Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Mcl-1 Interacts with Truncated Bid and Inhibits Its Induction of Cytochrome c Release and Its Role in Receptor-mediated Apoptosis *

Mcl-1 Interacts with Truncated Bid and Inhibits Its Induction of Cytochrome c Release and Its... THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 281, NO. 9, pp. 5750 –5759, March 3, 2006 © 2006 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. Mcl-1 Interacts with Truncated Bid and Inhibits Its Induction of Cytochrome c Release and Its Role in □ S Receptor-mediated Apoptosis Received for publication, May 24, 2005, and in revised form, November 11, 2005 Published, JBC Papers in Press, December 27, 2005, DOI 10.1074/jbc.M505688200 ‡1,2 ‡2,3 ‡ § ‡4 John G. Clohessy , Jianguo Zhuang , Jasper de Boer , Gabriel Gil-Go´mez , and Hugh J. M. Brady From the Molecular Haematology and Cancer Biology Unit, Institute of Child Health and Great Ormond Street Hospital for Children, University College London, London WC1N 1EH, United Kingdom and Unitat de Biologia Cellular i Molecular, Institut Municipal d’Investigacio´Me`dica-Universitat Pompeu Fabra (IMIM-UPF), E-03003 Barcelona, Spain Engagement of death receptors such as tumor necrosis factor-R1 The BH3 domain-only proteins require cooperation of other multido- and Fas brings about the cleavage of cytosolic Bid to truncated Bid main family members to induce apoptosis (3–7). (tBid), which translocates to mitochondria to activate Bax/Bak, In mammals two distinct apoptotic signaling pathways have been resulting in the release of cytochrome c. The mechanism underlying identified (8, 9). In the extrinsic pathway, apoptosis is initiated through the activation, however, is not fully understood. Here, we have iden- ligand binding to cell surface receptors of the tumor necrosis factor tified the anti-apoptotic Bcl-2 family member Mcl-1 as a potent (TNF) family such as TNF-R1 and Fas. Upon ligation these receptors tBid-binding partner. Site-directed mutagenesis reveals that the initiate the formation of a death-inducing signaling complex, which Bcl-2 homology (BH)3 domain of tBid is essential for binding to consists of adaptor molecules such as the Fas-associated death domain Mcl-1, whereas all three BH domains (BH1, BH2, and BH3) of Mcl-1 protein and procaspase-8. Within the complex, caspase-8 undergoes are required for interaction with tBid. In vitro studies using isolated autoproteolytic activation. Once activated, caspase-8 can activate mitochondria and recombinant proteins demonstrate that Mcl-1 downstream caspases, for example, caspase-3 and -7, leading to orderly strongly inhibits tBid-induced cytochrome c release. In addition to degradation of intracellular substrates and cell death. The cell-intrinsic its ability to interact directly with Bax and Bak, tBid also binds pathway is initiated when the integrity of the outer mitochondrial mem- Mcl-1 and displaces Bak from the Mcl-1-Bak complex. Importantly, brane is lost in response to diverse apoptotic stimuli. This results in the overexpression of Mcl-1 confers resistance to the induction of apo- release of cytochrome c and other apoptotic proteins into the cyto- ptosis by both TRAIL and tumor necrosis factor- in HeLa cells, plasm, where cytochrome c binds to apoptotic protease-activating fac- whereas targeting Mcl-1 by RNA interference sensitizes HeLa cells tor 1 (Apaf-1). Apaf-1 in turn recruits procaspase-9 to form a multim- to TRAIL-induced apoptosis. Therefore, our study demonstrates a eric complex, which leads to the autoproteolytic activation of caspase-9. novel regulation of tBid by Mcl-1 through protein-protein interac- The active caspase-9 then efficiently activates other downstream tion in apoptotic signaling from death receptors to mitochondria. caspases, bringing about the morphological changes characteristic of apoptosis. This intrinsic pathway is thus mitochondria-dependent and tightly controlled by the Bcl-2 family proteins. The commitment of cells to apoptosis in response to diverse physio- Although the two apoptotic pathways can function independently, an logical cues and cytotoxic agents is primarily regulated by proteins of the existing link between them is the BH3 domain-only protein Bid that is Bcl-2 family that are evolutionarily conserved from nematodes to cleaved by active caspase-8 following engagement of death receptor Fas humans (1, 2). Bcl-2 family proteins share one or more Bcl-2 homology (10, 11). Cleaved Bid, also known as truncated Bid (tBid), translocates to (BH) domains and are divided into two main groups based on their pro- mitochondria to induce oligomerization of Bax and/or Bak and cyto- / / or anti-apoptotic activities. The anti-apoptotic members include Bcl-2, chrome c release (4, 12). In both Bax and Bak cells, tBid fails to Bcl-x , A1, Bcl-w, and Mcl-1. The pro-apoptotic family members are induce cytochrome c release and apoptosis, suggesting that it requires further divided according to whether they contain multiple BH domains Bax and/or Bak to exert its mitochondrial pro-apoptotic activity (5). (such as Bax and Bak) or only the BH3 domain (such as Bid and Bim). However, the underlying mechanism and, in particular, the sequence of events that occurs after Bid cleavage and prior to cytochrome c release * This research was supported in part by the Medical Research Council (United Kingdom) are not completely defined. In this study we investigated the mechanism Grant G9900172 (to H. J. M. B.). The costs of publication of this article were defrayed in of tBid-induced activation of the mitochondrial apoptotic pathway by part by the payment of page charges. This article must therefore be hereby marked searching for novel tBid-interacting proteins using a yeast two-hybrid “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains supple- screen. We identified the anti-apoptotic Bcl-2 family protein Mcl-1 as a mental data. genuine tBid-binding partner. Further studies demonstrate that Mcl-1 Funded by a Child Health Research Appeal Trust (CHRAT) studentship from the Insti- tute of Child Health and Great Ormond Street Hospital Special Trustees. effectively inhibits tBid-induced cytochrome c release from mitochon- Both authors contributed equally to this work. dria and protects HeLa cells from apoptosis induced by both tumor Funded by the Great Ormond Street Hospital for Children REACH Fund. To whom correspondence should be addressed: Molecular Haematology and Cancer necrosis factor-related apoptosis-inducing ligand (TRAIL) and TNF-. Biology Unit, Institute of Child Health and Great Ormond Street Hospital for Children, University College London, 30 Guilford St., London WC1N 1EH, UK. Tel.: 44-20- EXPERIMENTAL PROCEDURES 79052731; Fax: 44-20-78138100; E-mail: [email protected]. The abbreviations used are: BH, Bcl-2 homology; TNF, tumor necrosis factor; tBid, trun- Materials and Cell Culture—All media and cell culture reagents were cated Bid; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand; HA, hemagglutinin; GST, glutathione S-transferase; Z-VAD-fmk, benzyloxycarbonyl-VAD- purchased from Invitrogen. Other chemicals, unless otherwise stated, fluoromethyl ketone; CHAPS, 3-[(3-cholamidopropyl)-dimethyl-ammonio]-1-pro- were obtained from Sigma. Human cervical carcinoma HeLa cells were panesulfonate; PARP, poly(ADP-ribose)polymerase; shRNA, short hairpin RNA; mtBH3, mutant BH3. obtained from American Type Culture Collection (Rockville, MD) and 5750 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 9 •MARCH 3, 2006 This is an Open Access article under the CC BY license. Mcl-1 Binds tBid and Blocks Cytochrome c Release cultured in Dulbecco’s modified Eagle’s medium supplemented with [ S]methionine were prepared using the TNT T7 Quick-Coupled 10% fetal bovine serum, 2 mM glutamine, 0.5 units/ml penicillin, and 0.5 Transcription/Translation System (Promega, Madison, WI), using mg/ml streptomycin. pcDNA3.1-HA constructs as templates. The S-labeled prey proteins Cloning of cDNAs and Plasmids Construction—Mouse tBid cDNA were incubated with GST, GST-Bid, or GST-tBid fusion proteins bound was amplified by PCR from a murine pcDNA3-Bid template (a gift from to glutathione-Sepharose beads in bead-binding buffer (50 mM potas- Dr. S. Korsmeyer, Harvard Medical School, Boston, MA) and inserted sium phosphate, pH 7.5, 150 mM KCl, 1 mM MgCl , 10% glycerol, 1% into EcoRI-BamHI sites downstream of the GAL4-DBD in the vector Triton X-100) and protease inhibitors mixture from Roche Applied pGBKT7 (Clontech). A cDNA encoding mouse Bax lacking the C-ter- Science. The mixtures were incubated at 4 °C for 2 h with rotation. The minal hydrophobic domain (Bax21) was amplified by PCR from a beads were then pelleted and washed five times in ice-cold bead-binding pcDNA3-Bax template (13) and cloned downstream of the GAL4-AD buffer. Finally, beads were resuspended in SDS sample buffer, and the in pGADT7 vector (Clontech). A pUC18-bcl-2 construct (Clonexpress, proteins were resolved on SDS-polyacrylamide gels, which were fixed, Gaithersburg, MD) was used to clone the human Bcl-2 cDNA. This vacuum-dried onto 3MM paper, and then visualized using a Phospho- construct was digested with EcoRI and HindIII to release the Bcl-2 rImager (Typhoon 8600, Amersham Biosciences). cDNA, which was then ligated into the corresponding sites in Immunoprecipitation—For each immunoprecipitation experiment, pcDNA3.1/myc-HIS©(-) (Invitrogen). Human Bid and tBid cDNA were HeLa cells were transfected with HA-tBid in pcDNA3.1 or empty vector generated by reverse transcriptase PCR using oligo(dT) primer with in the presence of 75 M Z-VAD-fmk (Enzyme System, Dublin, CA). total RNA obtained from human leukemic Jurkat T cells. Both cDNAs Immunoprecipitation was essentially carried out as described (15). were then cloned at the EcoRI-BamHI sites downstream of the HA-tag Briefly, 24 h after transfection, cells were harvested and resuspended in in pcDNA3.1-HA (Invitrogen) (14). The construct expressing ice-cold lysis buffer containing 2% CHAPS, 20 mM Tris/HCl (pH 7.4), HA-tagged human Mcl-1 was generated as previously described (14). 137 mM NaCl, 2 mM EDTA, 10% glycerol, 1% Triton X-100, and prote- GST fusion protein constructs were generated by PCR using ase inhibitor mixture (Roche Applied Science). Lysates were precleared pcDNA3.1-HA-Bid and -tBid as templates and cloned into the EcoRI- and then incubated with either mouse anti-HA monoclonal antibody XhoI sites in the pGEX-6P-2 vector (Amersham Biosciences). Muta- (clone 12CA5, Roche) or rabbit anti-Mcl-1 polyclonal antibody (Santa tions in the BH domains of tBid and Mcl-1 proteins were generated by Cruz Biotechnology, Santa Cruz, CA) at 4 °C for 1 h, and protein TM site-directed mutagenesis using the GeneTailor site-directed mutagen- A-Sepharose beads (Pharmacia, Piscataway, NJ) were added to pull esis system according to manufacturer’s instruction (Invitrogen). The down the immunocomplexes. The beads were washed five times in G94E mutation of BH3 domain of tBid was generated by PCR using the washing buffer containing 0.2% Triton X-100, 20 mM Tris/HCl (pH 7.4), pGEX-tBid construct as a template. Using pcDNA3.1-HA-Mcl-1 as a 137 mM NaCl, 2 mM EDTA, 10% glycerol before being resuspended in template for PCR of all Mcl-1 mutants, we generated the G262E muta- SDS sample buffer and subjected to SDS-PAGE. Immunoblotting was tion at BH1 domain, the W305A and W312A double mutations for BH2 performed using, where appropriate, goat anti-Bid polyclonal antibody domain, and the G217E and D218A double mutations of BH3 domain of (R&D Systems, Minneapolis, MN), rabbit anti-Mcl-1 polyclonal anti- Mcl-1. The cDNA fragment encoding human Mcl-1 generated by PCR body (Santa Cruz), rabbit anti-Bcl-2 polyclonal antibody (Santa Cruz), was inserted into the NdeI and SapI sites of the pTYB1 vector for the mouse anti-Mcl-1 monoclonal antibody (Chemicon International, expression of recombinant human Mcl-1 protein (New England Bio- Temecula, CA), rabbit anti-Bak (NT) polyclonal antibody (Upstate Bio- labs, Beverly, MA). The accuracy of the molecular identity of all con- technology, Lake Placid, NY), or rabbit anti-Bax (NT) polyclonal anti- structs was confirmed by sequencing. For details of the PCR primers body (Upstate). For quantification of Bak and Mcl-1 levels, bands rep- used see the supplemental data. resenting respective Bak and Mcl-1 from membranes of three Yeast Two-hybrid Assay—All yeast two-hybrid procedures were car- independent immunoprecipitation experiments were scanned and ana- ried out according to the manufacturer’s protocol (Clontech). The lyzed using a GS-800 Calibrated Densitometer with Quantity One soft- cDNA for mouse-truncated Bid was cloned into the pGBKT7 vector as ware (Bio-Rad). described earlier. The cDNA library was generated from mouse primary Immunoblotting—SDS-PAGE and immunoblotting were performed thymocytes that had been treated with 5 gray of -irradiation and cul- essentially as described (16). Briefly, cellular proteins were resolved on tured for5hto induce apoptosis. The RNA was then isolated from these the polyacrylamide gels and transferred to nitrocellulose membrane cells, and the cDNA library was prepared by oligo(dT) priming and (Amersham Biosciences). The membranes were probed with, where directionally cloned in the EcoRI-XhoI sites of the prey vector pAD- appropriate, goat anti-Bid polyclonal antibody, rabbit anti-Mcl-1 poly- Gal4-2.1 (Stratagene, La Jolla, CA). The library was amplified once and clonal antibody, mouse anti-cytochrome c monoclonal antibody (clone found to have over 90% recombinants with an average insert size of 1.5 7H8.2C12, Pharmingen), mouse anti-cytochrome oxidase subunit II Kb. Screening was carried out by sequential transformation of the tBid monoclonal antibody (clone 12C4-F12, Molecular Probes, Eugene, OR), construct followed by the cDNA library into yeast strain Y190. After mouse anti-PARP monoclonal antibody (clone C2-10, R&D Systems), transformation, the yeast were grown for 15 days on selection plates or rat anti--tubulin monoclonal antibody (Serotec, Oxford, UK). After containing 17 mM 