Ajoene, an Experimental Anti-Leukemic Drug: Mechanism of Cell Death VM Dirsch1, DSM Antlsperger1, H Hentze2,3 and AM Vollmar1

Leukemia (2002) 16, 74–83  2002 Nature Publishing Group All rights reserved 0887-6924/02 $25.00 www.nature.com/leu Ajoene, an experimental anti-leukemic drug: mechanism of cell death VM Dirsch1, DSM Antlsperger1, H Hentze2,3 and AM Vollmar1

1Department of Pharmacy, Center of Drug Research, University of Munich, Munich, Germany; and 2Biochemical Pharmacology, University of Konstanz, Konstanz, Germany

The organosulfur compound ajoene, a constitutent of garlic, Thus, evidence is provided that certain OSC may have anti- has been shown to induce apoptosis in a leukemic cell line as leukemic potential. The molecular pathway, however, by well as in blood cells of a leukemic patient. The mechanisms of action of ajoene, however, are unknown. The present study which OSC trigger apoptosis, needs to be clarified. aims to characterize the molecular events leading to ajoene- Two major pathways mediating drug-induced apoptosis triggered apoptosis. We show here that ajoene (20 ␮M) leads to have been characterized. One involves the triggering of cell a time-dependent activation of caspase-3-like activity as well surface death receptors and the other targets mitochondria as to the proteolytic processing of procaspase-3 and -8. Acti- without the involvement of a receptor/ligand system.13 vation of caspases was necessary for ajoene-induced apoptosis since the broad-range caspase inhibitor zVAD-fmk In certain cell types, the CD95 (APO-1/Fas)receptor system completely abrogated ajoene-mediated DNA fragmentation. was suggested to be involved in apoptosis evoked by chemo- Although the initiator caspase-8 was activated, the CD95 death therapeutic drugs. Drug-mediated activation of the CD95 sig- receptor was not involved in death signaling since the HL-60 naling pathway was shown to occur by induction of the CD95 clone used was shown to express a functionally inactive CD95 ligand (CD95-L)or increased expression of the CD95 recep- receptor. Furthermore, ajoene induced the release of cyto- 14–17 chrome c, which was not inhibited by zVAD-fmk indicating that tor. CD95 is a cell surface receptor belonging to the cytochrome c release precedes caspase activation. Ajoene also TNF/nerve growth factor-receptor superfamily. Activation of led to a dissipation of the mitochondrial transmembrane poten- this receptor leads to trimerization and subsequent recruit- tial. Overexpression of Bcl-xL clearly diminished ajoene- ment of FADD (Fas-associated protein with death domain)and induced caspase activation as well as apoptosis. These results procaspase-8 forming a death-inducing signaling complex indicate that apoptosis in leukemia cells triggered by ajoene is 18 based on the activation of a mitochondria-dependent caspase (DISC). Depending on the cell type, either large amounts of cascade which includes also the activation of the initiator cas- procaspase-8 are activated directly at the DISC leading to the pase-8. activation of a caspase cascade independent of mitochondria Leukemia (2002) 16, 74–83. DOI: 10.1038/sj/leu/2402337 (type I cells)or small amounts of activated caspase-8 lead to Keywords: apoptosis; organosulfur compound; caspases; mito- a pathway which needs to be amplified by a mitochondria- chondria; HL-60 controlled caspase cascade (type II cells).19 Besides the view that chemotherapeutic drugs act through triggering the CD95 receptor/ligand system, there are also Introduction studies demonstrating that anticancer drug-induced apoptosis occurs in the absence of a functional CD95 receptor/ligand Organosulfur compounds (OSC)from natural sources (eg gar- system.20–23 These apoptotic signals seem to be integrated by lic, onion, broccoli)as well as related synthetic compounds mitochondria leading to a pathway involving a caspase cas- have been shown to possess anticarcinogenic activity.1–4 The mechanism discussed to mediate the chemopreventive effect cade downstream of these organelles. A central event in mito- of OSC is an increase in the activity of phase II detoxication chondria-controlled cell death is the mitochondrial membrane enzymes.1–4 On the other hand, OSC were also demonstrated permeabilization (MMP). MMP of the inner mitochondrial to have direct antiproliferative effects on tumor cells.5–7 membrane causes loss of the mitochondrial transmembrane ⌬⌿ Importantly, several reports demonstrated triggering of potential ( m). MMP of the outer membrane leads to the apoptosis by diverse OSC in leukemia cells.8–11 Wong et al11 release of apoptogenic factors such as cytochrome c into the studied the structural requirements of synthetic OSC to induce cytosol. Released cytochrome c binds Apaf-1 forming a com- apoptosis in human leukemia cells. They found that a disulfide plex which in turn recruits and subsequently activates procas- moiety seems to be necessary to trigger cell death whereas pase-9. Activated caspase-9 is released from the complex to adjacent groups contribute to the specificity of cell killing. cleave and activate effector caspases, eg caspase-3. Important Studies from our own group demonstrated that ajoene, an regulators of cytochrome c release and therefore of mitochon- allylic disulfide originally isolated from garlic extracts,12 dria-mediated caspase activation are members of the Bcl-2 induces apoptosis in the human acute myeloid leukemia cell family proteins. Overexpression of Bcl-2 or Bcl-xL were shown line HL-60 as well as in peripheral blood mononuclear cells to block cytochrome c release in response to various (PBMC)isolated from a patient with a chronic myelogenous apoptotic stimuli.24,25 leukemia undergoing a myeloid blast crisis. In contrast, To date, the molecular pathways leading to OSC-triggered quiescent as well as proliferating PBMCs isolated from healthy apoptosis in leukemia cells remain unknown. The OSC ajoene donors remained unaffected.9 is a promising experimental anti-leukemic agent since it selec- tively induces apoptosis in human promyeloleukemic cells.9 Therefore, the aim of the present study was to characterize Correspondence: V Dirsch, Department of Pharmacy, Center of Drug the molecular signaling involved in ajoene-induced cell Research, Butenandtstraβe 5–13, D-81377 Munich, Germany; Fax: death. +49 +89 2180–7173 3Current address: Institute of Molecular and Cell Biology (IMBC), 30 Medical Drive, 117609 Singapore Received 30 May 2001; accepted 5 September 2001 Ajoene-induced apoptosis VM Dirsch et al 75 Materials and methods fluorogenic substrate DEVD-afc in 50 mM Hepes, pH 7.4, 1% sucrose, 0.1% CHAPS, 10 mM DTT). Blanks contained 10 ␮l Cells and culture conditions extraction buffer and 90 ␮l substrate buffer. Generation of free 7-amino-4-trifluoromethylcoumarin (afc)at 37 °C was deter- The human acute myeloid leukemia cell line HL-60 (DSM mined by fluorescence measurement at t = 0/t = 30 min using ACC 3; DSMZ, German Collection of Microorganisms and the fluorometer plate reader Victor2 (Wallac Instruments, Cell Cultures, Braunschweig, Germany)was cultured (37 °C Turku, Finland)set at an excitation wavelength of 385 nm and

