An Anti-apoptotic Viral Protein That Recruits Bax to Mitochondria*

Received for publication, July 31, 2003, and in revised form, March 4, 2004Published, JBC Papers in Press, March 5, 2004, DOI 10.1074/jbc.M308408200

Delphine Poncet,a,b Nathanael Larochette,a Anne-Laure Pauleau,a Patricia Boya,a,c

Abdel-Ali Jalil,d Pierre-Francois Cartron,e Francois Vallette,e Celine Schnebelen,f

Laura M. Bartle,g Anna Skaletskaya,g David Boutolleau,h Jean-Claude Martinou,f

Victor S. Goldmacher,g Guido Kroemer,a,i and Naoufal Zamzamia, j

From the aCentre National de la Recherche Scientifique, UMR 8125, Institut Gustave Roussy, 39 rue Camille-Desmoulins,F-94805 Villejuif, France, eIFR26, INSERM U 419, 9 quai Moncousu, 44035 Nantes, France, dU487 INSERM, InstitutGustave Roussy, 39 rue Camille-Desmoulins, F-94805 Villejuif, France, the fDepartement de Biologie Cellulaire, SciencesIII, Quai Ernest-Ansermet 30, CH-1211 Geneve 4, Switzerland, gImmunoGen, Inc, Cambridge, Massachusetts 02139, andthe hBacteriologie-Virologie-Parasitologie-Hygene, Hopital de Bicetre, 78, rue du General Leclerc,94277 Le Kremlin-Bicetre, France

The viral mitochondria-localized inhibitor of apopto-sis (vMIA), encoded by the UL37 gene of human cytomeg-alovirus, inhibits apoptosis-associated mitochondrialmembrane permeabilization by a mechanism differentfrom that of Bcl-2. Here we show that vMIA inducesseveral changes in Bax that resemble those found inapoptotic cells yet take place in unstimulated, non-apoptotic vMIA-expressing cells. These changes includethe constitutive localization of Bax at mitochondria,where it associates tightly with the mitochondrial mem-brane, forming high molecular weight aggregates thatcontain vMIA. vMIA recruits Bax to mitochondria butdelays relocation of caspase-8-activated truncated Bid-green fluorescent protein (GFP) (t-Bid-GFP) to mito-chondria. The ability of vMIA and its deletion mutantsto associate with Bax and to induce relocation of Bax tomitochondria correlates with their anti-apoptotic activ-ity and with their ability to suppress mitochondrialmembrane permeabilization. Taken together, our dataindicate that vMIA blocks apoptosis via its interactionwith Bax. vMIA neutralizes Bax by recruiting it to mito-chondria and “freezing” its pro-apoptotic activity. Thesedata unravel a novel strategy of subverting an intrinsicpathway of apoptotic signaling.

Apoptosis is mediated through two main pathways, the ex-trinsic (death receptor) pathway and the intrinsic (mitochon-drial) pathway. The extrinsic pathway is initiated by ligation ofa plasma membrane death receptor, which results in a stepwiserecruitment of adaptors and initiator caspases (in particular,caspase-8) into the death-inducing signaling complex. In somecells (type I), activation of caspase-8 directly triggers activationof the caspase cascade (1), whereas in other cells (type II),

caspase-8 mediates apoptosis only via proteolytic processing ofBID (2, 3), which in turn leads to mitochondrial membranepermeabilization (MMP)1 and consequently to the amplifica-tion of the apoptotic signal. The intrinsic pathway is induced byvarious apoptotic stimuli that converge on mitochondria totrigger MMP (4). This permeabilization causes the release ofmitochondrial pro-apoptotic factors, which in turn activate thecaspase pathway.

MMP is tightly controlled, negatively and positively, by theproteins of the Bcl-2 family. Anti-apoptotic proteins such asBcl-2, Bcl-xL, and Mcl-1 inhibit MMP, whereas pro-apoptoticBcl-2 homologues such as Bax, Bak, and BH3-only proteins suchas Bid and Bad enhance MMP (3–5). Recent evidence suggeststhat many stress signals trigger apoptosis, at least in part, byactivation of pro-apoptotic BH3-only proteins (3). BH3-only pro-teins seem to induce MMP by two different pathways: either bydirect activation of Bax/Bak (e.g. Bid or Bim) or by inhibition ofanti-apoptotic activity of Bcl-2/Bcl-XL (e.g. Bad, Bik) (3, 6).

Bax is one of the key inducers of MMP (7). This protein ispredominantly cytosolic in a conformation in which its hydro-phobic �-5/6 helix and its N-terminal helix-1 are hidden withina hydrophilic globular structure. Upon pro-apoptotic signaling,for instance following exposure to t-Bid, Bax undergoes a con-formation change that exposes its N terminus and, possibly, itsBH3 domain (8, 9). This allows its translocation and its tightassociation with the mitochondria, apparently through integra-tion into the outer membrane, forming homo-oligomers andhetero-oligomers with Bak (3, 10, 11). The mechanism of theassociation of Bax with the mitochondrial outer membraneduring apoptosis is not well understood. Some data suggestthat Bax interaction with cardiolipin (a mitochondria-specificlipid) is sufficient to trigger MMP (12). On the other hand, Baxhas been shown to interact (directly or indirectly) with sessilemitochondrial proteins, voltage-dependent anion channel (13)and the adenine nucleotide translocase (14), although thesefindings are controversial (12, 15–17). In any case, most of theavailable evidence indicates that oligomerization of Bax and itstight association with the outer mitochondrial membrane leadto MMP (12, 15, 16, 18).

* This work was supported by a special grant from Ligue NationaleContre le Cancer, as well as by grants from Agence Nationale deRecherches sur le SIDA, Fondation pour la Recherche Medicale, andEuropean Commission (Grant QLG1-CT-1999-00739) (to G. K.). Thecosts of publication of this article were defrayed in part by the paymentof page charges. This article must therefore be hereby marked “adver-tisement” in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

b Received a fellowship from the French Ministry of Education.c Received a fellowship from the European Commission (MCFI-200-

00943).i To whom correspondence may be addressed. Tel.: 33-1-42-11-60-46;

Fax: 33-1-42-11-60-47; E-mail: kroemer@igr.fr.j To whom correspondence may be addressed. Tel.: 33-1-42-11-60-48;

Fax: 33-1-42-11-5244; E-mail: zamzami@igr.fr.

1 The abbreviations used are: MMP, mitochondrial membrane perme-abilization; CMV, human cytomegalovirus; CHX, cyclohexamide; GFP,green fluorescent protein; EGFP, enhanced GFP; MEF, murine embryonicfibroblasts, STS, staurosporine; t-Bid, truncated Bid; vMIA, viral mito-chondria-localized inhibitor of apoptosis; vMIAt, Myc-tagged vMIA; TES,2-{[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]amino}ethanesulfonic acid;CHAPS, 3-((cholamidopropyl)dimethylammonio)-1-propane sulfonate; Z,benzyloxycarbonyl; fmk, fluoromethyl ketone.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 21, Issue of May 21, pp. 22605–22614, 2004© 2004 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org 22605

