GD3 ganglioside directly targets mitochondria in abcl-2-controlled fashion
MARIA RITA RIPPO,* FLORENCE MALISAN,* LUIGI RAVAGNAN,† BARBARA TOMASSINI,*IVANO CONDO,*,1 PAOLA COSTANTINI,† SANTOS A. SUSIN,† ALESSANDRA RUFINI,*MATILDE TODARO,‡ GUIDO KROEMER,† AND ROBERTO TESTI*,1
*Laboratory of Signal Transduction, Department of Experimental Medicine and BiochemicalSciences, University of Rome ‘Tor Vergata’, 00133 Rome; †Centre National de la RechercheScientifique, UMR 1599, Institut Gustave Roussy, F94805 Villejuif, France; and ‡Section ofAnatomical Sciences, University of Palermo, Italy
ABSTRACT Lipid and glycolipid diffusible media-tors are involved in the intracellular progression andamplification of apoptotic signals. GD3 gangliosideis rapidly synthesized from accumulated ceramideafter the clustering of death-inducing receptors andtriggers apoptosis. Here we show that GD3 inducesdissipation of DCm and swelling of isolated mito-chondria, which results in the mitochondrial releaseof cytochrome c, apoptosis inducing factor, andcaspase 9. Soluble factors released from GD3-treated mitochondria are sufficient to trigger DNAfragmentation in isolated nuclei. All these effects canbe blocked by cyclosporin A, suggesting that GD3 isacting at the level of the permeability transition porecomplex. We found that endogenous GD3 accumu-lates within mitochondria of cells undergoing apo-ptosis after ceramide exposure. Accordingly, sup-pression of GD3 synthase (ST8) expression in intactcells substantially prevents ceramide-induced DCm
dissipation, indicating that endogenously synthesizedGD3 induces mitochondrial changes in vivo. Finally,enforced expression of bcl-2 significantly preventsGD3-induced mitochondrial changes, caspase 9 acti-vation, and apoptosis. These results show that mito-chondria are a key destination for apoptogenic GD3ganglioside along the lipid pathway to programmedcell death and indicate that relevant GD3 targets areunder bcl-2 control.—Rippo, M. R., Malisan, F.,Ravagnan, L., Tomassini, B., Condo, I., Costantini,P., Susin, S. A., Rufini, A., Todaro, M., Kroemer, G.,Testi, R. GD3 ganglioside directly targets mitochon-dria in a bcl-2-controlled fashion. FASEB J. 14,2047–2054 (2000)
Key Words: apoptosis z permeability transition z AIF z cyto-chrome c z caspase 9
Unbearable stress due to internal disfunctions orexternal stimuli triggers programmed cell death byapoptosis. Whether the initiation of the apoptoticprogram is inside the cell or at the cell surface,signals are generated and rapidly spread to multiple
cellular compartments. Long-distance signal propa-gation is largely mediated by diffusible hydrolyticproducts of either protein substrates or phospholip-ids, mostly generated by the action of caspases andphospholipases.
Ceramide is a diffusible lipid mediator that isgenerated early during apoptosis from the hydrolysisof membrane sphingomyelin (SM) by sphingomyeli-nases (1–3). When the apoptotic program is initiatedat the cell surface from death receptors, a deathdomain/FADD/caspases-dependent acidic sphingo-myelinase (ASM) is responsible for SM hydrolysisand the transient accumulation of apoptogenic cer-amide within 5 to 15 min (4–9). Ceramide is thenrapidly converted, through the stepwise addition ofsugars and sialic residues, to gangliosides. About 10to 30 min after CD95 cross-linking, in fact, GD3ganglioside is neosynthesized and prominently accu-mulates (10). The neosynthesis of GD3 ganglioside,mediated by a GD3 synthase (a2,8-sialyltransferaseor ST8) (11–14), is critical to signal progression,since GD3 can directly trigger apoptosis and suppres-sion of ST8 expression substantially prevents cer-amide- and CD95-induced cell death (10). In hemo-poietic cells, the early accumulation of GD3 requiresASM-derived ceramide, since ASM-deficient cells failto accumulate GD3 and fail to effectively execute theapoptotic program on CD95 cross-linking, whereastransfer of ASM into ASM-deficient cells reconsti-tutes GD3 accumulation and efficient apoptosis (15).Therefore, ASM and ST8 belong to a single pathwaythat promotes the progression of apoptotic signals byultimately generating GD3 ganglioside.
GD3 targets, however, remain unclear. Cells un-dergoing apoptosis on GD3 exposure display earlyloss of mitochondrial transmembrane potential(DCm) (10). Accordingly, recent evidences indicate
1 Correspondence: Laboratory of Signal Transduction, De-partment of Experimental Medicine and Biochemical Sci-ences, University of Rome ‘Tor Vergata’, via di Tor Vergata135, 00133 Rome. E-mail: [email protected]
that GD3 contributes to the opening of the perme-ability transition pore complex (PTPC) in isolatedmitochondria (16–18). Mitochondria play a centralrole in the apoptotic program by directing theactivation of executioner caspases once irreversiblecell damage occurs (19). Early during the apoptoticprocess the DCm is dissolved, likely due to theopening of the PTPC, causing mitochondrial swell-ing and rupture of the outer mitochondrial mem-brane (20). This is associated with the release ofapoptogenic factors, normally confined between themitochondrial membranes, including cytochrome c(21, 22), apoptosis inducing factor (AIF) (23, 24),and selected caspases (25). Cytochrome c and AIF,through different pathways, are both capable oftriggering key nuclear events such as chromatincondensation, activation of endonucleases, and DNAfragmentation.
