Chemotherapy: Targeting the Mitochondrial Cell Death Pathway

Chemotherapy: targeting the mitochondrial cell death pathway

Klaus-Michael Debatin*,1, Delphine Poncet2 and Guido Kroemer*,2

1University Childrens Hospital, Prittwitzstrasse 43, D-89075 Ulm, Germany; 2Centre National de la Recherche Scientiﬁque, UMR1599, Institut Gustave Roussy, 39 rue Camille Desmoulins, F-94805 Villejuif, France

One of the mechanisms by which chemotherapeutics multidrug resistance phenotype (Kroemer, 1997), and destroy cancer cells is by inducing apoptosis. Apoptosis the major Bcl-2-controlled checkpoint is mitochondrial can be activated through several different signalling membrane permeabilization (MMP) (Kroemer, 1997). pathways, but these all appear to converge at a single One of the signature events of apoptotic autolysis event – mitochondrial membrane permeabilization consists in the activation of a cascade of proteolytic (MMP). This ‘point-of-no-return’ in the cell death enzyme, the so-called caspases, which are synthesized as program is a complex process that is regulated by the zymogens (pro-caspases) and undergo proteolytic composition of the mitochondrial membrane and pre- maturation. Caspase activation can be achieved by the mitochondrial signal-transduction events. MMP is occupation of so-called ‘death receptors’ (DR) from the subject to a complex regulation, and local alterations tumour necrosis factor (TNF) receptor superfamily (e.g. in the composition of mitochondrial membranes, as well CD95/Apo-1/Fas, DR3, DR4), which, once ligated, as alterations in pre-mitochondrial signal-transducing stimulate the recruitment of a specific subclass of events, can determine chemotherapy resistance in cancer activator pro-caspases (in particular pro-caspases -8 cells. Detecting MMP might thus be useful for detecting and -10), leading to their receptor-proximal activation chemotherapy responses in vivo. Several cytotoxic drugs (Box 2). In some cell types (‘type 1 cells’), the resulting induce MMP by a direct action on mitochondria. This caspase-8 activation fully turns on the caspase activa- type of agents can enforce death in cells in which tion cascade by proteolytic activation of pro-caspase-3, upstream signals normally leading to apoptosis have been the principal effector caspase. In most cell types (‘type 2 disabled. Cytotoxic components acting on mitochondria cells’), however, death receptor crosslinking fails to can specifically target proteins from the Bcl-2 family, the cause a caspase-8 activation that would be sufficiently peripheral benzodiazepin receptor, or the adenine strong to activate caspase-3. Rather, the pathway nucleotide translocase, and/or act by virtue of their involves the caspase-8 mediated digestion of the Bcl-2 physicochemical properties as steroid analogues, cationic family protein Bid (which yields truncated Bid, t-Bid), ampholytes, redox-active compounds or photosensitizers. t-Bid-mediated mitochondrial membrane permeabiliza- Some compounds acting on mitochondria can overcome tion (MMP), release of cytochrome c from the cytoprotective effect of Bcl-2-like proteins. Several mitochondria, and cytochrome c-dependent formation agents which are already used in anti-cancer chemother- of a caspase-3 activation complex, the apoptosome. apy can induce MMP, and new drugs specifically Chemotherapy can induce the expression of death designed to target mitochondria are being developed. receptors and/or their ligands and thus activate the Oncogene (2002) 21, 8786 – 8803. doi:10.1038/sj.onc. ‘extrinsic’ pathway which then kills cells in an MMP- 1206039 independent (‘type 1’) or in an MMP-dependent (‘type 2’) fashion (Fulda et al., 2001). In addition and Keywords: anti-cancer therapy; drug targets; mitochondria probably more important, chemotherapy can elicit cellular stress which ultimately leads to caspase- independent MMP, through the ‘intrinsic’ pathway Introduction (Costantini et al., 2000b). In this latter paradigm, Bcl-2- inhibitable MMP seals the cell’s fate, leading to death The most commonly used cancer therapeutics eliminate with or without post-mitochondrial caspase activation tumours by inducing apoptosis (Herr and Debatin, (Kroemer et al., 1995; Kroemer and Reed, 2000). 2001; Johnstone et al., 2002). Overexpression or loss of Tumours become resistant to treatment by modify- function of some anti-apoptotic proteins, notably those ing expression levels or function of proteins involved in from the Bcl-2 family (Box 1), have a rather large this apoptosis signalling pathway. Inactivation of p53, spectrum of apoptosis-inhibitory effects and confer a either directly through p53 mutations or indirectly through p53 modulators such as the mdm2 protein, occurs in most human cancers. Without p53 function, tumour cells evade apoptosis and can continue to *Correspondence: G Kroemer; E-mail: [email protected] or K-M Debatin; E-mail: [email protected] proliferate, despite genetic instabilities caused by Received 23 July 2002; revised 9 September 2002; accepted 11 chemotherapy or irradiation. Many cancer cells over- September 2002 express anti-apoptotic proteins such as Bcl-2 (Kroemer, Targeting mitochondria K-M Debatin et al 8787

Box 1 Proteins from the Bcl-2 family. Proteins from the Bcl-2 family share homology with Bcl-2 in the Bcl-2 homology regions (BH1 to BH4). Two classes of Bcl-2 family members can be distinguished. Anti-apoptotic Bcl-2 homologs possess all four BH1 homology regions. These proteins are primarily localized in mitochondria and stabilize their membranes. Pro-apoptotic Bcl-2 homologs can lack the BH4 domain (Bax, Bak, Bok/Mtd), BH2 (Bcl-XS), or BH1, BH2, and BH4. These latter proteins are referred to as ‘BH-3-only’ proteins and can possess a C-terminal transmembrane (TM) region.

During apoptosis induction pro-apoptotic Bcl-2 family members can translocate from an extra-mitochondrial to mitochondrial membranes, causing their permeabilization. This applies to Bax, which translocates from the cytosol to mitochondria as a result of cytosolic alkalinization. Bid is also normally found in the cytosol and translocates to mitochondria upon digestion by proteases (in particular by caspase-8 but also by calpain, cathepsin L, or granzyme B), yielding truncated Bid (t-Bid). Similar translocation reactions have been described for Bad (which is cytosolic when phosphorylated, for instance by the pro-survival kinase Akt/PKB or other kinases and mitochondrial when dephosphorylated), Bim (which is normally associated with the dynein light chain and hence associated with microtubuli), Bmf (normally associated with the myosin V actin motor complex), as well as NoxA and PUMA (all transcriptionally activated by p53, as Bax). Bak and BNIP3 (which is induced by hypoxia) are constitutively present in mitochondrial membranes.

Oncogene Targeting mitochondria K-M Debatin et al 8788

Box 2 Two alternative pathways elicited by chemotherapy. Intrinsic pathway (or stress pathway): Through a variety of direct or indirect mechanisms, chemotherapy leads to mitochondrial membrane permeabilization, which is regulated, at least to some extent, by proteins from the Bcl-2 family. Mitochondrial membrane permeabilization is rate limiting for cell death. Caspase activation is elicted by the intrinsic pathway yet is mostly dispensable for cell death to occur. Experimental chemotherapeutic agents directly inducing MMP are being developed (Table 4). Extrinsic pathway: (DR) of the tumour necrosis factor (TNF) receptor superfamily (e.g. CD95/Apo-1/Fas, DR3, DR4) and/or their ligands are induced at the transcriptional and/or post-transcriptional levels. DR ligation stimulates the recruitment and activation of a speciﬁc subclass of activator pro-caspases, in particular caspases 8 and 10. Inhibition of the activation of these initiator caspases results in cell death inhibition. The subsequent steps leading to apoptosis depend on the cell type: * Type 1 cells : DR binding leads to caspase-8 (or -10) activation, resulting in direct proteolytic activation of pro-caspase-3 – the principal effector caspase – and initiation of the caspase cascade. In type 1 cells, overexpression fails to prevent caspase activation and apoptosis. Type 1 cell death elicited via DR reportedly occurs independent from mitochondrial membrane permeabilization. * Type 2 cells : DR crosslinking fails to induce a level of caspase-8 activation sufﬁcient to activate caspase-3. Rather, the pathway involves the caspase-8 mediated digestion of the Bcl-2 family protein Bid (which yields truncated Bid, t-Bid), t-Bid-mediates mitochondrial membrane permeabilization, release of cytochrome c from mitochondria, and cytochrome c-dependent formation of a caspase-3 activation complex, the apoptosome. Inhibition of mitochondrial membrane permeabilization by Bcl-2 prevents apoptosis in type 2 cells.

1997). Studies in breast and gastrointestinal tumours, tion of Bid by casein kinases 1 or 2 prevents its as well as other cancer types, have shown that loss of cleavage by caspase-8 (Desagher et al., 2001). t-Bid, Bax expression, or increased expression of Bcl-2 or Bcl- which is rapidly N-myristoylated (Zha et al., 2000), XL, are associated with poor prognosis and decreased then translocates to mitochondrial membranes, exem- response to chemotherapy (Ionov et al., 2000; Paradiso plifying how pre-mitochondrial signal transduction can et al., 2001; Sturm et al., 2001; Tai et al., 1998). Since be controlled at several levels. these pro-apoptotic signalling pathways eventually lead Similar mitochondrial translocation events have to MMP, direct activation of this process might be able been reported for other pro-apoptotic Bcl-2 family to circumvent chemotherapy resistance caused by members including Bad (which is cytosolic when mutations in pre-mitochondrial signal-transducing phosphorylated, for instance by the pro-survival machinery. But how does MMP actually occur, and kinase Akt/PKB and mitochondrial when depho- how can we induce it specifically in cancer cells? sphorylated) (Chao and Korsmeyer, 1998), Bim (which is normally associated with the dynein light chain) (Puthalalath et al., 1999), Bmf (normally Mitochondrial membrane permeabilization (MMP): the associated with the myosin V actin motor complex) apoptotic checkpoint (Puthalakath et al., 2001), BNIP3 (which is transcrip- Numerous pro-apoptotic pathways converge at the tionally activated by hypoxia) (Sowter et al., 2001), as level of MMP induction, and MMP may have multiple well as NoxA, PUMA and Bbc3 (all transcriptionally lethal consequences for cellular physiology, suggesting activated by p53, as Bax) (Han et al., 2001; Oda et al., that MMP may function as the ‘central executioner’ of 2000a). Other interesting pro-apoptotic, p53-induced apoptosis (Green and Kroemer, 1998). gene products acting on mitochondria are oxidore- ductases (proline oxidase and ferredoxin reductase), Mitochondria as death signal integrators Mitochondria which are imported into mitochondria and locally integrate numerous different pro-apoptotic signal increase the production of ROS, the mitochondrial transducing pathways (Table 1). Prominent proteins matrix protein p53AIP1, which can dissipate the which translocate to mitochondria upon apoptosis mitochondrial tnansmembrane potential (DCm) induction include the pro-apoptotic members of the (Donald et al., 2001), and the Peutz-Jegher gene Bcl-2 family (Adams and Cory, 2001; Gross et al., product LKB1, a serine/threonine kinase translocating 1999; Kroemer, 1997; Martinou and Green, 2001; to mitochondria (Karuman et al., 2001). In addition, Vander Heiden and Thompson, 1999; Zamzami and p53 (and other transcription factors such as TR3), as Kroemer, 2001). Thus, Bax translocates from the well as different kinases (SEK1, JNK, PKC-d, c-Abl) cytosol to mitochondria in response to multiple reportedly translocate to mitochondria during early different apoptosis inducers. One mechanism triggering apoptosis (Aoki et al., 2002; Kumar et al., 2001; Li et Bax translocation is cytosolic alkalinization (which can al., 1999). All these proteins may induce MMP. be triggered by growth factor withdrawal) (Khaled et Divalent cations (in particular Ca2+), reactive oxygen al., 2001). Bax translocation is linked to a conforma- species (ROS), nitric oxide, and a number of different tional change, insertion into the mitochondrial (sphingo)lipid second messengers (in particular gang- membranes, and oligomerization. Bid is also normally lioside GD3) can also act on mitochondria to cause found in the cytosol and can be cleaved by proteases MMP (Bernardi et al., 2001). In view of this multitude (in particular by caspase-8 but also by calpain, of pro-apoptotic stimuli acting on mitochondria, it is cathepsin L, or granzyme B) to become truncated Bid not surprising that a number of different proteins and (t-Bid), a pro-apoptotic molecule. This reaction metabolites exert local mitochondrioprotective effects depends on the activation state of caspase-8, the (Table 2). This applies in particular to the anti- presence of specific caspase-8 inhibitors, as well as on apoptotic proteins from the Bcl-2 family (Bcl-2, Bcl- post-translational modifications of Bid. Phosphoryla- XL), certain enzymes involved in intermediate meta-