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from

Apoptotic elimination of virally infected cells is one of thebasic anti-viral responses of multicellular organisms. Duringevolution, viruses have developed a variety of strategies tosuppress the extrinsic and/or the intrinsic pathways. Thus,some viral proteins block the extrinsic pathway by suppressingthe activation of initiator caspases such as viral Flice-like in-hibitory proteins (FLIPs) encoded by �-herpesviruses (19),vICA homologues of �-herpesviruses that associate with andprevent death-inducing signaling complex-dependent activa-tion of pro-caspase-8, (20), CrmA of poxvirus, a serpin thatinhibits the enzymatic activity of caspase-8 (21), and the Serp-2serpin of myxoma virus (22). Other viruses encode inhibitors ofthe intrinsic pathway in particular structural and functionalBcl-2 homologues. Examples of structural homologues of Bcl-2include KSbcl-2, a Bcl-2 homologue from human herpesvirus(HHV8), which resembles Bcl-2 and Bcl-xL in its three-dimen-sional structure and, likely, in its function (23). Another celldeath suppressor BHRF1, encoded by Epstein-Barr virus,shares some homologies with Bcl-2, but its mechanism of celldeath suppression remains unclear (24). The E1B19K celldeath suppressor of adenovirus shares only a modest aminoacid sequence homology with Bcl-2 and appears to suppress celldeath via its interaction with Bax and Bak, preventing theiroligomerization (25, 26). Finally, a number of viruses interceptboth the intrinsic and the extrinsic pathways. This applies toMyxoma virus, which encodes both M11L (27) and Serp-2 (22)and the cytomegalovirus with viral mitochondria-localized in-hibitor of apoptosis (vMIA) (28) and vICA (20).

The vMIA is encoded by exon 1 of the immediate early UL37gene of human cytomegalovirus (CMV). Although vMIA sharesno sequence homology with Bcl-2, it has functional similaritieswith Bcl-2: it localizes at mitochondria, inhibits MMP, and is apotent cell death suppressor (29–31). Two domains of vMIA arenecessary and, together, sufficient for its cell death-suppress-ing activity, its N-terminal domain (amino acids 5–34), and asegment between amino acids 117 and 147 (32). The N-termi-nal domain targets vMIA to mitochondria, whereas the func-tion of the second domain as well as the molecular mechanismof cell death suppression by vMIA until now remainedunknown.

Here we report that vMIA inhibits apoptosis through aunique mechanism distinct from that of any previously char-acterized cell death suppressor-targeting MMP. We show thatvMIA, through an association with Bax, recruits Bax to mito-chondria, inducing its oligomerization and tight associationwith mitochondrial outer membrane while inhibiting its pro-apoptotic function. Our findings thus reveal a yet undescribedviral strategy for interfering with the mitochondrial apoptoticsignaling pathway.

MATERIALS AND METHODS

Cell Lines, Culture Conditions, and Transfection/Infection—HeLaand BJAB cells were stably transfected with UL37 exon 1 (vMIA), adeletion mutant of vMIA (32), Bcl-2, or the empty vector, pcDNA3 (neo),as described elsewhere (29). Murine embryonic fibroblasts (MEF) werea generous gift of stanley Korsmeyer (Harvard University, Boston, MA),and human colorectal carcinoma HCT116/Bax�/� and HCT116/Bax�/�

otherwise isogenic cells were a generous gift of Bernd Vogelstein (TheJohn Hopkins Onclogy Center, Baltimore, MD). The MRC-5 humannormal fibroblasts and CMV strain AD169varATCC were purchasedfrom the American Type Culture Collection. CMV-infected cells werefixed at 24 h after infection for immunofluorescence analysis. Cells werecultured in Dulbecco’s modified Eagle’s medium and RPMI 1640 (PAAlaboratories), respectively, supplemented with 10% fetal calf serum, 1mM sodium pyruvate, 10 mM HEPES, and 100 units/ml penicillin/streptomycin at 37 °C under 5% CO2. Transient transfections of HeLacells, MEFs, and HCT116 cells with vectors for Myc-tagged vMIA or itsdeletion mutants (32), a Bax-GFP fusion gene (kindly provided by Dr.Shigemi Matsuyama (33)), pEGFP only and pEGFP-Bid (Clontech),vMIA mutants (32), or an empty vector, pcDNA3.1, were performed

using the LipofectAMINE 2000 reagent (Invitrogen).Immunofluorescence and Confocal Microscopy—Cells grown on cov-

erslips were fixed with paraformaldehyde (4% w/v) and permeabilizedwith 0.1% SDS in phosphate-buffered saline (34, 35). Cells were thenstained for the detection of cytochrome c (monoclonal antibody 6H2.B4from Pharmingen) and the c-Myc epitope of tagged vMIA (monoclonalantibody 9E10, Santa Cruz Biotechnology). To detect subcellular local-ization of Bax, cells were stained with either monoclonal anti-Bax (6A7,Pharmingen), rabbit polyclonal anti-Bax antiserum (N20, Santa CruzBiotechnology), or rabbit polyclonal vMIA (29). Epifluorescence analy-ses were performed using a Leica microscope (DMIRE2), and confocalanalyses were performed using an LSM 510 Zeiss microscope equippedwith a �63 objective (oil immersion).

Isolation of Mitochondria—Cells resuspended in a homogenizationbuffer (300 mM sucrose, 10 mM TES, 300 �M EGTA) supplemented witha mixture of protease inhibitors (Roche Applied Science) were disruptedby cavitation (PAR system; 150 p.s.i., 2 � 15 min) and centrifuged at900 � g for 15 min twice to remove nuclei and unbroken cells. Thesupernatant was subsequently centrifuged at 9000 � g for 15 min, andthe pellet was used as a fraction enriched in mitochondria (“mitochon-drial fraction”). The supernatant was further centrifuged at 100,000 �g for 1 h to yield the cytosolic cell fraction.

Preparation of Mitochondrial Extracts—The mitochondrial fractionwas resuspended in the homogenization buffer supplemented with 2%CHAPS (Sigma), incubated on ice for 1 h, sonicated, and centrifuged at100,000 � g for 30 min. The supernatant (“soluble mitochondrial frac-tion”) was then recovered. Protein concentrations were determinedusing the Bio-Rad DC protein assay.

Detection of Tight Association of Bax with Mitochondrial Mem-branes—The mitochondrial fraction was resuspended in 100 mM

Na2CO3, pH 12, to a final protein concentration of 1 mg/ml and incu-bated on ice for 20 min. Subsequently, the sample was centrifuged at100,000 � g for 1 h. The supernatant, which contained loosely attachedproteins, was recovered, and CHAPS was added to a final concentrationof 2%. The pellet, which contained tightly membrane-associated pro-teins, was suspended in the homogenization buffer containing 2%CHAPS, incubated on ice for 1 h, sonicated, and centrifuged at100,000 � g for 30 min.

Gel Filtration Analysis—Gel filtrations were performed at 4 °C on aSuperdex 200 column (16/60 and 10/30, Amersham Biosciences), asdescribed before (10). Briefly, the column was equilibrated in 25 mM

HEPES-NaOH, 300 mM NaCl, 200 �M dithiothreitol, 2% (w/v) CHAPS,pH 7.5, and run at a flow rate of 1 ml/min. The column was calibratedwith gel filtration standard proteins from Amersham Biosciences. Frac-tions of 2 ml were collected. The proteins present in these fractions wereprecipitated with trichloroacetic acid (10%) in water and subjected toimmunoblot analysis.

Cell-free Assay of Bax Incorporation into Mitochondria—The associ-ation of Bax or vMIA with purified rat liver mitochondria was deter-mined as described (36). Briefly, proteins labeled with [35S]Met (Amer-sham Biosciences) were synthesized from cDNA using the TNT-coupledtranscription/translation system (Promega). Labeled proteins were in-cubated with mitochondrial fractions at 37 °C for 1 h in homogenizationbuffer. Radiolabeled proteins bound to the mitochondria were recoveredin the pellet after centrifugation of the incubation mixture for 10 min at4 °C at 8000 � g. [35S]Met-labeled vMIA associated with isolated mito-chondria was subjected to SDS-PAGE followed by quantitation of theradioactivity in a PhosphorImager (Amersham Biosciences) using theIPLab program (Signal Analytics, Vienna, VA).

