Portal de Periódicos da CAPES

Bcl-2 on the Endoplasmic Reticulum Regulates Bax Activity by Binding to BH3-only Proteins

2003; Elsevier BV; Volume: 278; Issue: 8 Linguagem: Inglês

10.1074/jbc.m208878200

1083-351X

Michael J. Thomenius, Nancy S. Wang, Edmunds Reineks, Zhengqi Wang, Clark Distelhorst,

Mitochondrial Function and Pathology

Bcl-2 family members have been shown to be key mediators of apoptosis as either pro- or anti-apoptotic factors. It is thought that both classes of Bcl-2 family members act at the level of the mitochondria to regulate apoptosis, although the founding anti-apoptotic family member, Bcl-2 is localized to the endoplasmic reticulum (ER), mitochondrial, and nuclear membranes. In order to better understand the effect of Bcl-2 localization on its activity, we have utilized a Bcl-2 mutant that localizes only to the ER membrane, designated Bcl-2Cb5. Bcl-2Cb5 was expressed in MDA-MB-468 cells, which protected against apoptosis induced by the kinase inhibitor, staurosporine. Data presented here show that Bcl-2Cb5 inhibits this process by blocking Bax activation and cytochromec release. Furthermore, we show that Bcl-2Cb5 can inhibit the activation of a constitutively mitochondrial mutant of Bax, indicating that an intermediate between Bcl-2 on the ER and Bax on the mitochondria must exist. We demonstrate that this intermediate is likely a BH3-only subfamily member. Data presented here show that Bcl-2Cb5 can sequester a constitutively active form of Bad (Bad3A) from the mitochondria and prevent it from activating Bax. These data suggest that Bcl-2 indirectly protects mitochondrial membranes from Bax, via BH3-only proteins. Bcl-2 family members have been shown to be key mediators of apoptosis as either pro- or anti-apoptotic factors. It is thought that both classes of Bcl-2 family members act at the level of the mitochondria to regulate apoptosis, although the founding anti-apoptotic family member, Bcl-2 is localized to the endoplasmic reticulum (ER), mitochondrial, and nuclear membranes. In order to better understand the effect of Bcl-2 localization on its activity, we have utilized a Bcl-2 mutant that localizes only to the ER membrane, designated Bcl-2Cb5. Bcl-2Cb5 was expressed in MDA-MB-468 cells, which protected against apoptosis induced by the kinase inhibitor, staurosporine. Data presented here show that Bcl-2Cb5 inhibits this process by blocking Bax activation and cytochromec release. Furthermore, we show that Bcl-2Cb5 can inhibit the activation of a constitutively mitochondrial mutant of Bax, indicating that an intermediate between Bcl-2 on the ER and Bax on the mitochondria must exist. We demonstrate that this intermediate is likely a BH3-only subfamily member. Data presented here show that Bcl-2Cb5 can sequester a constitutively active form of Bad (Bad3A) from the mitochondria and prevent it from activating Bax. These data suggest that Bcl-2 indirectly protects mitochondrial membranes from Bax, via BH3-only proteins. Bcl-2 family members comprise a group of proteins that regulate apoptosis. Bcl-2 family members can be grouped into three categories, the anti-apoptotic members including Bcl-2, Bcl-xL, and Mcl-1, the multidomain pro-apoptotic members such as Bax and Bak, and the BH3 domain only proteins such as Bim, Bid, Bad, and Bik (1Adams J.M. Cory S. Science. 1998; 281: 1322-1326Google Scholar). This family of proteins converges on the mitochondria to regulate events like cytochrome c release or mitochondrial membrane depolarization that ultimately determine cell fate. The Bax subfamily has been implicated as the “gateway” to apoptosis (2Wei M.C. Zong W.X. Cheng E.H. Lindsten T. Panoutsakopoulou V. Ross A.J. Roth K.A. MacGregor G.R. Thompson C.B. Korsmeyer S.J. Science. 2001; 292: 727-730Google Scholar). These proteins are required to initiate most forms of apoptosis (2Wei M.C. Zong W.X. Cheng E.H. Lindsten T. Panoutsakopoulou V. Ross A.J. Roth K.A. MacGregor G.R. Thompson C.B. Korsmeyer S.J. Science. 2001; 292: 727-730Google Scholar, 3Zong W.X. Lindsten T. Ross A.J. MacGregor G.R. Thompson C.B. Genes Dev. 2001; 15: 1481-1486Google Scholar). During apoptosis, Bax translocates to the mitochondria and oligomerizes, causing cytochrome c release from the mitochondria (4Antonsson B. Montessuit S. Lauper S. Eskes R. Martinou J.C. Biochem. J. 2000; 345: 271-278Google Scholar, 5Jurgensmeier J.M. Xie Z. Deveraux Q. Ellerby L. Bredesen D. Reed J.C. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 4997-5002Google Scholar, 6Wei M.C. Lindsten T. Mootha V.K. Weiler S. Gross A. Ashiya M. Thompson C.B. Korsmeyer S.J. Genes Dev. 2000; 14: 2060-2071Google Scholar). Translocation and oligomerization of Bax are preceded by a conformational change (7Nechushtan A. Smith C.L. Hsu Y.T. Youle R.J. EMBO J. 1999; 18: 2330-2341Google Scholar). Little is known about what induces this conformational change, although it has been shown that certain members of the BH3 subfamily can bind to and activate Bax and Bak (6Wei M.C. Lindsten T. Mootha V.K. Weiler S. Gross A. Ashiya M. Thompson C.B. Korsmeyer S.J. Genes Dev. 2000; 14: 2060-2071Google Scholar, 8Eskes R. Desagher S. Antonsson B. Martinou J.C. Mol. Cell. Biol. 2000; 20: 929-935Google Scholar). Some evidence suggests that changes in mitochondrial membrane potential induce the activation of Bax (9De Giorgi F. Lartigue L. Bauer M.K. Schubert A. Grimm S. Hanson G.T. Remington S.J. Youle R.J. Ichas F. FASEB J. 2002; 16: 607-609Google Scholar). A prominent theory for Bax activation suggests that anti-apoptotic Bcl-2 family members inhibit Bax conformational change by direct interaction (10Oltvai Z.N. Milliman C.L. Korsmeyer S.J. Cell. 1993; 74: 609-619Google Scholar), although some studies suggest that Bcl-2 and Bcl-xL can inhibit apoptosis without directly interacting with Bax (11Cheng E.H. Levine B. Boise L.H. Thompson C.B. Hardwick J.M. Nature. 1996; 379: 554-556Google Scholar, 12Zha H. Reed J.C. J. Biol. Chem. 1997; 272: 31482-31488Google Scholar). The BH3 subfamily members are characterized by having only one of the four Bcl-2 homology domains. These proteins induce cell death at a point upstream of mitochondria and Bax subfamily activation (13Cheng E.H. Wei M.C. Weiler S. Flavell R.A. Mak T.W. Lindsten T. Korsmeyer S.J. Mol. Cell. 2001; 8: 705-711Google Scholar, 3Zong W.X. Lindsten T. Ross A.J. MacGregor G.R. Thompson C.B. Genes Dev. 2001; 15: 1481-1486Google Scholar). BH3 proteins are either induced transcriptionally (14Puthalakath H. Huang D.C. O'Reilly L.A. King S.M. Strasser A. Mol. Cell. 1999; 3: 287-296Google Scholar) or activated when an apoptotic signal is received (15Zha J. Harada H. Yang E. Jockel J. Korsmeyer S.J. Cell. 1996; 87: 619-628Google Scholar, 16Wang H.G. Pathan N. Ethell I.M. Krajewski S. Yamaguchi Y. Shibasaki F. McKeon F. Bobo T. Franke T.F. Reed J.C. Science. 1999; 284: 339-343Google Scholar, 17Gross A. Yin X.M. Wang K. Wei M.C. Jockel J. Milliman C. Erdjument-Bromage H. Tempst P. Korsmeyer S.J. J. Biol. Chem. 1999; 274: 1156-1163Google Scholar). They are thought to induce apoptosis by binding to anti-apoptotic Bcl-2 family members and inhibiting their activity (18O'Connor L. Strasser A. O'Reilly L.A. Hausmann G. Adams J.M. Cory S. Huang D.C. EMBO J. 1998; 17: 384-395Google Scholar, 19Huang D.C. Strasser A. Cell. 2000; 103: 839-842Google Scholar) or by binding pro-apoptotic family members and inducing their activity (8Eskes R. Desagher S. Antonsson B. Martinou J.C. Mol. Cell. Biol. 2000; 20: 929-935Google Scholar, 20Marani M. Tenev T. Hancock D. Downward J. Lemoine N.R. Mol. Cell. Biol. 2002; 22: 3577-3589Google Scholar). Anti-apoptotic Bcl-2 family members function at least in part by inhibiting cytochrome c release from the mitochondria. They perform this task by preventing translocation and/or activation of Bax-like proteins on the mitochondria (21Mikhailov V. Mikhailova M. Pulkrabek D.J. Dong Z. Venkatachalam M.A. Saikumar P. J. Biol. Chem. 2001; 276: 18361-18374Google Scholar, 22Murphy K.M. Streips U.N. Lock R.B. J. Biol. Chem. 2000; 275: 17225-17228Google Scholar), however the mechanism of this inhibition is not entirely clear. In addition, it is not certain that the role of anti-apoptotic Bcl-2 family members is limited to the mitochondria. Much emphasis has been placed on Bcl-2 function on the mitochondria, although it has been reported that wild type Bcl-2 localizes to the mitochondria, endoplasmic reticulum (ER), 1The abbreviations used are: ER, endoplasmic reticulum; PBS, phosphate-buffered saline; STS, staurosporine; CoxIV, cytochrome c oxidase IV; HA, hemagglutinin; PF, paraformaldehyde; FL, FLAG; GFP, green fluorescent protein; Cb5, cytochrome b 5; z-VAD-FMK, z-Val-AlaDL-Asp-fluoromethylketone 1The abbreviations used are: ER, endoplasmic reticulum; PBS, phosphate-buffered saline; STS, staurosporine; CoxIV, cytochrome c oxidase IV; HA, hemagglutinin; PF, paraformaldehyde; FL, FLAG; GFP, green fluorescent protein; Cb5, cytochrome b 5; z-VAD-FMK, z-Val-AlaDL-Asp-fluoromethylketone and nuclear membranes (23Akao Y. Otsuki Y. Kataoka S. Ito Y. Tsujimoto Y. Cancer Res. 1994; 54: 2468-2471Google Scholar) and there is growing evidence that the ER is important in apoptosis. Significant findings have been gathered suggesting a role for Bcl-2 family members in regulating ER calcium during apoptosis (24Foyouzi-Youssefi R. Arnaudeau S. Borner C. Kelley W.L. Tschopp J. Lew D.P. Demaurex N. Krause K.H. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 5723-5728Google Scholar, 25He H. Lam M. McCormick T.S. Distelhorst C.W. J. Cell Biol. 1997; 138: 1219-1228Google Scholar, 26Lam M. Dubyak G. Chen L. Nunez G. Miesfeld R.L. Distelhorst C.W. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6569-6573Google Scholar, 27Nutt L.K. Pataer A. Pahler J. Fang B. Roth J. McConkey D.J. Swisher S.G. J. Biol. Chem. 2002; 277: 9219-9225Google Scholar, 28Pinton P. Ferrari D. Magalhaes P. Schulze-Osthoff K. Di Virgilio F. Pozzan T. Rizzuto R. J. Cell Biol. 2000; 148: 857-862Google Scholar). Recent evidence suggests that pro-apoptotic Bcl-2 family members like Bik (29Germain M. Mathai J.P. Shore G.C. J. Biol. Chem. 2002; 277: 18053-18060Google Scholar) or Bax and Bak (27Nutt L.K. Pataer A. Pahler J. Fang B. Roth J. McConkey D.J. Swisher S.G. J. Biol. Chem. 2002; 277: 9219-9225Google Scholar) can reside on the ER to regulate apoptosis. Also, many apoptosis-regulating proteins such as caspase-12 (30Nakagawa T. Zhu H. Morishima N. Li E. Xu J. Yankner B.A. Yuan J. Nature. 2000; 403: 98-103Google Scholar) and Bap-31 (31Ng F.W. Nguyen M. Kwan T. Branton P.E. Nicholson D.W. Cromlish J.A. Shore G.C. J. Cell Biol. 1997; 139: 327-338Google Scholar) reside on or in the ER. It has also been reported that Bcl-2 targeted solely to the ER is functional (32Zhu W. Cowie A. Wasfy G.W. Penn L.Z. Leber B. Andrews D.W. EMBO J. 1996; 15: 4130-4141Google Scholar). This ER-targeted form of Bcl-2, Bcl-2Cb5 is protective against many forms of apoptosis and has been shown to inhibit caspase activation and cytochrome c release (33Hacki J. Egger L. Monney L. Conus S. Rosse T. Fellay I. Borner C. Oncogene. 2000; 19: 2286-2295Google Scholar). Previous work has shown that Bcl-2Cb5 can inhibit apoptosis induced by overexpression of Bax (34Wang N.S. Unkila M.T. Reineks E.Z. Distelhorst C.W. J. Biol. Chem. 2001; 276: 44117-44128Google Scholar), suggesting that Bcl-2 can still inhibit the actions of pro-apoptotic family members when localized to the ER. These data demonstrate that Bcl-2, localized to the ER can inhibit mitochondrial events of apoptosis, which point to the presence of an intermediate between Bcl-2 and the mitochondria. In the work reported here, we investigated the mechanism by which Bcl-2 on the ER membrane can inhibit mitochondrial events of apoptosis. We utilized the cytochrome b 5 (Cb5) targeting sequence to target Bcl-2 to the ER, to examine changes in activation and subcellular distribution of pro-apoptotic Bcl-2 family members. We report that Bcl-2Cb5 can inhibit the conformational change of Bax from the ER, suggesting that Bcl-2 inhibits Bax by a mechanism other than direct interaction. Also, evidence presented here indicates that Bax activation, induced by a constitutively active form of Bad (Bad3A) is inhibited by Bcl-2 on the ER. We demonstrate that this inhibition is due to a direct interaction between Bad3A and Bcl-2Cb5. The Bcl-2Cb5 chimera was generated in both FLAG (FL) and GFP tag vectors as described in Ref. 34Wang N.S. Unkila M.T. Reineks E.Z. Distelhorst C.W. J. Biol. Chem. 2001; 276: 44117-44128Google Scholar. pCMV-tag2B, pCMV-tag3A (Stratagene), and pEMD-C1 parental vectors were used. Myc-tagged Bax was generated by subcloning the murine Bax cDNA into the EcoRI and XhoI sites of the pCMV-tag3A vector. Mutations in FL-Bcl-2Cb5 and Myc-Bax were generated using the QuickChangeTM site-directed mutagenesis kit (Stratagene). HA-Bad3A was generated as described (35Zhou X.M. Liu Y. Payne G. Lutz R.J. Chittenden T. J. Biol. Chem. 2000; 275: 25046-25051Google Scholar) in pCDNA3 vector (Stratagene). MDA-MB-468 cells were obtained from American Type Culture Collection. They were cultured in IMEM (BIOSOURCE) with 10% fetal bovine serum, 1%l-glutamine, and 1% non-essential amino acids. MCF7 cells were obtained from American Type Culture Collection. They were cultured in minimal essential medium (MEM) (Invitrogen) with 10% fetal bovine serum, 1% l-glutamine, 1 mm sodium pyruvate. All cells were grown in 7% CO2 at 37 °C. All transfections were performed with the FuGENE 6 transfection reagent (Roche Molecular Biochemicals) following the manufacturer's protocols. A FuGENE 6 to DNA ratio of 3 μl:2 μg was maintained for all experiments. The FuGENE was diluted in Optimem (Invitrogen) prior to reaction with DNA. Stable transfectants of pCMV-tag2B, FL-Bcl-2Cb5, and FL-Bcl-2 vectors into MDA-MB-468 were generated by transient transfection followed by geneticin selection. Following transfection, cells were grown in 0.6 mg/ml geneticin and kept under selective pressure while in culture. Transient transfections of MCF7 cells were performed using FuGENE 6. Cells for co-immunoprecipitation and cross-linking experiments were grown on 100-mm dishes to 50–60% confluency. They were transfected with 8.4 μl of FuGENE 6, 280 μl of Optimem, and 5.6 μg of DNA and left overnight. Cells for microscopy were grown on 4-chamber or 2-chamber Lab-Tek II Chambered Coverglass Slides (Nalge Nunc) to 50–60% confluency. They were transfected using 1.5 μl of FuGENE 6, 50 μl of Optimem, and 1 μg of DNA and incubated overnight. Fluorescence images were taken on a Zeiss Axiovert S100 epifluorescence microscope (Zeiss), with a Zeiss 63Xoil/1.4N.A. Plan Apochromat objective or a Zeiss 40Xoil/1.3N.A. Fluar objective. GFP fluorescence and Alexa 488 (Molecular Probes) fluorescence were detected with an XF23 filter cube (excitation = 485 nm, emission = 535 nm). Filter sets were obtained from Omega Optical. MitoTracker Red and Alexa 594 (Molecular Probes) fluorescence was detected with XF67 dichroic/emission filter and 575DF25 excitation filter (excitation = 575 nm, emission = 630 nm). Images were taken on a Hamamatsu ORCA C4742–95-cooled CCD camera operating with Simple PCI software (Compix, Inc.). Nearest neighbors deconvolution was performed using Simple PCI software. Confocal Microscopy was performed using a Zeiss LSM 510 confocal microscope. Images were captured with the Zeiss LSM510 v3.0 software and viewed using the Zeiss LSM 5 Image Browser. Images were processed using Adobe Photoshop 7.0. MDA-MB-468 cells were grown on 100-mm dishes and treated with 500 nm STS for 6 h. The media were collected, and the remaining cells were washed with PBS. The wash was collected. The remaining cells were trypsinized and combined with the media and wash. The cells were spun down, washed in PBS and resuspended in 4% paraformaldehyde (PF) and placed in 1.5-ml Eppendorf tubes. They were then incubated for 10 min at room temperature. The cells were then spun down, and the PF was removed. The cells were resuspended in ice-cold methanol and incubated for 20 min. The tubes were spun down, the methanol was removed, and the cells were washed in PBS. The cells were incubated in PBS with 5% goat serum for 30 min, followed by a 1-h incubation with a hamster monoclonal anti-human Bcl-2 antibody (6C8 clone, BD PharMingen) and a rabbit anti-active caspase 3 antibody (BD PharMingen). Cells were spun down, washed, and incubated with goat anti-hamster Alexa 647 and goat anti-rabbit Alexa 488. Cells were analyzed on a Becton-Dickinson LSR Flow Cytometer. Flow cytometry data were analyzed using Winlist (Verify Software House) and Microsoft Excel 2000. MDA-MB-468 cells stably transfected with FL-Bcl-2Cb5 or FL-Bcl-2 were grown in 2-chamber Lab-Tek II Chambered Coverglass Slides to 30% confluency. They were then fixed with 4% PF for 10 min. PF was removed, and cells were permeabilized with 0.5% Triton X-100 for 10 min. The cells were washed and incubated with PBS with 5% goat serum for 30 min. The cells were then incubated for 1 h with a 1:500 dilution of anti-Bcl-2 antibody (6C8 clone), followed by two washes in PBS with 5% goat serum. They were then incubated with goat anti-hamster Alexa 594 for 30 min. The cells were washed twice with PBS with 5% goat serum and then incubated with a monoclonal mouse anti-cytochrome c antibody (6H2.B4 clone, BD PharMingen) directly conjugated to Alexa 488. Conjugation was performed using the ZenonTM labeling kit (Molecular Probes). Cells were imaged using confocal microscopy as described previously. MCF7 cells were transfected with MycBaxS184V or cotransfected with GFPBcl-2Cb5 and MycBaxS184V and incubated overnight. The cells were fixed with PF and stained with a monoclonal anti-Myc antibody (Clontech) at a 1:500 dilution as described previously. Cells were also stained with an antibody to cytochromec oxidase IV (CoxIV) (Molecular Probes) as a mitochondrial marker. MycBaxS184V was visualized using a goat anti-mouse-Alexa 594 secondary antibody. CoxIV was visualized by directly binding the primary antibody with Alexa 488 using the Zenon™ labeling kit. Images of the cells were taken using epifluorescence microscopy as described earlier. MCF7 cells were cotransfected with GFP-Cb5/HA-Bad3A, GFP-Cb5/pcDNA3, GFP-Bcl-2Cb5/HA-Bad3A, or GFP-Bcl-2Cb5/ pcDNA3 in 1:1 ratios. The cells were grown overnight (about 15 h) following the transfection. The cells were then fixed and permeabilized as described previously. They were then incubated with a mouse anti-HA antibody, conjugated to Alexa 594 at a 1:500 dilution. The cells were washed with PBS and imaged with epifluorescence microscopy as described previously. Colocalization images were taken with the Zeiss 63Xoil/1.4N.A. Plan Apochromat objective as 0.3-μm increment Z-stacks with as many as 10 images. The images were then run through a nearest neighbors deconvolution algorithm (Compix Inc.) to remove out of focus light. MDA-MB-468 cells transfected with Bcl-2Cb5, Bcl-2, or pCMV-tag2B were treated with 500 nm STS for 6 h. The cells were then fixed and stained as described for colocalization experiments. The cells were incubated with the anti-Bcl-2 antibody (6C8 clone) and anti-cytochrome cantibody (6H2.B4 clone), followed by incubation with goat anti-hamster Alexa 488 and goat anti-mouse Alexa 594. The cells were then imaged with epifluoresence illumination with a Zeiss 40Xoil/1.3N.A. Fluar objective. MCF7 cells were grown to 50–60% confluency on a 100-mm dish and were cotransfected with Myc-BaxS184V/GFPCb5, Myc-BaxS184V/GFP-Bcl-2Cb5, pCMV-tag3A/GFPCb5, or pCMV-tag3A/GFP-Bcl-2Cb5 in a 3:2 ratio. After 12 h of transfection, they were incubated for 30 min with 1 mm BMH (Pierce Chemical) in media at 37 °C. The cells were trypsinized from the dish and washed