and 5% CO2)in RPMI 1640 medium supplemented with 10% an emission wavelength of 505 nm. Protein concentrations of fetal bovine serum and L-glutamine (2 mM)(Life Technologies, the corresponding samples were estimated with the Pierce- Eggenstein, Germany). HL-60 transfected with the pSFFVneo- Assay (Pierce, Rockford, IL, USA)and the activity was calcu- ␮ bcl-xL plasmid (HL-60/bcl-xL)or pSFFV-neo plasmid (HL- lated using serially diluted standards (0–5 M afc). Control 60/neo)(a gift from Dr KN Bhalla, Moffitt Cancer Center and experiments confirmed that the activity was linear with time Research Institute, University of South Florida, Tampa, FL, and with protein concentration under the conditions USA)26 was cultivated as described for HL-60 cells except that described above. 0.1 mM non-essential amino acids and 1 mM sodium pyruvate (Life Technologies)were added to the culture medium. One mg/ml G 418 (Life Technologies)was added every fifth pass- Expression of the CD95 receptor age. Cells exposed to G418 were not used for experiments. The Jurkat clone J16, JurkatR cells27 and Jurkat T cells trans- Cells were seeded (1 × 106/ml)in a 24-well plate. To detect 28 fected with vector control or bcl-xL (kindly provided by Drs CD95 surface expression by flow cytometry cells were incu- PH Krammer and H Walczak, DKFZ, Heidelberg, Germany) bated at 4°C for 30 min with mouse anti-CD95 antibody were cultured as described.29 Ajoene ((E,Z)-4,5,9-trithiado- (clone ZB4, 10 ␮g/ml; MBL, Nagoya, Japan)or the isotype

deca-1,6,11-triene-9-oxide)was obtained from Lichtwer control (mouse IgG1; Pharmingen, Becton Dickinson, Heidel- Pharma (Berlin, Germany)and dissolved in PBS. N-acetyl- berg, Germany). After washing the cells twice with phosphate- cysteine (NAC)was obtained from Sigma (Deisenhofen, buffered saline containing 1% fetal calf serum (PBS/1% FCS), Germany), dissolved in PBS and the pH adjusted to pH 7.4. cells were incubated at 4°C for 30 min with ﬂuorescein isothi-

ocyanate (FITC)-conjugated rat anti-mouse IgG1 (Pharmingen, Becton Dickinson). Cells were washed again (2 × PBS/1% Quantification of apoptosis FCS)and analyzed by a FACSCalibur flow cytometer (Becton Dickinson). Quantification of apoptosis was performed by flow cytometric analysis of propidium iodide-stained nuclei using CELLQuest software (Becton Dickinson, Heidelberg, Germany)according Analysis of mitochondrial transmembrane potential to Nicoletti et al.30 Briefly, cells were incubated for 12 h in a (⌬⌿) hypotonic buffer, containing 1% sodium citrate, 0.1% Triton- X-100, and 50 ␮g/ml propidium iodide, in order to lyse cells HL-60 were treated with ajoene (20 ␮M)for various time per- and stain nuclei, which were analyzed by flow cytometry. iods (0–24 h)and stained with the fluorochrome 5,5 Ј,6–6’te- Nuclei to the left of the ‘G1-peak’ containing hypodiploid trachloro-1,1Ј,3,3Ј-tetraethylbenzimidazolcarbocyanine iod- DNA were considered apoptotic. Susceptibility of HL-60 cells ide (JC-1; 1,25 ␮M). After incubation at 37°C for 20 min in towards CD95-mediated apoptosis was examined using sol- the dark, the membrane potential was measured by flow cyto- uble recombinant CD95-L (330 ng/ml; Alexis, Gru¨nberg, metry (FACSCalibur; Becton Dickinson). JC-1 aggregates were Germany). To examine whether apoptosis is dependent on detected at 625 nm (red fluorescence, FL-2), monomers at 534 caspase activation, cells were preincubated with the pan-cas- nm (green fluorescence, FL-1). JC-1 (Molecular Probes, Eug- pase inhibitor zVAD-fmk or the caspase-8 inhibitor zIETD-fmk ene, OR, USA)was dissolved in dimethylformamide (5 mg/ml) (both from Calbiochem, Bad Soden, Germany)for 1 h. and further diluted in 70% ethanol. For fluorescence Apoptosis was quantified as described above. microscopy cells were treated as described above and pictures were taken on a CLSM 510 microscope with a LSM 510 software (Zeiss, Jena, Germany). Analysis of caspase-3-like activity

Cells were seeded in 24-well plates (4 × 105/ml), stimulated Immunoblot analysis with ajoene (20 ␮M)or etoposide (10 ␮g/ml)for different periods of time (0–22 h). Cells were collected by centrifugation, Cells (1 × 106/ml)were incubated in the absence or presence ␮ washed with ice-cold PBS and lysed in 5 mM MgCl2,1mM of ajoene (20 M)for various time periods (0–24 h),collected EGTA, 0.1% Triton X-100, 25 mM Hepes pH 7.5 sup- by centrifugation, washed with ice-cold PBS and lysed plemented with 1 ␮g/ml aprotinin, leupeptin, pepstatin and 1 immediately in 1% Triton X-100, 0.15 M NaCl and 10 mM mM pefabloc (Roche, Mannheim, Germany)by four cycles of Tris-HCl pH 7.4 supplemented with the protease inhibitor freeze–thawing. Lysates were stored at −70°C until further complete (Roche, Mannheim, Germany)for 30 min at 4 °C. analysis. Cytosols were prepared by centrifugation at 14000 Lysates were homogenized through a 22-G needle and centri- g (15 min, 4°C)and transferred to 96-well microtiter plates fuged at 10000 g for 10 min at 4°C. The supernatants were (Greiner, Nu¨rtingen, Germany). The ﬂuorometric DEVD-afc collected and protein was measured by the method according cleavage assay was carried out according to the method to Bradford.32 Cell lysates, containing equal amounts of pro- described by Thornberry.31 Cytosolic extracts (10 ␮l, approxi- tein, were boiled in SDS-sample buffer for 5 min before run- mately 1 mg/ml protein)or recombinant caspase-3 (10 ␮l, 30 ning on a SDS-polyacrylamide gel (10% for PARP, caspase- ␮ ng/ml protein)were diluted 1:10 with substrate buffer (55 M 8, Bcl-xL, 15% for caspase-3, cytochrome c). Proteins were

