produced by LPA. DISCUSSION In this report, we have presented a novel method to monitor real-time changes of intracellular ROS using CMH2TMRos (the reduced form of MitoTracker orange) in Swiss 3T3 cells. We could successfully apply the fluorophore to the determination of real-time ROS changes induced by arachidonic acid, LPA, glucose oxidase, and exogenous H2O2 and KO2 . The fluorophore was photo-resistant, which is essential for real-time measurement using confocal microscopes. H2DCFDA and its derivatives, most widely used, are light-sensitive. Thus, CMH2TMRos is a very useful system to determine real changes of intracellular ROS. The responses of CMTMRos fluorescence to arachidonic acid and LPA were much faster than expected, because the maximal increases of DCF fluorescence by arachidonic acid and LPA were observed at 5 min (data not shown for LPA). Arachidonic acid induced a significant increase of CMTMRos fluorescence at 20–30 sec and the fluorescence intensity reached the plateau at 120–150 sec. LPA induced a faster increase of CMTMRos fluorescence than arachidonic acid, with a maximal increase at 30–50 sec. DCF fluorescence is known to represent the level of intracellular H2O2 [4, 18, 22, 23]. Our results suggested that the increase of CMTMRos fluorescence was primarily caused by intracellular superoxide, even though CMH2TMRos can detect both superoxide and H2O2 (Figs. 3 and 4). Thus, it is likely that the production of superoxide in response to agonists is a very early event and
Real-time Measurement of ROS
351
at least a couple of minutes are required to convert superoxide to H2O2 in cells. These results also showed the significance of real-time monitoring system to provide the precise information of intracellular ROS. Interestingly, the increased level of CMTMRos fluorescence did not decline after reaching the plateau, indicating that the oxidation of CMH2TMRos was not reversible. The irreversible reaction oxidation may be a general characteristic of ROS-sensitive fluorophores including CMH2TMRos, MitoTracker red, dihydrorhodamine and dihydroethidium, since they are sequestered in cell organelles after being oxidized by intracellular ROS. It is known that MitoTracker red is oxidized by ROS and trapped in mitochondria by its chloromethyl moiety [18]. Dihydrorhodamine is also oxidized to fluorescent rhodamine 123 and stain mitochondria [17]. Dihydroethidium exhibits blue fluorescent in cytosol and is oxidized by ROS into fluorescent ethidium, which stains nuclei by its binding to DNA [20]. Thus, ROS-sensitive fluorophores, which are sequestered in cell organelles, may be useful to determine the production of intracellular ROS. It has been reported that ROS is produced by electron transport chain in mitochondria [22, 18, 29, 30]. Thus, we have tested the possible role of mitochondria in the production of superoxide by arachidonic acid and LPA in Swiss 3T3 cells. However, rotenone and antimycin A, inhibitors of mitochondrial electron transport chain, had no effect on the ROS production by arachidonic acid and LPA (data not shown). In addition, carbonyl cyanide ρ-trifluoromethoxyphenyl-hydrazine, disrupting mitochondrial membrane potential, had no effect on the ROS formation (data not shown). These results suggested that the primary source of intracellular ROS produced by arachidonic acid and LPA was not electron transport chain. Recently, it has been reported that arachidonic acid activated NADPH oxidase in human neutrophils [31]. It has also been reported that intracellular ROS is produced by various enzymes including monoamine oxidases, tyrosine hydrolase, lipoxygenase, cyclo-oxygenase and xanthin oxidase [29]. Thus, it is possible to propose that arachidonic acid and LPA may produce ROS by the activation of oxidases or hydroxylases in Swiss 3T3 fibroblasts, even though it is necessary to elucidate the enzymes. ACKNOWLEDGMENTS This work was supported in part by the grant from the Korea Science and Engineering Foundation (1999-2-20700-004-5). REFERENCES 1. Thannickal, V. J., Day, R. M., Klinz, S. G., Bastien, M. C., Larios, J. M., and Fanburg, B. L. (2000) FASEB J. 14:1741–1748. 2. Rhee, S. G. (1999) Exp. Mol. Med. 31:53–59. 3. Schreck, R., Rieber, R., and Baeurele, P. A. (1991) EMBO J. 10:2247–2258. 4. Sundaresen, M., Yu, Z. X., Ferrans, V. J., Irani, K., and Finkel, T. (1995) Science 270:296–299. 5. Min, D. S., Kim, E.-G., and Exton, J. H. (1998) J. Biol. Chem. 273:29986–29994. 6. Klann, E., Robertson, E. D., Knapp, L. T., and Sweatt, J. D. (1998) J. Biol. Chem. 273:4516–4522. 7. Ohba, M., Shibanuma, M., Kuroki, T., and Nose, K. (1994) J. Cell. Biol. 126:1079–1088.