Oncogene Targeting mitochondria K-M Debatin et al 8789 Table 1 Protein factors and second messengers that permeabilize mitochondria Class of molecules Examples Putative target References

Sessile membrane proteins VDAC Bax, Bcl-2 Shimizu and Tsujimoto, 2000 ANT Bax, Bcl-2 Marzo et al., 1998) Bak VDAC, ANT Shimizu and Tsujimoto, 2000 BNIP3 Bcl-2/Bcl-XL Vande Velde et al., 2000 Siva-1 Bcl-XL Xue et al., 2002 Proteins translocating to mitochondria Pro-apoptotic Bcl-2 like proteins Bax ANT, VDAC Marzo et al., 1998; Shimizu et al., 2000a t-Bid Bax, Bak, cardiolipin Cheng et al., 2001; Wei et al., 2001 Bim Bcl-2 Cheng et al., 2001 p53-induced proteins Bax VDAC, ANT Vogelstein et al., 2000 BH-3-only proteins: NoxA, PUMA/Bbc3 Bax, Bak Han et al., 2001 Proline oxidase Locally generates ROS Donald et al., 2001 Ferredoxin reductase Locally generates ROS Hwang et al., 2001 p53AIP1 ? Oda et al., 2000b mtCLIC/CLIC4 Is a chloride channel Fernandez-Salas et al., 2002 p53-regulated protein LKB1 Serine/threonine kinase Karuman et al., 2001 Fission-regulatory proteins Dynamin-related protein 1 ? Frank et al., 2001 Transcription factors p53 Hsp70 Vogelstein et al., 2000 Nur77/TR3/NGFIB ? Li et al., 2000 Kinases JNK Bcl-2 Aoki et al., 2002 SEK1 JNK? Aoki et al., 2002 PKC-d ?Liet al., 1999 c-Abl ? Kumar et al., 2001 Caspases Caspase-2 ? Guo et al., 2002 Caspase-3 Bcl-XL Kirsch et al., 1999; Marzo et al., 1998 Cytotoxic granule components Granzyme B ? Barry and Bleackley, 2002 Granulysin Negatively charged lipids Pardo et al., 2001 Non-proteaceous factors Sphingolipid messengers Palmitate ANT de Pablo et al., 1999 Arachidonic acid ? Scorrano et al., 2001 Ganglioside GD3 ? Rippo et al., 2000 Divalent cations Ca2+ ANT, VDAC Gincel et al., 2001 Cd2+,Hg2+,Zn2+ ? Pro-oxidants Superoxide ANT, VDAC Madesh and Hajnoczky, 2001; Marzo et al., 1998 4-hydroxyhexenal ANT Vieira et al., 2001 Nitrix oxide ANT Vieira et al., 2001

Table 2 Protein factors and second messengers acting on mitochondria to prevent membrane permeabilization Class of molecules Examples Putative target References

Sessile membrane Bcl-2, Bcl-XL VDAC, ANT, BH3 proteins Kroemer and Reed, 2000 proteins Hexokinase II VDAC Pastorino et al., 2002 Creatine kinase ANT, VDAC Schlattner et al., 2001 PBR ANT, VDAC Stoebner et al., 2001 Matrix protein associated Cyclophilin D ANT ? Clarke et al., 2002; with PTPC Lin and Lechleiter, 2002 Antioxidant enzymes Mn-Superoxide dismutase Motoori et al., 2001 Phospholipide hydroperoxide Nomura et al., 1999 glutathione peroxidase Signal trasnducers cGMP-dependent ? Takuma et al., 2001 protein kinase (PKG) PKA Bad Harada et al., 1999 Metabolites Glucose Hexokinase II Beutner et al., 1998 ATP, ADP ANT Vieira et al., 2000; Zoratti and Szabo` , 1995 NADH+ VDAC Zoratti and Szabo` , 1995 Glutathione ? Zoratti and Szabo` , 1995 bolism (e.g. hexokinase II, which hinders Bax from How MMP coordinates multiple aspects of apopto- inserting into mitochondrial membranes), certain sis MMP variably aﬀects the outer (OM) and inner kinases (e.g. protein kinase A, which phosphorylates mitochondrial membranes (IM), respectively (Box 3). and inactivates Bad), anti-oxidant enzymes, as well as IM permeabilization is rather partial causing the to energy-rich metabolites (glucose, ATP, ADP, retention of most if not all matrix proteins, while NADH+), implying a major cross-talk between signal OM permeabilization leads to the release of soluble transduction, redox regulation and intermediate meta- apoptogenic proteins from the intermembrane space bolism at the mitochondrial level. (Table 3). Several among these proteins are involved in

Oncogene Targeting mitochondria K-M Debatin et al 8790

Box 3 Detection of mitochondrial alterations in apoptosis. There are several methods for detecting mitochondrial alterations that occur during apoptosis. (a) One strategy is to assay for translocation of intermembrane proteins. This can be achieved through in situ immunofluorescence staining with antibodies specific for cytochrome c, AIF, after Smac/DIABLO, after fixation and permeabilization of cells (see example 1). Another method is to transfect cells with a fusion protein that contains green fluorescent protein (GFP) fused into the C-terminus of cytochrome c or AIF. Translocation of such proteins to the mitochondrial membrane can be observed by time lapse videomicroscopy. (b) Another strategy is to monitor the inner mitochondrial (IM) membrane permeability. This can be done by labelling cells with DCm-sensitive dyes such as tetra- methylrhodamine ester or DiOC(6)3 (see example 2). A red-green shift in the emission spectrum of the DCm-sensitive, ratiometric fluorochrome JC-1 indicates a fall in the DCm. IM permeability can also be measured by monitoring the quenching of matrix calcein fluorescence after loading of cells with calcein plus Co2+.Co2+ quenches extra-mitochondrial calcein, yet does not penetrate into the matrix compartment as long as the IM membrane remains intact. Thus, the presence of a matix calcein signal indicates an impermeable IM and loss of the signal indicates IM permeabilization. (c) Changes in mitochondrial membrane lipid content can be detected by staining with the cardiolipin-specific dye nonyl acridine orange (NAO). (d) Ultrastructural analysis of cells with conventional transmission electron microscopy (two-dimensional vision) or computer-aided tomographic electron microscopy (three-dimensional reconstruction) is the most definitive method of detecting MMP.

Example 1: Redistribution of AIF from the mitochondrial intermembrane space to the nucleus during apoptosis. Normal or etoposide-treated apoptotic Hela cells were stained with an antibody speciﬁc for the resident mitochondrial protein Hsp60, an antibody speciﬁc for AIF, and Hoechst 33324 which stains chromatin.

Example 2: Analysis of the DCm in circulating tumour cells from a patient with acute lymphoblastic leukaemia (ALL) before (a) and after (b) treatment with prednisolon for 3 days (BFM protocol). Leukemia cells stained ex vivo with phycoerythrine (PE)-labelled anti-CD19 and with the DCm-sensitive dye DiOC(6)3 were analysed immediately after isolation from the patient.

Oncogene Targeting mitochondria K-M Debatin et al 8791 Table 3 Consequences of mitochondrial membrane permeabilization Type Examples Mode of action References

Apoptogenic proteins Cytochrome c Activates apoptosome Budijardjo et al., 1999 released from mitochondria Hsp10 Co-activates apoptosome Samali et al., 1999 Pro-caspase-2 Caspase Susin et al., 1999a Pro-caspase-3 Caspase Mannick et al., 2001 Pro-caspase-8 Caspase Quin et al., 2001 Pro-caspase-9 Caspase Susin et al., 1999a ARTS Translocates to nucleus and activates caspase 3 Larisch et al., 2000 Smac/DIABLO Inhibits IAP and thus indirectly activates caspases Du et al., 2000; Verhagen et al., 2000 Htra2/Omi Inbibits IAP and has serine protease activity Suzuki et al., 2001 AIF NADH reductase and activates DNAse in nucleus Susin et al., 1999b Endonuclease G Translocates to nucleus Li et al., 2001 IkBa Inhibits NF-kB Bottero et al., 2001 Metabolic alterations ROS generation Uncoupling and inhibited electron transfer Zorov et al., 2000 due to cytochrome c release 2+ Ca release Loss of Dcm Ichas et al., 1997 the activation of caspases. Cytochrome c, a heme transcription factor NF-kB (Bottero et al., 2001). protein responsible for electron transfer between Altogether, it appears that the mitochondrion is a sort complexes III and IV of the respiratory chain, is of poison cupboard replete with potentially lethal usually confined to mitochondria where it undergoes proteins which, after MMP, translocate to different electrostatic interactions with the outer side of IM. cellular localizations and participate in cellular demise. Cytochrome c release from mitochondria does not only As a consequence of apoptotic MMP, mitochondria rely on MMP but also requires its detachment from the sooner or later lose their vital metabolic and redox IM-specific lipid, cardiolipin (Ott et al., 2002). Once functions, their mitochondrial inner transmembrane released to the cytosol, cytochrome c interacts with potential (DCm), as well as their capacity to act Apaf-1, ATP, and pro-caspase-9 to form a molecular as Ca2+ storage organelles. The consequent depletion caspase activation complex, the apoptosome, which of ATP and oxidative reactions contribute to cell catalyzes the proteolytic maturation of caspase-3, the death. principal effector caspase (Budijardjo et al., 1999). Depending on the cell type, mitochondria contain MMP as an apoptotic checkpoint One particularly variable amounts of pro-caspases (e.g. pro-caspases-2, intriguing feature of MMP consists in the fact that -3, and -9), which are released upon apoptosis several of its consequences can stimulate MMP, induction (Susin et al., 1999a). One of these pro- implying a positive amplification loop. Thus, Ca2+ caspases, pro-caspase-3, is normally S-nitrosylated on can trigger mitochondria to release Ca2+ (Ichas et al., its catalytic Cys residue, when present in mitochondria. 1997). ROS induce mitochondria to produce ROS Upon apoptosis induction, pro-caspase-3 is de-nitrosy- (Zorov et al., 2000). Caspases activated as a conse- lated and released (Mannick et al., 2001). Smac/ quence of MMP digest the latent MMP inducers Bid DIABLO is another interesting pro-apoptotic protein and Bad, thereby activating them, and degrade the normally confined to mitochondria. Upon import of MMP inhibitors Bcl-2 and Bcl-XL, converting them the Smac/DIABLO precursor into the intermembrane into MMP inducers (Basanez et al., 2001). AIF space, the N-terminal mitochondrial localization released from mitochondria can stimulate MMP (Susin sequence (MLS) is removed, thereby unmasking a et al., 1999b). These features may explain why MMP new N-terminus (AVPIAQ...). Once released, the N- occurs mostly in an all-or-nothing fashion, as a terminus of Smac/DIABLO interacts with and neutra- consequence of a rapid irreversible process that, in lizes the protein XIAP, an inhibitor of caspases-3 and - large cells, can proceed in a wave-like fashion. 9. Thus Smac/DIABLO facilitates caspase activation Altogether, it thus appears that MMP is induced by (Du et al., 2000; Verhagen et al., 2000). A similar N- a cornucopia of different apoptosis inducers, that it has terminus (AVPSPP....) found in the intermembrane multiple lethal consequences and that it is closely protein Htra2/Omi also binds and neutralizes to XIAP. linked to or, perhaps, identical with the point-of-no- In addition, Htra2/Omi is a serine protease and can return of the cell death process. MMP is rate-limiting induce cell death through its catalytic properties for cell death induction in most situations, even when (Suzuki et al., 2001). AIF is an intermembrane the primary cell death inducer affects plasma flavoprotein with NADH+ reductase activity which, membranes (e.g. in type II cells in death receptor after apoptosis induction, translocates to the nucleus signalling) (Scaffidi et al., 1998), the nucleus (e.g. in where it can trigger peripheral chromatin condensation p53-dependent stress pathways) (Vogelstein et al., through a yet unknown mechanism (Susin et al., 2000) or other organelles including the endoplasmic 1999b). Another mitochondrial protein which translo- reticulum (Boya et al., 2002; Wei et al., 2001) and cates to the nucleus is endonuclease G (Li et al., 2001). lysosomes (Ferri and Kroemer, 2001; Stoka et al., Finally, it may be speculated that the release of IkBa 2001). MMP appears to be the (or one of the) may contribute to the inactivation of the cytoprotective checkpoint(s) determining entry into apoptosis.