Leukemia Ajoene-induced apoptosis VM Dirsch et al 76 transferred to polyvinylidene difluoride membranes formation of sub-diploid DNA as well as DNA fragmentation (Immobilon-P, Millipore, Eschborn, Germany). Membranes detected by gel electrophoresis.9 In this study, initially, the were blocked with 5% fat-free dry milk in TBS-T pH 8.0 (Tris- time dependency (1–24 h)of ajoene (20 ␮M)-induced cell buffered saline (50 mM Tris, pH 8.0, 150 mM NaCl)with 0.1% death was confirmed. Figure 1a shows that significant Tween 20)and then incubated with specific antibodies against apoptosis is detected as early as 5 h after ajoene treatment. In caspase-8 (mouse monoclonal antibody C-15, 1:5 dilution of order to characterize the molecular events involved in ajoene- hybridoma supernatant,33 kindly provided by Prof PH induced apoptosis, we first examined whether caspases are Krammer), caspase-3 (mouse monoclonal antibody, clone 19, activated upon ajoene treatment. As measured by the fluoro- Transduction Laboratories, Becton Dickinson, Heidelberg, metric DEVD-cleavage assay caspase-3-like activity was Germany), cytochrome c (mouse monoclonal antibody detectable at 4 h and significantly increased 8 h after cell 7H8.2C12, Pharmingen, Becton Dickinson), PARP (mouse exposure to ajoene (Figure 1b). Time-dependent activation of monoclonal antibody Ab-2, Calbiochem-Novabiochem, Bad caspase-3-like proteases was confirmed by monitoring cleav-

Soden, Germany)or Bcl-x S/L (rabbit polyclonal antibody S-18; age of poly(ADP-ribose)polymerase (PARP). The 116-kDa pro- Santa Cruz Biotechnology, Heidelberg, Germany)overnight at tein PARP, a nuclear protein involved in DNA repair, is 4°C. After washing three times with TBS-T, specific proteins specifically cleaved by caspases during apoptosis.35 The West- were visualized by secondary antibody-horseradish peroxi- ern blot in Figure 1c shows the appearance of the character- dase conjugates and the chemiluminiscence reagent Renaiss- istic 85-kDa fragment in the time course of ajoene treatment. ance Plus (NEN Life Science Products, Zaventem, Belgium). In accordance with the results obtained with the fluorometric Chemiluminiscence was detected on a Kodak Digital Science DEVD-cleavage assay, processed PARP was detected 8 h after Image station 440CF (NEN Life Science Products, Zaventem, cell exposure to ajoene. Importantly, pretreatment of cells Belgium). with the broad-spectrum caspase inhibitor zVAD-fmk completely abrogated DNA fragmentation indicating that activation of caspases is necessary for ajoene-triggered apoptosis Measurement of cytochrome c release (Figure 1d). To further specify the involved caspases, we monitored the HL-60 cells (3 × 106)seeded in 24-well plates were treated processing of the effector caspase-3 and initiator caspase-8 with 20 ␮M ajoene for the indicated time. Cells were har- by immunoblot analysis using specific antibodies against the vested by centrifugation (300 g). Release of cytochrome c from individual active proteases. Treatment of HL-60 cells with mitochondria was analyzed according to Leist et al.34 Briefly, ajoene (1–24 h)resulted in the proteolytic cleavage of cas- pellets were resuspended in permeabilization buffer (210 mM pase-3 to the active p17 subunit. As shown in Figure 2a, p17 D-mannitol, 70 mM sucrose, 10 mM HEPES, 5 mM succinate, was detectable as early as 4 h culminating at 5 to 8 hours, 0.2 mM Na-EGTA, 0.15% BSA, 80 ␮g/ml digitonin, pH 7.2) indicating that the effector caspase-3 is activated during and gently shaken at 4°C for 20 min. Permeabilized cells were ajoene-induced apoptosis. Caspase-8 is expressed as inactive centrifuged (300 g), the supernatant removed and centrifuged precursor in two isoforms, caspase-8/a and caspase-8/b, again (10 min, 13000 g). The obtained cytosol was separated which appear as two protein bands of 55 and 53 kDa.33 Upon on a 15% SDS-PAGE and probed for cytochrome c as activation, caspase-8 is cleaved to the p43 and p41 intermedi- described above. The remaining pellet of permeabilized cells ates, which are further processed to the active subunit p18. was lysed in 0.1% Triton/PBS (15 min, 4°C, 5× ultrasonic Figure 2b shows the time-dependent appearance of the p41 treatment), centrifuged (12000 g,4°C, 10 min)and the super- and p43 intermediates after ajoene treatment. Interestingly, natant containing cytochrome c from mitochondria analyzed caspase-8 cleavage occurred concomitantly with processing by SDS-PAGE. Protein was quantified according to Bradford.32 of caspase-3 (Figure 2a and b). These results demonstrate that caspase-3 and caspase-8 are involved in ajoene-induced apoptotic signaling. Quantification of reactive oxygen species (ROS)

Peroxide production in HL-60 cells was quantiﬁed as Ajoene-induced apoptosis occurs independently of the described previously.9 CD95 receptor system

The initiator caspase-8 was shown to be recruited and acti- Statistical analysis vated upon engagement of the CD95 (APO-1/Fas)receptor leading to the activation of a downstream caspase pathway.36 All experiments were performed at least three times. Results We, therefore, examined whether ajoene-mediated apoptosis are expressed as mean value ± s.e.m. Statistical comparisons in HL-60 cells might involve the CD95/CD95-L system. were made by ANOVA followed by a Dunnett or Bonferroni’s Although HL-60 cells were reported to express high levels of multiple comparisons test or by an unpaired two-tailed Stud- CD95 receptor and to possess a functional CD95/CD95-L sig- ent’s t-test. P values Ͻ0.05 were considered significant. naling system,37,38 we first checked whether this is also the case for the HL-60 clone used in the present study. Expression of CD95 receptor protein was assessed by immunofluoresc- Results ence flow cytometry. As shown in Figure 3a, HL-60 cells indeed expressed considerable levels of CD95 receptor. Com- Ajoene-induced apoptosis is dependent on activation paring published data on HL-60 cells it is evident that the of caspases susceptibility of these cells towards CD95-mediated apoptosis varies.37–39 Therefore, we analyzed whether soluble CD95-L As reported previously, ajoene was shown to induce apoptosis is able to induce apoptosis in the used HL-60 clone. Interest- in promyeloleukemic cells as judged by cell morphology, the ingly, HL-60 cells did not undergo apoptosis upon stimulation

Leukemia Ajoene-induced apoptosis VM Dirsch et al 77

Figure 2 Ajoene-induced processing of caspase-3 and caspase-8. Representative immunoblots showing the time-dependent cleavage of caspase-3 to the active subunit p17 (a)and the cleavage of caspase- 8 (appearing as double bond corresponding to the two isoforms caspase-8/a and -8/b)to the p43 and p41 intermediates (b).HL-60 cells were treated with ajoene (20 ␮M)for the indicated time (0.5–24 h). Fifty ␮g protein per lane were separated by 15% (a)or 10% (b)SDS- PAGE. Caspase-3 (a)and caspase-8 (b)processing was detected using monoclonal antibodies against caspase-3 and caspase-8, respectively.