Cell-free Association of vMIA Bax �2–37 and Bax �2–20—His-taggedBax (4 fmol) was incubated overnight with 10 �l of nickel-nitrilotriace-tic acid agarose beads (Qiagen) at 4 °C. The reaction mix was centri-fuged at 15,000 � g for 10 min at 4 °C, and the pellet was resuspendedin 10 �l of LB buffer (8 M urea, 100 mM NaP04, 10 mM Tris-Cl, pH 8) inthe presence of in vitro translated [35S]Met-labeled vMIA (4 fmol) andthe incubated for 2 h at 37 °C, and the beads were then washed in LBbuffer. Proteins bound to the beads were eluted from the beads with 250�M imidazole (20 min, 4 °C) and subjected to SDS-PAGE andautoradiography.

Immunoprecipitation—BJAB cell clones stably transfected withvMIA, a vMIA deletion mutant, or an empty vector were used in theseexperiments. Cells were lysed in lysis buffer (150 mM NaCl, 50 mM TrisHCl, pH 7.4, 5 mM EDTA, supplemented with either 1% Triton X-100 or2% CHAPS) in the presence of protease inhibitors for 30 min on ice andthen centrifuged (5 min at 1000 � g) to remove nuclei and cell debris,and the supernatants were rotated overnight at 4 °C with the 9E10anti-Myc antibody covalently bound to Affi-prep10 beads (29). Thebeads were then washed in lysis buffer, and proteins were eluted from

vMIA and Bax22606

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from

beads in non-reducing NuPAGE (Invitrogen) sample buffer and thenseparated by SDS-PAGE (Invitrogen) under reducing conditions. Celllysates were also run as controls. The gels were then analyzed by astandard immunoblot protocol using anti-Bax antibody (N20, SantaCruz Biotechnology) and the ECL (enhanced chemiluminescence) de-tection system (Amersham Biosciences).

RESULTS AND DISCUSSION

Bax Is Predominantly Localized at Mitochondria of Non-apoptotic vMIA Cells—We examined the localization of Bax intwo clones (clones 3 and 8) of HeLa cells stably transfected withvMIA by immunofluorescence with the anti-Bax N20 antibody.A representative experiment with HeLa/vMIA#3 and the con-trol HeLa/pcDNA3 (Neo) cells stably transfected with theempty vector is shown in Fig. 1A. In control HeLa/pcDNA3cells, Bax was diffusely distributed and did not co-localize withcytochrome c, consistent with cytoplasmic localization of Bax.In the HeLa clones constitutively expressing vMIA, Bax largelyco-localized with both cytochrome c (Fig. 1A) and vMIA (not

shown), two proteins that are localized at mitochondria (20, 29,32, 37). The finding that Bax was predominantly localized atmitochondria in HeLa/vMIA cells was unexpected since reloca-tion of Bax to mitochondria is considered to constitute an earlystep in apoptosis (38) followed by MMP. Instead, vMIA-ex-pressing HeLa cells are resistant to apoptosis due to the abilityof vMIA to prevent MMP (29).

To find out whether vMIA-induced recruitment of Bax tomitochondria is limited to HeLa cells or is a general phenom-enon, we examined localization of Bax in MEF transientlytransfected with vMIA/pcDNA3 (or with pcDNA3 as a control)and in MRC5 normal human fibroblasts infected with theAD169varATCC strain of human cytomegalovirus. We ob-served in both experimental systems that ectopic expression ofvMIA induces relocation of Bax to mitochondria (Fig. 1, B andC), whereas in control MEFs (Fig. 1B) and non-infected MRC5fibroblasts (data not shown), Bax is diffusely distributed.

FIG. 1. Bax constitutively localizes at mitochondria in non-apoptotic vMIA-expressing cells. A, HeLa/vMIA#3 cells stably transfectedwith Myc-tagged vMIA and HeLa/pcDNA3 cells (Neo) stably transfected with the empty vector were stained with an antibody against Bax (redfluorescence) and an antibody against cytochrome c (cyt.c) (green fluorescence) and analyzed by confocal microscopy. Yellow color in the overlay ofthese two images indicates co-localization of Bax and cytochrome c (presumably in mitochondria). Another HeLa/vMIA clone 8 was similarlyanalyzed by confocal microscopy and showed the same pattern of staining. B, Bax colocalizes with vMIA in transiently transfected MEFs. Cellswere transiently transfected with vMIAt or an empty vector (wild type (WT)), stained with an antibody against Bax (red fluorescence) and anantibody against vMIA (green fluorescence), and analyzed by confocal microscopy. The patterns of intracellular distribution of vMIA and of Bax andtheir colocalization is representative of the whole transfected population. Efficiency of transfection varies between 35 and 45% in two independentexperiments. C, co-localization of Bax and vMIA in MRC5 normal human fibroblasts infected with human cytomegalovirus (AD169varATCC) 24 hafter infection. Cells were permeabilized and stained with a polyclonal anti-vMIA antibody. No immunofluorescent staining was observed withpreimmune serum (data not shown). D, vMIA expression correlates with the mitochondrial localization of Bax. Transiently transfected HeLa cellswere collected at 2, 4, 6, and 17 h after transfection and fixed for the staining. At the indicated times, the percentage of cells showing a positivestaining for vMIA narrowed 10, 20, 55, and 60%, respectively. Of note, we observed a clear correlation between the vMIA mitochondrial expressionin individual cells and the mitochondrial localization of Bax. The patterns of intracellular distribution of vMIA and of Bax, and their colocalization,are representative of the whole transfected population.

vMIA and Bax 22607

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from

To study the kinetics of vMIA expression and Bax relocationto mitochondria, we examined the time course of both vMIAand Bax staining in transiently transfected HeLa cells. Asshown in Fig. 1D, 2 h after transfection in the absence of vMIAprotein, Bax shows a diffuse distribution in the cytoplasm,whereas from 4 h after transfection to 17 h, we noted that evenlow expression level of vMIA (at 4 h) was sufficient to triggersignificant relocation of Bax to mitochondria.

Next, we determined whether exposure of vMIA-expressingcells to various apoptotic stimuli induced apoptosis. HeLa/vMIA#3, were exposed to staurosporin (STS) and analyzed byimmunofluorescence confocal microscopy for the intracellulardistribution of Bax and cytochrome c. Representative images ofthe distribution of Bax (Fig. 2A, red fluorescence) and cyto-chrome c (Fig. 2A, green fluorescence) in HeLa/vMIA#3 cells(vMIA) and HeLa/pcDNA3 cells (Neo) that have or have notbeen exposed to STS are shown in Fig. 2A. In non-treated Neocells, Bax staining is diffuse and is not colocalized with cyto-chrome c. Following the exposure of Neo cells to STS, Baxdistribution becomes punctate and co-localized with cyto-chrome c (Fig. 2A, yellow) in mitochondria, which induces cy-tochrome c release (Fig. 2A, inset). In HeLa/vMIA#3 cells, Baxis constitutively co-localized with cytochrome c in granularstructures irrespective of whether the cells were or were notexposed to STS (Fig. 2A). Cells were then exposed to STS,cis-platin, oligomycin � carbonyl cyanide p-trifluoromethoxy-phenylhydrazone, or anti-CD-95 � CHX. To quantitate thepercentage of cells undergoing apoptosis, cytochrome c stainingand chromatin condensation were examined either 3 or 16 hlater. In vMIA-expressing cells, both cytochrome c release andchromatin condensation were blocked (Fig. 2B). Anti-apoptoticeffects of vMIA were subsequently tested in transiently trans-fected cells (vMIAt) in comparison with vMIA#3. As shown inFig. 2, C and D, the transient expression of vMIA was effectivein inhibiting both cytochrome c release and chromatin conden-sation induced by CD-95 ligation.