Oncogene Targeting mitochondria K-M Debatin et al 8792 Bax has been proposed to interact with a molecular Mechanisms and regulation of MMP complex including the putative CsA target cyclophilin In view of the role of MMP for apoptosis control, it is D – a mitochondrion-specific peptidyl prolyl isomer- of the utmost importance to understand its regulation ase. This interaction would explain why, at least in at the molecular level. Although many of the molecular some cases, CsA can inhibit Bax-induced membrane actors regulating MMP are known (Tables 1 and 2), permeabilization (Ju¨ rgensmeier et al., 1998; Marzo et the exact mechanisms of MMP are highly controversial al., 1998; Zamzami and Kroemer, 2001). (Martinou and Green, 2001; Zamzami and Kroemer, 2001). This applies in particular to the question of Other Bcl-2 family members t-Bid is thought to induce which mitochondrial proteins are actually functioning apoptosis through the induction of a conformational as non-specific pores and which are mere pore change in Bax (or Bak), thereby triggering its full regulators. Even worse, there is an ongoing controversy membrane insertion and oligomerization (Eskes et al., whether and to what extent mitochondrial ultrastruc- 2000). Accordingly, Bax7/7Bak7/7 MEF are rela- ture is altered during apoptosis. Moreover, the tively resistant to t-Bid-induced cytochrome c release temporal and functional order OM and IM permeabi- (Wei et al., 2001; Zong et al., 2001). Nevertheless, it lize may well depend on the primary apoptosis inducer. has also been suggested that t-Bid forms membrane- For the purpose of this review, we will briefly discuss inserted homotrimers not interacting with other the most important MMP-regulatory molecules (Figure mitochondrial proteins (Grinberg et al., 2002). t-Bid 1). can form channels in planar lipid bilayers (Schendel et al., 1999) and destabilize them (Kudla et al., 2000). The Bax and Bak Mouse embryonic fibroblasts (MEF) question of whether t-Bid’s effects on mitochondria lacking both Bax and Bak, two tumour suppressor depend on interactions with sessile mitochondrial gene products, are resistant to a plethora of different proteins, including a CsA target, has been controversial apoptotic stimuli including t-Bid, Bim, Bad, Noxa, (Shimizu and Tsujimoto, 2000; Zamzami et al., 2000). staurosporine, ultraviolet radiation, growth factor A recent study indicates that t-Bid-mediated permea- deprivation, etoposide, oxygen deprivation, and the bilization occurs in a two-step process, the first step endoplasmic reticulum stress stimuli thapsigargin and (CsA-independent) leading to only a partial (*15%) tunicamycin (Cheng et al., 2001; McClintock et al., cytochrome c release, and the second step (CsA- 2002; Wei et al., 2001; Zong et al., 2001). Such cells are inhibited) causing total cytochrome c release and a more susceptible to Ras-mediated transformation than remodelling of mitochondrial ultrastructure (Scorrano control cells (Zong et al., 2001). However, when treated et al., 2002). Other pro-apoptotic proteins from the with granzyme B, Bax7/7Bak7/7 MEF undergo Bcl-2 family such as Bad, Bim, and Noxa are thought apoptosis and DCm loss without cytochrome c release, to induce apoptosis via an interaction with either Bax indicating the existence of an independent mitochon- or Bak and/or by generating stable complexes with drial pathway that may participate in cell death anti-apoptotic Bcl-2/Bcl-XL (Adams and Cory, 2001; induction (Thomas et al., 2001). Overexpression of Cheng et al., 2001). Bax and Bak suffices to kill cells. Bak is normally associated with OM and fully inserts into OM, Voltage-dependent anion channel (VDAC) VDAC, the accompanied by a conformational change, upon most abundant mitochondrial OM protein, has been apoptosis induction. In contrast to Bak, Bax is mostly reported to be required for apoptosis induction (Figure found in the cytosol of healthy cells. Upon apoptosis 1b). VDAC is normally responsible for the transport of induction, Bax translocates from the cytosol to OM metabolites between the cytosol and the mitochondrial where it inserts and oligomerizes (Figure 1a). Recom- intermembrane space. Microinjection of VDAC-specific binant Bax can permeabilize liposomal membranes and antibodies or of a VDAC-specific inhibitor (Ko¨ nigs’s form channels in planar lipid bilayers. The size of these polyanion) can inhibit apoptosis induced by micro- channels is reportedly sufficient to allow for the release injected Bax (Ferri et al., 2000), etoposide, paclitaxel, of proteins such as cytochrome c (Saito et al., 2000). and staurosporine treatment (Shimizu et al., 2001a). Moreover, Bax has also been proposed to lipid bilayers VDAC might be sufficient to permeabilize membranes (Basan˜ ez et al., 1999). The mechanisms of membrane since VDAC proteoliposomes release cytochrome c in permeabilization induced in isolated mitochondria the presence of ROS (Madesh and Hajnoczky, 2001). exposed to recombinant Bax depend on the oligomer- Bcl-2 and Bcl-XL have been reported to close the ization status of Bax. Bax monomers cause VDAC, both in liposomes and in plain lipid bilayers. permeabilization through a pathway that can be This effect can be mimicked by peptides corresponding inhibited by cyclosporin A (CsA), while Bax oligomers to the BH4 domains, which are present in Bcl-2 and trigger CsA-resistant cytochrome c release (Gogvadze Bcl-XL yet are absent from the pro-apoptotic proteins et al., 2001) (Figure 1a). These findings are important Bax, Bak, Bid, Bim and Noxa (Shimizu et al., 2000b) in view of a long standing polemic whether Bax- (Box 1). Bax or Bak (but not Bid) might interact with mediated mitochondrial membrane permeabilization VDAC to create composite channels sufficiently large relies on the capacity of Bax alone to oligomerize to release cytochrome c (Shimizu and Tsujimoto, and to form channels (Martinou and Green, 2001) or 2000). VDAC therefore appears to be the prime whether Bax has to interact with other proteins. Thus, cytochrome c release channel in the OM, and its

Oncogene Targeting mitochondria K-M Debatin et al 8793

Figure 1 Permeabilization of the mitochondrial inner membrane (IM) or outer membrane (OM). Inhibitors of permeabilization are denoted in red, inducers in green. (a) Bax (in the cytosol) translocates to mitochondria in response to diverse signals. The incorporation of monomeric Bax into mitochondria is guided by VDAC (and presumably VDAC-interacting proteins such as ANT and cyclophilin D), explaining why the ANT ligand bongkrekate and cyclosporin A (CsA) can inhibit Bax effects. Hexo- kinase, which binds to VDAC, also blocks the (presumably VDAC-mediated) interaction of Bax with OM. Aided by BH3-only members of the Bcl-2 family, Bax oligomerization creates protein-permeant conduits in OM. (b) The voltage dependent anion channel (VDAC) is normally responsible for the exchange of metabolites (e.g. ATP, ADP, NADH etc.) on the outer membrane. VDAC might be converted into a cytochrome c-permeant conduit when interacting with Bax (right) and this reaction can be inhibited by the BH4 domain of Bcl-2/BclXL. The voltage-dependent closure of VDAC, which is lethal (left), is inhibited by Bcl-2 by virtue of its ion channel function. (c) The adenine nucleotide translocator (ANT) normally acts as a ATP/ADP specific antiporter. During an early step of apoptosis, the translocase activity of ANT is inhibited and later ANT can convert into a non- specific pore. Pore formation is aided by interaction with cyclophilin D (CypD), Bax, as well as interactions with ANT ligands affecting its conformation. It is possible that other members from the mitochondrial carrier protein family fulfil a similar role as does ANT

activity might be suppressed by Bcl-2 or Bcl-XL,or to a closure of VDAC (Vander Heiden et al., 2000; increased by Bax or Bak. Vander Heiden et al., 2001a). Although the exact There is another completely diﬀerent mechanism reasons for VDAC closure have not been elucidated, it through which VDAC might participate in apoptosis has been suggested that an increased IM transmembrane induction. Upon growth factor withdrawal, the potential (DCm) would directly aﬀect OM, creating a exchange of metabolites (ATP, ADP, NADH, creatine local voltage gradient and closure of VDAC. This OM and creatine phosphate) on OM might be reduced, due DC causing closure of VDAC would be dissipated by

Oncogene Targeting mitochondria K-M Debatin et al 8794 Bcl-2 by virtue of its capacity to create ion-permeant that occurs because the surface area of the IM (with its channels (Vander Heiden et al., 2001a). Further studies folded christae) exceeds that of the OM. Alternatively, are required to determine whether these two hypotheses the conformation of ANT, which is modulated by regarding the contribution of VDAC to mitochondrial ANT ligands and CsA-inhibited interactions with permeabilization are compatible (Figure 1b). cyclophilin D, might indirectly impinge on the function of VDAC. Adenine nucleotide translocator (ANT) ANT, the Several pieces of data indicate that ANT is involved most abundant IM protein, is responsible for the in apoptosis induction. Pro-apoptotic molecules such exchange of ATP and APD. ANT interacts with as Ca2+, palmitate, thiol oxidants, nitric oxide (NO), VDAC and other proteins, such as cyclophilin D, to 4-hydroxynonenal, the protein Vpr encoded by HIV-1, form the permeability transition pore complex (PTPC) as well as a synthetic peptide derived from the BH3 at the contact sites between IM and OM (Brenner et domain of Bax, cause ANT (which normally is a al., 2000; Crompton, 2000; Marzo et al., 1998) (Figure strictly specific antiporter: a transport protein specifi- 2). ANT forms pores (Figure 1c) that have been cally exchanging two molecules, in this case ATP and proposed to mediate the IM permeabilization, osmotic ADP, at a 1 : 1 stoichiometry) to form a non-specific matrix swelling, and consequent selective OM rupture pore (Vieira et al., 2001, 2002). When reconstituted

Figure 2 Hypothetical molecular architecture of the PTPC. The PTPC involves several transmembrane proteins (adenine nucleotide translocase, ANT; voltage dependent anion channel, VDAC; peripheral benzodiazepin receptor, PBR), members of the Bax/Bcl-2 family), as well as several associated proteins: hexokinase (HK), mitochondrial creatine kinase (mtCK), and the peptidyl-prolyl isomerase cyclophilin D (CypD). The reactions catalyzed by mtCK and HK are depicted. Pro-apoptotic second messengers, metabolites or processes labelled in green facilitate PTPC opening; agents in red inhibit pore opening. Proteins or peptides carrying the Bcl-2 homology region-3 (BH-3) motif may act on either Bax or Bcl-2 (or their homologues) in the outer mitochondrial membrane. Ca2+ may act on VDAC, ANT, or ANT-associated cardiolipin (CL) molecules. Note that the exact stoichiometry and composition of the PTPC are still elusive