with soluble CD95-L, even if high concentrations were used. In contrast, CD95-L at the same or much lower concentrations was confirmed to induce apoptosis in Jurkat T cells (data not shown). Figure 3b shows the percent of apoptotic cells obtained after treatment with ajoene (20 ␮M)and soluble CD95-L (330 ng/ml). Whereas ajoene-induced apoptosis was highly significant after 6 as well as 24 h, soluble CD95-L did not induce apoptosis at all, suggesting that the HL-60 clone used here has no functionally active CD95 system. Although the used HL-60 clone did not respond to soluble CD95-L ajoene might sensitize HL-60 cells towards the CD95 receptor system. We therefore preincubated HL-60 cells with ajoene (20 ␮M)for 1 h and treated cells with soluble CD95-L (330 Figure 1 Ajoene-induced apoptosis is dependent on the activation ng/ml). The apoptotic response after ajoene/CD95-L treatment of caspases. (a)Time-dependent induction of apoptosis in HL-60 cells was not significantly different to the treatment with ajoene by ajoene (20 ␮M). Apoptosis was quantified by counting nuclei with a subdiploid DNA content. Bars represent the mean ± s.e.m. of three alone (data not shown), making it unlikely that ajoene sensit- independent experiments performed in triplicate. **P Ͻ 0.01, ***P izes HL-60 cells towards the CD95 receptor. The initiator cas- Ͻ 0.001; (ANOVA/Bonferroni). (b) Caspase-3-like activity determined pase-8 is essential for the apoptotic response downstream of upon ajoene (20 ␮M)treatment for 0.5–22 h by the DEVD-afc cleav- the CD95 receptor. We, therefore, investigated whether cas- ␮ age assay. Etoposide (Et, 10 g/ml, 3 h)was applied as control. Bars pase-8 is essential for ajoene-mediated apoptosis. We prein- represent the mean ± s.e.m. of three independent experiments per- Ͻ Ͻ cubated cells with the caspase-8 inhibitor zIETD-fmk at con- formed in triplicate. *P 0.05, **P 0.01; (ANOVA/Dunnett). (c) ␮ Time-dependent cleavage of PARP. HL-60 cells were treated with centrations (10 and 50 M)that were confirmed to block ajoene (20 ␮M)for the indicated time. Forty ␮g protein per lane were CD95-L-induced apoptosis in Jurkat T cells effectively (data separated by 10% SDS-PAGE. PARP was detected by an anti-PARP not shown). As shown in Figure 3c, neither of these concen- monoclonal antibody. (d)The caspase inhibitor zVAD-fmk inhibits trations was able to block or even reduce ajoene-mediated ajoene-mediated apoptosis. HL-60 cells were incubated with 100 ␮M ␮ apoptosis significantly. Together these experiments suggest zVAD-fmk (control)or with ajoene (20 M, 8 h)or preincubated with that ajoene-mediated apoptosis occurs independently of the 100 ␮M zVAD-fmk for 1 h and then treated with ajoene (20 ␮M,8 h). Nuclei were stained by propidium iodide and analyzed for DNA CD95 receptor system. content by flow cytometry. Ajoene induces mitochondrial membrane permeabilization (MPP)

Many apoptotic signals acting independently of a receptor/ligand system use primarily mitochondria to trans-

Leukemia Ajoene-induced apoptosis VM Dirsch et al 78 duce their death-inducing message. Mitochondrial membrane permeabilization (MMP)affecting the outer and/or inner membrane is a key event in mitochondria-controlled cell death.24,40,41 Outer membrane permeabilization involves the release of proteins, such as cytochrome c, which are normally confined to the intermembrane space. Inner membrane permeabilization is indicated by a dissipation of the electro- chemical gradient (⌬⌿)created by the proteins of the respir- atory chain located on the inner mitochondrial membrane. We therefore examined whether ajoene treatment leads to outer and/or inner MMP of HL-60 cells. Figure 4a shows the time-dependent release of cytochrome c into the cytosol upon exposure to ajoene (20 ␮M). Significant increased levels of cytochrome c in the cytosol were detectable as early as 4 h after ajoene treatment indicating that outer MMP is an early event in ajoene-induced apoptosis which seems to precede caspase activation. Indeed, the broad-range caspase-inhibitor zVAD-fmk was unable to block cytochrome c release (Figure 4a, lower panel)providing evidence that outer MMP occurs independent of caspases. Involvement of inner membrane permeabilization was ⌬⌿ assessed indirectly by determining a reduction in m by incubating cells with the fluorochrome JC-1 which forms ⌬⌿ aggregates (red)or monomers (green)depending on m. Dis- ⌬⌿ sipation of m can be judged by flow cytometry measuring a decrease of FL-2 which correlates with an impaired formation of JC-1 aggregates.42 As shown in Figure 4b, a diminished FL-2 fluorescence became evident as early as 4 h after cell ⌬⌿ exposure to ajoene. To confirm the occurrence of m reduction, changes in JC-1 fluorescence (ajoene 20 ␮M,8h) were monitored by confocal microscopy. The left panel of Fig- ure 4c shows that untreated cells loaded with JC-1 displayed predominantly red fluorescing mitochondria referring to an ⌬⌿ intact m. In contrast, cells treated with ajoene (right panel) exhibited, to a large extent, green fluorescing cells corre- ⌬⌿ sponding to JC-1 monomer formation and loss of m. Thus, ajoene-induced apoptosis seemed to be associated with inner and outer MMP.

Ajoene-mediated apoptosis and activation of caspases in Bcl-xL overexpressing HL-60 cells

Anti-apoptotic members of the Bcl-2 family, eg Bcl-2 and Bcl-

xL, are able to prevent mitochondrial changes leading to apoptosis by ‘guarding the mitochondrial gate’.43 Over- Figure 3 Ajoene-induced apoptosis is neither dependent on a expression of these proteins can block cytochrome c release functional CD95 receptor nor on caspase-8 activation. (a)HL-60 cells and thereby the activation of caspases downstream of mito- express the CD95 receptor. Immunofluorescence flow cytometric chondria.25 Ajoene-induced caspase activation seemed to fol- analysis of HL-60 cells stained with a monoclonal anti-CD95 antibody low redistribution of cytochrome c suggesting that their acti- (thick line)or an isotype control (regular line)and a FITC-conjugated vation might occur downstream of mitochondria. To provide secondary antibody. (b)Soluble CD95-L is unable to induce apoptosis evidence for this notion, ajoene was applied to HL-60 cells in HL-60 cells. Apoptosis induced by ajoene (20 ␮M)and soluble recombinant CD95-L (330 ng/ml)was quantified by flow cytometric overexpressing Bcl-xL (HL-60/bcl-xL; Figure 5a). Dose– analysis of propidium iodide stained sub-diploid nuclei after 6 h (left response studies performed for 8 as well as 24 h showed that panel)and 24 h (right panel).(c)The caspase-8 inhibitor zIETD-fmk HL-60/bcl-xL cells underwent significant reduced apoptosis is unable to abrogate ajoene-induced DNA fragmentation. Apoptosis compared to control cells (HL-60/neo)(Figure 5b).Caspase- induced by ajoene (20 ␮M)with or without preincubation (1 h)of ␮ ␮ 3-like activity was considerably delayed in HL-60/bcl-xL cells zIETD-fmk (10 M,50 M)was quantified as in (b).Bars represent the compared to HL-60/neo cells. A significant activity was mean ± s.e.m. of three independent experiments performed in triplicate. ***P Ͻ 0.001; n.s. = not significant (ANOVA/Dunnett). observed as late as 22 hours after ajoene treatment (Figure 5c). Accordingly, processed caspase-3 in HL-60/bcl-xL appeared strongly delayed and was hardly detectable by immunoblot analysis. In contrast, in HL-60/neo cells p17 could be detected as early as 3 h after cell stimulation (Figure 5d). Interestingly,

ajoene completely failed to cleave caspase-8 in HL-60/bcl-xL cells whereas caspase-8 processing readily occurred in HL-