in PBS. The cells were lysed (1% Triton, 0.1% SDS, 50 mm Tris pH 8.0, 150 mm NaCl, 1 mm EDTA, 1 mm EGTA, phenylmethylsulfonyl fluoride, aprotinin) and then incubated on ice for 30 min. They were then spun down at 300 × g for 5 min at 4 °C to spin down nuclei and unbroken cells. Protein concentration was measured using the Bio-Rad protein assay. The lysate was boiled in sample buffer and run on a 10% SDS-PAGE gel. The gel was transferred onto a polyvinylidene difluoride membrane overnight. The membrane was blocked in 5% nonfat milk in Tris-buffered saline with 0.1% Tween 20 (TBST). The membrane was then probed with a 1:1000 dilution of anti-Myc antibody (Clontech) in TBST with 5% nonfat milk. It was then washed three times in TBST, followed by a 30-min incubation with a 1:1000 dilution of goat anti-mouse horseradish peroxidase-conjugated secondary antibody (Amersham Biosciences). The blot was then exposed on x-ray film (Fujifilm) and scanned into tiff format and processed using Adobe Photoshop 7.0. MCF7 cells were transfected and harvested as described in previous experiments with HA-Bad3A/FL-Bcl-2Cb5, pcDNA3/FL-Bcl-2Cb5, pcDNA3/pCMV-tag2B, HA-Bad3A/pCMV-tag2B, pcDNA3/FL-G145Ecb5, or HA-Bad3A/FL-G145Ecb5. Harvested cells were placed in lysis buffer and homogenized using a Tight Dounce homogenizer (Wheaton) with 25 strokes. The lysates were pre-incubated with 40 μl of protein G-agarose (Roche Molecular Biochemicals) for 3 h to remove nonspecific binding. The lysates were then incubated with anti-Bcl-2 antibody (6C8 clone) for 1 h. 40 μl of protein G-agarose was then added and incubated overnight. The protein G-agarose was spun down and washed once in Buffer 1 (Lysis Buffer) (50 mm Tris-HCl, pH 7.5, 150 mm NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, and protease inhibitor mixture (Roche Molecular Biochemicals)). The samples were washed again in Buffer 2 (50 mm Tris-HCl, pH7.5, 500 mm NaCl, 0.1% Nonidet P-40, and 0.05% sodium deoxycholate) and then once more in Buffer 3 (10 mm Tris-HCl, pH 7.5, and 0.1% Nonidet P-40, and 0.05% sodium deoxycholate). The protein G-agarose was then boiled in sample buffer. The sample was run on an SDS-PAGE gel along with a cell lysate. Western blots were performed as described previously with a 1:2500 dilution of mouse anti-FLAG antibody (Stratagene) and 1:1000 dilution of anti-HA antibody (Covance, HA.11). An anti-mouse horseradish peroxidase secondary antibody was used. In order to study the role of Bcl-2 on the ER, the C-terminal insertion sequence of Bcl-2 was replaced with that of cytochrome b 5, which was shown to restrict the subcellular localization of Bcl-2 to the ER (32Zhu W. Cowie A. Wasfy G.W. Penn L.Z. Leber B. Andrews D.W. EMBO J. 1996; 15: 4130-4141Google Scholar). This finding was confirmed using confocal microscopy (Fig.1). Fig. 1 shows MDA-MB-468 cells stably transfected with either wild type Bcl-2 or ER-targeted Bcl-2 (Bcl-2Cb5). Bcl-2 was imaged using a hamster monoclonal antibody and an Alexa 488-conjugated secondary antibody. In addition, these cells were stained with antibodies to calnexin, which is found in the ER lumen (Fig. 1, A and C) and cytochrome c, which is found in the intermembrane space of the mitochondria in healthy cells (Fig. 1, B and D). These images show that stably overexpressed Bcl-2 has strong colocalization with cytochrome c, but also localizes to the ER. Bcl-2Cb5 does not colocalize with cytochrome c but colocalizes strongly with calnexin, indicating that Bcl-2Cb5 localizes primarily to the ER and not the mitochondria. To study the effects of Bcl-2Cb5 on apoptosis, we assessed caspase-3 activation in cells undergoing apoptosis following treatment with staurosporine (STS). MDA-MB-468 cells, overexpressing Bcl-2 or Bcl-2Cb5 in a mixed population were treated with 500 nm STS for 6 h. These cells were fixed and stained with antibodies to Bcl-2 and active caspase 3. Transfected cells were identified by Bcl-2 antibody staining. Our findings show that expression of Bcl-2 or Bcl-2Cb5 protects cells against STS-induced cell death. Fig. 2shows that cells transfected with Bcl-2Cb5 or Bcl-2 have low active caspase 3 levels (Fig. 2, A and B) and do not have apoptotic nuclei (Fig. 2 A) following STS treatment in comparison to untransfected cells. This finding is consistent with previous results showing the protective effect of Bcl-2Cb5 (32Zhu W. Cowie A. Wasfy G.W. Penn L.Z. Leber B. Andrews D.W. EMBO J. 1996; 15: 4130-4141Google Scholar). We next assessed cytochrome c release in cells treated with STS to demonstrate the effect of ER-Bcl-2 on mitochondrial events of apoptosis. MDA-MB-468 cells were treated with STS and the caspase inhibitor, z-VAD-FMK. z-VAD-FMK was used to maintain cell integrity during the experiment. Cytochrome c release was assessed using immunocytochemistry with paraformaldehyde fixation. Transfected cells were identified by costaining for Bcl-2. Cells overexpressing Bcl-2Cb5 or Bcl-2 did not release cytochromec following STS treatment (Fig.3 A). These data indicate that mitochondrial events of apoptosis are inhibited by Bcl-2Cb5 and are consistent with the hypothesis that there may be some intermediary between the ER and mitochondria, as suggested by Hacki et al. (33Hacki J. Egger L. Monney L. Conus S. Rosse T. Fellay I. Borner C. Oncogene. 