Oncogene Targeting mitochondria K-M Debatin et al 8795 into liposomes or into planar lipid bilayers, ANT can also due to the membrane presence of acidic form non-specific channels in response to pro-apoptotic phospholipids. It is important to note that the ligands such as atractryloside (a pro-apoptotic ANT composition of mitochondrial lipids changes during inhibitor binding to the external site of ANT and apoptosis, with cardiolipin oxidation as a prominent favouring its conversion into a non-specific pore) and feature (Matsko et al., 2001). A local phospholipase A2 Vpr (a protein encoded by the HIV-1 genome). (Williams and Gottlieb, 2002) as well as a sphingo- ANT channel formation is also enhanced by Bax, but myelinase (Birbes et al., 2001) may increase the inhibited by Bcl-2 (Brenner et al., 2000; Jacotot et al., mitochondrial concentration of two potentially MMP- 2001; Marzo et al., 1998). The ANT ligand bongkrekic inducing agents, namely arachidonic acid and cera- acid (BA) inhibits cell death in numerous pathways mide, respectively. It remains an open question whether (Zamzami and Kroemer, 2001), and CsA, which can ceramide-induced lipid rafts in OM facilitate the prevent mitochondrial permeabilization in some circum- insertion and oligomerization of Bax-like proteins, in stances, has also been proposed to exert its effect analogy to what has been reported for CD95 in the indirectly via ANT, by inhibiting its interaction with plasma membrane (Gajate and Mollinedo, 2001). the matrix protein cyclophilin D (Crompton, 2000). Changes in mitochondrial ultrastructure As outlined Anti-apoptotic mitochondrial proteins It is well estab- above, the molecular control of MMP is complex and lished that the anti-apoptotic oncoproteins Bcl-2 and several distinct modes of MMP induction may exist. Bcl-XL are mitochondrial membrane proteins that act One particularly interesting feature that emerged from as barriers to membrane permeabilization (Adams and recent studies concerns mitochondrial ultrastructure. Cory, 2001; Kroemer, 1997; Martinou and Green, Thus, apoptosis is often associated with a disintegra- 2001; Zamzami and Kroemer, 2001). It is controversial, tion of the mitochondrial reticulum into multiple however, how they stabilize the membrane – it could punctiform organelles, a process which may require be by neutralizing the function of pro-apoptotic Bcl-2 mitochondrial fission (Frank et al., 2001). In addition, family proteins, by physically or functionally interact- it appears that, within mitochondria, the junctions ing with VDAC, by neutralizing ANT channel activity, between cristae and the intermembrane space are or a combination of all of these. Some PTPC proteins opened (Scorrano et al., 2002), at least in some (Figure 2) have also been reported to inhibit apoptosis. paradigms of apoptosis induction. Thus linking the One of these is the peripheral benzodiazepine receptor exact topology of the multiple pore-forming and pore- (PBR) – an 18-kDa mitochondrial membrane protein regulatory molecules (proteins+lipids) to the ultra- that interacts with VDAC and ANT. PBR over- structural architecture of MMP regulation constitutes a expression has been shown to be cytoprotective challenge for future investigation. (Stoebner et al., 2001). Hexokinase II, a mitochondrial protein that interacts with VDAC and catalyzes the Mitochondrial alterations in cancer first step of glycolysis, has also been shown to inhibit apoptosis. Hexokinase II transcription and mitochon- Do cancer cells also possess mitochondrial alterations drial localization are both induced by the Akt- that contribute to apoptosis resistance? A comparison mediated anti-apoptotic signalling pathway (Gottlob of mitochondria isolated from tumour cells and normal et al., 2001; Vander Heiden et al., 2001b). The cells revealed that 2-chlorodeoxyadenosine-resistant enzymatic activity of hexokinase II might somehow leukaemia cells harbour mitochondria that are intrinsi- inhibit apoptosis (Gottlob et al., 2001), and, alterna- cally resistant to Ca2+-induced MMP (Chandra et al., tively, this enzyme has also been proposed to impede 2002). In fact, it has been known since the 1930’s that VDAC-mediated membrane insertion of Bax (Pastor- cancer cell metabolism is different from that of normal ino et al., 2002). cells. Cancer cells exhibit enhanced glycolytic ATP generation and decreased respiratory phosphorylation, Mitochondrial lipids The lipid composition of the IM even under normal oxygen tension – a phenomenon is very different from that of other intracellular known as the Warburg effect (Warburg, 1956). This membranes. The IM is essentially cholesterol-free, property is clinically important because it is taken and is the only eukaryotic membrane to contain advantage of in fluorine-18-deoxyglucose-positron cardiolipin. IM contains a relatively high portion emission tomography to detect tumours in vivo. The of negatively charged phospholipids such as mechanism of tumour-associated aerobic glycolysis has phosphatidylserine. These characteristics might deter- not been fully elucidated, although it has been linked, mine the preferential insertion of certain proteins into at least speculatively, to an increase in hexokinase II mitochondrial membranes. For example, t-Bid prefer- expression (Golshani-Hebroni and Bessman, 1997). In entially associates with cardiolipin-containing addition, tumour cells may have a higher DCm than membranes (Lutter et al., 2000), which might explain control cells, although the biochemical basis of this why t-Bid is enriched in the contact site between IM phenomenon remains elusive (Chen, 1988). These and OM. Cationic ampholytes – molecules that metabolic alterations might be linked to an overall contain both a hydrophobic and a hydrophilic cationic change in the composition and/or regulation of the moiety – might also be expected to preferentially PTPC which would prevent mitochondrial apoptosis. redistribute to IM, driven not only by the DCm, but Indeed, cancer cells reportedly overexpress some

Oncogene Targeting mitochondria K-M Debatin et al 8796 apoptosis-inhibitory PTPC components, such as hexo- only methodology which proves that a given drug kinase II and PBR) (Casellas et al., 2002; Smith, 2000). exerts direct effects on mitochondria, is to show that They also undergo a shift in the ANT isoenzyme the compounds actually induces MMP in a cell-free expression pattern, reducing expression of the poten- system, when added to purified mitochondria. Using tially apoptogenic ANT-1 isoform and increasing the such a system, a number of conventional chemother- expression of ANT-2 (Giraud et al., 1998). apeutics have already been shown to induce cell death Cancer cells also increase expression of anti-apoptotic by permeabilizing mitochondrial membranes. These Bcl-2-like proteins and reduce expression of pro- include 2-chloro-2’-deoxyadenosine, etoposide, and apoptotic proteins from the Bcl-2 family. Bax expression, paclitaxel (Genini et al., 2000; Kidd et al., 2002; for example, is downregulated as a consequence of Robertson et al., 2000). Paclitaxel has been suggested genomic instability in colon carcinoma (Ionov et al., to act on VDAC-associated tubulin (Carre et al., 2002). 2000). A correlation between the expression levels of High doses of these drugs, however, are often required various apoptosis signalling proteins and clinical to obtain such effects. A number of other mitochon- outcome has been well established (Kroemer, 1997). drion-permeabilizing drugs are being developed (Table The oncogenic property of some Bcl-2-like proteins, and 4). What are the physicochemical and pharmacological the tumour-suppressive properties of Bcl-2 antagonists properties of such mitochondrion-targeted agents? has been validated in animal models (Joza et al., 2002). In addition, a number of alterations in the expression Bcl-2-targeted drugs A number of reagents have been level of proteins closely linked to mitochondrial developed to target Bcl-2 family members, and might apoptosis may be relevant to cancer biology. This applies be able to induce apoptosis in cancer cells. For to pro-apoptotic proteins that may translocate to example, a recombinant chimeric protein that contains mitochondrial membranes (e.g. p53, c-Abl, Table 1), as interleukin-2 fused to Bax has been shown to well as to proteins which modulate the function of selectively kill interleukin-2 receptor-bearing cells in proteins released from mitochondria. For instance, the vitro (Aqeilan et al., 1999). Similarly, Bak can be gene encoding Apaf-1, whose expression is required for targeted to the Fce receptor by fusing it to the constant cytochrome c to activate caspase-9, is frequently region of the immunoglobulin E molecule (Belostotsky hypermethylated and therefore silenced in melanoma and Lorberboum-Galski, 2001). This fusion protein has cells (Soengas et al., 2001). Hsp70, which inactivates been shown to selectively induce apoptosis in Fce Apaf-1 and/or AIF (Ravagnan et al., 2001) is over- receptor-expressing cells in vitro, such as mast cells and expressed in breast cancer. basophils. These findings illustrate the feasibility of The mitochondrial genome (mtDNA) is small targeting pro-apoptotic proteins to selected cellular (16.6 kb) and characterized by cytoplasmic transmis- populations, although they have not yet been applied sion, so mutations can be transmitted to only a to the experimental treatment of cancer. An antisense fraction of daughter cells after division (heteroplasmy) oligonucleotide designed to down-regulate the expres- or to all of them (homoplasmy). Little is known about sion of Bcl-2 (Genasense) has been shown to sensitize the association between mtDNA mutations and cancer. metastatic melanoma cells to apoptosis induction by Whether mtDNA mutations or deletions accumulate dacarbazine in a phase I/II trial (Jansen et al., 2000). due to active selection or due to random accumulation Genasense is also being evaluated in a phase I study of one particular mitochondrial genome in rapidly for the treatment of metastatic androgen-independent dividing clones is a matter of debate (Coller et al., prostate cancer (Chi et al., 2001). Such anti-sense 2001). Normal cells that lack mtDNA gene expression oligonucleotides, which are relatively non-toxic to or carry mtDNA mutations may become hypersensitive normal cells, might be particularly useful for treatment to apoptosis induction (Danielson et al., 2002; Wang et of tumours that express high levels of Bcl-2. al., 2001). However, in several cancer cell lines, mtDNA mutations reportedly reduce the efficacy of BH3 mimetics Several small molecules that disrupt chemotherapeutic agents, in particular of those agents the interaction between Bcl-XL and Bak- have been which elicit the production of ROS by the respiratory identified and shown to occupy the BH3-binding chain (Hail et al., 2001; Singh et al., 1999). It is pocket (Box 1) of Bcl-XL. These molecules block the therefore possible that mtDNA mutations contribute to BH3-domain-mediated heterodimerization between chemotherapy resistance. Bcl-2 family members to induce apoptosis, presumably by acting as BH3 mimetics (Degterev et al., 2001). The small molecule HA-14-1, which binds to the surface Cytotoxic drugs that target the mitochondria pocket of Bcl-2 can induce apoptosis in a similar The mere description that a given cytotoxic compound manner (Wang et al., 2000). Antimycin A, a compound induces signs of MMP when added to intact cells does that inhibits complex III of the respiratory chain, and not yield any valuable information on its mode of its methoxy derivative (which does not block complex action. Indeed, as discussed above, most, if not all, III) also bind to the BH3-binding hydrophobic groove apoptosis inducers have such an effect, which mostly is of Bcl-2 and Bcl-XL, thereby permeabilizing purified mediated in an indirect fashion, through eliciting mitochondria expressing Bcl-XL (Tzung et al., 2001). different (‘intrinsic’) stress pathways and/or the death Peptides that correspond to the BH3 domains of Bax receptor/ligand-dependent (‘extrinsic’) pathway. The (BaxBH3) or Bcl-2 (Bcl2BH3) permeabilize mitochon-

Oncogene Targeting mitochondria K-M Debatin et al 8797 Table 4 Cytotoxic drugs capable of inducing MMP in isolated mitochondria MMP inhibited by Structural features anti- Cationic Steroid PBR ANT BH3 Agent Anti-tumour activity CsA BA Bcl-2 oxidants Rho8 lipophil like ligand effects like References