The subcellular distribution of Bax in HeLa/Neo and HeLa/vMIA that had or had not been exposed to STS was alsodetermined by cell fractionation. As shown in Fig. 2E, exposureof HeLa/Neo cells to STS resulted in redistribution of Bax fromthe cytosol to the mitochondria/heavy membrane-enriched frac-tion and release of cytochrome c, in accord with previous ob-servations (10). In contrast, both Bax and cytochrome c pre-dominantly associated with the mitochondria-enriched fractionof vMIA cells that had or had not been treated with STS. Oneimportant observation made in these experiments was that invMIA-expressing cells, Bax is constitutively localized to mito-chondria, but in response to an apoptotic stimulus, MMP is nottriggered, and the cells do not undergo apoptosis.

Then, we asked whether transient expression of ectopic Baxwould behave as endogenous Bax. We transiently transfectedHeLa/vMIA cells (clone 3) and control HeLa/Neo cells withhuman Bax fused in its C terminus to the green fluorescentprotein (GFP) (33). To avoid the spontaneous translocation ofBax-GFP to mitochondria, we titrated the Bax-GFP plasmid.Twenty-four hours after transfection of a low concentration ofBax-GFP plasmid, in HeLa Neo cells, Bax-GFP shows a diffusedistribution, whereas in vMIA cells, the pattern colocalizedwith mitochondria without cytochrome c release in both cases(Fig. 3A). Then, to test the ability of Bax-GFP to induce cyto-chrome c release, we added STS (1 �M) to some of the trans-fected cultures, further incubated cells for 3 h in the presenceof 50 �M Z-VAD-fmk (to delay cell death), and examined theintracellular distribution of Bax-GFP under a fluorescent mi-croscope (Fig. 3A). A large fraction of Bax-GFP was relocated tomitochondria, and in the majority of cells, cytochrome c was

released into the cytoplasm, indicating that these cells wereundergoing apoptosis (Fig. 3B). In vMIA-expressing cells, bothBax-GFP and cytochrome c were localized at mitochondria,irrespective of whether the cells had or had not been exposedto STS.

To further investigate whether vMIA can suppress apoptosisinduced by overexpression of Bax-GFP, we used higher concen-trations of the plasmid where Bax-GFP spontaneously relo-cated to mitochondria in 30% of HeLa/Neo cells. This phenom-enon coincided with cytochrome c release and chromatincondensation in HeLa/Neo cells, whereas in vMIA-expressingcells, cytochrome c retained its granular localization, and chro-matin remained intact (Fig. 3, C and D). These results indicatethat (i) ectopically expressed Bax translocates spontaneously tomitochondria in vMIA cells and (ii) high doses of Bax-GFP thatare sufficient to kill Neo cells were not toxic for vMIA cells.

Apoptosis Inhibition by vMIA Is Associated with Mitochon-drial Localization of Bax—Two segments of the vMIA protein(Tyr-5–Leu34 and Asp-118–Arg-147) contain domains that to-gether are essential for its anti-apoptotic function (32). Thefirst region (amino acids 2–23) is necessary and sufficient forthe mitochondrial localization of vMIA, whereas the molecularfunction of the second domain (amino acids 118–147) has notbeen elucidated prior to this study. A “mini-vMIA” consisting ofthese two segments was shown to be a functional cell deathsuppressor (32). We have tested the ability of functionallydeficient vMIA deletion mutants to induce relocation of Bax tomitochondria (Fig. 4A). These experiments were done in tran-sient transfection assays with HeLa cells. Apoptosis was de-tected by cytochrome c release and chromatin condensation.First, we confirmed that vMIA and mini-vMIA protected tran-siently transfected HeLa cells against STS-induced apoptosis,whereas three mutants with a deletion either within the mito-chondria-targeting domain (�2–23) or within the second anti-apoptotic domain (�115–130 and �131–147) were inefficient inpreventing apoptosis (Fig. 4B). In accord with the literature(10), we found that a certain amount of endogenous Bax inHeLa cells is normally associated to mitochondria (Figs. 2E and4C). Of note, this result is not in contradiction with immuno-fluorescence analysis (Fig. 1A). This only reflects the differencein the sensitivity of immunofluorescence versus subcellularfractionation in detecting low levels of Bax. Only the functionalcell death suppressors, full-length vMIA and mini-vMIA, in-duced a significant relocation of Bax to mitochondria, whereasfunctionally inactive vMIA �2–23, vMIA �115–130, and vMIA�131–147 failed to trigger the relocation of Bax, as shown bothin immunofluorescence microscopic observations (Fig. 4A) andby immunoblotting analysis of cellular fractions (Fig. 4C).These experiments demonstrate a strong correlation betweenthe anti-apoptotic function of vMIA and its ability to relocateBax to mitochondria.

vMIA Causes Tight Association of Oligomeric Bax with Mi-tochondria—Prior to inducing permeabilization of mitochon-dria in cells undergoing apoptosis, Bax becomes strongly asso-ciated with mitochondria, possibly integrating into the outermitochondrial membrane, and forms either homo- or hetero-oligomeric complexes (3, 8–11, 38–40). These events were notobserved in Bcl-2-expressing cells exposed to pro-apoptoticstimuli, indicating that Bcl-2 blocked apoptosis either at orupstream of this step (6, 38).

Since in vMIA-expressing cells Bax was predominantly lo-calized at mitochondria, we tested whether Bax was oligomer-ized in these cells and whether its association with mitochon-dria was tight or loose. For these experiments, we preparedmitochondria-enriched membrane fractions from HeLa orBJAB/vMIA and BJAB/pcDNA3 cells. In addition, to examine

vMIA and Bax22608

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from

whether the status of Bax oligomerization and association withmitochondria was changed during apoptosis, we also preparedmitochondria-enriched membrane fractions from BJAB/vMIA

and BJAB/pcDNA3 cells that had been exposed to STS. Toseparate loosely attached Bax (Fig. 5A, Att) from tightly asso-ciated Bax (Fig. 5A, Ins), possibly integrated into the mem-

FIG. 2. vMIA inhibits the release of cytochrome c (cyt.c) from mitochondria despite the local presence of Bax. A, HeLa/vMIA#3 cells andHeLa/pcDNA3 cells (Neo) were incubated or not with staurosporine (1 �M) for 3 h. Cells were then stained for immunodetection of Bax (red) andcytochrome c (green) and then examined by laser fluorescence confocal microscopy. Co, control. B, vMIA suppresses cytochrome c release frommitochondria and nuclear condensation in HeLa/vMIA#3 cells that were treated for 16 h with STS (250 nM), cisplatin (50 �M), or oligomycine (5 �M)plus carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) (1 �M) or anti-CD-95 (4 �g/ml) plus CHX (2 �g/ml). The percentage of cells exhibitinga diffuse cytochrome c staining was determined as in panel A. Results are representative of three independent experiments (400 cells were counted ineach sample). C, vMIA suppresses apoptosis in stably (HeLa/vMIA#3) or transiently (vMIAt) transfected HeLa cells. Cells were treated with either 3or 6 �g/ml anti-CD-95 and stained for immunodetection of cytochrome c (red) and Myc-tag of vMIA (green) and counterstained with Hoechst 33342. Thescale bar corresponds to 10 �m. D, the percentage of cytochrome c release and chromatin condensation was determined as described in the legend forpanel B. E, Bax is localized in mitochondria of HeLa/vMIA#3 cells before (Co) and after treatment with STS. Cells were subjected to subcellularfractionation, and the cytosolic as well as the mitochondrial (Mito) fractions were blotted for immunodetection of Bax and cytochrome c.

vMIA and Bax 22609

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from

brane (10, 42), these membrane fractions were treated with 0.1M Na2CO3, as described under “Materials and Methods,” andthen mitochondria-containing pellets were examined by immu-noblot analysis. As expected (Fig. 5A), Bax was only looselyattached to the mitochondrial membranes isolated fromBJAB/pcDNA3 cells prior to induction of apoptosis but becamestrongly associated with the membranes following exposure ofcells to STS. In contrast, in vMIA-expressing cells, Bax wasconstitutively tightly associated with mitochondrial mem-branes irrespective of whether the cells had or had not beenexposed to STS.