Leukemia Ajoene-induced apoptosis VM Dirsch et al 79 Figure 4 Ajoene induces mitochondrial membrane permeabilization (MMP). (a) upper panel: time-course of ajoene-mediated cytochrome c release from mitochondria into the cytosol. HL-60 cells were either left untreated or treated for the indicated time with ajoene (20 ␮M)or as positive control with etoposide (Et, 17 ␮M)for 5 h. Lower panel: Time-course of cytochrome c as it appears in the cytosol of HL-60 cells treated with ajoene (20 ␮M)for the indicated time. In parallel, at each time-point, cells were preincubated for 1 h with the caspase inhibitor zVAD-fmk (100 ␮M, ‘i’)and then treated with ajoene (20 ␮M). Cytosol and mitochondrial fractions were prepared as described in Materials and methods and separated by 15% SDS- PAGE. Cytochrome c was detected by immunoblot analysis using a monoclonal antibody against cytochrome c. (b)and (c)Ajoene- ⌬⌿ mediated reduction of m. (b)HL-60 cells untreated or treated with ajoene (20 ␮M)for 1–24 h were stained with the fluorochrome JC-1. Red fluorescing JC-1 aggregates were detected by flow cytometry (FL- 2). Impaired formation of JC-1 aggregates (reduced FL-2 intensity) cor- ⌬⌿ responds to a decrease of m. (c)HL-60 cells untreated (left panel) or treated with ajoene (20 ␮M)for 8 h (right panel)were stained with the fluorochrome JC-1 and analyzed by fluorescence microscopy (Zeiss CLSM 510 using LSM 510 software). The red fluorescence reflects the multimer form of JC-1 localized in mitochondria under ⌬⌿ control conditions (left panel, intact m). The green fluorescence corresponds to monomer JC-1 after ajoene-induced dissipation of ⌬⌿ m (right panel).

Ajoene-mediated generation of ROS occurs upstream of mitochondrial-dependent caspase activation

In our previous study we found that ajoene-induced apoptosis is accompanied by the generation of ROS and subsequent activation of the transcription factor NF-␬B.9 Having identiﬁed an intrinsic pathway of apoptosis for ajoene the question arises whether generation of ROS/activation of NF-␬Bisan initial event upstream of the mitochondrial-dependent caspase activation or whether ROS production occurs downstream of or simultaneously with other mitochondrial events. Previous results demonstrated that increase of ROS and activation of NF-␬B are early events detected as early as 20 min and 1 h after activation with ajoene, respectively9 (data not shown). Thus, ROS and activation of NF-␬B clearly precede mitochondrial events like cytochrome c release and loss of ⌬⌿ m. However, in order to provide more direct evidence that ROS/activation of NF-␬B lies upstream of mitochondria additional experiments were performed. Figure 6a shows that the antioxidant NAC that was demonstrated to inhibit ajoene- induced ROS generation, NF-␬B activation as well as DNA fragmentation,9 is able to reduce caspase-3 activation considerably, indicating that ROS lies upstream of mitochondrial- dependent caspase-3 activation. Furthermore, if ajoene-mediated ROS generation is a mito-

chondrial event HL-60 cells overexpressing Bcl-xL should be less sensitive towards ajoene.24,43 Figure 6b, however, demon- strates that there is no signiﬁcant difference between HL-

60/neo and HL-60/bcl-xL cells regarding their ROS response 60/neo cells (Figure 5e). These data indicate that ajoene treat- after ajoene exposure. ment leads to the activation of the effector caspase-3 as well These data suggest that previously shown ROS production as the initiator caspase-8 downstream of mitochondrial events and subsequent activation of NF-␬B most likely lie upstream in HL-60 cells (intrinsic pathway of apoptosis). of mitochondrial-dependent caspase activation. To address the question whether ajoene also induces in other leukemic cells an intrinsic pathway of apoptosis, Jurkat leukemia T cells, JurkatR cells, a mutant that does not express Discussion a CD95 receptor27 as well as Jurkat T cells transfected with a 28 control vector or the bcl-xL gene were exposed to ajoene. Organosulfur compounds have been demonstrated to possess Ajoene induced apoptosis in Jurkat as well as in JurkatR cells potential as anti-leukemic drugs since several OSC including and Jurkat/bcl-xL cells showed a similar protection to ajoene ajoene have been shown to induce apoptosis in leukemia 8–11 as did HL-60/bcl-xL cells suggesting also an intrinsic pathway cells. The mechanism, however, by which these in Jurkat leukemia T cells. compounds trigger apoptosis is largely unknown.

Leukemia Ajoene-induced apoptosis VM Dirsch et al 80 Figure 5 Ajoene-mediated apoptosis and activation of caspases in

HL-60/bcl-xL cells compared to HL-60/neo cells. (a)Mitochondrial Bcl-xL expression in HL-60, HL-60/neo and HL-60/bcl-xL cells. Mito- chondrial fractions were analyzed by immunoblot analysis (10% SDS- = PAGE)using a polyclonal antibody against Bcl-x L/S. (us. unspecific). ␮ (b)Apoptosis induced by ajoene (10, 20 and 40 M)in HL-60/bcl-x L cells (black bars)was significantly reduced compared to HL-60/neo cells (white bars)after 8 (left panel)as well as 24 h (right panel). Apoptosis was quantified by flow cytometric analysis of propidium iodide stained nuclei. Bars represent the mean ± s.e.m. of three independent experiments performed in triplicate. ***P Ͻ 0.001; (Student’s t-test). (c) Caspase-3-like activity determined upon ajoene (20 ␮M)

treatment for 0.5–22 h was considerably delayed in HL-60/bcl-xL cells (right panel)compared to HL-60/neo cells (left panel).Etoposide (Et, 10 ␮g/ml, 3 h)was applied as control. Caspase-3-like activity was measured by the DEVD-afc cleavage assay. Bars represent the mean ± s.e.m. of three independent experiments performed in triplicate. ** P Ͻ 0.01; (ANOVA/Dunnett). (d) Representative immunoblots showing the time-dependent processing of caspase-3 to the active subunit

p17 in HL-60/neo cells (upper panel)and HL-60/bcl-x L cells (lower panel). (e) Time-dependent cleavage of caspase-8 to the p43 and p41

intermediates in HL-60/neo cells (upper panel)and HL-60/bcl-x L cells (lower panel). Cells (3 × 106/ml)were treated with ajoene (20 ␮M)for the indicated times (1–24 h). Seventy ␮g protein per lane were separated by 15% (d)or 10% (e)SDS-PAGE. Caspase-3 (d)and caspase-8 (e)was detected using monoclonal antibodies against caspase-3 and caspase-8, respectively.

