BH3-only proteins Mcl-1 and Bim as well as endonuclease G are targeted in spongistatin 1–induced apoptosis in breast cancer cells

A common problem in chemotherapy is the often developed resistance of cancer cells to anticancer agents. Because cytotoxic effects of many anticancer drugs are mediated via the apoptotic pathway, resistance to chemotherapy often reflects an inability of tumor cells to undergo apoptosis (1). The principal mediators of apoptosis are caspases and failure to activate caspases account for resistance to apoptosis (2, 3). Therefore, the triggering of caspase-independent apoptotic pathways is an attractive therapeutical strategy to combat chemoresistance (4). The proapoptotic mitochondrial proteins apoptosis-inducing factor (AIF) and endonuclease G (EndoG) are well-described death effectors working independently of caspases during cell death. On apoptotic stimuli, AIF and EndoG translocate from mitochondria to the nucleus, inducing chromatin condensation and DNA fragmentation (5, 6). Furthermore, sensitivity as well as resistance to apoptosis is regulated by members of the Bcl-2 protein family (1, 7–9). The Bcl-2 family possesses both antiapoptotic and proapoptotic members, which are divided into three subclasses defined by structural and functional similarities within four conserved Bcl-2 homology domains (BH1-4; ref. 10). The BH3-only proteins, containing only the short BH3 domain, constitute a key group of proapoptotic proteins. The BH3-only protein Bim (Bcl-2–interacting mediator of cell death) executes a highly apoptotic function by antagonizing all of the prosurvival Bcl-2 family members (10). Under physiologic conditions, Bim is bound to the dynein light chain (LC8) of the microtubular complex and thereby sequestered from other Bcl-2 family proteins (11). Apoptotic stimuli are thought to disrupt this interaction, freeing Bim to translocate to the mitochondria releasing proapoptotic factors from the intermembrane space of the mitochondria to the cytosol (12). Moreover, the activity of Bim was shown to be regulated by the antiapoptotic Bcl-2 family member Mcl-1, possessing a high-affinity binding capacity for Bim. On apoptotic stimuli, the Mcl-1/Bim complex is disrupted, allowing Bim to mediate the apoptotic cascade (13–15). Altered expression of Bcl-2 proteins, either overexpression of antiapoptotic proteins or decreased/altered expression of proapoptotic members of this family, has been identified in various human tumors, contributing to oncogenesis and also tumor cell resistance to anticancer drugs.

Thus, proapoptotic factors working independently of caspases as well as proapoptotic Bcl-2 proteins, such as the BH3-only protein Bim (16), are attractive intracellular targets for inducing tumor cell death or sensitizing tumor cells to chemotherapeutic drug-induced apoptosis. The identification of potent drugs affecting these proapoptotic proteins might be valuable to overcome chemoresistance.

Spongistatin 1 isolated from the sponges Spirastrella spinispirulifera and Hyrtios erecta was shown to bind to microtubules (17, 18) and is an emerging candidate in this therapeutical strategy. Recently, we characterized spongistatin 1 as a novel promising therapeutic agent for the treatment of leukemic tumor cells especially in the clinically relevant situation of chemoresistance due to overexpression of X-linked inhibitor of apoptosis protein (XIAP; ref. 19). These impressive results in leukemic cells encourage elucidating the activity and underlying cytotoxic mechanisms of spongistatin 1 in carcinoma cancer cells.

We present here that spongistatin 1 potently induces apoptosis of MCF-7 cells, an epithelial breast cancer cell line deficient in caspase-3 and quite insensitive to many chemotherapeutic agents (20), and elucidate the underlying apoptotic signaling pathways.

Spongistatin 1 (kindly provided by G.R. Pettit) has been isolated from the sponges S. spinispirulifera and H. erecta (17). Spongistatin 1 was dissolved and further diluted in DMSO and did not exceed 1%. Staurosporine and the caspase inhibitor N-(2-quinolyl)valyl-aspartyl-(2,6-difluorophenoxy)methyl ketone (Q-VD-OPh) were purchased from Merck Biosciences, and propidium iodide and Taxol were from Sigma.