Here, we provide evidence that GD3 ganglioside,which accumulates in cells undergoing apoptosis,can directly interact with mitochondria, causingDCm dissipation and the release of cytochrome c,AIF, and caspase 9. These events can be largelyprevented by bcl-2. Thus, the lipid pathway recruitsmitochondria to the apoptotic program throughGD3 ganglioside.
MATERIALS AND METHODS
Assessment of mitochondrial parameters in vitro
Mitochondria were purified from rat liver, as described(26), and resuspended in 250 mM sucrose, 0.1 mM EGTA,10 mM N-tris[hydroxymethyl]methyl-2-aminoethanesulfonicacid, pH 7.4. For the induction of PT, mitochondria (1 mgprotein/ml) were resuspended in PT buffer (200 mM su-crose, 10 mM Tris-MOPS, pH 7.4, 5 mM Tris-succinate, 1 mMTris-phosphate, 2 mM rotenone, and 10 mM EGTA-Tris) andmonitored in an F4500 fluorescence spectrometer (Hitachi,Tokyo, Japan) for the 90° light scattering (545 nm) todetermine large amplitude swelling after addition of 5 mMatractyloside, 1 mM cyclosporin A (CsA; Novartis, Basel,Switzerland), and/or 20 mM GD3, GD1a, GM3, GM1, orC2-ceramide (Sigma, St. Louis, Mo.).
Western blotting
Supernatants from mitochondria (6800 g for 15 min; then20,000 g for 1 h; 4°C) were analyzed by sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) andblotted onto nitrocellulose membranes. Cytochrome c wasdetected using a monoclonal antibody (clone 7H8.2C12,PharMingen, San Diego, Calif.), AIF by means of a rabbitantiserum directed against the amino acid residues 151–200(24), and caspase 9 by a rabbit antiserum directed to the p18subunit (kindly provided by Dr. D. Nicholson, Merck, Rah-way, N.J.). Blots were revealed by ECL (Amersham, Bucking-hamshire, U.K.), according to manufacturer instructions.
T cell lymphoma CEM cells (13106) stably transfected withBcl-2 cDNA or with the corresponding empty vector ascontrol were treated with GD3 200 mM for 6 h. Cytosolic
lysates were subjected to SDS-PAGE and then transferred to anitrocellulose membrane. The membranes were incubatedwith a rabbit polyclonal antibody directed to the caspase-9carboxyl terminus (Santa Cruz, Santa Cruz, Calif.) and re-vealed by ECL.
Evaluation of nuclear changes in a cell-free system
DNA fragmentation activity in the supernatant of mitochon-dria was tested on HeLa cell nuclei, as described (27). Briefly,purified HeLa nuclei were resuspended in 50 mM Tris-HCl,pH 7.2, and supernatants of mitochondria were added (90min, 37°C). Nuclei were then stained with propidium iodide(Sigma) and analyzed with a FACScan (Becton Dickinson,Rutherford, N.J.) to determine the frequency of hypodiploidnuclei.
Immunoelectron microscopy
Thin sections (80 nm) were prepared from cells treated withC2-ceramide, fixed in 2% paraformaldehyde and 0.5% glu-taraldehyde, in 0.1 M Sorensen’s phosphate buffer, embed-ded in LR white acrylic resin using gelatin embedding cap-sules (EMS, Ft. Washington, Pa.). Polymerization wasaccomplished at 50°C for 48 h. Sections mounted on forma-var pretreated gold grids were incubated for 10 min with 10%H2O2, rinsed in distilled water, and treated with 1% bovineserum albumin (BSA) for 10 min to minimize nonspecificstaining. Sections were then incubated overnight at 4°C withanti-GD3 mAb (clone R24, IgG3, gift from Dr. L. L. Old,Ludwig Institute, New York) and isotype matched control,followed by a 10 nm gold-conjugated goat antiserum tomouse (Aurion BSA-C kit, Aurion Wageningen, The Nether-lands). Sections were counterstained with 2% uranyl acetate(5 min) and lead citrate (1 min), then analyzed by electronmicroscopy (JEOL Jem 1220).
GD3 synthase antisense experiments
T cell lymphoma CEM cells were incubated for 66 h in RPMI1640 medium containing 10 mM HEPES pH 7.4, 1.0 mMsodium pyruvate, and 10% fetal bovine serum (FBS), with 40mM of GD3 synthase antisense phosphorothioate oligode-oxynucleotides (59-CAGTACAGCCATGGCCCCTCT-39) (28).As control, a scrambled sequence of the same oligo-deoxynucleotides (59-CGACCTACCTATGCGCTACCG-39) oranother irrelevant sequence was used at the same concentra-tions. After GD3 synthase antisense treatment, the cells wereviable but unable to synthesize GD3 in response to ceramide(ref 10 and data not shown). Cells were washed once andresuspended in the above medium, then treated with 40 mMC2-ceramide.
Measurement of mitochondrial transmembrane potential invivo
Dissipation of the mitochondrial transmembrane potential(DCm) and generation of reactive oxygen species (ROS) wereassessed by staining cells with 16 nM 3,39-dihexyloxacarbocya-nine iodide (DiOC6 (2), Molecular Probes) combined with 4mM dihydroethidium (HE, Molecular Probes, Eugene, Oreg.)for 20 min at 37°C, followed by FACS analysis (29).