3-amino-1,2,4-triazole. Colonies that grew on the incubating with respective secondary antibodies conjugated with horse- plates were tested for activity of the -galactosidase reporter gene by radish peroxidase, the membranes were visualized by ECL Kit (Amer- filter-lift assay. Plasmids from the positive colonies were isolated and sham Biosciences). subjected to PCR and sequencing to identify the prey cDNAs. The spec- In Vitro Assay for Mitochondrial Cytochrome c Release—This assay ificity of the interaction was confirmed by retransformation. was performed as described (17) with minimum modification. In brief, GST Fusion Protein Production and Binding Assay—GST fusion pro- 20  10 HeLa cells were harvested and washed once in ice-cold phos- teins were produced in BL21 Escherichia coli following the induction of phate-buffered saline. The cell pellet was resuspended in 5 volume of expression by isopropyl 1-thio--D-galactopyranoside (Insight Biotech- buffer A (20 mM HEPES, pH 7.4, 250 mM sucrose, 10 mM KCl, 1.5 mM nology, Middlesex, UK) and purified using glutathione-Sepharose beads MgCl ,1mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, and protease (Amersham Biosciences). In vitro translated proteins labeled with inhibitor mixture from Roche Applied Science) and incubated on ice for MARCH 3, 2006• VOLUME 281 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5751 Mcl-1 Binds tBid and Blocks Cytochrome c Release 15 min. Cells were then disrupted by passing them through a 23-gauge needle 25 times before undergoing centrifugation in two sequential steps: 1000  g and 10,000  g. The 10,000  g pellet was collected as mitochondrial fraction and resuspended at a 5 g/l concentration in buffer A supplemented with 150 mM NaCl. Resuspended mitochondria were incubated either alone or with caspase-8-cleaved recombinant human Bid (R&D Systems) at indicated concentrations at 37 °C for 15 min. Following incubation, the mitochondria were centrifuged, with the resulting supernatant collected for examination of cytochrome c release by immunoblotting and the pellet was cross-examined for loss of cyto- chrome c and cytochrome oxidase subunit II as sample loading control. Preparation of Recombinant Human Mcl-1—The cDNAs of full- length human Mcl-1 and mutant Mcl-1mtBH3 were cloned into pTYB1 vector (New England Biolabs), which were used to transform BL21 cells, respectively. The recombinant proteins were induced with the addition of isopropyl 1-thio--D-galactopyranoside and purified according to manufacturer’s instruction (New England Biolabs). The proteins were further concentrated using centrifugal filter devices (Amicon Ultra-4 30,000 MWCO) (Millipore, Bedford, MA). Generation of Stably Transfected Cell Lines and Induction of Apoptosis—HeLa cells were split to 40–50% confluence in 10-cm dishes the day prior to transfection and transfected with 10 gof pcDNA3.1-HA or pcDNA3.1-HA-Mcl-1 using Lipofectamine 2000 reagent (Invitrogen), according to the manufacturer’s instructions. 24 h after transfection cells were split and cultured under selection with 1 mg/ml G418 (Invitrogen). Single cell clones were picked and expanded to establish the stable Mcl-1 overexpressing cells. Overexpression of the Mcl-1 protein was confirmed by immunoblotting. To induce apoptosis, wild type, vector-only, and HA-Mcl-1 overexpressing stable HeLa cells were all treated with soluble recombinant human TRAIL (Alexis Bio- chemicals, San Diego, CA) at indicated concentrations for 14 h. Cells were also treated with TNF- (15 ng/ml) in the presence of cyclohexi- mide (30 g/ml) for 22 h. Apoptosis was assessed by flow cytometry for cells with sub-G DNA content following propidium iodide staining and PARP cleavage as previous described (16). Lentivirus Generation and Expression of Mcl-1 Short Hairpin RNA— FIGURE 1. Identification of interaction of tBid with Mcl-1 by yeast two-hybrid A BLOCK-iT Lentiviral RNAi Expression System (Invitrogen) was used screening. A, alignment of in-frame amino acid sequences from three clones containing according to manufacturer’s instruction. Briefly, the RNA interference cDNA encoding Mcl-1 with full-length murine Mcl-1 protein (mMcl-1). B, the activity of the -galactosidase reporter gene was examined by filter-lift assay after retransforma- sequence for human Mcl-1, GGACTGGCTAGTTAAACAAAG, was tion of yeast containing either vector alone or tBid cDNA with plasmids from positive identified using manufacturer’s RNAi Designer program, and the cor- clones representing cDNA encoding Mcl-1 (upper panel) and Bcl-2 (middle panel). As a positive control, mouse BaxcDNA was used in the retransformation (lower panel). responding oligonucleotides were cloned into pENTR™/U6 vector (Invitrogen). The RNA interference sequence for mouse eleven-nine- teen leukemia (ENL) gene, GCTGTGAGAAGCTCACCTTCA, was a number of tBid-interacting proteins. These included two anti-apop- used to produce control short hairpin RNA (shRNA), and its oligonu- totic Bcl-2 family proteins Mcl-1 and Bcl-2. Only one positive clone cleotides were also cloned into the vector. DH5 E. coli (Invitrogen) representing a cDNA of Bcl-2 was found to interact with tBid, consist- were transformed, and clones were verified by sequencing. The cor- ent with a previous report that Bid interacted with Bcl-2 (18). There rectly identified clones were transferred via a gateway reaction to a were, however, three independent positive clones identified represent- TM modified pLenti6/BLOCK-iT -DEST vector (Invitrogen), a promot- ing cDNAs encoding Mcl-1, and the in-frame amino acid sequences erless lentiviral destination vector in which the blasticidin resistance from each clones were aligned against the full-length murine Mcl-1 (19) marker is replaced with tailless human CD2 as a marker. 293 cells were as shown in Fig. 1A. One positive clone (clone 85) containing the short- transfected with the plasmids using Lipofectamine reagent to produce est Mcl-1 cDNA lacked the 5-region encoding for the first 144 amino the virus. 48 h later, the lentivirus-containing supernatants were har- acids, suggesting that the C terminus fragment of the Mcl-1 protein vested to infect HeLa cells in the presence of Polybrene. The infected containing the BH1, BH2, and BH3 domains was responsible for inter- HeLa cells were harvested 48 h later for the analysis of Mcl-1 expression acting with tBid. To exclude the possibility of false interaction, the plas- by immunoblotting. Induction of apoptosis by TRAIL was performed mids from the positive clones were isolated and retransformed into essentially as described above. yeast containing the plasmid of tBid cDNA, and the activity of the -ga- lactosidase reporter gene was examined using the filter-lift assay. As Bid RESULTS has also been reported to interact with Bax (18), we used Bax as positive Truncated Bid Interacts with Mcl-1—We initially performed a yeast control with the mouse Bax cDNA cloned in the pGADT7 vector two-hybrid screen using truncated Bid as the bait protein and identified during retransformation. The filter-lift assay detected the -galactosid- 5752 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 9 •MARCH 3, 2006 Mcl-1 Binds tBid and Blocks Cytochrome c Release FIGURE 2. Truncated Bid interacts with Mcl-1 in vitro and in vivo. A, in vitro translated [ S]methionine-labeled Mcl-1 (upper panel) and Bcl-2 (lower panel) were incubated with FIGURE 3. BH domains of both tBid and Mcl-1 are required for interaction. A, require- GST alone, GST-Bid, or GST-tBid immobilized on glutathione-Sepharose beads. Bound ment of BH3 domain of tBid to bind Mcl-1. In vitro translated [ S]methionine-labeled proteins were visualized using a PhosphorImager after protein separation by SDS-PAGE. Mcl-1 was incubated with GST alone, GST-tBid, or mutant GST-tBid (mtBH3) (G94E) B, HeLa cells were transfected with either pcDNA3.1 vector alone or a HA-tBid construct immobilized on glutathione-Sepharose beads. B, BH1, BH2, and BH3 domains of Mcl-1 in the presence of 75 M Z-VAD-fmk. 24 h after transfection, cells were lysed and subject contribute to interaction with tBid. GST alone and GST-tBid were immobilized on gluta- to immunoprecipitation (IP) using anti-HA antibody (upper panel) or anti-Mcl-1 antibody thione-Sepharose beads and incubated with in vitro translated [ S]methionine-labeled (lower panel). Precipitated immunocomplexes were analyzed by SDS-PAGE and immu- Mcl-1 (upper panel) or Mcl-1mtBH1 (G262E mutation) (upper middle panel) or Mcl- noblotting (IB) using anti-Mcl-1 antibody (upper panel), anti-Bcl-2 antibody (middle 1mtBH2 (W305A and W312A double mutations) (lower middle panel) or Mcl-1mtBH3 panel), or anti-Bid antibody (lower panel). (G217E and D218A double mutations) (lower panel). All bound proteins were analyzed using a PhosphorImager. ase activity in the clone transformed with plasmid containing Bax cDNA in the presence of tBid (Fig. 1B, lower right panel). Interestingly, the BH3 Domain of tBid and All Three BH Domains of Mcl-1 Are clone identified as expressing the Mcl-1 fusion protein showed a much Required for Interaction—We then investigated the binding sites greater intensity of -galactosidase staining than the clone expressing responsible for the interaction between tBid and Mcl-1. Based on pre- the Bcl-2 fusion protein (Fig. 1B, compare upper right panel with middle vious reports describing the amino acids critical for interactions right panel). between Bcl-2 family proteins (18, 20–22) and our observation from the To see whether the interaction of Mcl-1 with tBid could be confirmed yeast two-hybrid experiment that the C terminus fragment of Mcl-1 was by an independent method, we performed GST fusion protein pull- required for interaction, we hypothesized that the interaction between down experiments. GST-Bid and GST-tBid fusion proteins were immo- Mcl-1 and tBid involved the BH domains of both proteins. To test this bilized on glutathione-Sepharose beads, respectively, and incubated hypothesis we generated a number of constructs that contained cDNAs with either [ S]-methionine labeled Mcl-1 or Bcl-2 (Fig. 2A). This assay encoding for Mcl-1 and tBid but with their BH domains mutated. First, identified Mcl-1 as a protein specifically interacting with tBid rather we compared the abilities of GST-tBid with mutant GST-tBid (mtBH3) than the full-length Bid (Fig. 2A, compare lane 4 with lane 3, upper (containing G94E mutation at BH3 domain) to interact with [ S]me- panel). Bcl-2 was also seen to preferentially interact with the truncated thionine-labeled Mcl-1. GST-tBid, but not mutant GST-tBid (mtBH3), form of Bid (Fig. 2A, compare lane 4 with lane 3, lower panel). Similar to was shown to interact with Mcl-1 (Fig. 3A, compare lane 3 with lane 4). the result shown in Fig. 1B, tBid fusion protein again showed a higher This indicated that the BH3 domain of tBid was required for interaction affinity for Mcl-1 than Bcl-2 (Fig. 2A). with Mcl-1. Next, we compared the binding ability of wild type Mcl-1 to Next, we carried out co-immunoprecipitation experiments to check GST-tBid with that of Mcl-1 proteins that had mutations at BH1 whether this interaction occurs within cells. A HA-tagged tBid con- (G262E mutation), BH2 (W305A and W312A double mutations), or struct was generated and subsequently used to overexpress tBid in HeLa BH3 (G217E and D218A double mutations) domain, respectively. As cells in the presence of the pan caspase inhibitor Z-VAD-fmk to delay shown in Fig. 3B, apart from wild type Mcl-1, all three Mcl-1 proteins cell death. Cells were lysed 24 h after transfection, and immunoprecipi- with their respective BH domain mutated failed to interact with GST- tation was done using an anti-HA antibody. The endogenous Mcl-1 was tBid. Mutations in BH1 and BH3 domains resulted in a complete loss of detected in complex with HA-tBid by immunoblotting (Fig. 2B, upper interaction, whereas mutations in BH2 domains severely impaired the panel). In addition, endogenous Bcl-2 was also detected in complex with binding ability of Mcl-1 to GST-tBid. These results demonstrated that HA-tBid (Fig. 2B, middle panel), in agreement with an early report (18). all three BH domains of Mcl-1 contribute to its ability to interact with Similarly, when an anti-Mcl-1 antibody was used to immunoprecipitate tBid. Mcl-1 complex from lysates of cells overexpressing HA-tBid, tBid was Mcl-1 Prevents tBid-mediated Cytochrome c Release—We then stud- clearly observed to interact with Mcl-1 (Fig. 2B, lower panel). It is ied the functional significance of the interaction between tBid and important to mention that the endogenous full-length Bid was not Mcl-1. As tBid has been shown to possess potent cytochrome c release immunoprecipitated by the anti-Mcl-1 antibody under the experimen- activity (10, 11), an in vitro assay was set up to examine this activity. tal conditions used (data not shown), further underlining the specificity Mitochondria isolated from HeLa cells were treated with increasing of the interaction between Mcl-1 and the truncated form of the Bid amount of recombinant tBid, and as shown in Fig. 4A (upper panel), tBid protein. induced cytochrome c release into the supernatant in a dose-dependent MARCH 3, 2006• VOLUME 281 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5753 Mcl-1 Binds tBid and Blocks Cytochrome c Release of cytochrome c (Fig. 4B, lane 4, supernatant). At 2 ng/l concentration, Mcl-1 significantly inhibited the release (Fig. 4B, lane 5, supernatant), whereas at 5 ng/l Mcl-1 completely prevented tBid-induced cyto- chrome c release (Fig. 4B, lane 6, supernatant). The complete protection by Mcl-1 (5 ng/l) was also confirmed by the full retention of cyto- chrome c in the mitochondrial fraction (Fig. 4B, lane 6, upper panel, pellet). The same membrane from the mitochondrial fraction was also reprobed for tBid to confirm its presence (Fig. 4B, middle panel, pellet) and for cytochrome oxidase as sample loading control (Fig. 4B, lower panel, pellet). Thus, recombinant Mcl-1 inhibited tBid-induced cyto- chrome c release from the mitochondria in a dose-dependent manner. To further confirm that Mcl-1 