We also examined the oligomerization status of Bax in HeLaand BJAB cells stably transfected with pcDNA3 or vMIA bysize exclusion chromatography of CHAPS-solubilized mito-chondrial fractions (Fig. 5B and not shown). Bax extracted fromhealthy HeLa/pcDNA3 cells migrated as a monomer with theapparent molecular size of slightly above 20 kDa. Upon treat-ment with STS, most of the extracted Bax was found in largecomplexes with an apparent molecular mass mostly above 90kDa. In contrast, a major fraction of Bax extracted fromhealthy vMIA-expressing cells eluted as part of large com-

plexes, and this pattern of Bax migration did not change afterexposure to STS. These experiments demonstrated that vMIA-induced relocation of Bax at mitochondria was accompanied byeither homo-, or possibly, hetero-oligomerization of Bax and itstight association with mitochondria.

vMIA Interacts with Bax and Facilitates Its Tight Associationwith Mitochondria—To test whether Bax physically associateswith vMIA, we immunoprecipitated vMIA with anti-Myc anti-body from CHAPS or Triton X-100 lysates of BJAB or HeLacells stably transfected either with Myc-tagged vMIA or withthe empty vector and then examined these immunoprecipitatesfor the presence of Bax by immunoblot analysis (Fig. 6A). Theseexperiments detected Bax in either CHAPS or Triton X-100immunoprecipitates from HeLa or BJAB/vMIA cell lysates butnot from control HeLa or BJAB/pcDNA3 cell lysates (Fig. 6A).This indicated that, indeed, Bax formed a complex with vMIAdetected with either of these two detergents. Furthermore,when we immunoprecipitated vMIA deletion mutants fromTriton X-100 lysates of BJAB cells, only wild type vMIA andfunctionally active mini-vMIA (�35–112/�148–163 deletionmutant) formed a complex with Bax. The ability of mini-vMIA

FIG. 3. vMIA triggers relocation of Bax-GFP to mitochondria. A, Neo or vMIA cells were transfected with a low concentration (50ng/100,000 cells) of Bax-GFP plasmid (green) that does not trigger spontaneous translocation of Bax to mitochondria in HeLa Neo cells. Cells weretreated or not treated with staurosporine (1 �M) in the presence of z-VAD-fmk (50 �M) for 3 h, stained for detection of cytochrome c (cyt.c) (red),and examined by laser fluorescence confocal microscopy. Note that Bax is either in the cytosol (in non-treated Neo cells) or in a punctate,cytoplasmic (mitochondrial) localization (in Neo cells exposed to STS). Co, control. B, quantification of cytochrome c release in transfected cells. Thepercentage of cells exhibiting a diffuse localization of cytochrome c throughout the cytoplasm was determined among the Bax-GFP-expressing cellstreated with STS, as indicated. Results are representative of three independent experiments. C, Neo or vMIA cells were transfected with pEGFPor a high concentration (400 ng/100,000 cells) of Bax-GFP plasmid (green) that triggers spontaneous translocation of Bax in HeLa Neo cells. D, thepercentage of cytochrome c release and chromatin condensation was determined on transfected cells as in panel A. Transfection efficiencies variedbetween 40 and 50% of cells. Error bars represent standard deviation of three independent experiments (400 cells were counted in each sample).

vMIA and Bax22610

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from

consisting essentially of only the two anti-apoptotic domains toassociate with Bax indicated that this vMIA-Bax associationwas mediated by either the first (mitochondria-targeting) do-main or the second anti-apoptotic domain or the two domainstogether. Then, we tested three other deletion mutants that areinactive as cell death suppressors but do localize at mitochon-dria (32), �31–163vMIA, �115–130vMIA, and �131–147vMIA.Both tested vMIA mutants with deletions in the second domainfailed to associate with Bax, demonstrating that this domainwas required for the association of vMIA with Bax (Fig. 6B).

To further characterize the putative physical interaction be-

tween Bax and vMIA, we used an in vitro translation system ofvMIA and two different forms of Bax. Bax �2–20 (also calledBax �) is a newly described splice variant of Bax lacking theN-terminal 20 amino acids. This Bax variant is constitutivelyassociated with mitochondria and highly apoptogenic (43). Bax�2–37, which lacks the mitochondrial-targeting signal con-tained in the �-1 helix (amino acids 19–37) (44), is constitu-tively cytosolic, fails to induce apoptosis, and cannot be acti-vated to relocate to mitochondria. These two forms of Bax thusrepresent the two opposites of the spectrum of the conformationadopted by wild type Bax, activated (Bax �2–20), and inactiveBax (�2–37). Both forms of Bax interact with in vitro trans-lated vMIA (Fig. 6C). Next, we analyzed the capacity of vMIAto stimulate the tight association of Bax with mitochondria. Invitro translated vMIA readily incorporated into the mitochon-drial fraction, and this was also observed for Bax �2–20 (Fig.6D, middle row). In contrast, in vitro translated Bax �2–37 didnot incorporate into the mitochondrial fraction. However, whenco-incubated with vMIA (Fig. 6D, bottom row), Bax �2–37 wasincorporated into the mitochondrial fractions as efficiently asBax �2–20 (Fig. 6D). As a result, it appears that vMIA canstimulate tight association of Bax with mitochondria.

Bax Modified by vMIA Becomes Refractory to the Action oft-Bid—vMIA does not prevent caspase-8-mediated cleavage ofBid during CD-95-mediated apoptosis of HeLa cells and actsdownstream of this event but upstream of permeabilization ofmitochondria (29). Formation of the t-BID/Bax complex, a crit-

FIG. 4. The apoptosis inhibitory effect of vMIA correlates withmitochondrial localization of Bax. A, subcellular localization ofvMIA and its impact on Bax. HeLa cells were transiently transfectedwith various vMIA mutants. Twenty-four hours after transfection, cellswere fixed, permeabilized, and stained for detection of active Bax (red)and vMIA (green) followed by laser fluorescence confocal microscopy.The scale bar indicates 10 �m. WT, wild type. B, inhibition of apoptosisby different vMIA mutants. Cells transfected as in panel A were cul-tured in the absence or presence of staurosporine (1 �M, 3 h), and thepercentage of cells exhibiting cytochrome c (Cyt.c) release, the translo-cation of Bax to mitochondria, or chromatin condensation was deter-mined in transfected cells (30–40% of total cells). 400 cells werecounted in each sample. Results are presented as means of three ex-periments. C, mitochondrial (Mito) localization of Bax in vMIA mutantscorrelates with the anti-apoptotic function of vMIA. Cells were trans-fected with vector only (Neo), wild type vMIA, or various vMIA mu-tants. Twenty-four hours later, the transfected cells were fractionatedinto the cytosol and mitochondria. These fractions were blotted forimmunodetection of Bax. Equal loading was confirmed by Coomassiestaining (not shown).