Transfections of Bcl-2 overexpressing cells
T lymphoma CEM cells stably overexpressing bcl-2 (30),kindly provided by Dr. R. Kofler (University of Innsbruck),
2048 Vol. 14 October 2000 RIPPO ET AL.The FASEB Journal
were treated with GD3 (200 mM) in RPMI 1640, 10% FBS.Hypodiploid nuclei were assessed by staining cells with ahypotonic fluorochrome solution (propidium iodide 50mg/ml (Sigma) in 0.1% sodium citrate plus 0.1% tritonX-100) for 4–8 h at 4°C in the dark and analyzed by aFACScan.
The GD3 synthase cDNA was cloned in pEGFP-C3 expres-sion vector (Clontech, Palo Alto, Calif.), fused to the GFP bythe amino-terminal portion at the ApaI site. In 0.5 ml of RPMI1640, 10 3 106 cells were incubated for 10 min on ice with 20mg of pEGFP-ST8 cDNA or pEGFP empty vector. Cells werethen electrophoresed (Gene Pulser, Bio-Rad, Hercules, Cal-if.) at 290V, 950 mF, left 30 min on ice, and resuspended in 5ml of RPMI 1640, 10% FBS. After 4 h, live cells were recoveredby lymphoprep density gradient centrifugation and replated.After 24 h, apoptotic cells, among green fluorescent cells,were evaluated by fluorescence microscopy.
RESULTS
GD3 directly induces dissipation of mitochondrialtransmembrane potential (DCm)
Treatment of intact cells with exogenous GD3 resultsin apoptosis preceded by a rapid and dramatic loss ofDCm (10). To clarify whether GD3 can be directlyresponsible for dissipation of DCm, isolated rat livermitochondria were exposed to GD3 and mitochon-drial swelling was measured. GD3, but not GD1a (orother commercially available gangliosides such asGM1 and GM3, not shown), was able to induce adramatic swelling of mitochondria within minutes.On the contrary, ceramide was unable to induceswelling of isolated mitochondria, suggesting thatceramide-induced mitochondrial changes duringapoptosis might require GD3 neosynthesis. GD3-induced effects were completely prevented in thepresence of CsA, indicating that GD3 is causing theopening of the mitochondrial PTPC (Fig. 1).
GD3-induced DCm collapse causes the release ofapoptogenic factors from mitochondria
Dissipation of DCm and swelling is often associatedwith rupture of the outer mitochondrial membrane
and the release of multiple factors from the inter-membrane space into the cytosol (19, 25). We there-fore investigated whether GD3-induced DCm loss wascausing the release of apoptogenic factors and mito-chondrial caspases. Isolated mitochondria exposedfor 15 min to GD3, but not to GM3 (or GD1a, notshown), released cytochrome c, AIF (24), and thep32 cleavage product of caspase 9 (25) (Fig. 2).Cytochrome c, AIF, and p32 caspase 9 release wascompletely prevented by pretreating mitochondriawith CsA, indicating that GD3-induced mitochon-drial PTPC opening and consequent DCm collapsewere responsible for the cytosolic release of apopto-genic factors.
GD3-released mitochondrial factors are sufficientto drive DNA fragmentation
In cells undergoing apoptosis, released cyto-chrome c and AIF trigger caspase-dependent andindependent events that eventually result in theinduction of nuclear DNA fragmentation (23, 24,31). The ability of soluble factors released frommitochondria on GD3 contact to trigger nuclearDNA fragmentation was therefore investigated in acell-free system. Isolated HeLa nuclei were ex-posed to supernatants derived from isolated mito-chondria forced to DCm collapse by in vitro GD3treatment and DNA content analyzed by flowcytometry. Only supernatants from GD3-treatedmitochondria, but not from GM3 or GD1a-treated
Figure 2. GD3 induces the release of cytochrome c, AIF, andcaspase 9 (p32 cleavage product) from mitochondria. Super-natants from mitochondria treated as indicated were sub-jected to immunoblot detection of cytochrome c, AIF, andcaspase 9. Supernatants from mitochondria treated with Atr 5mM and/or Ca21 (200 mM) were used as control of maxi-mum release of cytochrome c or AIF. One representativeexperiment out of three performed is shown.
Figure 1. Mitochondrial largeamplitude swelling is inducedby GD3. Purified mitochon-dria (1 mg/ml) were treatedwith buffer alone, GD3 (20mM), GD3 (20 mM) 1 cyclo-sporine A (CsA,1 mM), GD1a(20 mM), C2-ceramide (20mM), positive control atractylo-side (Atr, 5 mM); diffraction oflight was measured at 545 nmwith a fluorescence spectrome-ter. One representative experi-ment out of five performed isshown.
2049GD3 TARGETING OF MITOCHONDRIA PREVENTED BY BCL-2
mitochondria, were effectively inducing DNAfragmentation in isolated nuclei (Fig. 3). Thiseffect was completely prevented by cyclosporinA pretreatment of mitochondria. Direct expo-sure of isolated nuclei to GD3 did not result inDNA fragmentation or nuclear damage (datanot shown). Thus, apoptogenic GD3 is sufficientto drive mitochondrial changes, and subsequentnuclear events, which are associated with apop-tosis.