inhibits tBid function through its inter- action with tBid, we generated a recombinant Mcl-1 protein containing G217E and D218A double mutations in its BH3 domain (Mcl-1mtBH3), which, we have shown previously, failed to interact with tBid (Fig. 3B, lower panel). Pre-incubating mitochondria with the recombinant Mcl- 1mtBH3 protein itself (5 ng/l) had no effect on the release of cyto- chrome c (Fig. 4C, lane 5, supernatant). However, it did not protect mitochondria from tBid (1 ng/l)-mediated cytochrome c release (Fig. 4C, lane 4, supernatant), whereas recombinant wild type Mcl-1 (5 ng/l) effectively blocked this release (Fig. 4C, lane 3, supernatant). This is also confirmed by the observation that the loss of cytochrome c in the mitochondria occurred when both recombinant Mcl-1mtBH3 protein and tBid were present (Fig. 4C, lane 4, upper panel, pellet) but was prevented by recombinant wild type Mcl-1 (Fig. 4C, lane 3, upper panel, pellet). Thus, our data provided compelling evidence for a functional effect of the interaction between Mcl-1 and tBid. Truncated Bid Displaces Bak from Mcl-1-Bak Complex—Recently it has been shown that Mcl-1 forms a complex with Bak in healthy, FIGURE 4. Mcl-1 prevents tBid-mediated cytochrome c release from isolated mito- unstressed cells (23, 24), we therefore wanted to examine what happens chondria. A, mitochondria isolated from HeLa cells were incubated with recombinant to this complex when tBid was present. We used a HA-tagged tBid tBid at the indicated concentrations. At the end of incubation, supernatant of the mito- construct to overexpress tBid in HeLa cells in the presence of the pan chondrial suspension was collected and subjected to SDS-PAGE and immunoblotting for cytochrome c (Cyt. c, upper panel). Untreated mitochondria (Mito.) were used as positive caspase inhibitor Z-VAD-fmk as described previously. Immunoprecipi- control for the detection of cytochrome c (lane 1). The same membrane was reprobed for tation using an anti-HA antibody showed that tBid interacted with both cytochrome oxidase (subunit II) to check whether the supernatant contains mitochon- drial contamination (Cyt. Oxid., lower panel). B, mitochondria were pre-incubated with endogenous Bak and Bax (Fig. 5A, upper and lower panels, respectively), recombinant Mcl-1 at the indicated concentrations before the addition of tBid. Both which was consistent with the published reports (3, 4). Immunoprecipi- supernatant and pellet of the mitochondrial suspension were collected at the end of tation using anti-Mcl-1 antibody showed that Bak indeed interacted incubation and subject to SDS-PAGE and immunoblotting for cytochrome c (Cyt. c, super- natant, and upper panel, pellet). Membrane from the pellet samples was also probed for with Mcl-1 in cells transfected with empty vector, as detected by immu- tBid for its presence (middle panel, pellet) and for cytochrome oxidase (subunit II) as noblotting (Fig. 5B, lane 1, upper panel). However, when cells were sample loading control (Cyt. Oxid., lower panel, pellet). C, mitochondria were pre-incu- bated with recombinant Mcl-1 (5 ng/l) or mutant Mcl-1mtBH3 (G217E and D218A dou- transfected with the HA-tBid construct the level of Bak in complex with ble mutations) (5 ng/l) before the addition of tBid. At the end of incubation, both Mcl-1 was greatly reduced (Fig. 5B, lane 2, upper panel), whereas the supernatant and pellet of the mitochondrial suspension were collected and subject to SDS-PAGE and immunoblotting as described in B. total amount of Mcl-1 in both immunoprecipitated samples was similar (Fig. 5B, middle panel). Interaction of Mcl-1 with tBid was again con- firmed by reprobing Bid on the same membrane (Fig. 5B, lane 2, lower manner. As a positive control, 30 g of untreated mitochondria was panel). We also probed for Bax and could not detect Bax in the samples used for the detection of cytochrome c (Fig. 4A, lane 1, upper panel). immunoprecipitated by anti-Mcl-1 antibody under the experimental The same blot was reprobed for cytochrome oxidase (subunit II), a conditions used (data not shown). To test the possibility that overex- mitochondrial membrane protein, and it confirmed that the superna- pression of HA-tBid may alter the expression levels of endogenous Bak tant samples were free from mitochondrial contamination (Fig. 4A, and/or Mcl-1, resulting in the reduction of Bak in complex with Mcl-1, lanes 2–5, lower panel). we also checked the levels of Bak, Mcl-1, and tBid in the total cell lysates As Mcl-1 is an anti-apoptotic protein, we reasoned that its ability to prior to the immunoprecipitation. Immunoblotting analysis showed interact with tBid would possibly interfere with the function of the tBid that the expression levels of both Bak and Mcl-1 remained unchanged protein. To test this, isolated mitochondria from HeLa cells were pre- (Fig. 5B, lanes 3 and 4, upper and middle panels), regardless of the incubated with increasing amounts of recombinant Mcl-1 protein before the addition of tBid. Recombinant Mcl-1 alone (5 ng/l) had no presence of HA-tBid (Fig. 5B, lane 4, lower panel). Densitometric anal- effect on the release of cytochrome c (Fig. 4B, lane 2, supernatant). The ysis of bands representing respective Bak and Mcl-1 on membranes from three independent immunoprecipitation experiments indicated treatment of mitochondria with tBid (1 ng/l) resulted in the release of cytochrome c from mitochondria (Fig. 4B, lane 3, supernatant), which is that the relative level of Bak to Mcl-1 was about 4-fold less in an immu- further confirmed by the disappearance of cytochrome c in the mito- noprecipitated sample from cells overexpressing tBid than that from chondrial fraction (Fig. 4B, lane 3, upper panel, pellet). At a concentra- control cells (Fig. 5C). Statistical analysis by Student’s t test showed that tion of 1 ng/l, recombinant Mcl-1 failed to block tBid-mediated release the difference was significant (p  0.05). Therefore, in addition to its 5754 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 9 •MARCH 3, 2006 Mcl-1 Binds tBid and Blocks Cytochrome c Release ability to interact directly with Bak and Bax, tBid can also bind Mcl-1 and displace Bak from the Mcl-1-Bak complex. Mcl-1 Inhibits Apoptosis Induced by TRAIL and TNF- in HeLa Cells—Because Bid has been shown to be cleaved early during the induction of apoptosis by TRAIL in HeLa cells (25, 26), we wished to investigate whether cells overexpressing Mcl-1 would be resistant to apoptosis induced by TRAIL. HeLa cells were transfected with either pcDNA3.1 vector alone or human Mcl-1 cDNA expression construct and selected in G418. Single cell clones were picked and expanded to establish the stable Mcl-1 overexpressing cells. The level of Mcl-1 expression was evaluated by immunoblotting, which showed that its level is higher in cells transfected with the Mcl-1 construct than those with vector-only or wild type HeLa cells (Fig. 6A). Induction of apopto- sis was assessed by flow cytometry for subdiploid DNA content (sub- G ) following propidium iodide staining and PARP cleavage, a biochem- ical marker of apoptosis. As shown in Fig. 6B, treatment of wild type HeLa cells with increasing doses of TRAIL resulted in a dose-dependent induction of apoptosis. This treatment also caused a similar dose-de- pendent induction of apoptosis in cells transfected with vector alone (Fig. 6B). Cells stably overexpressing Mcl-1 were, however, consistently protected from cell death following treatment with TRAIL at all con- centrations (Fig. 6B). Student’s t test analysis showed that the reduction in TRAIL (250 ng/ml)-induced apoptosis of cells overexpressing Mcl-1, when compared with cell death in wild type HeLa cells and in cells transfected with vector alone, was statistically significant (both p 0.01). Also, treatment with increasing amount of TRAIL resulted in a dose-dependent cleavage of PARP in wild type and vector-only trans- fected HeLa cells (Fig. 6B). PARP cleavage was less complete in Mcl-1 overexpressing cells than that seen in wild type and vector-only trans- fected HeLa cells following treatment of TRAIL at 250 ng/ml concen- tration (Fig. 6B, compare lane 15 with lanes 13 and 14, respectively, PARP). As TNF- has been shown to induce apoptosis in HeLa cells through a Bid-dependent pathway (15), we also treated the above cells with TNF- to see whether cells overexpressing Mcl-1 would be resistant to apoptosis induced by TNF-. Treatment with Me SO or cycloheximide (30 g/ml) alone did not cause significant increase in cell death in all three types of cells (data not shown). Treatment with TNF- (15 ng/ml) in the presence of cycloheximide for 22 h resulted in similar levels of apoptosis in wild type HeLa cells and cells stably transfected with vector alone (Fig. 6C, lanes 2 and 4, respectively). Cells stably overexpressing Mcl-1 were indeed partially resistant to TNF--induced cell death (Fig. 6C, lane 6). The reduction in TNF--induced apoptosis in cells overex- pressing Mcl-1, when compared with cell death in wild type HeLa cells and cells transfected with vector alone, was statistically significant (both p  0.05). Again, PARP cleavage was not as complete in Mcl-1 overex- pressing cells as that seen in wild type HeLa cells and cells transfected with vector alone (Fig. 6C, compare lane 6 with lanes 2 and 4, respec- tively, PARP). Inhibition of death receptor-mediated apoptosis was also FIGURE 5. tBid binds Mcl-1 and displaces Bak from Mcl-1-Bak complex. A, HeLa cells were transfected with either pcDNA3.1 vector alone or a HA-tBid construct in the pres- observed in another stable HeLa cell line overexpressing Mcl-1 (see Fig. ence of 75 M Z-VAD-fmk. Cells were lysed 24 h after transfection and subjected to 1 in supplemental data). Therefore, overexpressing Mcl-1 conferred immunoprecipitation (IP) using anti-HA antibody. Precipitated immunocomplexes were resistance to apoptosis induced by both TRAIL and TNF- in HeLa analyzed by SDS-PAGE and immunoblotting (IB) using anti-Bak antibody (upper panel)or anti-Bax antibody (lower panel). Asterisks denote nonspecific bands. B, the above cell cells. lysates were immunoprecipitated by anti-Mcl-1 antibody. The precipitates were subject Mcl-1 Silencing by RNA Interference Sensitizes HeLa Cells to TRAIL- to SDS-PAGE and immunoblotting using anti-Bak antibody (left upper panel), anti-Mcl-1 antibody (left middle panel), or anti-Bid antibody (left lower panel). Prior to immunopre- induced Apoptosis—To determine the effect of Mcl-1 down-regulation cipitation, 5% of total cell lysates (TCL) from cells transfected with either pcDNA3.1 vector on death receptor-mediated cell death, we used a lentiviral vector for alone or a HA-tBid construct were also analyzed by SDS-PAGE and immunoblotting for the expression levels of Bak (right upper panel), Mcl-1 (right middle panel), or Bid (right expression of shRNA to induce Mcl-1 silencing. HeLa cells were lower panel). C, relative level of Bak to Mcl-1 in immunoprecipitated samples by anti- infected with lentivirus containing vectors expressing either control Mcl-1 antibody was analyzed by densitometry after scanning the bands representing shRNA or Mcl-1 shRNA. The level of Mcl-1 expression was then eval- respective Bak and Mcl-1 on membranes from three independent immunoprecipitation experiments. uated by immunoblotting. As shown in Fig. 7A, the expression level of MARCH 3, 2006• VOLUME 281 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5755 Mcl-1 Binds tBid and Blocks Cytochrome c Release FIGURE 6. Mcl-1 inhibits apoptosis induced by TRAIL and TNF- in HeLa cells. A, HeLa cells were stably transfected with either pcDNA3.1 vector alone or plasmid containing the cDNA encoding human Mcl-1. The expression of Mcl-1 in wild type (w/t), pcDNA3.1 vector and Mcl-1 overexpressing HeLa cells was assessed by immunoblotting for Mcl-1 (upper panel). The same membrane was also probed for -tubulin as sample loading control (lower panel). B, apoptosis was induced by the treatment of the three HeLa cell lines with soluble human recombinant TRAIL at the indicated con- centrations for 14 h and assessed by both flow cytometry for cells with sub-G DNA content and PARP cleavage by immunoblotting. -Tubulin was also probed as sample loading control. C, apopto- sis was also induced by the treatment of the three HeLa cell lines with TNF- (15 ng/ml) in the pres- ence of cycloheximide (CHX,30 g/ml) for 22 h and assessed as described in B. Mcl-1 is clearly reduced in cells infected with lentivirus expressing control shRNA following the treatment with TRAIL (Fig. 7B, compare Mcl-1 shRNA. We then treated these cells with TRAIL (100 ng/ml) for lane 6 with lanes 2 and 4, respectively, PARP). 