FIG. 5. vMIA causes tight association of oligomerized Bax withmitochondria. A, mitochondria were isolated from Neo- or vMIA-expressing BJAB B cells and treated with 100 mM Na2CO3 followed bycentrifugation to separate loosely attached (Att) Bax (in the superna-tant) from membrane-inserted (Ins) proteins (in the pellet). The sam-ples were mixed with lysis buffer and subjected to immunoblotting forthe detection of Bax, vMIA, and the inner mitochondrial membraneprotein COXII. B, detection of Bax monomers and Bax oligomers inmitochondria isolated from HeLa/vMIA#3 and HeLa/Neo cells beforeand after treatment with STS (250 nM, 16 h). Mitochondrial fractionswere solubilized with lysis buffer containing CHAPS and subjected tomolecular sieve gel filtration, as detailed under “Materials and Meth-ods,” and then analyzed by immunoblot for Bax detection. Arrowsindicate molecular masses from internal standards run in parallel.Co, control.

vMIA and Bax 22611

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from

ical event in triggering the mitochondrial apoptotic signalingpathway (12, 16, 45), also takes place downstream of Bid cleav-age but upstream of mitochondrial permeabilization. To deter-mine whether vMIA acts upstream or downstream of t-BID/Bax complex formation, we employed a Bid-GFP construct intransient transfection assays in HeLa cells. First, we checkedwhether the Bid-GFP fusion protein behaved similarly to na-tive Bid during CD-95-mediated apoptosis in HeLa cells. HeLa/vMIA, HeLa/Bcl-2, and control HeLa/Neo cells were transientlytransfected with Bid-GFP and exposed to anti-CD-95 in thepresence of CHX. The ligation of CD-95 caused a complete,presumably caspase-8-mediated, cleavage of Bid-GFP to gen-erate active t-Bid-GFP in all three cell lines (Fig. 7A), confirm-ing that this Bid-GFP construct was processed upstream of theaction of vMIA, i.e. in a manner similar to that of native Bid(29). Following CD-95 ligation, the GFP-dependent fluores-cence shifted from a diffuse (cytosolic) to a punctate (mitochon-drial) distribution in control (Neo) cells, whereas cytochrome cwas released from mitochondria, as indicated by its diffusepresence throughout the cell (Fig. 7B). As expected, cytochromec was retained in mitochondria of HeLa/Bcl-2 and HeLa/vMIAcells exposed to anti-CD-95 plus CHX. Again, the Bid-GFPconstruct was functioning in a manner similar to that of nativeBid and therefore represented a valid tool to examine interac-tion of t-Bid with Bax in HeLa cells undergoing apoptosis.

Next, we compared redistribution of Bid-GFP/t-Bid-GFPupon CD-95-ligation in HeLa/vMIA cells with that in HeLa/Neoand HeLa/Bcl-2 cells. In HeLa/Bcl-2 cells, the GFP-dependentfluorescence became associated with mitochondria, in a man-ner similar to that of the control HeLa/Neo cells. In contrast, invMIA-expressing cells, the GFP-dependent fluorescence asso-ciated with mitochondria was reduced (Fig. 7, B and C). Theseexperiments could be interpreted as follows: during apoptosis,in HeLa/Neo cells, t-Bid triggers Bax activation which inducesMMP, whereas in HeLa/Bcl-2 cells, t-Bid is presumably seques-tered by Bcl-2, which prevents its interaction with Bax. InHeLa/vMIA cells, Bax is sequestered by vMIA into an inactivecomplex at mitochondria, preventing its activation by t-Bid.

vMIA Inhibits CD-95-induced Apoptosis as Efficiently as BaxKnock-out Cells—We examined the anti-apoptotic activity ofvMIA in response to CD-95-ligation in two clones of the humancarcinoma cell line HCT116, a Bax-expressing clone HCT116/Bax�/� and a Bax-deficient clone HCT116/Bax�/�. HCT116/Bax�/� and HCT116/Bax�/� cells were transiently transfectedwith vMIA, and 24 h later, treated or not treated with anti-CD-95 � CHX for 5 h. Transfected cells (400 cells/sample) werethen scored under a confocal microscope for cytochrome c re-lease from mitochondria and chromatin condensation (Fig. 8).Ligation of CD-95 induced extensive cytochrome c release andchromatin condensation in HCT116 wild type cells but only amarginal, although detectable, cytochrome c release inHCT116/Bax�/� cells. This result was in accord with the pre-viously published reports that mitochondria-mediated apopto-sis is controlled by Bax in these cells (46–48). In HCT116/Bax�/� cells transiently transfected with vMIA (vMIAt),cytochrome c release was inhibited to a level comparable withthat induced by the knock-out of Bax. This result indicated thatvMIA inhibited the Bax-mediated apoptotic pathway inducedby CD-95 ligation.

Concluding Remarks—A major finding of our study is thatthe viral inhibitor of apoptosis vMIA neutralizes Bax by re-cruiting it to mitochondria and “freezing” its pro-apoptotic ac-tivity. We showed that the triggering of Bax relocation to mi-tochondria by vMIA is a general phenomenon, observed indifferent human and mouse cell lines and in human fibroblastsinfected with human CMV.

FIG. 6. vMIA associates with Bax. Co-immunoprecipitation ofvMIA and Bax is shown. A, HeLa and BJAB cells stably expressing ornot vMIA were lysed in lysis buffer containing either Triton X-100(TX 100) or CHAPS as described under “Materials and Methods” andsubjected to immunoprecipitation with anti-Myc antibody followed bySDS-PAGE and immunoblot detection of Bax. Samples of cell lysatesprior to immunoprecipitation are labeled as L, and samples of immu-noprecipitated vMIA are labeled as IP. Note that Bax co-immunopre-cipitated with vMIA equally well from CHAPS- and Triton X-100lysates. B, BJAB cells stably transfected with pcDNA3, Myc-taggedfull-length vMIA, or various deletion constructs of vMIA lysed withTriton X-100-containing lysis buffer and immunoprecipitated withanti-Myc antibody followed by SDS-PAGE and immunoblot detectionof Bax. Note that Bax co-immunoprecipitated only with full-lengthvMIA and mini-vMIA. As for full-length vMIA, mini-vMIA coimmu-noprecipitation with Bax was confirmed in CHAPS-containing lysisbuffer. WT, wild type. C, direct interaction between vMIA and Bax�2–20 or Bax �2–37. Four fmol of in vitro translated vMIA wasincubated with 4 fmol of His-tagged Bax �2–20 or Bax �2–37 proteinimmobilized on nickel-nitrilotriacetic acid agarose beads or the aga-rose beads alone (control, Co). The input of in vitro translated vMIA(25% of the amount added to the beads) and the amount of 35S-labeledproteins bound to the beads was determined by PhosphorImager afterSDS-PAGE. Similar data were obtained in three independent exper-iments. D, mitochondrial relocation of in vitro translated Bax proteinin the absence or presence of vMIA. The indicated Bax mutants weregenerated in an in vitro transcription/translation system, and 25% ofsamples were run as input controls. The remaining samples weresubjected to a mitochondrial association assay, either alone (single) orafter co-incubation with vMIA (co-inc.), as indicated in the panel. Theresults shown in panel D are representative for three independentassays. Note that mitochondrial association of the Bax �2–20 mutantis not influenced by vMIA, whereas that of Bax �2–37 strictly de-pends on vMIA. Bax �2–37, which lacks the mitochondrial-targetingsignal contained in the �-1 helix (amino acids 19–37) (44), fails toassociate with mitochondria and thus serves as internal control forthe specificity of association with mitochondria.