To investigate whether endogenous GD3 might tar-get mitochondria in vivo, the possible accumulationof GD3 within mitochondria was investigated byimmunoelectron microscopy. After ;30 min expo-sure to apoptogenic doses of ceramide, GD3 wasfrequently found associated with membranes ofswelling mitochondria (Fig. 4, black dots). We theninvestigated whether accumulation of GD3 was caus-ing mitochondrial damage in vivo. CEM cells werepretreated with ST8 antisense oligodeoxynucleotides(28) to suppress the expression of ST8 and prevent
the neosynthesis of GD3 on ceramide exposure (10).Cells were then treated with apoptogenic ceramideand mitochondrial changes evaluated. Pretreatmentwith ST8-antisense, but not with a scrambled se-quence of the same oligodeoxynucleotides, resultedin a substantial reduction of ceramide-induced DCm
loss and ROS generation (Fig. 5). ST8-antisensecould not, however, prevent GD3-induced mitochon-drial changes (Fig. 5) or GD3-induced apoptosis (notshown). These data strongly suggest that endog-enously generated GD3 induces mitochondrial fail-ure during apoptosis.
Figure 3. The apoptogenic factors released from GD3-treatedmitochondria induce nuclear apoptosis. Supernatants frommitochondria treated as indicated in the figure were tested onpurified HeLa nuclei for the capacity to induce DNA frag-mentation in vitro. DNA content and hypoploidy were deter-mined by propidium iodide (PI) staining and cytofluoromet-ric analysis. Percentage of hypodiploid nuclei is indicated.Fluorescence channels on the x axis, cell counts on the y axis.One representative experiment out of four performed isshown.
Figure 4. Ultrastructural analysis of representative mito-chondria from unstimulated (A) and stimulated cells with40 mM C2-ceramide for 30 min (B). Cells were fixed andanalyzed by immunoelectron microscopy after treatmentwith R24 mAb followed by gold-conjugated anti-mouse Igantibody. GD3 is detectable as black grains on mitochon-drial membranes. Original magnification: 330,000. A de-tailed morphometric analysis of several different images ofceramide-treated cells indicated that the number of thegrains associated with mitochondrial membranes was atleast 10-fold higher compared to membrane-free cytosolicareas (not shown).
Figure 5. Endogenously synthesized GD3 induces mitochon-drial changes. CEM cells were incubated with 40 mM of GD3synthase antisense (As) or scrambled (Sc) oligonucleotides,then treated with 40 mM C2-ceramide or 200 mM GD3 for 3 h.Mitochondrial changes (DCm dissipation and ROS genera-tion) were assayed by FACS analysis (as described in Materialsand Methods). Means 6 1 sd of five independent experi-ments are shown.
2050 Vol. 14 October 2000 RIPPO ET AL.The FASEB Journal
Enforced expression of bcl-2 prevents GD3-induced mitochondrial changes, caspase 9activation, and apoptosis
The above results provide both in vitro and in vivoevidence for a role of GD3 in the induction ofmitochondrial DCm disruption. Since anti-apoptotic
bcl-2 family members directly interfere with PTPCopening and DCm loss (32, 33), the ability of bcl-2 toprotect from GD3-induced mitochondrial damageand apoptosis was investigated. CEM cells stablytransfected with bcl-2 or with the correspondingempty vector as control were exposed to GD3, andthe induction of permeability transition and ROSgeneration was measured. Bcl-2 overexpressing CEMcells displayed resistance to GD3-induced mitochon-drial changes compared to the control cells (Fig. 6A).
Cytochrome c, released concomitantly to DCm
dissipation, interacts with APAF-1, causing the re-cruitment and activation of caspase 9, dictating irre-versible commitment to apoptosis (34). This eventcan be effectively counteracted by bcl-2 (35). Theability of GD3 to induce caspase 9 activation in vivoand the possible interference by bcl-2 were thereforeinvestigated. As shown in Fig. 6B, exposure of CEMcells to GD3 induced pro-caspase 9 degradation andthe appearance of a 32 kDa cleavage product within6 h, whereas no caspase 9 activation could be de-tected in CEM cells stably overexpressing bcl-2. Ac-cordingly, GD3-induced apoptosis was substantiallydelayed in bcl-2 overexpressing CEM cells (Fig. 6C).
Finally, to assess whether bcl-2 could protect cellsfrom endogenous GD3 overproduction, a GFP-tagged ST8 (GFP-ST8) was transiently expressed inCEM cells stably overexpressing bcl-2. Essentially allcell death induced by GFP-ST8 in control CEM cellscould be blocked in CEM overexpressing bcl-2 (Fig.7). Together, these data indicate that GD3-induceddamage is mostly confined at sites that can berestrained by bcl-2.
DISCUSSION
The data presented in this paper demonstrate thatGD3 ganglioside contributes to the apoptotic pro-
Figure 6. Bcl-2 prevents GD3-induced mitochondrial changes,caspase 9 activation, and apoptosis. CEM cells stably trans-fected with bcl-2 or with the empty-vector were incubated with200 mM GD3 for the indicated time points, and the effects onmitochondrial changes (DCm dissipation and ROS genera-tion) (A) and DNA fragmentation (C) were measured byFACS analysis. Means 6 1 sd of four independent experi-ments are shown. CEM cells stably transfected with bcl-2 orwith the empty-vector were incubated with 200 mM GD3 for6 h and caspase 9 analyzed by immunoblotting of cell lysates(B). One experiment out of two performed with identicalresults is shown.
Figure 7. Bcl-2 prevents apoptosis induced by endogenousGD3 accumulation. The effects of the overexpression ofpEGFP-ST8 or of the pEGFP empty vector in the bcl-2-transfected cells after 24 h were analyzed by evaluatingapoptotic cells among the green fluorescent cells by fluores-cence microscopy. Means 6 1 sd of three independentexperiments are shown.