Alloxan Cytotoxic for pancreatic + + Sakurai et al., 2001 b-cells (diabetogenic) Arsenite Acute promyelocytic + + + + + Belzacq et al., 2001a; leukaemia Multiple myeloma Larochette et al., 1999 Antimycin A and its Particularly toxic for Bcl-XL + + Tzung et al., 2001 methoxy derivative overexpressing cells, in vitro Avicins In vitro + Haridas et al., 2001 Bax- BH3-domain- In vitro + – + + Vieira et al., 2002 derived peptide Betulinic acid Neuroectodermal tumours + + + + Fulda et al., 1998a; in vitro and in mice Fulda et al., 1998b Benzyl isothiocyanate Chemoprevention Nakamura et al., 2002 Butylated Chemoprevention + Yu et al., 2000 hydroxylanisole Butyrate Chemoprevention + + + Jan et al., 2002 CD437 In vitro + + + + + + + Belzacq et al., 2001a; Marchetti et al., 1999 2-Chloro-2’deoxy- Hairy cell leukaemia + ? Genini et al., 2000 adenosine Chronic lymphocytic leukaemia Ciprofloxacin In vitro + + Aranha et al., 2002 Etoposide Multiple neoplasias + + Robertson et al., 2000 FTY720 Immunosuppressive + + + + Nagahara et al., 2000 Hyperporfin In vitro Schempp et al., 2002 (KLAKKLAK)2 In vivo (mice) + Ellerby et al., 1999 peptide Lonidamine Ovarian carcinoma + + + + Ravagnan et al., 1999 MT-21 In vitro – + + Machida et al., 2002; Watabe et al., 2000 Paclitaxel Multiple neoplasias + + Kidd et al., 2002 Photoactivated In vitro + Pervaiz et al., 1999 merocyanine 540 C1 PK11195 In vitro and in animals + – + Banker et al., 2002; Decaudin et al., 2002; Hirsch et al., 1998 Resveratrol Chemoprevention + + + Tinhofer et al., 2001 Verteporfin+light Photodynamic therapy + + + + + Belzacq et al., 2001b Vpr In vitro (lentivirus or + + (+) + + Jacotot et al., 2001; adenovirus delivery) Muthumani et al., 2002 dria in a CsA-inhibitable fashion. When fused to the but when used in combination with etoposide, Atennapedia plasma membrane translocation domain cisplatin, arsenite, lonidamine, anti-CD95 antibody, (Ant), which facilitates cellular uptake, BaxBH3Ant dexamethasone or ceramide, these agents were cyto- and Bcl2BH3Ant induce mitochondrial permeabiliza- toxic and neutralized the anti-apoptotic potential of tion via a mechanism that is not inhibited by Bcl-2 (Banker et al., 2002; Decaudin et al., 2002; overexpressed Bcl-2 or Bcl-XL (Vieira et al., 2002). Hirsch et al., 1998). Agents such as these might be used to overcome Whether the apoptogenic or chemosensitizing effects multidrug resistance in cancer cells that overexpress of PBR ligands are truly due to an effect on the PBR is Bcl-2-like anti-apoptotic proteins. a matter of debate (Banker et al., 2002). The doses of PBR ligands required to obtain such cytotoxic effects PBR ligands Synthetic ligands of the PBR (such as are several orders of magnitude higher than the Kd of the isoquinoline carboxamide PK 11195 or the the high affinity PBR. Hence, PBR ligands either act benzodiazepine 4’-chlorodiazepam, Ro5-4864) can on a low affinity binding site present on mitochondria induce apoptosis or sensitize cells to apoptosis or via a completely different mechanism. Indeed, the induction, thereby overcoming the cytoprotective pro-apoptotic action of PK11195 might involve the effects of Bcl-2 or Bcl-XL. This chemosensitizing effect induction of ROS generation (Banker et al., 2002). It has been demonstrated in several tumour cell lines, in will be important to resolve the issue as to whether vitro (Banker et al., 2002; Decaudin et al., 2002; Hirsch expression of high PBR levels confers a higher et al., 1998), and also in SCID mice transplanted with susceptibility to PBR ligand-mediated chemosensitiza- human tumour cells (Decaudin et al., 2002). In these tion. If this was the case, then PBR ligands might be models, treatment with PBR ligands (such as PK 11195 particularly useful for the treatment of PBR over- or Ro5-4864) could not induce cell death on their own; expressing breast cancers (Casellas et al., 2002).

Oncogene Targeting mitochondria K-M Debatin et al 8798 perturb the lipid composition of mitochondrial ANT ligands membranes, and induce the opening of the PTPC. A number of different reagents are able to Betulinic acid has the interesting (and unexplained) permeabilize ANT-containing proteoliposomes, but ability to selectively kill neuroectodermal tumour cells not plain liposomes, so they might be used to including neuroblastoma, medulloblastoma, glioblasto- specifically permeabilize mitochondrial membranes. ma and Ewing sarcoma cells (Fulda et al., 1998a,b; These include cytotoxic drugs such as lonidamine, 2001; Fulda and Debatin, 2000). Isolated mitochon- arsenite and CD437) (Belzacq et al., 2001a), a peptide dria from different cell types, including from normal corresponding to the BH3 domain of Bax (Vieira et mouse liver, are permeabilized by betulinic acid, and al., 2002), and the HIV-1 protein Vpr (Jacotot et al., this effect is prevented by CsA, the ANT ligand 2001). All these agents induce channel formation bongkrekate, as well as by Bcl-2 overexpression when ANT is reconstituted into planar lipid bilayers, (Fulda et al., 1998b). It is unclear, however, through but it is not known whether they act directly on which receptor (if any) betulinic acid acts on ANT or on the ANT/lipid interface. They do, mitochondria and how the relative cell type specificity however, appear to induce a conformational change of betulinic acid can be explained. Nonetheless, it in ANT, since ANT proteoliposome permeabilization appears that derivatives of betulinic acid endowed can be suppressed by the natural ANT ligands ATP with an enhanced in vivo half life might be useful in and ADP (Belzacq et al., 2001a). Whether the ATP treating neuroectodermal tumours (Debatin et al., analogue 2-chloro-2’-deoxyadenosine trisphosphate unpublished). (generated from 2-chloro-deoxyadenosine by the action of the mitochondrial deoxyguanosine kinase) Cationic ampholytes The specific physicochemical also acts on ANT is a matter of speculation. The properties of the IM and the net deficit of positive bisphonate clodronate, which kills osteoclasts and is charges on the inner side of the IM results in the DCm- used for the palliative treatment of bone metastases, driven accumulation of cationic lipophils in the apparently is transformed into an ATP analogue mitochondrial matrix. Following the Nernst equation, which targets the ANT (Lehenkari et al., 2002). MT- positively charged agents sufficiently lipophilic to insert 21, an ANT inhibitot, has been shown to dissociate in or to cross cellular membranes may be expected to the ANT-cylophilin complex in vitro and to induce enrich in IM and/or in the mitochondrial matrix. In cytochrome c in isolated mitochondria in a reaction fact, it might be expected that any compound with that is inhibited by bongkrekic acid (Machida et al., mild detergent-like properties would permeabilize 2002). Lonidamine has been shown to revert the mitochondria if it reached high enough concentrations resistance to cisplatin and to potentiate cisplatin in and around the IM. activity in experimental models. This drug is under- This molecular mode of action might apply to 2- going clinical phase II trials for the treatment of amino-2- [2-(4-octylphenyl)ethyl] propane-1,3-diol ovarian carcinoma (de Lena et al., 2001). Lonidamine hydrochloride (FTY720), an immunosuppressive agent. kills a diverse range of tumour cells in vitro (Angioli MKT-077, a cationic rhodacyanine dye that is et al., 1997; Belzacq et al., 2001a). However, the anti- selectively toxic to carcinoma cells in vitro and in vivo apoptotic proteins of the Bcl-2 family do confer a (Modica-Napolitano et al., 1997) is now undergoing relative resistance to lonidamine that can be over- clinical trials (Britten et al., 2000). Another cationic come by addition of PBR ligands (Miccoli et al., ampholyte is the peptide DKLAKLAKKLAKLAK or 1998b; Ravagnan et al., 1999). Moreover, lonidamine (KLAKLAK)2 (K=lysine, L=leucine, A=alanine) (and arsenite) can bypass the resistance of TPA- which directly permeabilizes mitochondria. D(KLAK- differentiated leukaemic cells to apoptosis in vitro LAK)2 can be fused to targeting peptides that interact (Sordet et al., 2001), underscoring that direct with surface receptors expressed on angiogenic targeting of mitochondria may be useful for the endothelial cells in tumours (Ellerby et al., 1999) or eradication of tumour cells. On theoretical grounds it on prostate vasculature (Arap et al., 2002a). These is possible that ANT inhibitors might exhibit a fusion peptides are translocated into the cytoplasm, selective toxicity on particular cell populations, and induce apoptosis via mitochondrial permeabiliza- depending on the expression of the ANT isoform tion (Ellerby et al., 1999), specifically in the targeted and/or the mode of ANT operation (which is cell population. Thus, a prostate-targeted D(KLAK- reversed in cells having respiratory defects). Thus, it LAK)2 peptide might be used to target prostate might be conceivable to create ANT-specific agents tumours in mice (Arap et al., 2002a). Treatment of that would selectively kill ANT2-overexpressing mice with this peptide postponed the development of cancer cells or tumour cells having accumulated cancer in a transgenic mouse model spontaneously mutations in the mitochondrial genome. developing prostate adenocarcinoma (Arap et al., 2002a). Attempts to create peptides that would home Steroid analogues A number of agents that act on to specific vascular beds are on the way in patients mitochondria possess a steroid-like core structure. (Arap et al., 2002b), and it is well possible that the This applies to avicins, betulinic acid, CD437 (a strategy of targeting hybrid molecules, first to surface retinoid analogue), and the bile acid glycochenode- receptors and then to mitochondria, will become soxycholic acid. As a possibility, such agents might applicable to a variety of different tumours.