vMIA and Bax22612

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from

Our data support a correlation between the anti-apoptoticfunction of vMIA and its ability to relocate/interact with Bax atmitochondria. Indeed, among different deletion mutants ofvMIA, only mini-vMIA (a truncated protein consisting of thetwo segments of vMIA that are required for mitochondriallocalization and anti-apoptotic function) maintained the abilityto relocate/interact with Bax at mitochondria. The two vMIA

mutants with deletions in the C terminus domain of vMIA(�115–130 and �131–147) showed mitochondrial localizationbut were ineffective in preventing apoptosis and failed to relo-cate/associate with Bax. These results suggest that the C ter-minus domain of vMIA-(115–147) was required for cell deathsuppressing by favoring the association of vMIA with Bax.

Previously published experiments with recombinant Baxprotein demonstrated that this protein is capable of forminghigher order structures with membrane-permeabilizing prop-erties and thus destabilizes lipid bilayers (49), induces theformation of non-specific ion channels in synthetic lipid bilay-ers, and forms cytochrome c-permeable conduits in liposomes(50). These effects produced by recombinant Bax protein coin-cided with its oligomerization, and it has been tacitly concludedfrom these studies that relocation of Bax to mitochondria ac-companied by its oligomerization within the mitochondrialmembrane is sufficient to mediate the MMP reaction (3, 12, 16,45, 51). The data reported in this study provide a seeminglycontradicting example. vMIA stimulates relocation of Bax tomitochondria, its oligomerization, and tight association withmitochondria, possibly due to its integration into the mitochon-drial outer membrane. Although this status of Bax normallycorrelates with apoptotic MMP, MMP is blocked in vMIA-ex-pressing cells. This is reminiscent of the description of thetranslocation and formation of dimer/complexes of Bax at mi-tochondria without MMP, occurring in a taxol-resistant cellline (41).

How vMIA recruits Bax to mitochondria is an ongoing co-nundrum. Newly translated vMIA could act as a co-chaperonefor Bax during its relocation to mitochondria. Alternatively,vMIA, which is already present on mitochondria, could recruitcytosolic Bax. How can we explain, in speculative terms, thatBax can oligomerize without spontaneously triggering MMP? Itis possible that vMIA keeps Bax in an inactive state. In favor ofthis hypothesis, we found that the recruitment of t-Bid intomitochondria is delayed in vMIA-expressing cells. However, itis also quite possible that oligomerization of Bax itself is notsufficient to create the pores responsible for MMP and thatinteractions with additional mitochondrial proteins are re-quired (16). At present, it is not clear whether Bax oligomersformed in the presence of vMIA have the same characteristics

FIG. 7. vMIA delays translocation of t-Bid to mitochondria. A,cell extracts were prepared from Bid-GFP transfected cells, treated ornot with anti-CD-95 � CHX, and subjected to anti-GFP immunoblot-ting. CD-95 ligation causes similar levels of Bid activation in differentcells. B, vMIA-mediated inhibition of t-Bid relocation to mitochondria.Cells treated as in panel A were examined by fluorescence microscopy todetect cytochrome c (cyt.c) (red) and the accumulation of the Bid/t-Bid-GFP fusion protein (green) in mitochondria. C, quantitation of t-Bidtranslocation. The percentage of cells exhibiting a clearly punctatepattern of GFP fluorescence among the transfected population wasdetermined as in panel B. The percentage of apoptosis was determinedby calculating the number of cells exhibiting a diffuse pattern of cyto-chrome c. Results are means of three different experiments.

FIG. 8. vMIA suppresses apoptosis as efficiently as Bax�/�

cells. HCT116/Bax�/� cells and HCT116/Bax�/� cells were transientlytransfected with vMIA, and 24 h later, the cells were treated withanti-CD-95 � CHX for 6 h and fixed for cytochrome c (Cyt.c) and nuclearstaining. Error bars represent standard deviation of three independentexperiments (400 cells were counted in each sample). WT, wild type.

vMIA and Bax 22613

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from

as those formed after Bax activation during apoptosis. Addi-tional studies are clearly needed to answer this question. More-over, since vMIA has no sequence homology with other anti-apoptotic members of the Bcl-2 family, we will address in thefuture whether vMIA specifically targets Bax or may interactwith other pro-apoptotic family members.

vMIA is an obligatory virulence factor for human cytomega-lovirus and cells infected by cytomegalovirus, in particular,connective tissues that are abundant in Bax. Based on this, onecan speculate that inhibitors of the Bax-vMIA interactionwould constitute a new kind of treatment of cytomegalovirusinfection, especially in immunocompromised individuals.

Acknowledgment—We thank Dr. Shigemi Matsuyama (Blood Re-search Institute, Milwaukee WI) for the Bax-GFP plasmid construct.

REFERENCES

1. Krammer, P. H. (2000) Nature 407, 789–7952. Li, H., Zhu, H., Xu, C., and Yuan, J. (1998) Cell 94, 491–5013. Letai, A., Bassik, M., Walensky, L., Sorcinelli, M., Weiler, S., and Korsmeyer,

S. (2002) Cancer Cell 2, 1834. Kroemer, G., and Reed, J. C. (2000) Nat. Med. 6, 513–5195. Vander Heiden, M. G., and Thompson, C. B. (1999) Nat. Cell Biol. 1,

E209-E2166. Cheng, E. H.-Y. A., Wei, M. C., Weiler, S., Flavell, R. A., Mak, T. W., Lindsten,

T., and Korsmeyer, S. J. (2001) Mol. Cell 8, 705–7117. Knudson, C. M., Tung, K. S., Tourtellotte, W. G., Brown, G. A., and Korsmeyer,

S. J. (1995) Science 270, 96–998. Nechushtan, A., Smith, C. L., Hsu, Y. T., and Youle, R. J. (1999) EMBO J. 18,

2330–23419. Suzuki, M., Youle, R. J., and Tjandra, N. (2000) Cell 103, 645–654

10. Antonsson, B., Montessuit, S., Sanchez, B., and Martinou, J. C. (2001) J. Biol.Chem. 276, 11615–11623

11. Gross, A., McDonnell, J. M., and Korsmeyer, S. J. (1999) Genes Dev. 13,1899–1911

12. Kuwana, T., Mackey, M. R., Perkins, G. A., Ellisman, M. H., Latterich, M.,Schneiter, R., Green, D. R., and Newmeyer, D. D. (2002) Cell 111, 1–12

13. Shimizu, S., Narita, M., and Tsujimoto, Y. (1999) Nature 399, 483–48714. Marzo, I., Brenner, C., Zamzami, N., Jurgensmeier, J., Susin, S. A., Vieira,

H. L. A., Prevost, M.-C., Xie, Z., Matsuyama, S., Reed, J. C., and Kroemer,G. (1998) Science 281, 2027–2031

15. Martinou, J.-C., and Green, D. R. (2001) Nat. Rev. Mol. Cell. Biol. 2, 63–6716. Roucou, X., Montessuit, S., Antonsson, B., and Martinou, J. C. (2002) Biochem.