2051GD3 TARGETING OF MITOCHONDRIA PREVENTED BY BCL-2
gram by directly targeting mitochondria in a bcl-2-controlled manner and inducing the mitochondrialrelease of apoptogenic factors, including cyto-chrome c, AIF, and caspase 9.
GD3 rapidly accumulates in cells undergoing cer-amide-dependent apoptosis by the action of ST8, apolysialyltransferase mostly resident in the endoplas-mic reticulum and early Golgi compartment (ER/EG) (11–14). ST8 generates GD3 from its immediateprecursor GM3 by adding a second sialic acid to theGM3 sialic residue. Although little information iscurrently available concerning ST8 regulation invivo, excess ceramide accelerates the rate of GD3neosynthesis, resulting in GD3 accumulation (10). Inhemopoietic cells, excess ceramide derives initiallyfrom the accelerated hydrolysis of sphingomyelin bythe action of acidic sphingomyelinase (ASM),boosted by membrane clustering of death receptorsand death domain/FADD-dependent caspases acti-vation (4–9). Therefore, excess ceramide generatedin acidic compartments by ASM feeds into the gan-glioside biosynthetic pathway down to GD3 accumu-lation (15), which occurs mostly in the ER/EGcompartment. Mitochondria may come into closecontact with the ER/EG, which they form a function-ally interconnected network with (36, 37). GD3accumulated in the ER/EG could therefore interactwith close-by mitochondrial membranes or be phys-ically redistributed to mitochondrial membranes.Further studies will clarify this issue.
GD3 efficiently disrupts DCm in intact cells (10).Moreover, recent reports indicate that GD3 is apotent inducer of mitochondrial permeability tran-sition in isolated mitochondria (16, 17). We showhere that GD3 is sufficient to cause the release ofapoptogenic cytochrome c, AIF, and caspase 9 fromisolated mitochondria and, remarkably, that GD3-treated mitochondria can release factors that aresufficient to induce DNA fragmentation in isolatednuclei. Moreover, selective suppression of ST8 ex-pression in vivo, and therefore of endogenous GD3neosynthesis, significantly prevented ceramide-in-duced DCm dissipation. Although ceramide-depen-dent pathways might affect mitochondrial functionthrough kinase/phosphatase cascades regulating Aktactivity and bad phosphorylation (38, 39), the evi-dence presented indicates that some relevant effectsof ceramide on mitochondria may require conver-sion to GD3.
Endogenously generated GD3 accumulates withinmitochondria and causes mitochondrial changes invivo in cells undergoing apoptosis. This findingallows the enlistment of GD3 ganglioside among the‘natural’ macromolecular inducers of mitochondrialpermeability transition, mobilized to recruit mito-chondria to the apoptotic program. They includepro-apoptotic bcl-2 family members such as bax
(40–43), bak (44) and p15bid (45–47) as well as AIF(24) and selected caspases (32).
Most known mitochondrial permeability transitioninducers seem to act at the level of PTPC, a multi-protein complex situated at contact sites between theinner and the outer mitochondrial membranes.Dimeric bax resides in mitochondrial membranes,where it participates to the regulation of the PTPC bydirectly interacting with the adenine nucleotidetranslocator (48) and, together with bak, with thevoltage-dependent anion channel (44, 49), bothcomponents of the PTPC. Monomeric bax translo-cates from the cytosol to the mitochondria in cellsundergoing apoptosis (50), and BH3-mediated ho-motypic interactions with resident bax might directlyaffect mitochondrial permeability transition. Simi-larly, p15bid, resulting from caspase 8-mediatedcleavage of bid (45–47), relocates to the mitochon-dria and very efficiently alters PTPC function (51).AIF is responsible for both the direct induction ofnuclear changes and the cytosolic amplification ofthe apoptotic response, by acting on nearby mito-chondria through yet unknown mechanisms (23,24). Caspases can lower the mitochondrial perme-ability transition threshold by processing mitochon-drial bcl-2 family members known to stabilize thePTPC, such as bcl-2 and bcl-XL (32). The protectiveeffects of bcl-2 on GD3-induced apoptosis stronglysuggest that relevant GD3 targets are controlled bybcl-2.
Different from cytosolic apoptotic effectors of pro-teic nature, GD3 ganglioside is likely to reach mito-chondria via physical continuity between ER/EG andmitochondrial membranes (36, 37). Recruitment ofavailable mitochondria through membrane connec-tions is expected to be a slower process compared todiffusion of soluble products within the cytosol. Thismight explain the generally slower kinetics of apo-ptosis experimentally induced in vitro by ceramides,or by GD3 itself, compared to death receptor cross-linking agents activating upstream caspases. Accu-mulation of sphingomyelin-derived ceramides dur-ing acute stress responses occurs in simple organismsthat lack caspases or bcl-2 family members (52, 53).Membrane-directed delivery of death messages maytherefore represent an evolutionary ancient mecha-nism for the recruitment of mitochondria to the celldeath program.
Mitochondria appear therefore to represent acritical destination of the lipid pathway. In fact, GD3is unable to directly affect nuclear membranesand/or cause nuclear events associated with apopto-sis, whereas supernatants from GD3-treated mito-chondria are entirely competent in inducing DNAfragmentation within isolated nuclei. This suggeststhat GD3 is not generically perturbing cellular mem-branes. Pretreatment of isolated mitochondria with
2052 Vol. 14 October 2000 RIPPO ET AL.The FASEB Journal
cyclosporin A completely suppressed GD3-inducedswelling and release of apoptogenic factors, indicat-ing that GD3 is acting at the level of the PTPC.Whether this is due to a direct interaction with any ofthe PTPC components, some of which are currentlyunknown, or to a local perturbation of specificmitochondrial membrane microdomains affectingelectrical and spatial constraints relevant to thePTPC physiology remains to be established.