14 h to induce apoptosis. Treatment with TRAIL resulted in similar DISCUSSION levels of apoptosis in wild type HeLa cells and cells infected with virus expressing control shRNA (Fig. 7B, lanes 2 and 4, respectively). How- Most of the studies on protein-protein interaction of Bid in apoptosis ever, cells infected with virus expressing Mcl-1 shRNA became more have, to date, focused on its interaction with the pro-apoptotic Bax and sensitive to TRAIL-induced cell death (Fig. 7B, lane 6). The increase in Bak proteins. It has been shown that tBid can directly interact with both TRAIL-induced apoptosis in cells infected with virus expressing Mcl-1 Bax and Bak and induce their oligomerization, resulting in cytochrome shRNA, when compared with cell death in wild type HeLa cells and cells c release (3, 4). This function of tBid is vital for the transmission of infected with virus expressing control shRNA, was statistically signifi- apoptotic signals from death receptors to mitochondria in certain tis- cant (both p  0.05). In addition, there is a greater loss of intact PARP sues, and, indeed, apoptosis in hepatocytes is dependent on tBid-medi- observed in cells infected with virus expressing Mcl-1 shRNA than that ated amplification of the apoptotic signal via the mitochondria after seen in wild type HeLa cells and cells infected with virus expressing the engagement of the death receptor Fas (27). However, there has been 5756 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 9 •MARCH 3, 2006 Mcl-1 Binds tBid and Blocks Cytochrome c Release FIGURE 7. Mcl-1 silencing by RNA interference sensitizes HeLa cells to TRAIL-induced apoptosis. A, HeLa cells were infected with lentivirus-containing vectors expressing either control shRNA or Mcl-1 shRNA. The expression of Mcl-1 in wild type (w/t) HeLa cells and cells infected with virus expressing either control shRNA or Mcl-1 shRNA was assessed by immunoblotting for Mcl-1 (upper panel). The same membrane was also probed for-tubulin as sample loading control (lower panel). B, apoptosis was induced by the treatment of the three HeLa cell lines with soluble human recombinant TRAIL (100 ng/ml) for 14 h and assessed by both flow cytometry for cells with sub-G DNA content and PARP cleavage by immunoblotting. -Tubulin was also probed as sample loading control. little information about the role of anti-apoptotic Bcl-2 proteins in reg- Mcl-1 proteins, whereas in our study tBid was shown to interact with ulating the function of tBid. Here we show that Mcl-1 can regulate endogenous Mcl-1. tBid-mediated apoptosis through its ability to interact with tBid. Our It has been proposed that the BH3 domain-only proteins can be fur- yeast two-hybrid screen identified Mcl-1 as a potent interaction partner ther divided into two subgroups, activating or sensitizing (29, 30). Those for tBid, and this interaction was confirmed by both GST fusion protein proteins with activating BH3 domains (e.g. Bid and Bim) appear to have pull-down experiments in vitro and co-immunoprecipitation in vivo.In a higher affinity for Bak or Bax than Bcl-2, perhaps directly activating addition, we have compared Mcl-1 with Bcl-2 for the ability to bind tBid Bak and Bax. The BH3-only proteins with sensitizing BH3 domains (e.g. in both the yeast two-hybrid and GST pull-down experiments, and tBid Bad) appear to preferentially bind Bcl-2 allowing Bak and Bax to be appears to have a much higher affinity for Mcl-1 than Bcl-2. The co- derepressed. In this study, we confirm that tBid can directly interact immunoprecipitation assay showed that tBid interacts with Mcl-1 as with Bak and Bax. More importantly, we have also shown that tBid can well as Bcl-2 in cells. Recently, it has been shown that C-terminal frag- bind to Mcl-1 and disrupt the Mcl-1-Bak complex. As a result, Bak is ments of Mcl-1 can also interact with tBid (28). However, in that study displaced from the complex. Recently, it has been reported that another the interaction was observed on the basis of exogenously expressed BH3-only protein Noxa can also displace Bak from Mcl-1/Bak complex MARCH 3, 2006• VOLUME 281 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5757 Mcl-1 Binds tBid and Blocks Cytochrome c Release (31), suggesting that this displacement is indeed a common feature for chondria and apoptosis in HeLa cells (17). Mcl-1 alone may not be able BH3-only proteins to activate Bak and/or Bax. Therefore, our study to offer complete protection of cells from apoptosis. Conversely, deple- tion of Mcl-1 alone may be insufficient to render all the cells sensitive to demonstrates dual modes of action by tBid, which is capable of both activating and sensitizing other pro-apoptotic Bcl-2 family proteins. On death receptor-mediated apoptosis. the other hand, Mcl-1 can act through direct binding to neutralize tBid, Mcl-1 was initially discovered as an early induction gene during dif- ferentiation of the myeloid cell line, ML-1 (39), and is widely expressed thus preventing it from activating Bak or Bax. In this respect Mcl-1 may in a variety of human tissues and cells as well as many tumors (40, 41). be playing an active role in counter balancing activating BH3 domain- only proteins. This notion is supported by the recent studies that Bim, Deletion of Mcl-1 in mice led to embryonic lethality during the peri- implantation stage, suggesting it is essential for embryonic development another activating BH3 domain-only protein, has a higher affinity for (42). Genetic studies with conditional knock-out approach also reveal Mcl-1 than Bcl-2 (32), and Mcl-1 effectively inhibits Bim-mediated that Mcl-1 is required both in early lymphoid development and in the release of mitochondrial cytochrome c (33). Our study shows that Mcl-1 maintenance of mature B and T lymphocytes, which are rapidly lost interacts with tBid and impairs the ability of tBid to induce cytochrome when Mcl-1 is deleted (32). Mcl-1 also plays physiologically important c release and apoptosis. Recently, two independent studies revealed dif- roles in regulating myeloid cell survival (43, 44). Given that Mcl-1 can ferential targeting of anti-apoptotic Bcl-2 proteins by BH3-only pro- interact strongly with tBid and other BH3-only proteins such as Bim and teins using individual BH3 peptides (34, 35). A Bim BH3 peptide appears inhibits their induction of cytochrome c release and activation of the capable of interacting with most anti-apoptotic proteins including mitochondrial apoptotic pathway, loss of Mcl-1 may render the cells Mcl-1 with high affinity, whereas the Bid BH3 peptide does not bind sensitive to apoptosis induced by a variety of apoptotic stimuli including Mcl-1. This is not unexpected, as we have shown that full-length Bid is the activation of death receptors of TNF family. Here we demonstrate a not able to interact with Mcl-1. Only the truncated Bid, which may biochemical link between Mcl-1, an essential regulator of lymphoid undergo further modification by postproteolytic N-myristoylation, is homeostasis, and receptor-mediated apoptosis, the pre-eminent path- able to bind Mcl-1, suggesting that there is a complex structural basis way controlling survival of the cells of the immune system (45, 46). underlying the interaction of Bid with other proteins, which is not dis- cernable in binding studies using Bid BH3 peptide alone. REFERENCES Our in vivo experiments in HeLa cells were prompted by the involve- 1. Gross, A., McDonnell, J. M., and Korsmeyer, S. J. (1999) Genes Dev. 13, 1899–1911 ment of Bid in apoptosis mediated by TRAIL and TNF- in HeLa cells 2. Cory, S., and Adams, J. M. (2002) Nat. Rev. Cancer 2, 647–656 (15, 25, 26). We have shown that Mcl-1 can indeed protect HeLa cells 3. Desagher, S., Osen-Sand, A., Nichols, A., Eskes, R., Montessuit, S., Lauper, S., Maun- drell, K., Antonsson, B., and Martinou, J.-C. (1999) J. Cell Biol. 144, 891–901 from apoptosis induced by TRAIL and TNF-. Mcl-1 silencing by a 4. Wei, M. C., Lindsten, T., Mootha, V. K., Weiler, S., Gross, A., Ashiya, M., Thompson, shRNA approach sensitizes HeLa cells to TRAIL-induced apoptosis. C. B., and Korsmeyer, S. J. (2000) Genes Dev. 14, 2060–2070 Recently, it has been reported that Mcl-1 mediates resistance to TRAIL- 5. Wei, M. C., Zong, W.-X., Cheng, E. H.-Y., Lindsten, T., Panoutsakopoulou, V., Ross, A. J., Roth, K. A., MacGregor, G. R., Thompson, C. B., and Korsmeyer, S. J. (2001) induced apoptosis in human cholangiocarcinoma cells (36). In their Science 292, 727–730 study, resistance was specifically associated with overexpression of 6. Cheng, E.H.-Y.A., Wei, M. C., Weiler, S., Flavell, R. A., Mak, T. W., Lindsten, T., and Mcl-1 and depletion of Mcl-1 by the small interfering RNA method also Korsmeyer, S. J. (2001) Mol. Cell 8, 705–711 sensitizes cells to TRAIL-mediated apoptosis despite Bcl-2 expression. 7. Zong, W.-X., Lindsten, T., Ross, A. J, MacGregor, G. R., and Thompson, C. B. (2001) Genes Dev. 15, 1481–1486 Furthermore, another study has also shown that hepatocyte growth 8. Green, D. R. (2000) Cell 102, 1–4 factor-mediated Mcl-1 induction inhibits apoptosis induced by Fas in 9. Strasser, A., O’Connor, L., and Dixit, V. M. (2000) Annu. Rev. Biochem. 69, 217–245 human primary hepatocytes (37). As TRAIL and Fas induce apoptosis 10. Li, H., Zhu, H., Xu, C.-J., and Yuan, J. (1998) Cell 94, 491–501 via a similar mechanism (38), it is conceivable that Mcl-1 could bind to 11. Luo, X., Budihardjo, I., Zou, H., Slaughter, C., and Wang, X. (1998) Cell 94, 481–490 12. Eskes, R., Desagher, S., Antonsson, B., and Martinou, J. C. (2000) Mol. Cell. Biol. 20, tBid and interfere with its function in activating Bax and/or Bak, result- 929–935 ing in protection of these cells from death receptor-mediated apoptosis. 13. Brady, H. J., Salomons, G. S., Bobeldijk, R. C., and Berns, A. J. (1996) EMBO J. 15, Our observations of interaction between Mcl-1 and tBid and subse- 1221–1230 quent prevention of cytochrome c release therefore provide a potential 14. Clohessy, J. G., Zhuang, J., and Brady, H. J. M. (2004) Br. J. Haematol. 125, 655–665 15. Perez, D., and White, E. (2000) Mol. Cell 6, 53–63 biochemical mechanism to explain the anti-apoptotic effect of Mcl-1 in 16. Zhuang, J., Ren, Y., Snowden, R. T., Zhu, H., Gogvadze, V., Savill, J. S., and Cohen, death receptor-mediated apoptosis through specific targeting of tBid. G. M. (1998) J. Biol. Chem. 273, 15628–15632 It is worth noting that under the experimental conditions used, over- 17. Nijhawan, D., Fang, M., Traer, E., Zhong, Q., Gao, W., Du, F., and Wang, X. (2003) expression of Mcl-1 does not fully protect cells from both TRAIL- and Genes Dev. 17, 1475–1486 18. Wang, K., Yin, X. M., Chao, D. T., Milliman, C. L., and Korsmeyer, S. J. (1996) Genes TNF--induced apoptosis although it significantly reduces the level of Dev. 10, 2859–2869 cell death in HeLa cells. It is possible that the level of Mcl-1 expression 19. Okita, H., Umezawa, A., Suzuki, A., and Hata, J.-i. (1998) Biochim. Biophys. Acta. in stably transfected cells is not high enough to prevent cell death com- 1398, 335–341 20. Cheng, E. H.-Y., Levine, B., Boise, L. H., Thompson, C. B., and Hardwick, J. M. (1996) pletely. In addition, in the case of TNF--induced apoptosis, the pres- Nature 379, 554–556 ence of protein synthesis inhibitor cycloheximide inevitably blocks pro- 21. Yin, X.-M., Oltvai, Z. N., and Korsmeyer, S. J. (1996) Nature 369, 321–323 duction of Mcl-1 protein and thus reduces the anti-apoptotic function 22. Zha, H., Aime-Sempe, C., Sato, T., and Reed, J. C. (1996) J. Biol. Chem. 271, of Mcl-1. However, there is emerging evidence to suggest that complete 7440–7444 23. Cuconati, A., Mukherjee, C., Perez, D., and White, E. (2003) Genes Dev. 17, protection of apoptosis requires multiple lines of resistance conferred 2922–2932 by anti-apoptotic proteins and Mcl-1 may provide the first line of resist- 24. Leu, J. I.-J., Dumont, P., Hafey, M., Murphy, M. E., and George, D. L. (2004) Nat. Cell ance to the induction of apoptosis by TRAIL and TNF-. This is con- Biol. 6, 443–450 sistent with the observation that in adenoviral protein E1A-induced 25. Harper, N., Farrow, S. N., Kaptein, A., Cohen, G. M., and MacFarlane, M. (2001) J. Biol. Chem. 276, 34743–34752 apoptosis in HeLa cells, loss of Mcl-1 is required to initiate the apoptotic 26. Seol, D.-W., Li, J., Seol, M.-H., Park, S.-Y., Talanian, R. V., and Billiar, T. R. (2001) pathway (23). The idea is further supported by a recent study demon- Cancer Res. 61, 1138–1143 strating that Mcl-1 functions upstream of and together with Bcl-x in 27. Yin, X.-M., Wang, K., Gross, A., Zhao, Y., Zinkel, S., Klocke, B., Roth, K. A., and preventing UV irradiation-induced cytochrome c release from mito- Korsmeyer, S. J. (1999) Nature 400, 886–891 5758 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 9 •MARCH 3, 2006 Mcl-1 Binds tBid and Blocks Cytochrome c Release 28. Weng, C., Li, Y., Yu, D., Shi, Y., and Tang, H. (2005). J. Biol. Chem. 280, 10491–10500 Dax, A., Zink, D., Buchler, P., Muller, M., and Krammer, P. H. (2004) Hepatology 39, 29. Letai, A., Bassik, M. C., Walensky, L. D., Sorcinelli, M. D., Weiler, S., and Korsmeyer, 645–654 S. J. (2002) Cancer Cell 2, 183–192 38. Wang, S., and El-Deiry, W. S. (2003) Oncogene. 22, 8628–8633 30. Scorrano, L., and Korsmeyer, S. J. (2003) Biochem. Biophys. Res. Commun. 304, 39. Kozopas, K. M., Yang, T., Buchan, H. L., Zhou, P., and Craig, R. W. (1993) Proc. Natl. 437–444 Acad. Sci. U. S. A. 90, 3516–3520 31. Willis, S. N., Chen, L., Dewson, G., Wei, A., Naik, E., Fletcher, J. I., Adams, J. M., and 40. Krajewski, S., Bodrug, S., Gascoyne, R., Berean, K., Krajewska, M., and Reed, J. C. Huang, D. C. S. (2005) Genes Dev. 19, 1294–1305 (1994) Am. J. Pathol. 145, 515–525 32. Opferman, J. T., Letai, A., Beard, C., Sorcinelli, M. D., Ong, C. C., and Korsmyer, S. J. 41. Krajewski, S., Bodrug, S., Krajewska, M., Shabaik, A., Gascoyne, R., Berean, K., and (2003) Nature 426, 671–676 Reed, J. C. (1995) Am. J. Pathol. 146, 1309–1319 33. Han, J., Goldstein, L. A., Gastman, B. R., Froelich, C. J., Yin, X.-M., and Rabinnowich, 42. Rinkenberger, J. L., Horning, S., Klocke, B., Roth, K., and Korsmeyer, S. J. (2000) Genes H. (2004) J. Biol. Chem. 279, 22020–22029 Dev. 14, 23–27 34. Chen, L., Wills, S. N., Wei, A., Smith, B. J., Fletcher, J. I., Hinds, M. G., Coleman, P. M., 43. Epling-Burnette, P. K., Zhong, B., Bai, F., Jiang, K., Bailey, R. D., Garcia, R., Jove, R., Day, C. L., Adams, J. M., and Huang, D. C. S. (2005) Mol. Cell 17, 393–403 Djeu, J. Y., Loughran, T. P. Jr., and Wei, S. (2001) J. Immunol. 166, 7486–7495 35. Kuwana, T., Bouchier-Hayes, L., Chipuk, J. C., Bonzon, C., Sullivan, B. A., Green, 44. Edwards, S. W., Derouet, M., Howse, M., and Moots, R. J. (2004) Biochem. Soc. Trans. D. R., and Newmeyer, D. D. (2005) Mol. Cell 17, 525–535 32, 489–492 36. Taniai, M., Grambihler, A., Higuchi, H., Werneburg, N., Bronk, S. F., Farrugia, D. J., 45. Nagata, S. (1999) Annu. Rev. Genet. 33, 29–55 Kaufmann, S. H., and Gores, G. J. (2004) Cancer Res. 64, 3517–3524 46. Marsden, V. S., and Strasser, A. (2003) Annu. Rev. Immunol. 21, 71–105 37. Schulze-Bergkamen, H., Brenner, D., Krueger, A., Suess, D., Fas, S. C., Freym C. R., MARCH 3, 2006• VOLUME 281 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5759 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Biological Chemistry American Society for Biochemistry and Molecular Biology

Mcl-1 Interacts with Truncated Bid and Inhibits Its Induction of Cytochrome c Release and Its Role in Receptor-mediated Apoptosis *

Download PDF
10 pages

Loading next page...