MCF-7 cells and caspase-3 reconstituted MCF-7 cells (kindly provided by K. Schulze-Osthoff, University of Münster; ref. 21) were cultured (37°C and 5% CO₂) in RPMI 1640 containing 2 mmol/L l-glutamine (PAN Biotech) supplemented with 10% heat-inactivated FCS (PAA Laboratories). SK-Mel-5 were obtained from the American Type Culture Collection and Panc-1 cells were from I.A.Z.; both cell lines were cultured in DMEM supplemented with 1 mmol/L sodium pyruvate and 10% FCS.

Quantification of apoptosis was done according to Nicoletti et al. (22). Briefly, cells were incubated in a hypotonic buffer (0.1% sodium citrate, 0.1% Triton X-100, and 50 μg/mL propidium iodide) overnight at 4°C and analyzed by flow cytometry on a FACSCalibur (Becton Dickinson). Nuclei to the left of the G₁ peak containing hypodiploid DNA were considered apoptotic. Because of the elevated spontaneous apoptosis rate (∼10%) of cells transfected with small interfering RNA (siRNA), results of experiments with transfected cells are represented as percent specific apoptosis. Thereby the spontaneous apoptosis rate of the control cells is considered as 0% and the apoptosis rate of the stimulated cells is set in correlation.

MCF-7 cells were left untreated or treated with spongistatin 1. After 48 h, Hoechst 33342 solution (final concentration 5 μg/mL) was added to the cells and incubated for 5 min. Subsequently, pictures were taken using a Zeiss Axiovert 35 microscope (Zeiss).

MCF-7 cells were left untreated or treated with spongistatin 1, staurosporine, or Taxol for 3 h. Subsequently, cells were harvested with T/E, washed with PBS, and seeded as triplicates in a 6-well plate (500 per well). After 7 days of culture, cells were stained with 0.5% crystal violet in 20% methanol and the colonies were scored.

Activation of Bax was measured by fluorescence-activated cell sorting analysis. MCF-7 cells were left untreated or treated with spongistatin 1 for 8 h. Cells were harvested with T/E, washed with PBS, and fixed in PBS/0.5% paraformaldehyde on ice for 30 min. Cells were then washed three times in PBS/1% FCS. Staining with 0.5 μg anti-Bax 6A7 was done in staining buffer (PBS, 1% FCS, 50 μg/mL digitonin). After washing, cells were resuspended in staining buffer containing 0.1 μg Alexa Fluor 488–labeled goat anti-mouse (Molecular Probes) and incubated on ice for 30 min in the dark. After three wash steps, conformational change of Bax was immediately measured in the FL-1 channel of a flow cytometer.

Release of cytochrome c, Smac/DIABLO, and Omi/HtrA2 from mitochondria was analyzed as described previously (23). Briefly, cells were collected by centrifugation and washed with PBS. Cell pellets were resuspended in permeabilization buffer [210 mmol/L d-mannitol, 70 mmol/L sucrose, 10 mmol/L HEPES, 5 mmol/L succinate, 0.2 mmol/L EGTA, 0.15% bovine serum albumin, 60 μg/mL digitonin (pH 7.2), protease inhibitor Complete (Roche)] and incubated for 20 min at 4°C. Permeabilized cells were centrifuged (10 min, 1,500 × g, 4°C), and the supernatant was removed and centrifuged again (10 min, 14,000 × g, 4°C). The obtained cytosol was separated on a 15% SDS-PAGE and probed for cytochrome c, Smac/DIABLO, and Omi/HtrA2 as described below. The remaining pellet of permeabilized cells was lysed in 0.1% Triton X-100/PBS (15 min, 4°C) and centrifuged (10 min, 14,000 × g, 4°C) and the supernatant containing mitochondrial cytochrome c, Smac/DIABLO, and Omi/HtrA2 was analyzed by SDS-PAGE.