We are grateful to Drs. G. Stassi (University of Palermo), R.De Maria (ISS, Rome), L. Lenti, R. Gradini, and F. d’Agostino(University of Rome ‘La Sapienza’) and to Prof. A. Modesti(University of Rome ‘Tor Vergata’) for helpful discussionsand advices. We are also indebted to Dr. R Kofler (Universityof Innsbruch) for providing the bcl-2-transfected CEM cellsand to Dr. D. Nicholson (Merck, Rahway, N.J.) for providingthe anti-caspase 9 antiserum. This work has been supportedby Associazione Italiana Ricerca sul Cancro, Ministero Uni-versita’ e Ricerca Scientifica e Tecnologica, Agenzia SpazialeItaliana, Consiglio Nazionale delle Ricerche Progetto Biotec-nologie, Istituto Superiore di Sanita’ Progetto Sclerosi Multi-pla, European Community Biomed 2 and TMR programs (toR.T), Agence Nationale pour la Recherche sur le SIDA,Association pour la Recherche sur le Cancer, Fondation pourla Recherche Medicale, Ligue Nationale contre le Cancer,and European Commission French Ministry for Science (toG.K.). F.M. and I.C. are AIRC fellowship holders. M.R.R. is aFondazione Adriano Buzzati-Traverso fellowship holder.
REFERENCES
1. Kolesnick, R. N., and Kronke, M. (1998) Regulation of ceramideproduction and apoptosis. Annu. Rev. Physiol. 60, 643–665
2. Perry, D. K., and Hannun, Y. A. (1998) The role of ceramide incell signaling. Biochim. Biophys. Acta 1436, 233–243
3. Testi, R. (1996) Sphingomyelin breakdown and cell fate. TrendsBiochem. Sci. 21, 468–471
4. Wiegmann, K., Schutze, S., Machleidt, T., Witte, D., and Krsnke,M. (1994) Functional dichotomy of neutral and acidic sphingo-myelinases in tumor necrosis factor signaling. Cell 78, 1005–1015
5. Cifone, M. G., De Maria, R., Roncaioli, P., Rippo, M. R., Azuma,M., Lanier, L. L., Santoni, A., and Testi, R. (1994) Apoptoticsignaling through CD95 (Fas/APO-1) activates an acidic sphin-gomyelinase. J. Exp. Med. 180, 1547–1552
6. Cifone, M. G., Roncaioli, P., De Maria, R., Camarda, G., Santoni,A., Ruberti, G., and Testi, R. (1995) Multiple signaling originateat the Fas/Apo-1 (CD95) receptor: sequential involvement ofphosphatidylcholine-specific phospholipase C and acidic sphin-gomyelinase in the propagation of the apoptotic signal. EMBO J.14, 5859–5868
7. De Maria, R., Boirivant, M., Cifone, M. G., Roncaioli, P., Hahne,M., Tschopp, J., Pallone, F., Santoni, A., and Testi, R. (1996)Functional expression of Fas and Fas Ligand on human gutlamina propria lymphocytes. A potential role for the acidicsphingomyelinase pathway in normal immunoregulation.J. Clin. Invest. 97, 316–322
8. Brenner, B., Ferlinz, K., Grassme, H., Weller, M., Koppenhoefer,U., Dichgans, J., Sandhoff, K., Lang, F., and Gulbins, E. (1998)Fas/CD95/Apo-I activates the acidic sphingomyelinase viacaspases. Cell Death. Differ. 5, 29–37
9. Wiegmann, K., Schwandner, R., Krut, O., Yeh, W. C., Mak, T. W.,and Kronke, M. (1999) Requirement of FADD for tumornecrosis factor-induced activation of acid sphingomyelinase.J. Biol. Chem. 274, 5267–5270
10. De Maria, R., Lenti, M. L., Malisan, F., d’Agostino, F., Tomassini,B., Zeuner, A., Rippo, M. R., and Testi, R. (1997) Requirementfor GD3 ganglioside in CD95- and ceramide-induced apoptosis.Science 277, 1652–1655
11. Haraguchi, M., Yamashiro, S., Yamamoto, A., Furukawa, K.,Takamiya, K., Lloyd, K. O., Shiku, H., and Furukawa, K. (1994)Isolation of GD3 synthase gene by expression cloning of GM3a-2,8-sialyltransferase cDNA using anti-GD2 monoclonal anti-body. Proc. Natl. Acad. Sci. USA 91, 10455–10459
12. Nara, K., Watanabe, Y., Maruyama, K., Kasahara, K., Nagai, Y.,and Sanai, Y. (1994) Expression cloning of a CMP-NeuAc:NeuAc a2–3Gal b1–4Glc b1–19Cer alpha 2,8-sialyltransferase(GD3 synthase) from human melanoma cells. Proc. Natl. Acad.Sci. USA 91, 7952–7256
13. Sasaki, K., Kurata, K., Kojima, N., Kurosawa, N., Ohta, S., Hanai,N., Tsuji, S., and Nishi, T. (1994) Expression cloning of aGM3-specific a-2,8-sialyltransferase (GD3 synthase). J. Biol.Chem. 269, 15950–15956
14. Nakayama, J., Fukuda, M. N., Hirabayashi, Y., Kanamori, A.,Sasaki, K., Nishi, T., and Fukuda, M. (1996) Expression cloningof a human GT3 synthase. GD3 and GT3 are synthesized by asingle enzyme. J. Biol. Chem. 271, 3684–3691