 
/lp/american-society-for-biochemistry-and-molecular-biology/mcl-1-interacts-with-truncated-bid-and-inhibits-its-induction-of-RznNDkZ0jw

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
American Society for Biochemistry and Molecular Biology
Copyright
Copyright © 2006 Elsevier Inc.
ISSN
0021-9258
eISSN
1083-351X
DOI
10.1074/jbc.m505688200
Publisher site
See Article on Publisher Site

Abstract

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 281, NO. 9, pp. 5750 –5759, March 3, 2006 © 2006 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. Mcl-1 Interacts with Truncated Bid and Inhibits Its Induction of Cytochrome c Release and Its Role in □ S Receptor-mediated Apoptosis Received for publication, May 24, 2005, and in revised form, November 11, 2005 Published, JBC Papers in Press, December 27, 2005, DOI 10.1074/jbc.M505688200 ‡1,2 ‡2,3 ‡ § ‡4 John G. Clohessy , Jianguo Zhuang , Jasper de Boer , Gabriel Gil-Go´mez , and Hugh J. M. Brady From the Molecular Haematology and Cancer Biology Unit, Institute of Child Health and Great Ormond Street Hospital for Children, University College London, London WC1N 1EH, United Kingdom and Unitat de Biologia Cellular i Molecular, Institut Municipal d’Investigacio´Me`dica-Universitat Pompeu Fabra (IMIM-UPF), E-03003 Barcelona, Spain Engagement of death receptors such as tumor necrosis factor-R1 The BH3 domain-only proteins require cooperation of other multido- and Fas brings about the cleavage of cytosolic Bid to truncated Bid main family members to induce apoptosis (3–7). (tBid), which translocates to mitochondria to activate Bax/Bak, In mammals two distinct apoptotic signaling pathways have been resulting in the release of cytochrome c. The mechanism underlying identified (8, 9). In the extrinsic pathway, apoptosis is initiated through the activation, however, is not fully understood. Here, we have iden- ligand binding to cell surface receptors of the tumor necrosis factor tified the anti-apoptotic Bcl-2 family member Mcl-1 as a potent (TNF) family such as TNF-R1 and Fas. Upon ligation these receptors tBid-binding partner. Site-directed mutagenesis reveals that the initiate the formation of a death-inducing signaling complex, which Bcl-2 homology (BH)3 domain of tBid is essential for binding to consists of adaptor molecules such as the Fas-associated death domain Mcl-1, whereas all three BH domains (BH1, BH2, and BH3) of Mcl-1 protein and procaspase-8. Within the complex, caspase-8 undergoes are required for interaction with tBid. In vitro studies using isolated autoproteolytic activation. Once activated, caspase-8 can activate mitochondria and recombinant proteins demonstrate that Mcl-1 downstream caspases, for example, caspase-3 and -7, leading to orderly strongly inhibits tBid-induced cytochrome c release. In addition to degradation of intracellular substrates and cell death. The cell-intrinsic its ability to interact directly with Bax and Bak, tBid also binds pathway is initiated when the integrity of the outer mitochondrial mem- Mcl-1 and displaces Bak from the Mcl-1-Bak complex. Importantly, brane is lost in response to diverse apoptotic stimuli. This results in the overexpression of Mcl-1 confers resistance to the induction of apo- release of cytochrome c and other apoptotic proteins into the cyto- ptosis by both TRAIL and tumor necrosis factor- in HeLa cells, plasm, where cytochrome c binds to apoptotic protease-activating fac- whereas targeting Mcl-1 by RNA interference sensitizes HeLa cells tor 1 (Apaf-1). Apaf-1 in turn recruits procaspase-9 to form a multim- to TRAIL-induced apoptosis. Therefore, our study demonstrates a eric complex, which leads to the autoproteolytic activation of caspase-9. novel regulation of tBid by Mcl-1 through protein-protein interac- The active caspase-9 then efficiently activates other downstream tion in apoptotic signaling from death receptors to mitochondria. caspases, bringing about the morphological changes characteristic of apoptosis. This intrinsic pathway is thus mitochondria-dependent and tightly controlled by the Bcl-2 family proteins. The commitment of cells to apoptosis in response to diverse physio- Although the two apoptotic pathways can function independently, an logical cues and cytotoxic agents is primarily regulated by proteins of the existing link between them is the BH3 domain-only protein Bid that is Bcl-2 family that are evolutionarily conserved from nematodes to cleaved by active caspase-8 following engagement of death receptor Fas humans (1, 2). Bcl-2 family proteins share one or more Bcl-2 homology (10, 11). Cleaved Bid, also known as truncated Bid (tBid), translocates to (BH) domains and are divided into two main groups based on their pro- mitochondria to induce oligomerization of Bax and/or Bak and cyto- / / or anti-apoptotic activities. The anti-apoptotic members include Bcl-2, chrome c release (4, 12). In both Bax and Bak cells, tBid fails to Bcl-x , A1, Bcl-w, and Mcl-1. The pro-apoptotic family members are induce cytochrome c release and apoptosis, suggesting that it requires further divided according to whether they contain multiple BH domains Bax and/or Bak to exert its mitochondrial pro-apoptotic activity (5). (such as Bax and Bak) or only the BH3 domain (such as Bid and Bim). However, the underlying mechanism and, in particular, the sequence of events that occurs after Bid cleavage and prior to cytochrome c release * This research was supported in part by the Medical Research Council (United Kingdom) are not completely defined. In this study we investigated the mechanism Grant G9900172 (to H. J. M. B.). The costs of publication of this article were defrayed in of tBid-induced activation of the mitochondrial apoptotic pathway by part by the payment of page charges. This article must therefore be hereby marked searching for novel tBid-interacting proteins using a yeast two-hybrid “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains supple- screen. We identified the anti-apoptotic Bcl-2 family protein Mcl-1 as a mental data. genuine tBid-binding partner. Further studies demonstrate that Mcl-1 Funded by a Child Health Research Appeal Trust (CHRAT) studentship from the Insti- tute of Child Health and Great Ormond Street Hospital Special Trustees. effectively inhibits tBid-induced cytochrome c release from mitochon- Both authors contributed equally to this work. dria and protects HeLa cells from apoptosis induced by both tumor Funded by the Great Ormond Street Hospital for Children REACH Fund. To whom correspondence should be addressed: Molecular Haematology and Cancer necrosis factor-related apoptosis-inducing ligand (TRAIL) and TNF-. Biology Unit, Institute of Child Health and Great Ormond Street Hospital for Children, University College London, 30 Guilford St., London WC1N 1EH, UK. Tel.: 44-20- EXPERIMENTAL PROCEDURES 79052731; Fax: 44-20-78138100; E-mail: [email protected]. The abbreviations used are: BH, Bcl-2 homology; TNF, tumor necrosis factor; tBid, trun- Materials and Cell Culture—All media and cell culture reagents were cated Bid; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand; HA, hemagglutinin; GST, glutathione S-transferase; Z-VAD-fmk, benzyloxycarbonyl-VAD- purchased from Invitrogen. Other chemicals, unless otherwise stated, fluoromethyl ketone; CHAPS, 3-[(3-cholamidopropyl)-dimethyl-ammonio]-1-pro- were obtained from Sigma. Human cervical carcinoma HeLa cells were panesulfonate; PARP, poly(ADP-ribose)polymerase; shRNA, short hairpin RNA; mtBH3, mutant BH3. obtained from American Type Culture Collection (Rockville, MD) and 5750 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 9 •MARCH 3, 2006 This is an Open Access article under the CC BY license. Mcl-1 Binds tBid and Blocks Cytochrome c Release cultured in Dulbecco’s modified Eagle’s medium supplemented with [ S]methionine were prepared using the TNT T7 Quick-Coupled 10% fetal bovine serum, 2 mM glutamine, 0.5 units/ml penicillin, and 0.5 Transcription/Translation System (Promega, Madison, WI), using mg/ml streptomycin. pcDNA3.1-HA constructs as templates. The S-labeled prey proteins Cloning of cDNAs and Plasmids Construction—Mouse tBid cDNA were incubated with GST, GST-Bid, or GST-tBid fusion proteins bound was amplified by PCR from a murine pcDNA3-Bid template (a gift from to glutathione-Sepharose beads in bead-binding buffer (50 mM potas- Dr. S. Korsmeyer, Harvard Medical School, Boston, MA) and inserted sium phosphate, pH 7.5, 150 mM KCl, 1 mM MgCl , 10% glycerol, 1% into EcoRI-BamHI sites downstream of the GAL4-DBD in the vector Triton X-100) and protease inhibitors mixture from Roche Applied pGBKT7 (Clontech). A cDNA encoding mouse Bax lacking the C-ter- Science. The mixtures were incubated at 4 °C for 2 h with rotation. The minal hydrophobic domain (Bax21) was amplified by PCR from a beads were then pelleted and washed five times in ice-cold bead-binding pcDNA3-Bax template (13) and cloned downstream of the GAL4-AD buffer. Finally, beads were resuspended in SDS sample buffer, and the in pGADT7 vector (Clontech). A pUC18-bcl-2 construct (Clonexpress, proteins were resolved on SDS-polyacrylamide gels, which were fixed, Gaithersburg, MD) was used to clone the human Bcl-2 cDNA. This vacuum-dried onto 3MM paper, and then visualized using a Phospho- construct was digested with EcoRI and HindIII to release the Bcl-2 rImager (Typhoon 8600, Amersham Biosciences). cDNA, which was then ligated into the corresponding sites in Immunoprecipitation—For each immunoprecipitation experiment, pcDNA3.1/myc-HIS©(-) (Invitrogen). Human Bid and tBid cDNA were HeLa cells were transfected with HA-tBid in pcDNA3.1 or empty vector generated by reverse transcriptase PCR using oligo(dT) primer with in the presence of 75 M Z-VAD-fmk (Enzyme System, Dublin, CA). total RNA obtained from human leukemic Jurkat T cells. Both cDNAs Immunoprecipitation was essentially carried out as described (15). were then cloned at the EcoRI-BamHI sites downstream of the HA-tag Briefly, 24 h after transfection, cells were harvested and resuspended in in pcDNA3.1-HA (Invitrogen) (14). The construct expressing ice-cold lysis buffer containing 2% CHAPS, 20 mM Tris/HCl (pH 7.4), HA-tagged human Mcl-1 was generated as previously described (14). 137 mM NaCl, 2 mM EDTA, 10% glycerol, 1% Triton X-100, and prote- GST fusion protein constructs were generated by PCR using ase inhibitor mixture (Roche Applied Science). Lysates were precleared pcDNA3.1-HA-Bid and -tBid as templates and cloned into the EcoRI- and then incubated with either mouse anti-HA monoclonal antibody XhoI sites in the pGEX-6P-2 vector (Amersham Biosciences). Muta- (clone 12CA5, Roche) or rabbit anti-Mcl-1 polyclonal antibody (Santa tions in the BH domains of tBid and Mcl-1 proteins were generated by Cruz Biotechnology, Santa Cruz, CA) at 4 °C for 1 h, and protein TM site-directed mutagenesis using the GeneTailor site-directed mutagen- A-Sepharose beads (Pharmacia, Piscataway, NJ) were added to pull esis system according to manufacturer’s instruction (Invitrogen). The down the immunocomplexes. The beads were washed five times in G94E mutation of BH3 domain of tBid was generated by PCR using the washing buffer containing 0.2% Triton X-100, 20 mM Tris/HCl (pH 7.4), pGEX-tBid construct as a template. Using pcDNA3.1-HA-Mcl-1 as a 137 mM NaCl, 2 mM EDTA, 10% glycerol before being resuspended in template for PCR of all Mcl-1 mutants, we generated the G262E muta- SDS sample buffer and subjected to SDS-PAGE. Immunoblotting was tion at BH1 domain, the W305A and W312A double mutations for BH2 performed using, where appropriate, goat anti-Bid polyclonal antibody domain, and the G217E and D218A double mutations of BH3 domain of (R&D Systems, Minneapolis, MN), rabbit anti-Mcl-1 polyclonal anti- Mcl-1. The cDNA fragment encoding human Mcl-1 generated by PCR body (Santa Cruz), rabbit anti-Bcl-2 polyclonal antibody (Santa Cruz), was inserted into the NdeI and SapI sites of the pTYB1 vector for the mouse anti-Mcl-1 monoclonal antibody (Chemicon International, expression of recombinant human Mcl-1 protein (New England Bio- Temecula, CA), rabbit anti-Bak (NT) polyclonal antibody (Upstate Bio- labs, Beverly, MA). The accuracy of the molecular identity of all con- technology, Lake Placid, NY), or rabbit anti-Bax (NT) polyclonal anti- structs was confirmed by sequencing. For details of the PCR primers body (Upstate). For quantification of Bak and Mcl-1 levels, bands rep- used see the supplemental data. resenting respective Bak and Mcl-1 from membranes of three Yeast Two-hybrid Assay—All yeast two-hybrid procedures were car- independent immunoprecipitation experiments were scanned and ana- ried out according to the manufacturer’s protocol (Clontech). The lyzed using a GS-800 Calibrated Densitometer with Quantity One soft- cDNA for mouse-truncated Bid was cloned into the pGBKT7 vector as ware (Bio-Rad). described earlier. The cDNA library was generated from mouse primary Immunoblotting—SDS-PAGE and immunoblotting were performed thymocytes that had been treated with 5 gray of -irradiation and cul- essentially as described (16). Briefly, cellular proteins were resolved on tured for5hto induce apoptosis. The RNA was then isolated from these the polyacrylamide gels and transferred to nitrocellulose membrane cells, and the cDNA library was prepared by oligo(dT) priming and (Amersham Biosciences). The membranes were probed with, where directionally cloned in the EcoRI-XhoI sites of the prey vector pAD- appropriate, goat anti-Bid polyclonal antibody, rabbit anti-Mcl-1 poly- Gal4-2.1 (Stratagene, La Jolla, CA). The library was amplified once and clonal antibody, mouse anti-cytochrome c monoclonal antibody (clone found to have over 90% recombinants with an average insert size of 1.5 7H8.2C12, Pharmingen), mouse anti-cytochrome oxidase subunit II Kb. Screening was carried out by sequential transformation of the tBid monoclonal antibody (clone 12C4-F12, Molecular Probes, Eugene, OR), construct followed by the cDNA library into yeast strain Y190. After mouse anti-PARP monoclonal antibody (clone C2-10, R&D Systems), transformation, the yeast were