Cells were collected by centrifugation, washed with ice-cold PBS, and lysed for 30 min in 1% Triton X-100, 150 mmol/L NaCl, and 30 mmol/L Tris-HCl (pH 7.5) with the protease inhibitor complete (Roche). Lysates were centrifuged at 10,000 × g for 10 min at 4°C. Equal amounts of protein were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes (Immobilon-P; Millipore). Membranes were blocked with 5% fat-free milk powder in PBS containing 0.05% Tween 20 (1 h) and incubated with specific antibodies against caspase-8 and AIF (rabbit polyclonal IgG; Upstate); caspase-9, caspase-6, Bcl-x_L, Mcl-1, and β-tubulin (rabbit polyclonal; Cell Signaling); caspase-7 (mouse monoclonal IgG1) and cytochrome c (mouse monoclonal IgG2b, clone 7H8.2C12; BD Pharmingen); caspase-2 (mouse IgG1), Bax clone 6A7 (mouse monoclonal IgG1), and XIAP clone 28 (mouse IgG1; BD Transduction Laboratories); Smac/DIABLO (rabbit; Biozol); Omi/HtrA2 (rabbit polyclonal; R&D Systems); EndoG (rabbit polyclonal; Prosci); Bim (rabbit polyclonal) and Bcl-2 (mouse monoclonal IgG₁; Merck Biosciences); Bax (rabbit polyclonal; Santa Cruz Biotechnology); cytochrome c oxidase (mouse monoclonal IgG_2a, clone 1D6; Molecular Probes Invitrogen); and β-actin (mouse monoclonal IgG₁; Chemicon International) overnight at 4°C. Proteins were visualized by secondary antibodies conjugated to horseradish peroxidase and the ECL Plus Western blotting detection reagent (Amersham Biosciences). Membranes were exposed to X-ray film and developed in a tabletop film processor (Curix 60 Agfa).

Translocation of AIF and EndoG to the nucleus was analyzed using the Nuclear Extract Kit (Active Motif) according to the manufacturer's protocol. Briefly, cells were collected by centrifugation and washed with PBS. Cell pellets were resuspended in hypotonic buffer and incubated for 15 min at 4°C. Cells were permeabilized with detergent and centrifuged (1 min, 13,000 × g, 4°C) and the supernatants were removed (nonnucleic fraction). Pellets were resuspended in lysis buffer, incubated for 30 min at 4°C, and centrifuged (10 min, 13,000 × g, 4°C). The supernatants contain the nuclear proteins.

Soluble and unsoluble cell fractions were obtained according to Puthalakath et al. (11). Cells were collected by centrifugation in ice-cold PBS and cell pellets were resuspended in lysis buffer [PIPES (pH 6.9), 2 mol/L glycerin, 0.5% Triton X-100, 2 mmol/L MgCl₂, 2 mmol/L EGTA, 1 mmol/L GTP, 5 μmol/L Taxol, 50 mmol/L phenylmethylsulfonyl fluoride, protease inhibitor Complete (Roche)], incubated for 20 min at room temperature, and centrifuged (45 min, 100,000 × g, 4°C) and the supernatants were removed (soluble fraction). Pellets were resuspended in disassembling buffer [Tris-HCl (pH 6.8), 1 mmol/L MgCl₂, 10 mmol/L CaCl₂], incubated for 30 min at 4°C, and centrifuged (10 min, 1,500 × g, 4°C). The supernatants contain the unsoluble fractions.

Protein A-Sepharose beads (Sigma) were incubated with the primary antibodies and gently shaken overnight at 4°C. Cells were treated and collected by centrifugation (10 min, 360 × g, 4°C), washed with ice-cold PBS, and lysed in 1% Triton X-100, 150 mmol/L NaCl, 2 mmol/L EDTA, and 30 mmol/L Tris-HCl (pH 7.5) with the protease inhibitor Complete (Roche) for 30 min. Lysates were centrifuged (10 min, 10,000 × g, 4°C). Equal amounts of protein were added to the antibody beads mixture and incubated for 3 h at 4°C. After centrifugation (10 min, 14,000 × g, 4°C), probes were washed with lysis buffer and separated by SDS-PAGE as described above.