Figure 6 Ajoene-induced release of ROS lies upstream of mitochondrial-dependent caspase-activation. (a)Representative immunoblot showing caspase-3 and the active subunit p17 in untreated cells (control), cells treated with NAC only, ajoene (20 ␮M, 6 h), and cells The data presented here provide insight into the signaling pretreated with NAC (15 mM, 3 h)and then activated with ajoene. pathway leading to ajoene-mediated apoptosis in the human Separation and detection of caspase-3 was performed as in Figure 2. promyeloic cell line HL-60. We demonstrated that ajoene led (b)Detection of reactive oxygen species in HL-60/neo (white bars) to the proteolytic processing of procaspases. Ajoene also and HL-60/bcl-xL (black bars)cells. Cells were treated with ajoene (20 ␮M)for the indicated time and ROS was measured using dihydror- induced caspase activity, as determined by the DEVD-afc hodamine-123 as described in Materials and methods. Bars represent cleavage assay and the proteolytic degradation of the caspase- the mean ± s.e.m. of three independent experiments performed in 3 substrate PARP. Activation of caspases was necessary for triplicate. n.s. = not signiﬁcant (Student’s t-test).

Leukemia Ajoene-induced apoptosis VM Dirsch et al 81 ajoene-mediated apoptotic cell death since DNA fragmen- caspase-3 processing. Moreover, Bcl-xL overexpression com- tation was completely inhibited by the broad-spectrum cas- pletely blocked proteolytic cleavage of caspase-8 giving clear pase inhibitor zVAD-fmk. Caspases were activated as early as evidence for a mitochondria-controlled caspase-3 as well as 5 h after cell exposure to ajoene. Interestingly, the initiator caspase-8 activation. caspase-8 was processed concomitantly with the effector Processing of the apical procaspase-8 downstream of mito- caspase-3. chondria was shown to require, besides caspase-9, caspase-3 Caspase-8 is well known to be activated upon CD95 recep- and possibly caspase-6.41,45,46 The similar time-frame of tor triggering. The enzyme is classiﬁed as initiator caspase ajoene-mediated caspase-3 and caspase-8 processing, how- containing the characteristic large prodomain with two death ever, would favor a model in which both caspases are acti- effector domains (DED). DEDs of caspase-8 and signaling vated by a common preceding event, ie activation of procas- adapter proteins such as FADD can interact and thus initiate pase-9. On the other hand, in HL-60/bcl-xL cells caspase-8 death receptor-mediated apoptosis.25,44 Since exposure of processing was completely blocked whereas activation of cas- cells to ajoene caused activation of caspase-8, a CD95- pase-3 appeared to be reduced and strongly delayed. This mediated pathway was initially presumed to be involved in argues for the possibilities that either caspase-8 indeed suc- ajoene-induced apoptosis. In fact, various other chemothera- ceeds caspase-3 whose activation is too weak in HL-60/bcl- peutic drugs have been shown to activate this death receptor xL cells to further activate caspase-8, or caspase-8 is directly pathway by upregulating CD95 and/or CD95-L expression.14–17 activated by a mitochondrial factor which is blocked by Bcl-

Unexpectedly, analysis of the HL-60 clone used in this study xL. The latter possibility fits into a model which was proposed clearly demonstrated that although the cells express CD95 for the CD95 triggered apoptotic pathway in type II cells.19 protein, the receptor could not be activated by its ligand Further studies using specific caspase inhibitors will clarify (CD95-L), indicating that the HL-60 cells used do not possess this issue in ajoene-induced apoptosis. a functional CD95 system. Although this finding rendered a In ajoene-mediated apoptosis caspase-3-like proteases, participation of the CD95 receptor unlikely we investigated particularly caspase-3 and caspase-8, were demonstrated to whether activation of the CD95-coupled initiator caspase-8 is be activated downstream of mitochondria. However, other necessary for ajoene-induced DNA fragmentation. Applying caspases might be active in the signaling pathway from ajoene the caspase-8 inhibitor zIETD-fmk at concentrations that effec- to mitochondria. In this respect, we showed, that the broad- tively block CD95-L-induced apoptosis in Jurkat T cells, the spectrum caspase inhibitor zVAD-fmk was not able to abro- inhibitor was unable to inhibit ajoene-mediated apoptosis. gate ajoene-induced cytochrome c release, suggesting that The CD95 receptor system, thus, seemed not to be involved initial redistribution of cytochrome c is not mediated by in ajoene-mediated apoptosis. caspases. On the other hand, there are also non-receptor signal trans- Previously shown data demonstrated that ajoene-induced duction pathways leading to the activation of caspases. Those apoptosis is accompanied by the release of ROS and the acti- signals seem to be integrated by mitochondria involving the vation of the transcription factor NF-␬B.9 This causes the ques- release of mitochondrial factors such as cytochrome c into tion to arise of how these data are linked to the present results the cytosol. Released cytochrome c binds to the cytoplasmic showing an ajoene-mediated intrinsic pathway of apoptosis. protein Apaf-1 which then undergoes a nucleotide-dependent All data collected support the view that ROS/NF-␬B activation conformational change that allows binding and subsequent rather lies up- than downstream of mitochondria-dependent processing of procaspase-9. Active caspase-9 then leads to the caspase activation. Particularly the finding that overexpression 13,25,44 processing of downstream caspases like caspase-3. of Bcl-xL in HL-60 leads to an inhibition of caspases and DNA Redistribution of cytochrome c into the cytosol, indeed, fragmentation but not to an inhibition of ROS production sub- occurred in response to ajoene. Elevated levels in the cytosol stantiates this view. An observation that also fits into this were detectable as early as 4 h after cell exposure to ajoene, model is that inhibition of caspases, eg by zVAD-fmk or Bcl- indicating that cytochrome c release is an early event in the xL overexpression, prevents ajoene-induced DNA fragmen- signaling cascade preceding caspase activation. Release of tation as a hallmark of apoptosis but does not totally prevent cytochrome c from the mitochondrial intermembrane space is cell death (unpublished observation). This cell death may be connected to outer mitochondrial membrane permeabiliz- due to reactive oxygen species that might activate alternative ation (MMP). Outer MMP is mostly but not always associated pathways when caspases are blocked.47 with loss of the mitochondrial transmembrane potential In summary, we demonstrated that the apoptotic cell death ⌬⌿ 24 ( m), which occurs as a consequence of inner MMP. Mea- induced by the OSC ajoene is fully dependent on the acti- ⌬⌿ suring m we found that ajoene also induced inner MMP vation of caspases which are processed in response to apop- which occurred in a similar time-frame as outer MMP. totic mitochondrial events. Signaling via the CD95 receptor Further evidence for a central role of mitochondria in as well as the activation of caspases upstream of mitochondria ajoene-induced apoptosis came from experiments with HL-60 could be excluded. Further experiments will clarify whether cells overexpressing Bcl-xL. Overexpression of anti-apoptotic ajoene directly affects mitochondria or whether ajoene causes Bcl-2 or Bcl-xL was shown to prevent apoptotic mitochondrial mitochondrial permeabilization by an indirect fashion. events like cytochrome c release in response to a wide variety The organosulfur compound ajoene, thus, clearly shows 25,43 of stimuli. Indeed, Bcl-xL was found to strongly reduce anti-leukemic potential which might be of use especially in ajoene-induced DNA fragmentation indicating that ajoene- cells with low chemosensitivity due to a lack in CD95 mediated apoptosis is dependent on the ‘activation’ of signaling. mitochondria. Processing of the effector caspase-3 and the initiator caspase-8 occurred simultaneously, both succeeding cytochrome Acknowledgements c release. This led us to assume that both caspases are activated downstream of mitochondria. Indeed, HL-60/bcl-xL cells This work was supported by the Deutsche Forschungsgemein- showed significant delayed caspase-3-like activity as well as schaft (SFB 369). We thank Dr PH Krammer and Dr H Wal-