352 8. Brauchle, M., Funk, J. O., Kind, P., and Werner, S. (1996) J. Biol. Chem. 271:21793–21797. 9. Bae, G.-U. et al. (1999) J. Biol. Chem. 274:32596–32602. 10. Wang, X., McCullough, K. D., Franke, T. F., and Holbrook, N. J. (2000) J. Biol. Chem. 275: 14624–14631. 11. Nishida, M., Maruyama, Y., Kontani, K., Nagao, T., and Kurose, H. (2000) Nature 408:492–495. 12. Huot, J., Houle, F., Marceau, F., and Landry, J. (1997) Circ. Res. 80:383–392. 13. Koo, H.-Y. et al. (1999) Cell. Signal. 11:677–683. 14. Suzuki, Y. J., Forman, H. J., and Sevanian, A. (1997) Free Rad. Biol. and Med. 22:269–285. 15. Lee, Z. W. et al. (1998) J. Biol. Chem. 273:12710–12715. 16. Lee, Z. W. et al. (2000) Cell. Signal. 12:91–98. 17. Tanaka, K. et al. (1998) J. Biol. Chem. 273:25922–25928. 18. Degli Esposti, M., Hatzinisiriou, I., McLennan, H., and Ralph, S. (1999) J. Biol. Chem. 274:29831– 29837. 19. Royall, J. A. and Ischiropopulos, H. (1993) Arch. Biochem. Biophys. 302:348–355. 20. Haugland, R. P. (1996) in: Handbook of Fluorescent Probes and Research Chemicals, 6th ed., Molecular Probes Inc., Eugene, OR, p. 491–493. 21. Cai, J. and Jones, D. P. (1998) J. Biol. Chem. 273:11401–11404. 22. Vanden Hoek, T. J., Becker, L. B., Shao, Z., Li, C., and Schumacker, P. T. (1998) J. Biol. Chem. 273:18092–18098. 23. Ye, J. et al. (1999) J. Biol. Chem. 274:34974–34980. 24. Li, Y., Zhu, H., Kuppusamy, P., Roubaud, V., Zweier, J. L., and Trush, M. A. (1998) J. Biol. Chem. 273:2015–2023. 25. Smith, J., Ladi, E., Mayer-Pro¨ schel, M., and Noble, M. (2000) Proc. Natl. Acad. Sci. USA 97:10032– 10037. 26. Dana, R., Leto, T. L., Malech, H. L., and Levy, R. (1998) J. Biol. Chem. 273:441–445. 27. Shin, E. A. et al. (1997) FEBS Lett. 452:355–359. 28. Kaszuba, M. and Jones, M. N. (1999) Biochim. Biophys. Acta 1419:221–228. 29. Tan, S., Sagara, Y., Liu, Y., Maher, P., and Schubert, D. (1998) J. Cell. Biol. 141:1423–1432. 30. Deshpande, S. S., Angkeow, P., Huang, J., Ozaki, M., and Irani, K. (2000) FASEB J. 14:1705–1714. 31. Shiose, A. and Sumimoto, H. (2000) J. Biol. Chem. 275:13793–13801.