Oncogene Targeting mitochondria K-M Debatin et al 8799 Redox-active compounds and photosensitizing agents apoptotic cells are rapidly removed by phagocytosis. Agents that induce oxidative reactions can also be For example, in HIV-1-infected patients, a portion of used to permeabilize the mitochondrial membrane. circulating T lymphocytes exhibit a decreased DCm These agents include arsenite (which depletes that is detectable ex vivo, when phosphotidylserine glutathione), menadione (which undergoes futile redox exposure, caspase activation or DNA fragmentation cycles on the respiratory chain), as well as the thiol- have not yet occurred (Macho et al., 1995). Similarly, crosslinking agent diamide, which is able to overcome in pediatric patients receiving chemotherapy for the cytoprotection by Bcl-2 and to cause thiol oxidation of treatment of solid tumours, depletion of circulating ANT (Costantini et al., 2000a). Arsenite, which CD4+ and CD8+ lymphocytes is preceded by a loss of permeabilizes purified mitochondria in vitro (Laroch- DCm, detectable ex vivo (Stahnke et al., 2001). Assays ette et al., 1999), is used to treat acute pro- to measure mitochondrial membrane permeability myelomonocytic leukaemia and multiple myeloma. might therefore be used to more accurately monitor However, alternative mechanisms for the mode of the effects of chemotherapy. action of arsenite such as lipid raft-linked signalling Because MMP is the ‘point of no return’ in the have been proposed (Hossain et al., 2000). apoptotic pathway, agents that induce it might be able Photosensitizing agents that act on mitochondria – to circumvent chemotherapy resistance caused by some of which interact with PBR – generate ROS mutations in pre-mitochondrial signal-transducing when exposed to light (Kessel et al., 2001). Locally machinery, such as the tumour suppressors p53 or generated ROS might then act on mitochondrial PTPC PTEN. Accordingly, agents such as arsenite, betulinic components including VDAC and ANT causing the acid, CD437, lonidamine and Vpr induce cell death formation of protein-permeable conduits (in the case of independently of the p53 status, via a pathway that is VDAC) (Madesh and Hajnoczky, 2001) or non-specific not affected by caspase inhibitors (Fulda et al., 2001; ion channels (in the case of ANT) (Belzacq et al., Huang et al., 1999; Jacotot et al., 2001; Larochette et 2001b). The photosensitizer hypericin reportedly al., 1999; Marchetti et al., 1999; Ravagnan et al., detaches hexokinase from mitochondria (Miccoli et 1999; Sun et al., 1999). Moreover, several mitochon- al., 1988a), whereas the phthalocyanine photosensitizer drion-permeabilizing agents (in particular PBR ligands Pc 4 causes the photochemical destruction of Bcl-2 and BH3 mimetics) can induce apoptosis irrespective (Xue et al., 2001). Altogether, these data suggest that of the expression level of Bcl-2 (Banker et al., 2002; photosensitizing agents can target proteins contained in Decaudin et al., 2002; Hirsch et al., 1998; Tzung et the PTPC. al., 2001; Vieira et al., 2002). Given that the Photodynamic therapy employing drugs that target composition and function of the PTPC from cancer mitochondria are likely to have high specificity for cells differs from that of control, determining a cancer cells but few side effects. This is particularly relatively high DCm as well as the Warburg effect, important for the treatment of surface cancers as well even non-specific interventions on the PTPC (and as for tumours that can be attained by the endoscopic mitochondria at large) could have selective effects on delivery of light. transformed cells. Nonetheless, it will be a challenge to develop drugs with a strong mitochondrion-permeabilizing potential Clinical relevance of mitochondrial apoptotis that selectively target tumour cells. Such targeting In vitro, all chemotherapeutic agents have been shown might be possible, however, by generating hybrid to induce apoptosis by causing mitochondrial permea- molecules composed of a mitochondriotoxic moiety bilization – characterized by the redistribution of fused to ligands for tumour cell surface proteins. As mitochondrial intermembrane proteins (such as AIF or described above, this has been achieved, in principle, cytochrome c) and DCm collapse (Box 3). In vivo for the peptide D(KLAKLAK)2 which, when fused to studies of apoptosis induction in experimental models suitable ligands, specifically binds to target cells, or in patients undergoing therapy, however, have been becomes internalized and then acts on mitochondria. limited to histological evaluation, providing a static In the future, mitochondrion-targeted reagents might picture of end stage apoptosis, rather than a real time become a valuable tool for destroying cancer cells that observation of ongoing cell death. As a first attempt to have become resistant to the cytotoxic effects of monitor apoptosis induction in a real-time fashion, conventional cancer therapies. radioactive annexin-V has been employed to detect plasma membrane changes linked to apoptosis-induced chemotherapy in vivo. However, in vivo detection of apoptosis is hampered by the rapid clearance of apoptotic cells by phagocytes (Zhao et al., 2001). Acknowledgements Since mitochondrial permeabilization is a relatively Supported by a special grant from the Ligue Nationale contre le Cancer, as well as by grants from ANRS (to G early event in the apoptotic program, methods to Kroemer), European Commission (QLG1-CT-1999-00739) detect this process might be more useful in revealing (to G Kroemer and K-M Debatin), Deutsche Forschungs- the presence of apoptotic cells than other assays, such gemeinschaft and Sander Stiftung (to K-M Debatin). D as those that measure caspase-3 activation or advanced Poncet receives a fellowship from the Acade´ mie Nationale DNA fragmentation – presumably because late-stage de Me´ decine.

Oncogene Targeting mitochondria K-M Debatin et al 8800 References

Adams JM and Cory S. (2001). Trends Biochem. Sci., 26, 61 – Chandra J, Mansson E, Gogvadze V, Kaufmann SH, 66. Albertioni F and Orrenius S. (2002). Blood, 99, 655 – 663. Angioli R, Janicek M, Sevin BU, Estape R, Averette HE, Chao DT and Korsmeyer SJ. (1998). Annu. Rev. Immunol., Koechli OR, Untch M and Penalver M. (1997). Int. J. 16, 395 – 419. Oncol., 11, 777 – 780. Chen LB. (1988). Annu. Rev. Cell Biol., 4, 155 – 181. AokiH,YoshimuraK,KangPM,HampeJ,NomaT, Cheng EH-YA, Wei MC, Weiler S, Flavell RA, Mak TW, Matsuzaki M and Izumo S. (2002). J. Biol. Chem., 277, Lindsten T and Korsmeyer SJ. (2001). Mol. Cell, 8, 705 – 10244 – 10250. 711. Aqeilan R, Varkoni S and Lorberboum Galski H. (1999). Chi KN, Gleave ME, Klasa R, Murray N, Bryce C, Lopes de FEBS Lett., 457, 271 – 276. Menezes DE, d’Aloisio S and Tolcher AW. (2001). Clin. Aranha O, Zhu L, Alhasan S, Wood Jr DP, Kuo TH and Cancer Res., 7, 3920 – 3927. Sarkar FH. (2002). J. Urol., 167, 1288 – 1294. Clarke SJ, McStay GP and Halestrap AP. (2002). J. Biol. Arap W, Haedicke W, Bernasconi M, Kain R, Rajotte D, Chem., 2,2. KrajewskiS,EllerbyHM,BredesenDE,PasqualiniRand Coller HA, Khrapko K, Bodyak ND, Nekhaeva E, Herrero- Ruoslahti E. (2002a). Proc. Natl. Acad. Sci. USA, 99, Jimenez P and Thilly WG. (2001). Nat. Genet., 28, 147 – 1527 – 1531. 150. Arap W, Kolonin MG, Trepel M, Lahdenranta J, Cardo- Costantini P, Belzacq A-S, Vieira HLA, Larochette N, de Vila M, Giordano RJ, Mintz PJ, Ardelt PU, Yao VJ, Vidal Pablo M, Zamzami N, Susin SA, Brenner C and Kroemer CI, Chen L, Flamm A, Valtanen H, Weavind LM, Hicks G. (2000a). Oncogene, 19, 307 – 314. ME, Pollock RE, Botz GH, Bucana CD, Koivunen E, Costantini P, Jacotot E, Decaudin D and Kroemer G. Cahill D, Troncoso P, Baggerly KA, Pentz RD, Do KA, (2000b). J. Natl. Cancer Inst., 92, 1042 – 1053. Logothetis CJ and Pasqualini R. (2002b). Nat. Med., 8, Crompton M. (2000). J. Physiol., 15, 11 – 21. 121 – 127. Danielson SR, Wong A, Carelli V, Martinuzzi A, Schapira Banker DE, Cooper JJ, Fennell DA, Willman CL, AH and Cortopassi GA. (2002). J. Biol. Chem., 277, 5810 – Appelbaum FR and Cotter FE. (2002). Leuk. Res., 26, 5815. 91 – 106. De Lena M, Lorusso V, Latorre A, Fanizza G, Gargano G, Barry M and Bleackley RC. (2002). Nat. Rev. Immunol., 2, Caporusso L, Guida M, Catino A, Crucitta E, Sambiasi D 401 – 409. and Mazzei A. (2001). Eur. J. Cancer, 37, 364 – 368. Basan˜ ez G, Neshushtan A, Drozhinin O, A C, Choe E, Tuti de Pablo M, Susin SA, Jacotot E, Larochette N, Costantini S, Wood KA, Jsi Y-T, Zimmerberg J and Youle RJ. (1999). P, Ravagnan L, Zamzami N and Kroemer G. (1999). Proc. Natl. Acad. Sci. USA, 96, 5492 – 5497. Apoptosis, 4, 81 – 87. Basanez G, Zhang J, Chau BN, Maksaev GI, Frolov VA, Decaudin D, Castedo M, Beurdeley-Thomas A, Nemati F, Brandt TA, Burch J, Hardwick JM and Zimmerberg J. de Pinieux G, Caron A, Pouillart P, Wijdenes J, Rouillard (2001). J. Biol. Chem., 276, 31083 – 31091. D, Kroemer G and Poupon M-F. (2002). Cancer Res., 62, Belostotsky R and Lorberboum-Galski H. (2001). J. 1388 – 1393. Immunol., 167, 4719 – 4728. Degterev A, Lugovskoy A, Cardone M, Mulley B, Wagner Belzacq AS, El Hamel C, Vieira HLA, Cohen I, Haouzi D, G, Mitchison T and Yuan J. (2001). Nat. Cell Biol., 3, Metivier D, Marchetti P, Goldmacher V, Brenner C and 173 – 182. Kroemer G. (2001a). Oncogene, 20, 7579 – 7587. Desagher S, Osen-Sand A, Montessuit S, Magnenat E, Belzacq AS, Jacotot E, Vieira HLA, Mistro D, Granville DJ, Vilbois F, Hochmann A, Journot L, Antonsson B and Xie ZH, Reed JC, Kroemer G and Brenner C. (2001b). Martinou JC. (2001). Mol. Cell, 8, 601 – 611. Cancer Res., 61, 1260 – 1264. Donald SP, Sun XY, Hu CA, Yu J, Mei JM, Valle D and Bernardi P, Petronilli V, di Lisa F and Forte M. (2001). Phang JM. (2001). Cancer Res., 61, 1810 – 1815. Trends Biochem. Sci., 26, 112 – 117. Du C, Fang M, Li Y, Li L and Wang X. (2000). Cell, 102, Beutner G, Ruck A, Riede B and Brdiczka D. (1998). 33 – 42. Biochim. Biophys. Acta - Biomembr., 1368, 7 – 18. Ellerby HM, Arap W, Ellerby LM, Kain R, Andrusiak R, Birbes H, El Bawab S, Hannun YA and Obeid LM. (2001). DelRioG,KrajewskiS,LombardoCR,RaoR,Ruoslahti FASEB J., 15, 2669 – 2679. E, Bredesen DE and Pasqualini R. (1999). Nat. Med., 5, Bottero V, Rossi F, Samson M, Mari M, Hofman P and 1032 – 1038. Peyron JF. (2001). J. Biol. Chem., 276, 21317 – 21324. Eskes R, Desagher S, Antonsson B and Martinou JC. (2000). Boya P, Cohen I, Zamzami N, Vieira HLA and Kroemer G. Mol. Cell. Biol., 20, 929 – 935. (2002). Cell Death Differ. in press. Fernandez-Salas E, Suh KS, Speransky VV, Bowers WL, Brenner C, Cadiou H, Vieira HLA, Zamzami N, Marzo I, Levy JM, Adams T, Pathak KR, Edwards LE, Hayes DD, Xie Z, Leber B, Andrews D, Duclohier H, Reed JC and Cheng C, Steven AC, Weinberg WC and Yuspa SH. Kroemer G. (2000). Oncogene, 19, 329 – 336. (2002). Mol. Cell. Biol., 22, 3610 – 3620. Britten CD, Rowinsky EK, Baker SD, Weiss GR, Smith L, Ferri KF, Jacotot E, Blanco J, Este´ JA, Zamzami A, Susin Stephenson J, Rothenberg M, Smetzer L, Cramer J, SA, Brothers G, Reed JC, Penninger JM and Kroemer G. Collins W, Von Hoff DD and Eckhardt SG. (2000). Clin. (2000). J. Exp. Med., 192, 1081 – 1092. Cancer Res., 6, 42 – 49. Ferri KF and Kroemer GK. (2001). Nature Cell Biol., 3, Budijardjo I, Oliver H, Lutter M, Luo X and Wang X. E255 – E263. (1999). Annu. Rev. Cell Dev. Biol., 15, 269 – 290. Frank S, Guame B, Bergmann-Leitner ES, Leitner WW, Carre M, Andre N, Carles G, Borghi H, Brichese L, Briand C Robert EG, Catez F, Smith CL and Youle RJ. (2001). Dev. and Braguer D. (2002). J. Biol. Chem., 26,26. Cell, 1, 515 – 525. Casellas P, Galiegue S and Basile AS. (2002). Neurochem. Fulda B, Meyer E, Susin SA, Kroemer G and Debatin KM. Int., 40, 475 – 486. (2001). Oncogene, 20, 1073 – 1075.