Leukemia Ajoene-induced apoptosis VM Dirsch et al 82 czak (DKFZ, Heidelberg, Germany)for supplying the anti-cas- induced cell death and doxorubicin-induced apoptosis in leu- pase-8 antibody and all Jurkat cell lines, Dr KN Bhalla (Moffitt kemic T cells. Blood 2000; 95: 301–308. Cancer Center and Research Institute, University of South Flo- 18 Scaffidi C, Kirchhoff S, Krammer PH, Peter ME. Apoptosis signaling in lymphocytes. Curr Opin Immunol 1999; 11: 277–285. rida, Tampa, USA)for providing HL-60/bcl-x L and HL-60/neo 19 Scaffidi C, Fulda S, Srinivasan A, Friesen C, Li F, Tomaselli KJ, cells, T Meindl (GSF, Mu¨nchen, Germany)for taking confocal Debatin K-M, Krammer PH, Peter ME. Two CD95 (APO-1/Fas)sig- microscopy pictures and T Roos for excellent technical assist- naling pathways. EMBO J 1998; 17: 1675–1687. ance. 20 Villunger A, Egle A, Kos M, Hartmann BL, Geley S, Kofler R, Greil R. Drug-induced apoptosis is associated with enhanced Fas (APO- 1/CD95)ligand expression but occurs independently of Fas (APO- 1/CD95)signaling in human T-acute lymphatic leukemia cells. References Cancer Res 1997; 57: 3331–3334. 21 Eischen CM, Kottke TJ, Martins LM, Basi GS, Tung JS, Earnshaw 1 Sparnins VL, Barany G, Wattenberg LW. Effects of organosulfur WC, Leibson PJ, Kaufmann SH. Comparison of apoptosis in wild- compounds from garlic and onions on benzo[a]pyrene-induced type and Fas-resistant cells: chemotherapy-induced apoptosis is neoplasia and glutathione S-transferase activity in the mouse. Car- not dependent on Fas/Fas ligand interactions. Blood 1997; 90: cinogenesis 1988; 9: 131–134. 935–943. 2 Sumiyoshi H, Wargovich MJ. Chemoprevention of 1,2-dime- 22 Kataoka T, Schro¨ter M, Hahne M, Schneider P, Irmler M, Thome thylhydrazine-induced colon cancer in mice by naturally occur- M, Froelich CJ, Tschopp J. FLIP prevents apoptosis induced by ring organosulfur compounds. Cancer Res 1990; 50: 5084–5087. death receptors but not by perforin/granzyme B, chemothera- 3 Zhang Y, Talalay P, Cho C-G, Posner GH. A major inducer of peutic drugs, and gamma irradiation. J Immunol 1998; 161: anticarcinogenic protective enzymes from broccoli: isolation and 3936–3942. elucidation of structure. Proc Natl Acad Sci USA 1992; 89: 23 Wesselborg S, Engels IH, Rossmann E, Los M, Schulze-Osthoff K. 2399–2403. Anticancer drugs induce caspase-8/FLICE activation and apoptosis 4 Reddy BS, Rao CV, Rivenson A, Kelloff G. Chemoprevention of in the absence of CD95 receptor/ligand interaction. Blood 1999; colon carcinogenesis by organosulfur compounds. Cancer Res 93: 3053–3063. 1993; 53: 3493–3498. 24 Kroemer G, Reed JC. Mitochondrial control of cell death. Nat Med 5 Pinto JT, Qiao C, Xing J, Rivlin RS, Protomastro ML, Weissler ML, 2000; 6: 513–519. Tao Y, Thaler H, Heston WDW. Effects of garlic thioallyl deriva- 25 Budihardjo I, Oliver H, Lutter M, Luo X, Wang X. Biochemical tives on growth, glutathione concentration, and polyamine forma- pathways of caspase activation during apoptosis. Annu Rev Cell tion of human prostate carcinoma cells in culture. Am J Clin Nutr Dev Biol 1999; 15: 269–290. 1997; 66: 398–405. 26 Ibrado AM, Liu L, Bhalla K. Bcl-x overexpression inhibits pro- 6 Gamet-Payrastre L, Lumeau S, Gasc N, Cassar G, Rollin P, Tulliez L gression of molecular events leading to paclitaxel-induced J. Selective cytostatic and cytotoxic effects of glucosinolates apoptosis of human acute myeloid leukemia. Cancer Res 1997; hydrolysis products on human colon cancer cells in vitro. 57: 1109–1115. Anticancer Drugs 1998; 9: 141–148. 27 Peter ME, Dhein J, Ehret A, Hellbardt S, Walczak H, Moldenhauer 7 Gamet-Payrastre L, Li P, Lumeau S, Cassar G, Dupont M-A, Chev- olleau S, Gasc N, Tulliez J, Terce´ F. Sulforaphane, a naturally G, Krammer PH. APO-1 (CD95)-dependent and -independent occurring isothiocyanate, induces cell cycle arrest and apoptosis antigen receptor-induced apoptosis in human T and B cell lines. in HT29 human colon cancer cells. Cancer Res 2000; 60: Int Immunol 1995; 7: 1873–1877. 1426–1433. 28 Walczak H, Bouchon A, Stahl H, Krammer PH. Tumor necrosis 8 Sigounas G, Hooker JL, Li W, Anagnostou A, Steiner M. S- factor-related apoptosis-inducing ligand retains its apoptosis- allylmercaptocysteine, a stable thioallyl compound, induces inducing capacity on Bcl-2- or Bcl-xL-overexpressing chemo- apoptosis in erythroleukemia cell lines. Nutr Cancer 1997; 28: therapy-resistant tumor cells. Cancer Res 2000; 60: 3051–3057. 153–159. 29 Dirsch VM, Stuppner H, Vollmar AM. Helenalin triggers a CD95 9 Dirsch VM, Gerbes AL, Vollmar AM. Ajoene, a compound of gar- death receptor-independent apoptosis that is not affected by over- lic, induces apoptosis in human promyeloleukemic cells, expression of Bcl-xL or Bcl-2. Cancer Res 2001; 61: 5817–5823. accompanied by generation of reactive oxygen species and acti- 30 Nicoletti I, Migliorati G, Pagliacci MC, Grignani F, Riccardi CA. vation of nuclear factor ␬B. Mol Pharmacol 1998; 53: 402–407. A rapid and simple method for measuring thymocyte apoptosis by 10 Lee E, Kong G, Lee SJ, Kim ND, Surh YJ. 2-(allylthio)pyrazine sup- propidium iodide staining and flow cytometry. J Immunol presses the growth and proliferation of human promyelocytic leu- Methods 1991; 139: 271–279. kemia (HL-60)cells via induction of apoptosis. Anticancer Res 31 Thornberry NA. Interleukin-1 beta converting enzyme. Meth 1999; 19: 4073–4080. Enzymol 1994; 244: 615–631. 11 Wong WW-L, MacDonald S, Langler RF, Penn LZ. Novel synthetic 32 Bradford MM. A rapid and sensitive method for the quantitation organosulfur compounds induce apoptosis of human leukemic of microgram quantities of protein utilizing the principle of pro- cells. Anticancer Res 2000; 20: 1367–1374. tein-dye binding. Anal Biochem 1976; 72: 248–254. 12 Apitz-Castro R, Cabrera S, Cruz MR, Ledezma E, Jain MK. Effects 33 Scaffidi C, Medema JP, Krammer PH, Peter ME. FLICE is predomi- of garlic extract and of three pure components isolated from it nantly expressed as two functionally active isoforms, caspase 8/a on human platelet aggregation, arachidonate metabolism, release and caspase 8/b. J Biol Chem 1997; 272: 26953–26958. reaction, and platelet ultrastructure. ThrombRes 1983; 32: 155– 34 Leist M, Volbracht C, Fava E, Nicotera P. 1-Methyl-4-phenylpyridi- 169. nium induces autocrine excitotoxicity protease activation, and 13 Kaufmann SH, Earnshaw WC. Induction of apoptosis by cancer neuronal apoptosis. Mol Pharmacol 1998; 54: 789–801. chemotherapy. Exp Cell Res 2000, 256: 42–49. 35 Kidd VJ. Proteolytic activities that mediate apoptosis. Annu Rev 14 Friesen C, Herr I, Krammer PH, Debatin K-M. Involvement of the Physiol 1998; 60: 533–573. CD95 (APO-1/Fas)receptor/ligand system in drug-induced 36 Medema JP, Scaffidi C, Kischkel FC, Shevchenko A, Mann M, apoptosis in leukemia cells. Nat Med 1996; 2: 574–577. Krammer PH, Peter ME. FLICE is activated by association with the 15 Mu¨ller M, Strand S, Hug H, Heinemann E-M, Walczak H, Hof- CD95 death-inducing signaling complex (DISC). EMBO J 1997; mann WJ, Stremmel W, Krammer PH, Galle PR. Drug-induced 16: 2794–2804. apoptosis in hepatoma cells is mediated by the CD95 (APO-1/Fas) 37 Cle´ment M-V, Hirpara JL, Chawdhury S-H, Pervaiz S. Chemoprev- receptor/ligand system and involves activation of wild-type p53. entive agent resveratrol, a natural product derived from grapes, J Clin Invest 1997; 99: 403–413. triggers CD95 signaling-dependent apoptosis in human tumor 16 Micheau O, Solary E, Hammann A, Martin F, Dimanche-Boitrel cells. Blood 1998; 92: 996–1002. MT. Sensitization of cancer cells treated with cytotoxic drugs to 38 Gajate C, Fonteriz RI, Cabaner C, Alvarez-Noves G, Alvarez- Fas mediated cytotoxicity. J Natl Cancer Inst 1997; 89: 783–789. Rodriguez Y, Modolell M, Mollinedo F. Intracellular triggering of 17 Fulda S, Strauss G, Meyer E, Debatin K-M. Functional CD95 ligand Fas, independently of FasL, as a new mechanism of antitumor and CD95 death-inducing signaling complex in activation- ether lipid-induced apoptosis. Int J Cancer 2000; 85: 674–682.