grown for 15 days on selection plates or rat anti--tubulin monoclonal antibody (Serotec, Oxford, UK). After containing 17 mM 3-amino-1,2,4-triazole. Colonies that grew on the incubating with respective secondary antibodies conjugated with horse- plates were tested for activity of the -galactosidase reporter gene by radish peroxidase, the membranes were visualized by ECL Kit (Amer- filter-lift assay. Plasmids from the positive colonies were isolated and sham Biosciences). subjected to PCR and sequencing to identify the prey cDNAs. The spec- In Vitro Assay for Mitochondrial Cytochrome c Release—This assay ificity of the interaction was confirmed by retransformation. was performed as described (17) with minimum modification. In brief, GST Fusion Protein Production and Binding Assay—GST fusion pro- 20  10 HeLa cells were harvested and washed once in ice-cold phos- teins were produced in BL21 Escherichia coli following the induction of phate-buffered saline. The cell pellet was resuspended in 5 volume of expression by isopropyl 1-thio--D-galactopyranoside (Insight Biotech- buffer A (20 mM HEPES, pH 7.4, 250 mM sucrose, 10 mM KCl, 1.5 mM nology, Middlesex, UK) and purified using glutathione-Sepharose beads MgCl ,1mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, and protease (Amersham Biosciences). In vitro translated proteins labeled with inhibitor mixture from Roche Applied Science) and incubated on ice for MARCH 3, 2006• VOLUME 281 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5751 Mcl-1 Binds tBid and Blocks Cytochrome c Release 15 min. Cells were then disrupted by passing them through a 23-gauge needle 25 times before undergoing centrifugation in two sequential steps: 1000  g and 10,000  g. The 10,000  g pellet was collected as mitochondrial fraction and resuspended at a 5 g/l concentration in buffer A supplemented with 150 mM NaCl. Resuspended mitochondria were incubated either alone or with caspase-8-cleaved recombinant human Bid (R&D Systems) at indicated concentrations at 37 °C for 15 min. Following incubation, the mitochondria were centrifuged, with the resulting supernatant collected for examination of cytochrome c release by immunoblotting and the pellet was cross-examined for loss of cyto- chrome c and cytochrome oxidase subunit II as sample loading control. Preparation of Recombinant Human Mcl-1—The cDNAs of full- length human Mcl-1 and mutant Mcl-1mtBH3 were cloned into pTYB1 vector (New England Biolabs), which were used to transform BL21 cells, respectively. The recombinant proteins were induced with the addition of isopropyl 1-thio--D-galactopyranoside and purified according to manufacturer’s instruction (New England Biolabs). The proteins were further concentrated using centrifugal filter devices (Amicon Ultra-4 30,000 MWCO) (Millipore, Bedford, MA). Generation of Stably Transfected Cell Lines and Induction of Apoptosis—HeLa cells were split to 40–50% confluence in 10-cm dishes the day prior to transfection and transfected with 10 gof pcDNA3.1-HA or pcDNA3.1-HA-Mcl-1 using Lipofectamine 2000 reagent (Invitrogen), according to the manufacturer’s instructions. 24 h after transfection cells were split and cultured under selection with 1 mg/ml G418 (Invitrogen). Single cell clones were picked and expanded to establish the stable Mcl-1 overexpressing cells. Overexpression of the Mcl-1 protein was confirmed by immunoblotting. To induce apoptosis, wild type, vector-only, and HA-Mcl-1 overexpressing stable HeLa cells were all treated with soluble recombinant human TRAIL (Alexis Bio- chemicals, San Diego, CA) at indicated concentrations for 14 h. Cells were also treated with TNF- (15 ng/ml) in the presence of cyclohexi- mide (30 g/ml) for 22 h. Apoptosis was assessed by flow cytometry for cells with sub-G DNA content following propidium iodide staining and PARP cleavage as previous described (16). Lentivirus Generation and Expression of Mcl-1 Short Hairpin RNA— FIGURE 1. Identification of interaction of tBid with Mcl-1 by yeast two-hybrid A BLOCK-iT Lentiviral RNAi Expression System (Invitrogen) was used screening. A, alignment of in-frame amino acid sequences from three clones containing according to manufacturer’s instruction. Briefly, the RNA interference cDNA encoding Mcl-1 with full-length murine Mcl-1 protein (mMcl-1). B, the activity of the -galactosidase reporter gene was examined by filter-lift assay after retransforma- sequence for human Mcl-1, GGACTGGCTAGTTAAACAAAG, was tion of yeast containing either vector alone or tBid cDNA with plasmids from positive identified using manufacturer’s RNAi Designer program, and the cor- clones representing cDNA encoding Mcl-1 (upper panel) and Bcl-2 (middle panel). As a positive control, mouse BaxcDNA was used in the retransformation (lower panel). responding oligonucleotides were cloned into pENTR™/U6 vector (Invitrogen). The RNA interference sequence for mouse eleven-nine- teen leukemia (ENL) gene, GCTGTGAGAAGCTCACCTTCA, was a number of tBid-interacting proteins. These included two anti-apop- used to produce control short hairpin RNA (shRNA), and its oligonu- totic Bcl-2 family proteins Mcl-1 and Bcl-2. Only one positive clone cleotides were also cloned into the vector. DH5 E. coli (Invitrogen) representing a cDNA of Bcl-2 was found to interact with tBid, consist- were transformed, and clones were verified by sequencing. The cor- ent with a previous report that Bid interacted with Bcl-2 (18). There rectly identified clones were transferred via a gateway reaction to a were, however, three independent positive clones identified represent- TM modified pLenti6/BLOCK-iT -DEST vector (Invitrogen), a promot- ing cDNAs encoding Mcl-1, and the in-frame amino acid sequences erless lentiviral destination vector in which the blasticidin resistance from each clones were aligned against the full-length murine Mcl-1 (19) marker is replaced with tailless human CD2 as a marker. 293 cells were as shown in Fig. 1A. One positive clone (clone 85) containing the short- transfected with the plasmids using Lipofectamine reagent to produce est Mcl-1 cDNA lacked the 5-region encoding for the first 144 amino the virus. 48 h later, the lentivirus-containing supernatants were har- acids, suggesting that the C terminus fragment of the Mcl-1 protein vested to infect HeLa cells in the presence of Polybrene. The infected containing the BH1, BH2, and BH3 domains was responsible for inter- HeLa cells were harvested 48 h later for the analysis of Mcl-1 expression acting with tBid. To exclude the possibility of false interaction, the plas- by immunoblotting. Induction of apoptosis by TRAIL was performed mids from the positive clones were isolated and retransformed into essentially as described above. yeast containing the plasmid of tBid cDNA, and the activity of the -ga- lactosidase reporter gene was examined using the filter-lift assay. As Bid RESULTS has also been reported to interact with Bax (18), we used Bax as positive Truncated Bid Interacts with Mcl-1—We initially performed a yeast control with the mouse Bax cDNA cloned in the pGADT7 vector two-hybrid screen using truncated Bid as the bait protein and identified during retransformation. The filter-lift assay detected the -galactosid- 5752 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 9 •MARCH 3, 2006 Mcl-1 Binds tBid and Blocks Cytochrome c Release FIGURE 2. Truncated Bid interacts with Mcl-1 in vitro and in vivo. A, in vitro translated [ S]methionine-labeled Mcl-1 (upper panel) and Bcl-2 (lower panel) were incubated with FIGURE 3. BH domains of both tBid and Mcl-1 are required for interaction. A, require- GST alone, GST-Bid, or GST-tBid immobilized on glutathione-Sepharose beads. Bound ment of BH3 domain of tBid to bind Mcl-1. In vitro translated [ S]methionine-labeled proteins were visualized using a PhosphorImager after protein separation by SDS-PAGE. Mcl-1 was incubated with GST alone, GST-tBid, or mutant GST-tBid (mtBH3) (G94E) B, HeLa cells were transfected with either pcDNA3.1 vector alone or a HA-tBid construct immobilized on glutathione-Sepharose beads. B, BH1, BH2, and BH3 domains of Mcl-1 in the presence of 75 M Z-VAD-fmk. 24 h after transfection, cells were lysed and subject contribute to interaction with tBid. GST alone and GST-tBid were immobilized on gluta- to immunoprecipitation (IP) using anti-HA antibody (upper panel) or anti-Mcl-1 antibody thione-Sepharose beads and incubated with in vitro translated [ S]methionine-labeled (lower panel). Precipitated immunocomplexes were analyzed by SDS-PAGE and immu- Mcl-1 (upper panel) or Mcl-1mtBH1 (G262E mutation) (upper middle panel) or Mcl- noblotting (IB) using anti-Mcl-1 antibody (upper panel), anti-Bcl-2 antibody (middle 1mtBH2 (W305A and W312A double mutations) (lower middle panel) or Mcl-1mtBH3 panel), or anti-Bid antibody (lower panel). (G217E and D218A double mutations) (lower panel). All bound proteins were analyzed using a PhosphorImager. ase activity in the clone transformed with plasmid containing Bax cDNA in the presence of tBid (Fig. 1B, lower right panel). Interestingly, the BH3 Domain of tBid and All Three BH Domains of Mcl-1 Are clone identified as expressing the Mcl-1 fusion protein showed a much Required for Interaction—We then investigated the binding sites greater intensity of -galactosidase staining than the clone expressing responsible for the interaction between tBid and Mcl-1. Based on pre- the Bcl-2 fusion protein (Fig. 1B, compare upper right panel with middle vious reports describing the amino acids critical for interactions right panel). between Bcl-2 family proteins (18, 20–22) and our observation from the To see whether the interaction of Mcl-1 with tBid could be confirmed yeast two-hybrid experiment that the C terminus fragment of Mcl-1 was by an independent method, we performed GST fusion protein pull- required for interaction, we hypothesized that the interaction between down experiments. GST-Bid and GST-tBid fusion proteins were immo- Mcl-1 and tBid involved the BH domains of both proteins. To test this bilized on glutathione-Sepharose beads, respectively, and incubated hypothesis we generated a number of constructs that contained cDNAs with either [ S]-methionine labeled Mcl-1 or Bcl-2 (Fig. 2A). This assay encoding for Mcl-1 and tBid but with their BH domains mutated. First, identified Mcl-1 as a protein specifically interacting with tBid rather we compared the abilities of GST-tBid with mutant GST-tBid (mtBH3) than the full-length Bid (Fig. 2A, compare lane 4 with lane 3, upper (containing G94E mutation at BH3 domain) to interact with [ S]me- panel). Bcl-2 was also seen to preferentially interact with the truncated thionine-labeled Mcl-1. GST-tBid, but not mutant GST-tBid (mtBH3), form of Bid (Fig. 2A, compare lane 4 with lane 3, lower panel). Similar to was shown to interact with Mcl-1 (Fig. 3A, compare lane 3 with lane 4). the result shown in Fig. 1B, tBid fusion protein again showed a higher This indicated that the BH3 domain of tBid was required for interaction affinity for Mcl-1 than Bcl-2 (Fig. 2A). with Mcl-1. Next, we compared the binding ability of wild type Mcl-1 to Next, we carried out co-immunoprecipitation experiments to check GST-tBid with that of Mcl-1 proteins that had mutations at BH1 whether this interaction occurs within cells. A HA-tagged tBid con- (G262E mutation), BH2 (W305A and W312A double mutations), or struct was generated and subsequently used to overexpress tBid in HeLa BH3 (G217E and D218A double mutations) domain, respectively. As cells in the presence of the pan caspase inhibitor Z-VAD-fmk to delay shown in Fig. 3B, apart from wild type Mcl-1, all three Mcl-1 proteins cell death. Cells were lysed 24 h after transfection, and immunoprecipi- with their respective BH domain mutated failed to interact with GST- tation was done using an anti-HA antibody. The endogenous Mcl-1 was tBid. Mutations in BH1 and BH3 domains resulted in a complete loss of detected in complex with HA-tBid by immunoblotting (Fig. 2B, upper interaction, whereas mutations in BH2 domains severely impaired the panel). In addition, endogenous Bcl-2 was also detected in complex with binding ability of Mcl-1 to GST-tBid. These results demonstrated that HA-tBid (Fig. 2B, middle panel), in agreement with an early report (18). all three BH domains of Mcl-1 contribute to its ability to interact with Similarly, when an anti-Mcl-1 antibody was used to immunoprecipitate tBid. Mcl-1 complex from lysates of cells overexpressing HA-tBid, tBid was Mcl-1 Prevents tBid-mediated Cytochrome c Release—We then stud- clearly observed to interact with Mcl-1 (Fig. 2B, lower panel). It is ied the functional significance of the interaction between tBid and important to mention that the endogenous full-length Bid was not Mcl-1. As tBid has been shown to possess potent cytochrome c release immunoprecipitated by the anti-Mcl-1 antibody under the experimen- activity (10, 11), an in vitro assay was set up to examine this activity. tal conditions used (data not shown), further underlining the specificity Mitochondria isolated from HeLa cells were treated with increasing of the interaction between Mcl-1 and the truncated form of the Bid amount of recombinant tBid, and as shown in Fig. 4A (upper panel), tBid protein. induced cytochrome c release into the supernatant in a dose-dependent MARCH 3, 2006• VOLUME 281 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5753 Mcl-1 Binds tBid and Blocks Cytochrome c Release of cytochrome c (Fig. 4B, lane 4, supernatant). At 2 ng/l concentration, Mcl-1 significantly inhibited the release (Fig. 4B, lane 5, supernatant), whereas at 5 ng/l Mcl-1 completely prevented tBid-induced cyto- chrome c release (Fig. 4B, lane 6, supernatant). The complete protection by Mcl-1 (5 ng/l) was also confirmed by the full retention of cyto- chrome c in the mitochondrial fraction (Fig. 4B, lane 6, upper panel, pellet). The same membrane from the mitochondrial fraction was also reprobed for tBid to confirm its presence (Fig. 4B, middle panel, pellet) and for cytochrome oxidase as sample loading control (Fig. 4B, lower panel, pellet). Thus, recombinant Mcl-1 