Oncogene Targeting mitochondria K-M Debatin et al 8801 Fulda S and Debatin KM. (2000). Medical Pediatric Oncol., Johnstone RW, Ruefli AA and Lowe SW. (2002). Cell, 108, 35, 616 – 618. 153 – 164. Fulda S, Scaffidi C, Susin SA, Krammer PH, Kroemer G, Joza N, Kroemer G and Penninger JM. (2002). Trends Peter ME and Debatin KM. (1998a). J. Biol. Chem., 273, Genet., 18, 142 – 149. 33942 – 33948. Ju¨ rgensmeier JM, Xie Z, Deveraux Q, Ellerby L, Bredesen D Fulda S, Susin SA, Kroemer G and Debatin KM. (1998b). and Reed JC. (1998). Proc.Natl.Acad.Sci.USA,95, Cancer Res., 58, 4453 – 4460. 4997 – 5002. Gajate C and Mollinedo F. (2001). Blood, 98, 3860 – 3863. KarumanP,GozaniO,OdzeRD,ZhouXC,ZhuH,ShawR, Genini D, Adachi S, Chao Q, Rose DW, Carrera CJ, Cottam BrienTP,BozzutoCD,OoiD,CantleyLCandYuanJ. HB, Carson DA and Leoni LM. (2000). Blood, 96, 3537 – (2001). Mol. Cell, 7, 1307 – 1319. 3543. Kessel D, Antolovich M and Smith KM. (2001). Photochem. Gincel D, Zaid H and Shoshan-Barmatz V. (2001). Biochem. Photobiol., 74, 346 – 349. J., 358, 147 – 155. Khaled AR, Moor AN, Li A, Kim K, Ferris DK, Muegge K, Giraud S, Bonod-Bidaud C, Wesolowski-Louvel M and Fisher RJ, Fliegel L and Durum SK. (2001). Mol. Cell. Stepien D. (1998). J. Mol. Biol., 281, 409 – 418. Biol., 21, 7545 – 7557. Gogvadze V, Robertson JD, Zhivotovsky B and Orrenius S. KiddJF,PilkingtonMF,SchellMJ,FogartyKE,Skeppe (2001). J. Biol. Chem., 276, 19066 – 19071. JN, Taylor CW and Thorn P. (2002). J. Biol. Chem., 277, Golshani-Hebroni SG and Bessman SP. (1997). J. Bioenerg. 6504 – 6510. Biomembr., 29, 331 – 338. Kirsch DG, Doseff A, Chau BN, Lim DS, de Souza Gottlob K, Majewski N, Kennedy S, Kandel E, Robey RB PintoNC, Hansford R, Kastan MB, Lazebnik YA and and Hay N. (2001). Genes Dev., 15, 1406 – 1418. Hardwick JM. (1999). J. Biol. Chem., 274, 21155 – 21161. Green DR and Kroemer G. (1998). Trends Cell Biol., 8, 267 – Kroemer G. (1997). Nat. Med., 3, 614 – 620. 271. Kroemer G, Petit PX, Zamzami N, Vayssie` re J-L and Grinberg M, Sarig R, Zaltsman V, Frumkin D, Grammati- Mignotte B. (1995). FASEB J., 9, 1277 – 1287. kakis N, Reuveny E and Gross A. (2002). J. Biol. Chem., Kroemer G and Reed JC. (2000). Nat. Med., 6, 513 – 519. 277, 12237 – 12245. Kudla G, Montessuit S, Eskes R, Berrier C, Martinou J-C, Gross A, McDonnell JM and Korsmeyer SJ. (1999). Genes Ghazi A and Antonsson B. (2000). J. Biol. Chem., 275, Dev., 13, 1988 – 1911. 22713 – 22718. Guo Y, Srinivasula SM, Druilhe A, Fernandes-Alnemri T KumarS,BhartiA,MishraNC,RainaD,KharbandaS, and Alnemri ES. (2002). J. Biol. Chem., 6,6. Saxena S and Kufe D. (2001). J. Biol. Chem., 276, 17281 – Hail Jr N, Youssef EM and Lotan R. (2001). Cancer Res., 61, 17285. 6698 – 6702. Larisch S, Yi Y, Lotan R, Kerner H, Eimerl S, Tony Parks Han J, Flemington C, Houghton AB, Gu Z, Zambetti GP, W, Gottfried Y, Birkey Reffey S, de Caestecker MP, Lutz RJ, Zhu L and Chittenden T. (2001). Proc. Natl. Danielpour D, Book-Melamed N, Timberg R, Duckett CS, Acad. Sci. USA, 98, 11318 – 11323. Lechleider RJ, Steller H, Orly J, Kim SJ and Roberts AB. Harada H, Becknell B, Wilm M, Mann M, Huang LJS, (2000). Nat. Cell. Biol., 2, 915 – 921. Taylor SS, Scott JD and Korsmeyer SJ. (1999). Mol. Cell, Larochette N, Decaudin D, Jacotot E, Brenner C, Marzo I, 3, 413 – 422. Susin SA, Zamzami N, Xie Z, Reed JC and Kroemer G. Haridas V, Arntzen CJ and Gutterman JU. (2001). Proc. (1999). Exp. Cell Res., 249, 413 – 421. Natl. Acad. Sci. USA, 98, 11557 – 11562. Lehenkari PP, Kellinsalmi M, Napankangas JP, Ylitalo KV, Herr I and Debatin KM. (2001). Blood, 98, 2603 – 2614. Monkkonen J, Rogers MJ, Azhayev A, Vaananen HK and HirschT,DecaudinD,SusinSA,MarchettiP,LarochetteN, Hassinen IE. (2002). Mol. Pharmacol., 61, 1255 – 1262. Resche-Rigon M and Kroemer G. (1998). Exp. Cell Res., LiH,KolluriSK,GuJ,DaxsonMJ,CaoX,HobbsPD,Lin 241, 426 – 434. B, Chen G-Q, Lu J-S, Lin F, Xie Z, Fontana JA, Reed JC Hossain K, Akhand AA, Kato M, Du J, Takeda K, Wu J, and Zhang X-K. (2000). Science, 289, 1159 – 1164. Takeuchi K, Liu W, Suzuki H and Nakashima I. (2000). J. Li L, Lorenzo PS, Bogi K, Blumberg PM and Yuspa SH. Immunol., 165, 4290 – 4297. (1999). Mol. Cell. Biol., 19, 8547 – 8558. Huang CS, Ma WY, Li JX and Dong ZG. (1999). Cancer Li LY, Luo X and Wang X. (2001). Nature, 412, 95 – 99. Res., 59, 3053 – 3058. Lin DT and Lechleiter JD. (2002). J. Biol. Chem., 19,19. Hwang PM, Bunz F, Yu J, Rago C, Chan TA, Murphy MP, Lutter M, Fang M, Luo X, Nishijima M, Xie XS and Wang Kelso GF, Smith RA, Kinzler KW and Vogelstein B. X. (2000). Nat. Cell Biol., 2, 754 – 756. (2001). Nat. Med., 7, 1111 – 1117. Machida K, Hayashi Y and Osada H. (2002). J. Biol. Chem., Ichas F, Jouavill LS and Mazat J-P. (1997). Cell, 89, 1145 – 12, 12. 1153. MachoA,CastedoM,MarchettiP,AguilarJJ,DecaudinD, Ionov Y, Yamamoto H, Krajewski S, Reed JC and Perucho Zamzami N, Girard PM, Uriel J and Kroemer G. (1995). M. (2000). Proc. Natl. Acad. Sci. USA, 97, 10872 – 10877. Blood, 86, 2481 – 2487. Jacotot E, Ferri KF, El Hamel C, Brenner C, Druillenec S, Madesh M and Hajnoczky G. (2001). J. Cell Biol., 155, Hoebeke J, Rustin P, Me´ tivier D, Lenoir C, Geuskens M, 1003 – 1015. Vieira HLA, Loeffler M, Belzacq A-S, Briand J-P, Mannick JB, Schonhoff C, Papeta N, Ghafourifar P, Szibor Zamzami N, Edelman L, Xie ZH, Reed JC, Roques BP M, Fang K and Gaston B. (2001). J. Cell. Biol., 154, 1111 – and Kroemer G. (2001). J. Exp. Med., 193, 509 – 520. 1116. Jan G, Belzacq A-S, Haouzi D, Rouault A, Metivier D, Marchetti P, Zamzami N, Josph B, Schraen-Maschke S, Kroemer G and Brenner C. (2002). Cell Death Differ., 9, Mereau-Richard C, Costantini P, Metivier D, Susin SA, 179 – 188. Kroemer G and Formstecher P. (1999). Cancer Res., 54, Jansen B, Wacheck V, Heere-Ress E, Schlagbauer-Wadl H, 6257 – 6275. Hoeller C, Lucas T, Hoermann M, Hollenstein U, Wolff K Martinou J-C and Green DR. (2001). Nat. Rev. Mol. Cell. and Pehamberger H. (2000). Lancet, 356, 1728 – 1733. Biol., 2, 63 – 67.