15. De Maria, R., Rippo, M. R., Schuchman, H. E., and Testi, R.(1998) Acidic sphingomyelinase (ASM) is necessary for Fas-induced GD3 ganglioside accumulation and efficient apoptosisof lymphoid cells. J. Exp. Med. 187, 897–902
16. Scorrano, L., Petronilli, V., Di Lisa, F., and Bernardi, P. (1999)Commitment to apoptosis by GD3 ganglioside depends onopening of the mitochondrial permeability transition pore.J. Biol. Chem. 274, 22581–22585
17. Kristal, B. S., and Brown, A. M. (1999) Apoptogenic gangliosideGD3 directly induces the mitochondrial permeability transition.J. Biol. Chem. 274, 23169–23175
18. Pastorino, J. G., Tafani, M., Rothman, R. J., Marcineviciute, A.,Hoek, J. B., and Farber, J. L. (1999) Functional consequences ofthe sustained or transient activation by bax of the mitochondrialpermeability transition pore. J. Biol. Chem. 274, 31734–31739:
19. Green, D., and Kroemer, G. (1998) The central executioner ofapoptosis: caspases or mitochondria? Trends Cell Biol. 8, 267–271
20. Kroemer, G. (1998) The mitochondrial death/life regulator inapoptosis and necrosis. Annu. Rev. Physiol. 60, 619–642
21. Liu, X., Kim, C. N., Yang, J., Jemmerson, R., and Wang, X.(1996) Induction of apoptotic program in cell-free extracts:requirement for cytochrome c. Cell 86, 147–157
22. Reed, J. C. (1997) Cytochrome c: can’t live with it—can’t livewithout it. Cell 91, 559–562
23. Susin, S. A., Zamzami, N., Castedo, M., Hirsch, T., Marchetti, P.,Macho, A., Daugas, E., Geuskens, M., and Kroemer, G. (1996)Bcl-2 inhibits the mitochondrial release of an apoptogenicprotease. J. Exp. Med. 184, 1331–1341
24. Susin, S. A., Lorenzo, H. K., Zamzami, N., Marzo, I., Snow, B. E.,Brothers, G. M., Mangion, J., Jacotot, E., Costantini, P., Loeffler,M., Larochette, N., Goodlett, D. R., Aebersold, R., Siderovski,D. P., Penninger, J. M., and Kroemer, G. (1999) Molecularcharacterization of mitochondrial apoptosis-inducing factor.Nature (London) 397, 441–446
25. Susin, S. A., Lorenzo, H. K., Zamzami, N., Marzo, I., Brenner, C.,Larochette, N., Prevost, M. C., Alzari, P. M., and Kroemer, G.(1999) Mitochondrial release of caspase-2 and -9 during theapoptotic process. J. Exp. Med. 189, 381–394
26. Costantini, P., Petronilli, V., Colonna, R., and Bernardi, P.(1995) On the effects of paraquat on isolated mitochondria.Evidence that paraquat causes opening of the cyclosporinA-sensitive permeability transition pore synergistically with nitricoxide. Toxicology 99, 77–88
27. Susin, S. A., Zanzami, N., Castedo, M., Daugas, E., Wang, H.-G.,Geley, S., Fassy, F., Reed, J. C., and Kroemer, G. (1997) Thecentral executioner of apoptosis: multiple connections betweenprotease activation and mitochondria in Fas/APO-1/CD95- andceramide-induced apoptosis. J. Exp. Med. 186, 25–37
28. Zeng, G., Ariga, T., Gu, X.-b., and Yu, R. K. (1995) Regulation ofglycolipid synthesis in HL-60 cells by antisense oligodeoxynucle-otides to glycosyltransferase sequences: effect on cellular differ-entiation. Proc. Natl. Acad. Sci. USA 92, 8670–8674
29. Zamzami, N., Marchetti, P., Castedo, M., Zanin, C., Vayssiere,J. L., Petit, P. X., and Kroemer, G. (1995) Reduction inmitochondrial potential constitutes an early irreversible step ofprogrammed lymphocyte death in vivo. J. Exp. Med. 181, 1661–1672
30. Geley, S., Hartmann, B. L., and Kofler, R. (1997) Ceramidesinduce a form of apoptosis in human acute lymphoblastic
2053GD3 TARGETING OF MITOCHONDRIA PREVENTED BY BCL-2
leukemia cells that is inhibited by Bcl-2, but not by CrmA. FEBSLett. 400, 15–18
31. Zamzami, N., Susin, S. A., Marchetti, P., Hirsch, T., Gomez-Monterrey, I., Castedo, M., and Kroemer, G. (1996) Mitochon-drial control of nuclear apoptosis. J. Exp. Med. 183, 1533–1544
32. Marzo, I., Brenner, C., Zamzami, N., Susin, S. A., Beutner, G.,Brdiczka, D., Remy, R., Xie, Z. H., Reed, J. C., and Kroemer, G.(1998) The permeability transition pore complex: a target forapoptosis regulation by caspases and bcl-2-related proteins. J.Exp. Med. 187, 1261–1271
33. Gross, A., McDonnell, J. M., and Korsmeyer, S. J. (1999) BCL-2family members and the mitochondria in apoptosis. Genes Dev.13, 1899–1911
34. Li, P., Nijhawan, D., Budihardjo, I., Srinivasula, S. M., Ahmad,M., Alnemri, E. S., and Wang, X. (1997) Cytochrome c anddATP-dependent formation of Apaf-1/caspase-9 complex ini-tiates an apoptotic protease cascade. Cell 91, 479–489