inhibited tBid-induced cyto- chrome c release from the mitochondria in a dose-dependent manner. To further confirm that Mcl-1 inhibits tBid function through its inter- action with tBid, we generated a recombinant Mcl-1 protein containing G217E and D218A double mutations in its BH3 domain (Mcl-1mtBH3), which, we have shown previously, failed to interact with tBid (Fig. 3B, lower panel). Pre-incubating mitochondria with the recombinant Mcl- 1mtBH3 protein itself (5 ng/l) had no effect on the release of cyto- chrome c (Fig. 4C, lane 5, supernatant). However, it did not protect mitochondria from tBid (1 ng/l)-mediated cytochrome c release (Fig. 4C, lane 4, supernatant), whereas recombinant wild type Mcl-1 (5 ng/l) effectively blocked this release (Fig. 4C, lane 3, supernatant). This is also confirmed by the observation that the loss of cytochrome c in the mitochondria occurred when both recombinant Mcl-1mtBH3 protein and tBid were present (Fig. 4C, lane 4, upper panel, pellet) but was prevented by recombinant wild type Mcl-1 (Fig. 4C, lane 3, upper panel, pellet). Thus, our data provided compelling evidence for a functional effect of the interaction between Mcl-1 and tBid. Truncated Bid Displaces Bak from Mcl-1-Bak Complex—Recently it has been shown that Mcl-1 forms a complex with Bak in healthy, FIGURE 4. Mcl-1 prevents tBid-mediated cytochrome c release from isolated mito- unstressed cells (23, 24), we therefore wanted to examine what happens chondria. A, mitochondria isolated from HeLa cells were incubated with recombinant to this complex when tBid was present. We used a HA-tagged tBid tBid at the indicated concentrations. At the end of incubation, supernatant of the mito- construct to overexpress tBid in HeLa cells in the presence of the pan chondrial suspension was collected and subjected to SDS-PAGE and immunoblotting for cytochrome c (Cyt. c, upper panel). Untreated mitochondria (Mito.) were used as positive caspase inhibitor Z-VAD-fmk as described previously. Immunoprecipi- control for the detection of cytochrome c (lane 1). The same membrane was reprobed for tation using an anti-HA antibody showed that tBid interacted with both cytochrome oxidase (subunit II) to check whether the supernatant contains mitochon- drial contamination (Cyt. Oxid., lower panel). B, mitochondria were pre-incubated with endogenous Bak and Bax (Fig. 5A, upper and lower panels, respectively), recombinant Mcl-1 at the indicated concentrations before the addition of tBid. Both which was consistent with the published reports (3, 4). Immunoprecipi- supernatant and pellet of the mitochondrial suspension were collected at the end of tation using anti-Mcl-1 antibody showed that Bak indeed interacted incubation and subject to SDS-PAGE and immunoblotting for cytochrome c (Cyt. c, super- natant, and upper panel, pellet). Membrane from the pellet samples was also probed for with Mcl-1 in cells transfected with empty vector, as detected by immu- tBid for its presence (middle panel, pellet) and for cytochrome oxidase (subunit II) as noblotting (Fig. 5B, lane 1, upper panel). However, when cells were sample loading control (Cyt. Oxid., lower panel, pellet). C, mitochondria were pre-incu- bated with recombinant Mcl-1 (5 ng/l) or mutant Mcl-1mtBH3 (G217E and D218A dou- transfected with the HA-tBid construct the level of Bak in complex with ble mutations) (5 ng/l) before the addition of tBid. At the end of incubation, both Mcl-1 was greatly reduced (Fig. 5B, lane 2, upper panel), whereas the supernatant and pellet of the mitochondrial suspension were collected and subject to SDS-PAGE and immunoblotting as described in B. total amount of Mcl-1 in both immunoprecipitated samples was similar (Fig. 5B, middle panel). Interaction of Mcl-1 with tBid was again con- firmed by reprobing Bid on the same membrane (Fig. 5B, lane 2, lower manner. As a positive control, 30 g of untreated mitochondria was panel). We also probed for Bax and could not detect Bax in the samples used for the detection of cytochrome c (Fig. 4A, lane 1, upper panel). immunoprecipitated by anti-Mcl-1 antibody under the experimental The same blot was reprobed for cytochrome oxidase (subunit II), a conditions used (data not shown). To test the possibility that overex- mitochondrial membrane protein, and it confirmed that the superna- pression of HA-tBid may alter the expression levels of endogenous Bak tant samples were free from mitochondrial contamination (Fig. 4A, and/or Mcl-1, resulting in the reduction of Bak in complex with Mcl-1, lanes 2–5, lower panel). we also checked the levels of Bak, Mcl-1, and tBid in the total cell lysates As Mcl-1 is an anti-apoptotic protein, we reasoned that its ability to prior to the immunoprecipitation. Immunoblotting analysis showed interact with tBid would possibly interfere with the function of the tBid that the expression levels of both Bak and Mcl-1 remained unchanged protein. To test this, isolated mitochondria from HeLa cells were pre- (Fig. 5B, lanes 3 and 4, upper and middle panels), regardless of the incubated with increasing amounts of recombinant Mcl-1 protein before the addition of tBid. Recombinant Mcl-1 alone (5 ng/l) had no presence of HA-tBid (Fig. 5B, lane 4, lower panel). Densitometric anal- effect on the release of cytochrome c (Fig. 4B, lane 2, supernatant). The ysis of bands representing respective Bak and Mcl-1 on membranes from three independent immunoprecipitation experiments indicated treatment of mitochondria with tBid (1 ng/l) resulted in the release of cytochrome c from mitochondria (Fig. 4B, lane 3, supernatant), which is that the relative level of Bak to Mcl-1 was about 4-fold less in an immu- further confirmed by the disappearance of cytochrome c in the mito- noprecipitated sample from cells overexpressing tBid than that from chondrial fraction (Fig. 4B, lane 3, upper panel, pellet). At a concentra- control cells (Fig. 5C). Statistical analysis by Student’s t test showed that tion of 1 ng/l, recombinant Mcl-1 failed to block tBid-mediated release the difference was significant (p  0.05). Therefore, in addition to its 5754 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 9 •MARCH 3, 2006 Mcl-1 Binds tBid and Blocks Cytochrome c Release ability to interact directly with Bak and Bax, tBid can also bind Mcl-1 and displace Bak from the Mcl-1-Bak complex. Mcl-1 Inhibits Apoptosis Induced by TRAIL and TNF- in HeLa Cells—Because Bid has been shown to be cleaved early during the induction of apoptosis by TRAIL in HeLa cells (25, 26), we wished to investigate whether cells overexpressing Mcl-1 would be resistant to apoptosis induced by TRAIL. HeLa cells were transfected with either pcDNA3.1 vector alone or human Mcl-1 cDNA expression construct and selected in G418. Single cell clones were picked and expanded to establish the stable Mcl-1 overexpressing cells. The level of Mcl-1 expression was evaluated by immunoblotting, which showed that its level is higher in cells transfected with the Mcl-1 construct than those with vector-only or wild type HeLa cells (Fig. 6A). Induction of apopto- sis was assessed by flow cytometry for subdiploid DNA content (sub- G ) following propidium iodide staining and PARP cleavage, a biochem- ical marker of apoptosis. As shown in Fig. 6B, treatment of wild type HeLa cells with increasing doses of TRAIL resulted in a dose-dependent induction of apoptosis. This treatment also caused a similar dose-de- pendent induction of apoptosis in cells transfected with vector alone (Fig. 6B). Cells stably overexpressing Mcl-1 were, however, consistently protected from cell death following treatment with TRAIL at all con- centrations (Fig. 6B). Student’s t test analysis showed that the reduction in TRAIL (250 ng/ml)-induced apoptosis of cells overexpressing Mcl-1, when compared with cell death in wild type HeLa cells and in cells transfected with vector alone, was statistically significant (both p 0.01). Also, treatment with increasing amount of TRAIL resulted in a dose-dependent cleavage of PARP in wild type and vector-only trans- fected HeLa cells (Fig. 6B). PARP cleavage was less complete in Mcl-1 overexpressing cells than that seen in wild type and vector-only trans- fected HeLa cells following treatment of TRAIL at 250 ng/ml concen- tration (Fig. 6B, compare lane 15 with lanes 13 and 14, respectively, PARP). As TNF- has been shown to induce apoptosis in HeLa cells through a Bid-dependent pathway (15), we also treated the above cells with TNF- to see whether cells overexpressing Mcl-1 would be resistant to apoptosis induced by TNF-. Treatment with Me SO or cycloheximide (30 g/ml) alone did not cause significant increase in cell death in all three types of cells (data not shown). Treatment with TNF- (15 ng/ml) in the presence of cycloheximide for 22 h resulted in similar levels of apoptosis in wild type HeLa cells and cells stably transfected with vector alone (Fig. 6C, lanes 2 and 4, respectively). Cells stably overexpressing Mcl-1 were indeed partially resistant to TNF--induced cell death (Fig. 6C, lane 6). The reduction in TNF--induced apoptosis in cells overex- pressing Mcl-1, when compared with cell death in wild type HeLa cells and cells transfected with vector alone, was statistically significant (both p  0.05). Again, PARP cleavage was not as complete in Mcl-1 overex- pressing cells as that seen in wild type HeLa cells and cells transfected with vector alone (Fig. 6C, compare lane 6 with lanes 2 and 4, respec- tively, PARP). Inhibition of death receptor-mediated apoptosis was also FIGURE 5. tBid binds Mcl-1 and displaces Bak from Mcl-1-Bak complex. A, HeLa cells were transfected with either pcDNA3.1 vector alone or a HA-tBid construct in the pres- observed in another stable HeLa cell line overexpressing Mcl-1 (see Fig. ence of 75 M Z-VAD-fmk. Cells were lysed 24 h after transfection and subjected to 1 in supplemental data). Therefore, overexpressing Mcl-1 conferred immunoprecipitation (IP) using anti-HA antibody. Precipitated immunocomplexes were resistance to apoptosis induced by both TRAIL and TNF- in HeLa analyzed by SDS-PAGE and immunoblotting (IB) using anti-Bak antibody (upper panel)or anti-Bax antibody (lower panel). Asterisks denote nonspecific bands. B, the above cell cells. lysates were immunoprecipitated by anti-Mcl-1 antibody. The precipitates were subject Mcl-1 Silencing by RNA Interference Sensitizes HeLa Cells to TRAIL- to SDS-PAGE and immunoblotting using anti-Bak antibody (left upper panel), anti-Mcl-1 antibody (left middle panel), or anti-Bid antibody (left lower panel). Prior to immunopre- induced Apoptosis—To determine the effect of Mcl-1 down-regulation cipitation, 5% of total cell lysates (TCL) from cells transfected with either pcDNA3.1 vector on death receptor-mediated cell death, we used a lentiviral vector for alone or a HA-tBid construct were also analyzed by SDS-PAGE and immunoblotting for the expression levels of Bak (right upper panel), Mcl-1 (right middle panel), or Bid (right expression of shRNA to induce Mcl-1 silencing. HeLa cells were lower panel). C, relative level of Bak to Mcl-1 in immunoprecipitated samples by anti- infected with lentivirus containing vectors expressing either control Mcl-1 antibody was analyzed by densitometry after scanning the bands representing shRNA or Mcl-1 shRNA. The level of Mcl-1 expression was then eval- respective Bak and Mcl-1 on membranes from three independent immunoprecipitation experiments. uated by immunoblotting. As shown in Fig. 7A, the expression level of MARCH 3, 2006• VOLUME 281 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5755 Mcl-1 Binds tBid and Blocks Cytochrome c Release FIGURE 6. Mcl-1 inhibits apoptosis induced by TRAIL and TNF- in HeLa cells. A, HeLa cells were stably transfected with either pcDNA3.1 vector alone or plasmid containing the cDNA encoding human Mcl-1. The expression of Mcl-1 in wild type (w/t), pcDNA3.1 vector and Mcl-1 overexpressing HeLa cells was assessed by immunoblotting for Mcl-1 (upper panel). The same membrane was also probed for -tubulin as sample loading control (lower panel). B, apoptosis was induced by the treatment of the three HeLa cell lines with soluble human recombinant TRAIL at the indicated con- centrations for 14 h and assessed by both flow cytometry for cells with sub-G DNA content and PARP cleavage by immunoblotting. -Tubulin was also probed as sample loading control. C, apopto- sis was also induced by the treatment of the three HeLa cell lines with TNF- (15 ng/ml) in the pres- ence of cycloheximide (CHX,30 g/ml) for 22 h and assessed as described in B. Mcl-1 is clearly reduced in cells infected with lentivirus expressing control shRNA following the treatment with TRAIL (Fig. 7B, compare Mcl-1 shRNA. We then treated these cells with TRAIL (100 ng/ml) for lane 6 with lanes 2 and 4, respectively, PARP). 