J. 368, 915–92117. Zamzami, N., and Kroemer, G. (2003) Curr. Biol. 13, R71-R7318. Gross, A., Jockel, J., Wei, M. C., and Korsmeyer, S. J. (1998) EMBO J. 17,

3878–388519. Tschopp, J., Thome, M., Hofmann, K., and Meinl, E. (1998) Curr. Opin. Genet.

Dev. 8, 82–8720. McCormick, A. L., Smith, V. L., Chow, D., and Mocarski, E. S. (2003) J. Virol.

77, 631–64121. Zhou, Q., Snipas, S., Orth, K., Muzio, M., Dixit, V. M., and Salvesen, G. S.

(1997) J. Biol. Chem. 272, 7797–780022. Messud-Petit, F., Gelfi, J., Delverdier, M., Amardeilh, M., Py, R., Sutter, G.,

and Bertagnoli, S. (1998) J. Virol. 72, 7830–783923. Huang, Q., Petros, A., Virgin, H., Fesik, S., and Olejniczak, E. (2002) Proc.

Natl. Acad. Sci. U. S. A. 99, 3428–3433

24. Huang, Q., Petros, A., Virgin, H., Fesik, S., and Olejniczak, E. (2003) J. Mol.Biol. 332, 1123–1130

25. Sundararajan, R., Cuconati, A., D., N., and E., W. (2001) J. Biol. Chem. 276,45120–45127

26. Sundararajan, R., and E., W. (2001) J. Virol. 75, 7506–751627. Everett, H., Barry, M., Sun, X., Lee, S. F., Frantz, C., Berthiaume, L. G.,

McFadden, G., and Bleackley, R. C. (2002) J. Exp. Med. 196, 1127–113928. McCormick, A., Skaletskaya, A., Barry, P., Mocarski, E., and Goldmacher, V.

(2003) Virology 316, 221–23329. Goldmacher, V. S., Bartle, L. M., Skletskaya, S., Dionne, C. A., Kedersha,

N. L., Vater, C. A., Han, J. W., Lutz, R. J., Watanabe, S., McFarland,E. D. C., Kieff, E. D., Mocarski, E. S., and Chittenden, T. (1999) Proc. Natl.Acad. Sci. U. S. A. 96, 12536–12541

30. Vieira, H. L., Belzacq, A.-S., Haouzi, D., Bernassola, F., Cohen, I., Jacotot, E.,Ferri, K. F., Hamel, E. H., Bartle, L. M., Melino, G., Brenner, C., Goldma-cher, V., and Kroemer, G. (2001) Oncogene 20, 4305–4316

31. Belzacq, A. S., El Hamel, C., Vieira, H. L. A., Cohen, I., Haouzi, D., Metivier,D., Marchetti, P., Goldmacher, V., Brenner, C., and Kroemer, G. (2001)Oncogene 20, 7579–7587

32. Hayajneh, W. A., Colberg-Oley, A. M., Skaleskaya, A., Bartle, L. M., Lesper-ance, M. M., Contopoulos-Ionnidis, D. G., Kedersha, N. L., and Goldmacher,V. S. (2001) Virology 279, 233–240

33. Sawada, M., Hayes, P., and Matsuyama, S. (2003) Nat. Cell Biol. 4, 352–35734. Castedo, M., Ferri, K. F., Blanco, J., Roumier, T., Larochette, N., Barretina, J.,

Amendola, A., Nardacci, R., Metivier, D., Este, J. A., Piacentini, M., andKroemer, G. (2001) J. Exp. Med. 194, 1097–1110

35. Castedo, M., Ferri, K., Roumier, T., Metivier, D., Zamzami, N., and Kroemer,G. (2002) J. Immunol. Methods 265, 39–47

36. Cartron, P. F., Moreau, C., Oliver, E., Maya, C., Meflah, K., and Vallette, F. M.(2002) FEBS Lett. 512, 95–100

37. Daugas, E., Susin, S. A., Zamzami, N., Ferri, K., Irinopoulos, T., Larochette,N., Prevost, M. C., Leber, B., Andrews, D., Penninger, J., and Kroemer, G.(2000) FASEB J. 14, 729–739

38. Wolter, K. G., Hsu, Y.-T., Smith, C. L., Nechushtan, A., Xi, X. G., and Youle,R. J. (1997) J. Cell Biol. 139, 1281–1292

39. Goping, I. S., Gross, A., Lavoie, J. N., Nguyen, M., Jemmerson, R., Roth, K.,Korsmeyer, S. J., and Shore, G. C. (1998) J. Cell Biol. 143, 207–215

40. Wei, M. C., Lindsten, T., Mootha, V. K., Weiler, S., Gross, A., Ashiya, M.,Thompson, C. B., and Korsmeyer, S. J. (2000) Genes Dev. 14, 2060–2071

41. Makin, G. W., Corfe, B. M., Griffiths, G. J., Thistlethwaite, A., Hickman, J. A.,and Dive, C. (2001) EMBO J. 20, 6306–6315

42. Eskes, R., Desagher, S., Antonsson, B., and Martinou, J. C. (2000) Mol. Cell.Biol. 20, 929–935

43. Cartron, P. F., Oliver, L., Martin, S., Moreau, C., LeCabellec, M. T., Jezequel,P., Meflah, K., and Vallette, F. (2002) Hum. Mol. Genet. 11, 675–687

44. Cartron, P. F., Priaul, M., Olive, L., Meflah, K., Manon, S., and Vallette, F. M.(2003) J. Biol. Chem. 278, 11633–11641

45. Roucou, X., Rostovtseva, T., Montessuit, S., Martinou, J. C., and Antonsson, B.(2002) Biochem. J. 363, 547–552

46. Heidi, L., David, L., Eugene, V., Klara, T., John, M., Peter, S., Sharon, F.,Ralph, S., Dominick, S., and Avi, A. (2002) Nat. Med. 8, 274–281

47. Gillissen, B., Essmann, F., Graupner, V., Starck, L., Radetzki, S., Dorken, B.,Schulze-Osthoff, K., and Daniel, P. (2003) EMBO J. 22, 3580–3590

48. Yamaguchi, H., Bhalla, K., and Wang, H. (2003) Cancer Res. 63, 1483–148949. Basanez, G., Neshushtan, A., Drozhinin, O., A., C., Choe, E., Tuti, S., Wood,

K. A., Jsi, Y.-T., Zimmerberg, J., and Youle, R. J. (1999) Proc. Natl. Acad.Sci. U. S. A. 96, 5492–5497

50. Saito, M., Korsmeyer, S. J., and Schlesinger, P. H. (2000) Nat. Cell Biol. 2,553–555

51. Nechushtan, A., Smith, C. L., Lamensdorf, I., Yoon, S. H., and Youle, R. J.(2001) J. Cell Biol. 153, 1265–1276

vMIA and Bax22614

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from

Guido Kroemer and Naoufal ZamzamiAnna Skaletskaya, David Boutolleau, Jean-Claude Martinou, Victor S. Goldmacher,Jalil, Pierre-Francois Cartron, Francois Vallette, Céline Schnebelen, Laura M. Bartle,

Delphine Poncet, Nathanael Larochette, Anne-Laure Pauleau, Patricia Boya, Abdel-AliAn Anti-apoptotic Viral Protein That Recruits Bax to Mitochondria

doi: 10.1074/jbc.M308408200 originally published online March 5, 20042004, 279:22605-22614.J. Biol. Chem. 

  10.1074/jbc.M308408200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/279/21/22605.full.html#ref-list-1

This article cites 51 references, 24 of which can be accessed free at

by guest on January 30, 2020http://w

ww

.jbc.org/D

ownloaded from