35. Yang, J., Liu, X., Bhalla, K., Kim, C. N., Ibrado, A. M., Cai, J.,Peng, T. I., Jones, D. P., and Wang, X. (1997) Prevention ofapoptosis by Bcl-2: release of cytochrome c from mitochondriablocked. Science 275, 1129–1132
36. Rusinol, A. E., Cui, Z., Chen, M. H., and Vance, J. E. (1994) Aunique mitochondria-associated membrane fraction from ratliver has a high capacity for lipid synthesis and contains pre-Golgi secretory proteins including nascent lipoproteins. J. Biol.Chem. 269, 27494–27502
37. Rizzuto, R., Pinton, P., Carrington, W., Fay, F. S., Fogarty, K. E.,Lifshitz, L. M., Tuft, R. A., and Pozzan, T. (1998) Close contactswith the endoplasmic reticulum as determinants of mitochon-drial Ca21 responses. Science 280, 1763–1766
38. Zundel, W., and Giaccia, A. (1998) Inhibition of the anti-apoptotic PI(3)K/Akt/Bad pathway by stress. Genes Dev. 12,1941–1946
39. Basu, S., Bayoumy, S., Zhang, Y., Lozano, J., and Kolesnick, R.(1998) BAD enables ceramide to signal apoptosis via Ras andRaf-1. J. Biol. Chem. 273, 30419–30426
40. Pastorino, J. G., Chen, S. T., Tafani, M., Snyder, J. W., andFarber, J. L. (1998) The overexpression of Bax produces celldeath upon induction of the mitochondrial permeability tran-sition. J. Biol. Chem. 273, 7770–7775
41. Jurgensmeier, J. M., Xie, Z., Deveraux, Q., Ellerby, L., Bredesen,D., and Reed, J. C. (1998) Bax directly induces release ofcytochrome c from isolated mitochondria. Proc. Natl. Acad. Sci.USA 95, 4997–5002
42. Gross, A., Jockel, J., Wei, M. C., and Korsmeyer, S. J. (1998)Enforced dimerization of BAX results in its translocation,mitochondrial dysfunction and apoptosis. EMBO J. 17, 3878–3885
43. Finucane, D. M., Bossy-Wetzel, E., Waterhouse, N. J., Cotter,T. G., and Green, D. R. (1999) Bax-induced caspase activationand apoptosis via cytochrome c release from mitochondria isinhibitable by Bcl-xL. J. Biol. Chem. 274, 2225–2233
44. Narita, M., Shimizu, S., Ito, T., Chittenden, T., Lutz, R. J.,Matsuda, H., and Tsujimoto, Y. (1998) Bax interacts with thepermeability transition pore to induce permeability transitionand cytochrome c release in isolated mitochondria. Proc. Natl.Acad. Sci. USA 95, 14681–14686
45. Luo, X., Budihardjo, I., Zou, H., Slaughter, C., and Wang, X.(1998) Bid, a Bcl2 interacting protein, mediates cytochrome crelease from mitochondria in response to activation of cellsurface death receptors. Cell 94, 481–490
46. Li, H., Zhu, H., Xu, C. J., and Yuan, J. (1998) Cleavage of BIDby caspase 8 mediates the mitochondrial damage in the Faspathway of apoptosis. Cell 94, 491–501
47. Gross, A., Yin, X. M., Wang, K., Wei, M. C., Jockel, J., Milliman,C., Erdjument-Bromage, H., Tempst, P., and Korsmeyer, S. J.(1999) Caspase cleaved BID targets mitochondria and is re-quired for cytochrome c release, whereas BCL-XL prevents thisrelease but not tumor necrosis factor-R1/Fas death. J. Biol.Chem. 274, 1156–1163
48. Marzo, I., Brenner, C., Zamzami, N., Jurgensmeier, J. M., Susin,S. A., Vieira, H. L., Prevost, M. C., Xie, Z., Matsuyama, S., Reed,J. C., and Kroemer, G. (1998) Bax and adenine nucleotidetranslocator cooperate in the mitochondrial control of apopto-sis. Science 281, 2027–2031
49. Griffiths, G. J., Dubrez, L., Morgan, C. P., Jones, N. A., White-house, J., Corfe, B. M., Dive, C., and Hickman, J. A. (1999) Celldamage-induced conformational changes of the pro-apoptoticprotein Bak in vivo precede the onset of apoptosis. J. Cell Biol.144, 903–914
50. Wolter, K. G., Hsu, Y. T., Smith, C. L., Nechushtan, A., Xi,X. G., and Youle, R. (1997) Movement of Bax from thecytosol to mitochondria during apoptosis. J. Cell Biol. 139,1281–1292
51. Desagher, S., Osen-Sand, A., Nichols, A., Eskes, R., Montessuit,S., Lauper, S., Maundrell, K., Antonsson, B., and Martinou, J. C.(1999) Bid-induced conformational change of Bax is responsi-ble for mitochondrial cytochrome c release during apoptosis.J. Cell Biol. 144, 891–901
52. Basu, S., and Kolesnick, R. (1998) Stress signals for apoptosis:ceramide and c-Jun kinase. Oncogene 17, 3277–3285
53. Fraser, A., and James, C. (1998) Fermenting debate: do yeastundergo apoptosis? Trends Cell Biol. 8, 219–221
Received for publication December 9, 1999.Revised for publication April 4, 2000 .
2054 Vol. 14 October 2000 RIPPO ET AL.The FASEB Journal