14 h to induce apoptosis. Treatment with TRAIL resulted in similar DISCUSSION levels of apoptosis in wild type HeLa cells and cells infected with virus expressing control shRNA (Fig. 7B, lanes 2 and 4, respectively). How- Most of the studies on protein-protein interaction of Bid in apoptosis ever, cells infected with virus expressing Mcl-1 shRNA became more have, to date, focused on its interaction with the pro-apoptotic Bax and sensitive to TRAIL-induced cell death (Fig. 7B, lane 6). The increase in Bak proteins. It has been shown that tBid can directly interact with both TRAIL-induced apoptosis in cells infected with virus expressing Mcl-1 Bax and Bak and induce their oligomerization, resulting in cytochrome shRNA, when compared with cell death in wild type HeLa cells and cells c release (3, 4). This function of tBid is vital for the transmission of infected with virus expressing control shRNA, was statistically signifi- apoptotic signals from death receptors to mitochondria in certain tis- cant (both p  0.05). In addition, there is a greater loss of intact PARP sues, and, indeed, apoptosis in hepatocytes is dependent on tBid-medi- observed in cells infected with virus expressing Mcl-1 shRNA than that ated amplification of the apoptotic signal via the mitochondria after seen in wild type HeLa cells and cells infected with virus expressing the engagement of the death receptor Fas (27). However, there has been 5756 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 9 •MARCH 3, 2006 Mcl-1 Binds tBid and Blocks Cytochrome c Release FIGURE 7. Mcl-1 silencing by RNA interference sensitizes HeLa cells to TRAIL-induced apoptosis. A, HeLa cells were infected with lentivirus-containing vectors expressing either control shRNA or Mcl-1 shRNA. The expression of Mcl-1 in wild type (w/t) HeLa cells and cells infected with virus expressing either control shRNA or Mcl-1 shRNA was assessed by immunoblotting for Mcl-1 (upper panel). The same membrane was also probed for-tubulin as sample loading control (lower panel). B, apoptosis was induced by the treatment of the three HeLa cell lines with soluble human recombinant TRAIL (100 ng/ml) for 14 h and assessed by both flow cytometry for cells with sub-G DNA content and PARP cleavage by immunoblotting. -Tubulin was also probed as sample loading control. little information about the role of anti-apoptotic Bcl-2 proteins in reg- Mcl-1 proteins, whereas in our study tBid was shown to interact with ulating the function of tBid. Here we show that Mcl-1 can regulate endogenous Mcl-1. tBid-mediated apoptosis through its ability to interact with tBid. Our It has been proposed that the BH3 domain-only proteins can be fur- yeast two-hybrid screen identified Mcl-1 as a potent interaction partner ther divided into two subgroups, activating or sensitizing (29, 30). Those for tBid, and this interaction was confirmed by both GST fusion protein proteins with activating BH3 domains (e.g. Bid and Bim) appear to have pull-down experiments in vitro and co-immunoprecipitation in vivo.In a higher affinity for Bak or Bax than Bcl-2, perhaps directly activating addition, we have compared Mcl-1 with Bcl-2 for the ability to bind tBid Bak and Bax. The BH3-only proteins with sensitizing BH3 domains (e.g. in both the yeast two-hybrid and GST pull-down experiments, and tBid Bad) appear to preferentially bind Bcl-2 allowing Bak and Bax to be appears to have a much higher affinity for Mcl-1 than Bcl-2. The co- derepressed. In this study, we confirm that tBid can directly interact immunoprecipitation assay showed that tBid interacts with Mcl-1 as with Bak and Bax. More importantly, we have also shown that tBid can well as Bcl-2 in cells. Recently, it has been shown that C-terminal frag- bind to Mcl-1 and disrupt the Mcl-1-Bak complex. As a result, Bak is ments of Mcl-1 can also interact with tBid (28). However, in that study displaced from the complex. Recently, it has been reported that another the interaction was observed on the basis of exogenously expressed BH3-only protein Noxa can also displace Bak from Mcl-1/Bak complex MARCH 3, 2006• VOLUME 281 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5757 Mcl-1 Binds tBid and Blocks Cytochrome c Release (31), suggesting that this displacement is indeed a common feature for chondria and apoptosis in HeLa cells (17). Mcl-1 alone may not be able BH3-only proteins to activate Bak and/or Bax. Therefore, our study to offer complete protection of cells from apoptosis. Conversely, deple- tion of Mcl-1 alone may be insufficient to render all the cells sensitive to demonstrates dual modes of action by tBid, which is capable of both activating and sensitizing other pro-apoptotic Bcl-2 family proteins. On death receptor-mediated apoptosis. the other hand, Mcl-1 can act through direct binding to neutralize tBid, Mcl-1 was initially discovered as an early induction gene during dif- ferentiation of the myeloid cell line, ML-1 (39), and is widely expressed thus preventing it from activating Bak or Bax. In this respect Mcl-1 may in a variety of human tissues and cells as well as many tumors (40, 41). be playing an active role in counter balancing activating BH3 domain- only proteins. This notion is supported by the recent studies that Bim, Deletion of Mcl-1 in mice led to embryonic lethality during the peri- implantation stage, suggesting it is essential for embryonic development another activating BH3 domain-only protein, has a higher affinity for (42). Genetic studies with conditional knock-out approach also reveal Mcl-1 than Bcl-2 (32), and Mcl-1 effectively inhibits Bim-mediated that Mcl-1 is required both in early lymphoid development and in the release of mitochondrial cytochrome c (33). Our study shows that Mcl-1 maintenance of mature B and T lymphocytes, which are rapidly lost interacts with tBid and impairs the ability of tBid to induce cytochrome when Mcl-1 is deleted (32). Mcl-1 also plays physiologically important c release and apoptosis. Recently, two independent studies revealed dif- roles in regulating myeloid cell survival (43, 44). Given that Mcl-1 can ferential targeting of anti-apoptotic Bcl-2 proteins by BH3-only pro- interact strongly with tBid and other BH3-only proteins such as Bim and teins using individual BH3 peptides (34, 35). A Bim BH3 peptide appears inhibits their induction of cytochrome c release and activation of the capable of interacting with most anti-apoptotic proteins including mitochondrial apoptotic pathway, loss of Mcl-1 may render the cells Mcl-1 with high affinity, whereas the Bid BH3 peptide does not bind sensitive to apoptosis induced by a variety of apoptotic stimuli including Mcl-1. This is not unexpected, as we have shown that full-length Bid is the activation of death receptors of TNF family. Here we demonstrate a not able to interact with Mcl-1. Only the truncated Bid, which may biochemical link between Mcl-1, an essential regulator of lymphoid undergo further modification by postproteolytic N-myristoylation, is homeostasis, and receptor-mediated apoptosis, the pre-eminent path- able to bind Mcl-1, suggesting that there is a complex structural basis way controlling survival of the cells of the immune system (45, 46). underlying the interaction of Bid with other proteins, which is not dis- cernable in binding studies using Bid BH3 peptide alone. REFERENCES Our in vivo experiments in HeLa cells were prompted by the involve- 1. Gross, A., McDonnell, J. M., and Korsmeyer, S. J. (1999) Genes Dev. 13, 1899–1911 ment of Bid in apoptosis mediated by TRAIL and TNF- in HeLa cells 2. Cory, S., and Adams, J. M. (2002) Nat. Rev. Cancer 2, 647–656 (15, 25, 26). We have shown that Mcl-1 can indeed protect HeLa cells 3. Desagher, S., Osen-Sand, A., Nichols, A., Eskes, R., Montessuit, S., Lauper, S., Maun- drell, K., Antonsson, B., and Martinou, J.-C. (1999) J. Cell Biol. 144, 891–901 from apoptosis induced by TRAIL and TNF-. Mcl-1 silencing by a 4. Wei, M. C., Lindsten, T., Mootha, V. K., Weiler, S., Gross, A., Ashiya, M., Thompson, shRNA approach sensitizes HeLa cells to TRAIL-induced apoptosis. C. B., and Korsmeyer, S. J. (2000) Genes Dev. 14, 2060–2070 Recently, it has been reported that Mcl-1 mediates resistance to TRAIL- 5. Wei, M. C., Zong, W.-X., Cheng, E. H.-Y., Lindsten, T., Panoutsakopoulou, V., Ross, A. J., Roth, K. A., MacGregor, G. R., Thompson, C. B., and Korsmeyer, S. J. (2001) induced apoptosis in human cholangiocarcinoma cells (36). In their Science 292, 727–730 study, resistance was specifically associated with overexpression of 6. Cheng, E.H.-Y.A., Wei, M. C., Weiler, S., Flavell, R. A., Mak, T. W., Lindsten, T., and Mcl-1 and depletion of Mcl-1 by the small interfering RNA method also Korsmeyer, S. J. (2001) Mol. Cell 8, 705–711 sensitizes cells to TRAIL-mediated apoptosis despite Bcl-2 expression. 7. Zong, W.-X., Lindsten, T., Ross, A. J, MacGregor, G. R., and Thompson, C. B. (2001) Genes Dev. 15, 1481–1486 Furthermore, another study has also shown that hepatocyte growth 8. Green, D. R. (2000) Cell 102, 1–4 factor-mediated Mcl-1 induction inhibits apoptosis induced by Fas in 9. Strasser, A., O’Connor, L., and Dixit, V. M. (2000) Annu. Rev. Biochem. 69, 217–245 human primary hepatocytes (37). As TRAIL and Fas induce apoptosis 10. Li, H., Zhu, H., Xu, C.-J., and Yuan, J. (1998) Cell 94, 491–501 via a similar mechanism (38), it is conceivable that Mcl-1 could bind to 11. Luo, X., Budihardjo, I., Zou, H., Slaughter, C., and Wang, X. (1998) Cell 94, 481–490 12. Eskes, R., Desagher, S., Antonsson, B., and Martinou, J. C. (2000) Mol. Cell. Biol. 20, tBid and interfere with its function in activating Bax and/or Bak, result- 929–935 ing in protection of these cells from death receptor-mediated apoptosis. 13. Brady, H. J., Salomons, G. S., Bobeldijk, R. C., and Berns, A. J. (1996) EMBO J. 15, Our observations of interaction between Mcl-1 and tBid and subse- 1221–1230 quent prevention of cytochrome c release therefore provide a potential 14. Clohessy, J. G., Zhuang, J., and Brady, H. J. M. (2004) Br. J. Haematol. 125, 655–665 15. Perez, D., and White, E. (2000) Mol. Cell 6, 53–63 biochemical mechanism to explain the anti-apoptotic effect of Mcl-1 in 16. Zhuang, J., Ren, Y., Snowden, R. T., Zhu, H., Gogvadze, V., Savill, J. S., and Cohen, death receptor-mediated apoptosis through specific targeting of tBid. G. M. (1998) J. Biol. Chem. 273, 15628–15632 It is worth noting that under the experimental conditions used, over- 17. Nijhawan, D., Fang, M., Traer, E., Zhong, Q., Gao, W., Du, F., and Wang, X. (2003) expression of Mcl-1 does not fully protect cells from both TRAIL- and Genes Dev. 17, 1475–1486 18. Wang, K., Yin, X. M., Chao, D. T., Milliman, C. L., and Korsmeyer, S. J. (1996) Genes TNF--induced apoptosis although it significantly reduces the level of Dev. 10, 2859–2869 cell death in HeLa cells. It is possible that the level of Mcl-1 expression 19. Okita, H., Umezawa, A., Suzuki, A., and Hata, J.-i. (1998) Biochim. Biophys. Acta. in stably transfected cells is not high enough to prevent cell death com- 1398, 335–341 20. Cheng, E. H.-Y., Levine, B., Boise, L. H., Thompson, C. B., and Hardwick, J. M. (1996) pletely. In addition, in the case of TNF--induced apoptosis, the pres- Nature 379, 554–556 ence of protein synthesis inhibitor cycloheximide inevitably blocks pro- 21. Yin, X.-M., Oltvai, Z. N., and Korsmeyer, S. J. (1996) Nature 369, 321–323 duction of Mcl-1 protein and thus reduces the anti-apoptotic function 22. Zha, H., Aime-Sempe, C., Sato, T., and Reed, J. C. (1996) J. Biol. Chem. 271, of Mcl-1. However, there is emerging evidence to suggest that complete 7440–7444 23. Cuconati, A., Mukherjee, C., Perez, D., and White, E. (2003) Genes Dev. 17, protection of apoptosis requires multiple lines of resistance conferred 2922–2932 by anti-apoptotic proteins and Mcl-1 may provide the first line of resist- 24. Leu, J. I.-J., Dumont, P., Hafey, M., Murphy, M. E., and George, D. L. (2004) Nat. Cell ance to the induction of apoptosis by TRAIL and TNF-. This is con- Biol. 6, 443–450 sistent with the observation that in adenoviral protein E1A-induced 25. Harper, N., Farrow, S. N., Kaptein, A., Cohen, G. M., and MacFarlane, M. (2001) J. Biol. Chem. 276, 34743–34752 apoptosis in HeLa cells, loss of Mcl-1 is required to initiate the apoptotic 26. Seol, D.-W., Li, J., Seol, M.-H., Park, S.-Y., Talanian, R. V., and Billiar, T. R. (2001) pathway (23). The idea is further supported by a recent study demon- Cancer Res. 61, 1138–1143 strating that Mcl-1 functions upstream of and together with Bcl-x in 27. Yin, X.-M., Wang, K., Gross, A., Zhao, Y., Zinkel, S., Klocke, B., Roth, K. A., and preventing UV irradiation-induced cytochrome c release from mito- Korsmeyer, S. J. (1999) Nature 400, 886–891 5758 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 9 •MARCH 3, 2006 Mcl-1 Binds tBid and Blocks Cytochrome c Release 28. Weng, C., Li, Y., Yu, D., Shi, Y., and Tang, H. (2005). J. Biol. Chem. 280, 10491–10500 Dax, A., Zink, D., Buchler, P., Muller, M., and Krammer, P. H. (2004) Hepatology 39, 29. Letai, A., Bassik, M. C., Walensky, L. D., Sorcinelli, M. D., Weiler, S., and Korsmeyer, 645–654 S. J. (2002) Cancer Cell 2, 183–192 38. Wang, S., and El-Deiry, W. S. (2003) Oncogene. 22, 8628–8633 30. Scorrano, L., and Korsmeyer, S. J. (2003) Biochem. Biophys. Res. Commun. 304, 39. Kozopas, K. M., Yang, T., Buchan, H. L., Zhou, P., and Craig, R. W. (1993) Proc. Natl. 437–444 Acad. Sci. U. S. A. 90, 3516–3520 31. Willis, S. N., Chen, L., Dewson, G., Wei, A., Naik, E., Fletcher, J. I., Adams, J. M., and 40. Krajewski, S., Bodrug, S., Gascoyne, R., Berean, K., Krajewska, M., and Reed, J. C. Huang, D. C. S. (2005) Genes Dev. 19, 1294–1305 (1994) Am. J. Pathol. 145, 515–525 32. Opferman, J. T., Letai, A., Beard, C., Sorcinelli, M. D., Ong, C. C., and Korsmyer, S. J. 41. Krajewski, S., Bodrug, S., Krajewska, M., Shabaik, A., Gascoyne, R., Berean, K., and (2003) Nature 426, 671–676 Reed, J. C. (1995) Am. J. Pathol. 146, 1309–1319 33. Han, J., Goldstein, L. A., Gastman, B. R., Froelich, C. J., Yin, X.-M., and Rabinnowich, 42. Rinkenberger, J. L., Horning, S., Klocke, B., Roth, K., and Korsmeyer, S. J. (2000) Genes H. (2004) J. Biol. Chem. 279, 22020–22029 Dev. 14, 23–27 34. Chen, L., Wills, S. N., Wei, A., Smith, B. J., Fletcher, J. I., Hinds, M. G., Coleman, P. M., 43. Epling-Burnette, P. K., Zhong, B., Bai, F., Jiang, K., Bailey, R. D., Garcia, R., Jove, R., Day, C. L., Adams, J. M., and Huang, D. C. S. (2005) Mol. Cell 17, 393–403 Djeu, J. Y., Loughran, T. P. Jr., and Wei, S. (2001) J. Immunol. 166, 7486–7495 35. Kuwana, T., Bouchier-Hayes, L., Chipuk, J. C., Bonzon, C., Sullivan, B. A., Green, 44. Edwards, S. W., Derouet, M., Howse, M., and Moots, R. J. (2004) Biochem. Soc. Trans. D. R., and Newmeyer, D. D. (2005) Mol. Cell 17, 525–535 32, 489–492 36. Taniai, M., Grambihler, A., Higuchi, H., Werneburg, N., Bronk, S. F., Farrugia, D. J., 45. Nagata, S. (1999) Annu. Rev. Genet. 33, 29–55 Kaufmann, S. H., and Gores, G. J. (2004) Cancer Res. 64, 3517–3524 46. Marsden, V. S., and Strasser, A. (2003) Annu. Rev. Immunol. 21, 71–105 37. Schulze-Bergkamen, H., Brenner, D., Krueger, A., Suess, D., Fas, S. C., Freym C. R., MARCH 3, 2006• VOLUME 281 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5759

Journal

Journal of Biological Chemistry – American Society for Biochemistry and Molecular Biology

Published: Mar 3, 2006

There are no references for this article.