2000; 19: 2286-2295Google Scholar). To address the mechanism of Bcl-2Cb5-mediated protection against apoptosis, we measured the activity of the pro-apoptotic Bcl-2 family member Bax using the 6A7 antibody, which preferentially recognizes the active conformation of Bax (36Hsu Y.T. Youle R.J. J. Biol. Chem. 1997; 272: 13829-13834Google Scholar). MDA-MB-468 cells, stably transfected with Bcl-2Cb5 or Bcl-2 in mixed population were treated with 500 nm STS and z-VAD-FMK for 6 h. They were then stained with a Bcl-2 antibody and the 6A7 antibody. We found that vector control cells treated with STS and z-VAD-FMK stained positively for 6A7. Cells overexpressing Bcl-2Cb5 or Bcl-2 did not stain with the 6A7 antibody (Fig. 3 B). These data suggest that Bcl-2Cb5 is inhibiting the activation of Bax, indicating that there is an intermediate between Bcl-2Cb5 and Bax, as the two proteins do not colocalize. We next examined whether Bcl-2Cb5 could inhibit the activation of a constitutively mitochondrial form of Bax. This was done so that we could separate the proteins spatially, ensuring that they cannot interact with one another. For these experiments, we transiently transfected the S184V mutant of Bax into MCF7 cells. This form of Bax has a constitutive localization to the mitochondria (7Nechushtan A. Smith C.L. Hsu Y.T. Youle R.J. EMBO J. 1999; 18: 2330-2341Google Scholar). We first confirmed that Bcl-2Cb5 localized to the ER and not mitochondria following transient transfection in the MCF7 cells, by transfecting GFP-Bcl-2Cb5 and then staining for cytochrome c. GFP-Bcl-2Cb5 and GFP-Cb5 have ER patterns that are distinct from cytochrome c (Fig.4 A). Mitochondrial localization of BaxS184V was confirmed by transfecting Myc-tagged BaxS184V (Myc-BaxS184V) into MCF7 and costaining with an antibody to Myc and an antibody to the mitochondrial protein CoxIV (Fig.4 B). In addition, MCF7 cells were cotransfected with GFP-Bcl-2Cb5 and Myc-BaxS184V (Fig. 4 B). Fig. 4 Bshows that Myc-BaxS184V localizes to the mitochondria and has a different subcellular localization from GFPBcl-2Cb5. The homobifunctional sulfhydryl cross-linker, bismaleimidohexane (BMH) was used to assess Bax oligomerization in MCF7 cells following S184V transfection (Fig. 4 C). MCF7 cells were cotransfected with GFP-Cb5 and pCMV-tag3A (lane 1), GFP-Bcl-2Cb5 and pCMV-tag3A (lane 2), GFP-Cb5 and Myc-BaxS184V (lane 3), or GFP-Bcl-2Cb5 and Myc-BaxS184V (lane 4) and cultured overnight. The cells were then treated with 1 mm BMH and subsequently lysed and run on SDS-PAGE. A Western blot was then performed using an anti-Myc antibody. As seen in Fig. 4 C, there is a striking decrease in the amount of higher molecular weight Bax complexes in cells cotransfected with GFP-Bcl-2Cb5 than in cells transfected with GFP-Cb5. Based on these results, it appears that Bcl-2Cb5 can inhibit BaxS184V oligomerization without a direct interaction as the two proteins are spatially separated in the cell. Based on lanes 3 and 4 in the no BMH blot, it appears that there is an increased amount of MycBaxS184V with transfection of GFP-Bcl-2Cb5. This is most likely due to protection by GFP-Bcl-2Cb5 in cells over-expressing Myc-BaxS184V and only reaffirms the conclusion that GFP-Bcl-2Cb5 is inhibiting the action of Myc-BaxS184V. The inhibition of BaxS184V oligomerization by Bcl-2Cb5 suggests that an intermediate is involved. Likely candidates for this intermediate are the members of the BH3 subfamily. For the following experiments, we used a constitutively active mutant of Bad, Bad3A (35Zhou X.M. Liu Y. Payne G. Lutz R.J. Chittenden T. J. Biol. Chem. 2000; 275: 25046-25051Google Scholar). This mutant has serine to alanine mutations at three of its phosphorylation sites and therefore is not deactivated by phosphorylation. Initially, we assessed the ability of Bad3A to induce Bax activation, using the 6A7 antibody. MCF7 cells were transiently cotransfected with GFP-Cb5 or GFP-Bcl-2Cb5. The cells were then fixed and stained with the Bax 6A7 antibody. GFP expression was used as a marker of cotransfection. Cells were counted for GFP expression and 6A7 staining. Fig. 5 A shows images of cells stained with 6A7 antibody following cotransfection with HABad3A and GFP-Cb5 or GFP-Bcl-2Cb5. When cotransfected with GFP-Cb5, HA-Bad3A induces Bax activation. When cotransfected with GFP-Bcl-2Cb5, there is significantly less Bax activation (Fig. 5, A andB). This result shows that GFP-Bcl-2Cb5 expression inhibits the activation of Bax by HA-Bad3A, indicating that Bcl-2 on the ER can intercede in the activation of Bax by Bad. Next, we examined the