Leukemia Ajoene-induced apoptosis VM Dirsch et al 83 39 Ohashi M, Iwase M, Nagumo M. Changes in susceptibility to Fas- 44 Earnshaw WC, Martins LM, Kaufmann SH. Mammalian caspases: mediated apoptosis during differentiation of HL-60 cells. J Leukoc structure, activation, substrates, and functions during apoptosis. Biol 2000; 67: 374–380. Annu Rev Biochem 1999; 68: 383–424. 40 Loeffler M, Kroemer G. The mitochondrion in cell death control: 45 Slee EA, Harte MT, Kluck RM, Wolf BB, Casiano CA, Newmeyer certainties and incognita. Exp Cell Res 2000; 256: 19–26. DD, Wang H-G, Reed JC, Nicholson DW, Alnemri ES, Green DR, 41 Bratton SB, MacFarlane M, Cain K, Cohen GM. Protein complexes Martin SJ. Ordering the cytochrome c-initiated caspase cascade: activate distinct caspase cascades in death receptor and stress- hierarchical activation of caspases-2, -3, -6, -7, -8, and -10 in a induced apoptosis. Exp Cell Res 2000; 256: 27–33. caspase-9-dependent manner. J Cell Biol 1999; 144: 281–292. 42 Cossarizza A, Ceccarelli D, Masini A. Functional heterogeneity of 46 Tang D, Lahti JM, Kidd VJ. Caspase-8 activation and Bid cleavage an isolated mitochondrial population revealed by cytofluorometric contribute to MCF7 cellular execution in a caspase-3-dependent analysis at the single organelle level. Exp Cell Res 1996; 222: manner during staurosporine-mediated apoptosis. J Biol Chem 84–94. 2000; 275: 9303–9307. 43 Gross A, McDonnell JM, Korsmeyer SJ. Bcl-2 family members and 47 Leist M, Jaä¨ttela¨ M. Four deaths and a funeral: from caspases to the mitochondria in apoptosis. Genes Dev 1999; 13: 1899–1911. alternative mechanisms. Nature Rev Cell Biol 2001; 2: 589–598.

Leukemia