Oncogene Targeting mitochondria K-M Debatin et al 8802 MarzoI,BrennerC,ZamzamiN,Ju¨ rgensmeier J, Susin SA, Robertson JD, Gogvadze V, Zhivotovsky B and Orrenius S. Vieira HLA, Pre´ vost M-C, Xie Z, Matsuyama S, Reed JC (2000). J. Biol. Chem., 275, 32438 – 32443. and Kroemer G. (1998). Science, 281, 2027 – 2031. Saito M, Korsmeyer SJ and Schlesinger PH. (2000). Nat. Cell Matsko CM, Hunter OC, Rabinowich H, Lotze MT and Biol., 2, 553 – 555. Amoscato AA. (2001). Biochem. Biophys. Res. Commun., Sakurai K, Katoh M and Fujimoto Y. (2001). J. Biol. Chem, 287, 1112 – 1120. 276, 26942 – 26946. McClintock DS, Santore MT, Lee VY, Brunelle J, Budinger Samali A, Cai J, Zhivotovsky B, Jones DP and Orrenius S. GR, Zong WX, Thompson CB, Hay N and Chandel NS. (1999). EMBO J., 18, 2040 – 2048. (2002). Mol. Cell. Biol., 22, 94 – 104. Scaffidi C, Fulda S, Srinivasan A, Friesen C, Li F, Tomaselli Miccoli L, Beurdeley-Thomas A, De Pinieux G, Sureau F, KJ, Debatin K-M, Krammer PH and Peter ME. (1998). Oudard S, Dutrillaux B and Poupon MF. (1998a). Cancer EMBO J., 17, 1675 – 1687. Res., 58, 5777 – 5786. Schempp CM, Kirkin V, Simon-Haarhaus B, Kersten A, Miccoli L, Poirson-Bichat F, Sureau F, Gonçalves RB, Kiss J, Termeer CC, Gilb B, Kaufmann T, Borner C, Bourgeois Y, Dutrillaux B, Poupon M-F and Oudard S. Sleeman JP and Simon JC. (2002). Oncogene, 21, 1242 – (1998b). J. Natl. Cancer Inst., 90, 1400 – 1406. 1250. Modica-Napolitano JS, Koya K, Weisberg E, Brunelli BT, Schendel SL, Azimov R, Pawloski K, Godzik A, Kagan BL Li Y and Chen IB. (1997). Cancer Res., 56, 544 – 550. andReedJC.(1999).J. Biol. Chem., 274, 21932 – 21936. Motoori S, Majima HJ, Ebara M, Kato H, Hirai F, Schlattner U, Dolder M, Wallimann T and Tokarska- Kakinuma S, Yamaguchi C, Ozawa T, Nagano T, Tsujii Schlattner M. (2001). J. Biol. Chem., 276, 48027 – 48030. H and Saisho H. (2001). Cancer Res., 61, 5382 – 5388. Scorrano L, Ashiya M, Buttle K, Weiler S, Oakes SA, Muthumani K, Zhang D, Hwang DS, Kudchodkar S, Dayes Mannella CA and Korsmeyer SJ. (2002). Dev. Cell, 2, 55 – NS, Desai BM, Malik AS, Yang JS, Chattergoon MA, 67. Maguire Jr HC and Weiner DB. (2002). Oncogene, 21, Scorrano L, Penzo D, Petronilli V, Pagano F and Bernardi P. 4613 – 4625. (2001). J. Biol. Chem., 276, 12035 – 12040. Nagahara Y, Ikekita M and Shinomiya T. (2000). J. Shimizu S, Ide T, Yanagida T and Tsujimoti Y. (2000a). J. Immunol., 165, 3250 – 3259. Biol. Chem., 275, 12321 – 12325. Nakamura Y, Kawakami M, Yoshihiro A, Miyoshi N, Shimizu S, Konishi A, Kodama T and Tsujimoto Y. (2000b). Ohigashi H, Kawai K, Osawa T and Uchida K. (2002). J. Proc. Natl. Acad. Sci. USA, 97, 3100 – 3105. Biol. Chem., 277, 8492 – 8499. Shimizu S, Matsuoka Y, Shinohara Y, Yoneda Y and Nomura K, Imai H, Koumura T, Arai M and Nakagawa Y. Tsujimoto Y. (2001). J. Cell Biol., 152, 237 – 250. (1999). J. Biol. Chem., 274, 29294 – 29304. Shimizu S and Tsujimoto Y. (2000). Proc. Natl. Acad. Sci. Oda E, Ohki R, Murasawa H, Nemoto J, Shibue T, USA, 97, 577 – 582. Yamashita T, Tokino T, Taniguchi T and Tanaka N. Singh KK, Russel J, Sigala B, Zhang Y, Williams J and (2000a). Science, 288, 1053 – 1058. Keshav KF. (1999). Oncogene, 18, 6641 – 6646. Oda K, Arakawa H, Tanak T, Matsuda K, Tanikawa C, Smith TA. (2000). Br.J.Biomed.Sci.,57, 170 – 178. Mori T, Nishimori H, Tamai K, Tokino T, Nakamura Y Soengas MS, Capodieci P, Polsky D, Mora J, Esteller M, and Taya Y. (2000b). Cell, 102, 849 – 862. Opitz-Araya X, McCombie R, Herman JG, Gerald WL, Ott M, Robertson JD, Gogvadze V, Zhivotovsky B and Lazebnik YA, Cordon-Cardo C and Lowe SW. (2001). Orrenius S. (2002). Proc. Natl. Acad. Sci. USA, 99, 1259 – Nature, 409, 207 – 211. 1263. Sordet O, Rebe C, Leroy I, Bruey JM, Garrido C, Miguet C, Paradiso A, Simone G, Lena MD, Leone B, Vallejo C, Lizard G, Plenchette S, Corcos L and Solary E. (2001). Lacava J, Dellapasqua S, Daidone MG and Costa A. Blood, 97, 3931 – 3940. (2001). Br. J. Cancer, 84, 651 – 658. Sowter HM, Ratcliffe PJ, Watson P, Greenberg AH and Pardo J, Pe´ rez-Gala´ nP,GamenS,MarzoI,Monleo´ nI, Harris AL. (2001). Cancer Res., 61, 6669 – 6673. Kaspar AA, Susin SA, Kroemer G, Krensky AM, Naval J Stahnke K, Fulda S, Friesen C, Strauss G and Debatin KM. and Anel A. (2001). J. Immunol., 67, 1222 – 1229. (2001). Blood, 98, 3066 – 3073. Pastorino JG, Shulga N and Hoek JB. (2002). J. Biol. Chem., Stoebner PE, Carayon P, Casellas P, Portier M, Lavabre- 277, 7610 – 7618. Bertrand T, Cuq P, Cano JP, Meynadier J and Meunier L. PervaizS,SeyedMA,HirparaJL,ClementMVandLoh (2001). Cell Death Differ., 8, 747 – 753. KW. (1999). Blood, 93, 4096 – 4108. Stoka V, Turk B, Schendel SL, Kil TH, Cirman T, Snipas SJ, Puthalakath H, Villunger A, O’Reilly LA, Beaumont JG, Ellerby LM, Bredesen D, Freeze H, Abrahamson M, Coultas L, Cheney RE, Huang DC and Strasser A. (2001). Bromme D, Krajewski S, Reed JC, Yin XM, Turk V and Science, 293, 1829 – 1832. Salvesen GS. (2001). J. Biol. Chem., 276, 3149 – 3157. Puthalalath G, Huang DCS, O’Reilly LA, King SM and Sturm I, Petrowsky H, Volz R, Lorenz M, Radetzki S, Strasser A. (1999). Mol. Cell, 3, 287 – 296. Hillebrand T, Wolff G, Hauptmann S, Dorken B and Quin ZH, Wang Y, Kikly KK, Sapp E, Kegel KB, Aronin N Daniel PT. (2001). J. Clin. Oncol., 19, 2272 – 2281. and DiFiglia M. (2001). J. Biol. Chem., 276, 8079 – 8086. SunSY,YueP,WuGS,El-DeiryWS,ShrootB,HongWK Ravagnan L, Gurbuxani S, Susin SA, Maisse C, Daugas E, and Lotan R. (1999). Oncogene, 18, 2357 – 2365. Zamzami N, Mak T, Jaattela M, Penninger JM, Garrido C Susin SA, Lorenzo HK, Zamzami N, Marzo I, Larochette N, and Kroemer G. (2001). Nature Cell Biol., 3, 839 – 843. Alzari PM and Kroemer G. (1999a). J. Exp. Med., 189, Ravagnan L, Marzo I, Costantini P, Susin SA, Zamzami N, 381 – 394. Petit PX, Hirsch F, Poupon M-F, Miccoli L, Xie Z, Reed Susin SA, Lorenzo HK, Zamzami N, Marzo I, Snow BE, JC and Kroemer G. (1999). Oncogene, 18, 2537 – 2546. Brothers GM, Mangion J, Jacotot E, Costantini P, Rippo MR, Malisan F, Ravagnan L, Condo I, Tomassini B, Loeffler M, Larochette N, Goodlett DR, Aebersold R, Costantini P, Todaro M, Susin SA, Rufini A, Kroemer G Siderovski DP, Penninger JM and Kroemer G. (1999b). and Testi R. (2000). FASEB J., 14, 2047 – 2054. Nature, 397, 441 – 446.

Oncogene Targeting mitochondria K-M Debatin et al 8803 Suzuki Y, Imai Y, Nakayama H, Takahashi K, Takio K and Vogelstein B, Lane D and Levine AJ. (2000). Nature, 408, Takahashi R. (2001). Mol. Cell, 8, 613 – 621. 307 – 310. Tai YT, Lee S, Niloff E, Weisman C, Strobel T and Cannistra Wang J, Silva JP, Gustafsson CM, Rustin P and Larsson SA. (1998). J. Clin. Oncol., 16, 2583 – 2590. NG. (2001). Proc. Natl. Acad. Sci. USA, 98, 4038 – 4043. Takuma K, Phuagphong P, Lee E, Mori K, Baba A and WangJL,ZhangZJ,ChoksiS,ShanS,LuZ,CroceCM, Matsuda T. (2001). J. Biol. Chem., 276, 48093 – 48099. Alnemri ES, Korngold R and Huang Z. (2000). Cancer Thomas DA, Scorrano L, Putcha GV, Korsmeyer SJ and Ley Res., 15, 1498 – 1502. TJ. (2001). Proc. Natl. Acad. Sci. USA, 98, 14985 – 14990. Warburg O. (1956). Science, 123, 309 – 314. Tinhofer I, Bernard D, Senfter M, Anether G, Loefﬂer G, Watabe M, Machida K and Osada H. (2000). Cancer Res., Kroemer G, Koﬂer R, Csordas A and Greil R. (2001). 60, 5214 – 5222. FASEB J., 15, 1613 – 1615. Wei MC, Zong W-X, Cheng EH-Y, Lindsten T, Panoutsa- Tzung SP, Kim KM, Basanez G, Giedt CD, Simon J, kopoulou V, Ross AJ, Roth KA, MacGregor GR, Zimmerberg J, Zhang KY and Hockenbery DM. (2001). Thompson CB and Korsmeyer SJ. (2001). Science, 292, Nat. Cell. Biol., 3, 183 – 191. 727 – 730. Vande Velde C, Cizeau J, Dubik D, Alimonti J, Brown T, Williams SD and Gottlieb RA. (2002). Biochem. J., 362, 23 – Israels S, Hakem R and Greenberg AH. (2000). Mol. Cell. 32. Biol., 20, 5454 – 5468. XueL,ChuF,ChengY,SunX,BorthakurA,RamaraoM, Vander Heiden MG, Chandel NS, Li XX, Schumacker PT, Pandey P, Wu M, Schlossman SF and Prasad KV. (2002). Colombini M and Thompson CB. (2000). Proc. Natl. Acad. Proc. Natl. Acad. Sci. USA, 99, 6925 – 6930. Sci. USA, 97, 4666 – 4671. Xue LY, Chiu SM and Oleinick NL. (2001). Oncogene, 20, Vander Heiden MG, Li XX, Gottleib E, Hill RB, Thompson 3420 – 3427. CB and Colombini M. (2001a). J. Biol. Chem., 276, Yu R, Mandlekar S and Kong A-NT. (2000). Mol. 19414 – 19419. Pharmacol., 58, 431 – 437. Vander Heiden MG, Plas DR, Rathmell JC, Fox CJ, Harris ZamzamiN,ElHamelC,Mun˜ oz C, Brenner C, Belzacq A-S, MH and Thompson CB. (2001b). Mol. Cell. Biol., 21, Costantini P, Molle G and Kroemer G. (2000). Oncogene, 5899 – 5912. 19, 6342 – 6350. Vander Heiden MG and Thompson CB. (1999). Nat. Cell Zamzami N and Kroemer G. (2001). Nat. Rev. Mol. Cell. Biol., 1, E209 – E216. Biol., 2, 67 – 71. Verhagen AM, Ekert PG, Pakusch M, Silke J, Connolly LM, Zha J, Weiler S, Oh KJ, Wei MC and Korsmeyer SJ. (2000). Reid GE, Moritz RL, Simpson RJ and Vaux DL. (2000). Science, 290, 1761 – 1765. Cell, 102, 43 – 53. Zhao M, Beauregard DA, Loizou L, Davletov B and Brindle Vieira HL, Belzacq A-S, Haouzi D, Bernassola F, Cohen I, KM. (2001). Nat. Med., 7, 1241 – 1244. JacototE,FerriKF,HamelEH,BartleLM,MelinoG, ZongWX,LindstenT,RossAJ,MacGregorGRand Brenner C, Goldmacher V and Kroemer G. (2001). Thompson CB. (2001). Genes Dev., 15, 1481 – 1486. Oncogene, 20, 4305 – 4316. Zoratti M and Szabo` I. (1995). Biochem. Biophys. Acta - Rev. VieiraHL,HaouziD,ElHamelC,BelzacqAS,BrennerC Biomembranes, 1241, 139 – 176. and Kroemer G. (2000). Cell Death Differ., 7, 1146 – 1154. Zorov DB, Filburn CR, Klotz LO, Zweier JL and Sollott SJ. Vieira HLA, Boya P, Cohen I, El Hamel C, Haouzi D, (2000). J. Exp. Med., 192, 1001 – 1014. Druillenec S, Belzacq A-S, Roques B, Brenner C and Kroemer G. (2002). Oncogene, 21, 1963 – 1977.

Oncogene