subcellular localization of HA-Bad3A during GFP-Bcl-2Cb5 cotransfection. Following cotransfection of GFP-Bcl-2Cb5 or GFP-Cb5 with HA-Bad3A, MCF7 cells were stained with anti-HA antibody fused to Alexa 594. These cells were visualized using epifluorescence microscopy. HA-Bad3A cotransfected with GFP-Cb5 has a mitochondrial pattern seen in Fig. 6 A, distinct from the ER pattern associated with GFP-Cb5. This is in contrast to HA-Bad3A cotransfected with GFP-Bcl-2Cb5. When cotransfected with GFP-Bcl-2Cb5, HA-Bad3A has a reticular pattern and overlaps closely with GFP-Bcl-2Cb5 (Fig. 6, A andB). This indicates that these two proteins are colocalizing on the ER. HA-Bad3A also colocalized with stably expressed Bcl-2 following transient transfection in the 468 cells (data not shown). To assess whether this colocalization could be due to direct binding, we performed a coimmunoprecipitation between the HA-Bad3A and FL-Bcl-2Cb5 to determine if an interaction can occur between these two proteins. Fig. 6 C shows that Fl-Bcl-2Cb5 can bind to HA-Bad3A. This is evident in lane 5 of the blot. When cotransfected, the two proteins co-immunoprecipitate. To demonstrate this further, a mutant of Bcl-2Cb5 was generated that would inhibit binding. The G145E is a mutation in the BH1 domain of Bcl-2 shown to disrupt dimerization with pro-apoptotic Bcl-2 family members (13Cheng E.H. Wei M.C. Weiler S. Flavell R.A. Mak T.W. Lindsten T. Korsmeyer S.J. Mol. Cell. 2001; 8: 705-711Google Scholar). As shown in lane 6 of Fig. 6 C Bad3A does not co-immunoprecipitate with Bcl-2G145E. In addition, Fig.7 A shows that HA-Bad3A maintains mitochondrial localization when cotransfected with Bcl-2Cb5G145E. The cells in Fig. 7 A are stained with an antibody to CoxIV as a mitochondrial marker. Upon cotransfection of Fl-Bcl-2Cb5, HA-Bad3A appears to have a pattern distinct from CoxIV, while HA-Bad3A retains some co-localization with CoxIV following cotransfection with Fl-Bcl-2Cb5G145E. Fig. 7 B shows that the G145E mutant is not effective in inhibiting Bax activation induced by Bad3A overexpression. HA-Bad3A was cotransfected into MCF7 with either pCMV-tag2B, Fl-Bcl-2Cb5, or Fl-Bcl-2Cb5G145E. Cells cotransfected with Fl-Bcl-2Cb5 were resistant to HA-Bad3A-induced Bax activation as measured by 6A7 antibody staining. Cells cotransfected with the G145E mutant were not resistant. These data suggest that an interaction between Bad3A and Bcl-2Cb5 is likely and demonstrates a likely mechanism for how Bcl-2Cb5 inhibits Bad3A-induced Bax activation.Figure 7Interaction between Bcl-2Cb5 and Bad3A is required to inhibit Bax activation. A, HA-Bad3A and Fl-Bcl-2Cb5 or Fl-Bcl-2Cb5G145E were cotransfected into MCF7 cells and fixed with PF. They were stained with an antibody to CoxIV with an Alexa 594 secondary antibody (shown in red) and an antibody to HA, directly conjugated to Alexa 488 (shown in green).B, HA-Bad3A and pCMV-tag2B, Fl-Bcl-2Cb5, or FL-Bcl-2Cb5G145E were cotransfected into MCF7 cells and fixed with PF. The cells were then stained with an antibody to HA, directly conjugated to Alexa 488 and with the active Bax antibody, 6A7 with a 594 secondary antibody. Transfected cells were identified by HA staining. Bax activation was assessed by 6A7 staining. Approximately 100 cells were counted per dish. The data shown are averaged from three separate experiments.View Large Image Figure ViewerDownload (PPT) Previous Bcl-2 targeting studies have provided valuable information about the role of subcellular localization in the anti-apoptotic activity of Bcl-2. The first localization requirement for Bcl-2 was discovered from studies of a C-terminal truncation mutant that localized to the cytoplasm. It was found that the activity of Bcl-2 was dependent on insertion into an intracellular membrane, but there was some residual function of the cytoplasmic form (37Hockenbery D.M. Oltvai Z.N. Yin X.M. Milliman C.L. Korsmeyer S.J. Cell. 1993; 75: 241-251Google Scholar, 38Nguyen M. Millar D.G. Yong V.W. Korsmeyer S.J. Shore G.C. J. Biol. Chem. 1993; 268: 25265-25268Google Scholar). Later studies suggested that Bcl-2 could inhibit apoptosis induced by a variety of stimuli when targeted to either the ER or mitochondrial membranes (32Zhu W. Cowie A. Wasfy G.W. Penn L.Z. Leber B. Andrews D.W. EMBO J. 1996; 15: 4130-4141Google Scholar). These stimuli include ceramide, Myc overexpression, Bax overexpression, and staurosporine. Bcl-2Cb5 was shown to inhibit caspase 3 activation, mitochondrial membrane depolarization, and cytochrome c release (32Zhu W. Cowie A. Wasfy G.W. Penn L.Z. Leber B. Andrews D.W. EMBO J. 1996; 15: 4130-4141Google Scholar, 39Lee S.T. Hoeflich K.P. Wasfy G.W. Woodgett J.R. Leber B. Andrews D.W. Hedley D.W. Penn L.Z. Oncogene. 1999; 18: 3520-3528Google Scholar, 40Annis M.G. Zamzami N. Zhu W. Penn L.Z. Kroemer G. Leber B. Andrews D.W. Oncogene. 2001; 20: 1939-1952Google Scholar). The finding that Bcl-2Cb5 can inhibit apoptosis suggests that Bcl-2 can protect mitochondrial membranes by an indirect mechanism. The data presented in this paper are consistent with the growing evidence that Bcl-2 blocks apoptosis by inhibiting the pro-apoptotic activity of BH3-only proteins. These data show that Bcl-2 can inhibit the action of a BH3-only member at a location distinct from the mitochondria and Bax. Several studies have shown that mutants of Bcl-2 and Bcl-xL that are deficient in Bax binding are still functional (11Cheng E.H. Levine B. Boise L.H. Thompson C.B. Hardwick J.M. Nature. 1996; 379: 554-556Google Scholar,12Zha H. Reed J.C. J. Biol. Chem. 1997; 272: 31482-31488Google Scholar). Recent work by Cheng et al. (13Cheng E.H. Wei M.C. Weiler S. Flavell R.A. Mak T.W. Lindsten T. Korsmeyer S.J. Mol. Cell. 2001; 8: 705-711Google Scholar) showed that mutants of Bcl-xL that bind BH3 proteins but not Bax still retained anti-apoptotic activity, while mutants that could not bind BH3 proteins were not active. These data point to BH3 proteins as intermediates between Bcl-2 and Bax. We propose a model, whereby Bcl-2 protects the mitochondrial membrane by binding the active form of BH3 proteins. As an example, Bad is normally sequestered in the cytoplasm by 14-3-3 until it becomes dephosphorylated and dissociates during interleukin-3 withdrawal (15Zha J. Harada H. Yang E. Jockel J. Korsmeyer S.J. Cell. 1996; 87: 619-628Google Scholar) or disruption of calcium homeostasis (16Wang H.G. Pathan N. Ethell I.M. Krajewski S. Yamaguchi Y. Shibasaki F. McKeon F. Bobo T. Franke T.F. Reed J.C. Science. 1999; 284: 339-343Google Scholar). Bcl-2 most likely sequesters Bad following dephosphorylation, thus preventing unbound Bad from accumulating on the mitochondria and inducing apoptosis. Bcl-2 and Bcl-xL would therefore function as a checkpoint of BH3 protein-mediated apoptosis following initial activation. These data also provide mechanisms for divergent pathways of apoptosis suggested by Lee et al. (39Lee S.T. Hoeflich K.P. Wasfy G.W. Woodgett J.R. Leber B. Andrews D.W. Hedley D.W. Penn L.Z. Oncogene. 1999; 18: 3520-3528Google Scholar), who demonstrated that Bcl-2Cb5 could not inhibit apoptosis by etoposide, while wild-type or mitochondrial-targeted Bcl-2 could. It is likely that this is due to the number of different BH3 proteins with different properties and subcellular localizations. It is possible that some BH3 proteins require Bcl-2 to be on the mitochondria for their activity. For instance, Bbc3/Puma is constitutively on the mitochondria (41Han J. Flemington C. Houghton A.B. Gu Z. Zambetti G.P. Lutz R.J. Zhu L. Chittenden T. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 11318-11323Google Scholar, 42Nakano K. Vousden K.H. Mol Cell. 2001; 7: 683-694Google Scholar), which might make it impossible for Bcl-2Cb5 to intercede in its ability to activate Bax or Bak, as there would be little access of Bcl-2Cb5 to sequester this BH3 proteins. How BH3 domain only proteins induce Bax activation is still unclear. It is possible that BH3 proteins directly bind to Bax, although there are only a few examples of such interactions occurring. Another possibility is that BH3 proteins affect mitochondria or mitochondrial membranes in such a manner as to cause Bax activation. There is evidence that the BH3 protein, Bim induces apoptosis by binding to the VDAC channel, inducing loss of mitochondrial membrane potential (43Sugiyama T. Shimizu S. Matsuoka Y. Yoneda Y. Tsujimoto Y. Oncogene. 2002; 21: 4944-4956Google Scholar). This is supported by data suggesting that Bax oligomerization is preceded by disruption of mitochondrial membrane potential (9De Giorgi F. Lartigue L. Bauer M.K. Schubert A. Grimm S. Hanson G.T. Remington S.J. Youle R.J. Ichas F. FASEB J. 2002; 16: 607-609Google Scholar). This study provides a unique insight into Bcl-2, as it shows that Bcl-2 on the ER can inhibit mitochondrial events of apoptosis by inhibiting the oligomerization of Bax on the mitochondria. The ability of Bcl-2Cb5 to inhibit Bax oligomerization calls into question the notion that Bcl-2 inhibits Bax by a direct biochemical effect on Bax or on mitochondrial membranes. The more likely scenario is that Bcl-2 prevents the pro-apoptotic activity of BH3 proteins, suggesting a model where BH3 proteins activate Bax by either a direct interaction or by affecting mitochondria. More generally, this study begins to address the question of what Bcl-2 is doing on the ER membrane. We thank Dr. Minh Lam and the Confocal Microscopy Core Facility for assistance with confocal microscopy (P30 CA43703-12). We thank Michael Sramkowski, Megan Gottlieb, and the Flow Cytometry Core Facility (NCI CA43703) for assistance with flow cytometry. We are grateful to Dr. Tom Chittenden and Dr. Xiao-Mai Zhou (ImmunoGen, Inc.) for the use of their HA-Bad3A vector. We also thank Dr. Alex Almasan, Dr. John Mieyal, Dr. Yu-Chung Yang, Michael Malone, and Karen McColl for technical and intellectual assistance.