For localization of AIF and EndoG during apoptosis, MCF-7 were seeded on glass coverslips coated with collagen A (10% in PBS) in 24-well plates. Cells were stimulated for 24 h. After three wash steps with PBS, cells were fixed with 4% paraformaldehyde in PBS for 15 min at room temperature followed by permeabilization through incubation with 0.2% Triton X-100 in PBS for 2 min. Cells were blocked with 0.2% bovine serum albumin and incubated with specific antibodies against AIF (rabbit polyclonal IgG; Upstate) and EndoG (rabbit polyclonal antibody; Prosci). Specific proteins were visualized by secondary antibodies (Alexa 488). Nuclei were stained with Hoechst 33342 (bisbenzimide). Dual-channel images were taken by Zeiss Meta confocal laser scanning microscopy.

Sense and antisense siRNA oligonucleotides of Bim [sense: 5′-caauugaccuucucgg(dTdT)-3′ and antisense: 5′-ccgagaagguagacaauug(dTdT)-3′; ref. 24], sense and antisense siRNA oligonucleotides of EndoG [sense: 5′-augccuggaacaaccuggadTdT-3′ and antisense: 5′-uccagguuguuccaggcau(dTdT)-3′; ref. 25], sense and antisense siRNA oligonucleotides of AIF [sense: 5′-ggaaauaugggaaagaucc(dTdT)-3′ and antisense: 5′-ggaucuuucccauauuucc(dTdT)-3′; ref. 26], and sense and antisense siRNA oligonucleotides corresponding to nucleotides 978 to 998 of Apaf-1 (AATTGGTGCACTTTTACGTGA; ref. 27) and oligonucleotides corresponding to a nonsense sequence were purchased from Biomers.net. Sense and antisense siRNA oligonucleotides of Omi/HtrA2 [sense: 5′-aacggcucaggauucgugg(dTdT)-3′ and antisense: 5′-ccacgaauccugagccguu(dTdT)-3′] were obtained from Dharmacon. The oligonucleotides were annealed to create the double-stranded siRNAs. pSUPER.retro129 and pSUPER.retro1188 (Smac/DIABLO 129: 5′-gatccccgaagcggtgtttctcagaattcaagagattctgagaaacaccgcttctttttggaaa-3′ and Smac/DIABLO 1188: 5′-gatcccccctgtccagtttgtacgatttcaagagaatcgtacaaactggacaggtttttggaaa-3′) were kindly given by Simone Fulda. MCF-7 cells (2 × 10⁶) were transfected with 2.5 μg nonsense, Apaf-1 siRNA, Omi/HtrA2 siRNA, Bim siRNA, AIF siRNA, or EndoG siRNA or 2 μg vector using the Nucleofector II (Amaxa) according to the manufacturer's instructions. For Smac downregulation, 2 μg pSUPER.retro129 and pSUPER.retro1188 in a 1:1 mixture of the siRNA constructs were used. A vector containing a nonsense sequence and the vector alone were used as control. Cells were seeded and stimulated 24 h after nucleofection. Efficiency of RNA interference was checked by Western blot analysis using antibodies against Apaf-1, Smac/DIABLO, Omi/HtrA2, Bim, EndoG, and AIF.

All experiments were done at least three times in triplicate. Results are expressed as mean ± SE. One-way ANOVA with Bonferroni's post-test were done using GraphPad Prism version 3.0 for Windows (GraphPad Software). P values < 0.05 were considered significant.

Spongistatin 1 induces apoptosis in MCF-7 cells in a time- and dose-dependent manner. The appearance of apoptotic cells was significant 16 h after exposure to spongistatin 1 (500 pmol/L) and at a concentration as low as 200 pmol/L (48 h; Fig. 1A). Moreover, on treatment with spongistatin 1 (500 pmol/L) over 48 h, apoptotic MCF-7 cells detach and undergo representative morphologic changes such as cell shrinkage and formation of apoptotic bodies (Fig. 1B, top). The condensation of DNA, a characteristic feature of apoptotic cells, is clearly observed on staining spongistatin 1–treated MCF-7 (48 h) cells with the vital dye Hoechst 33342 (Fig. 1B, bottom).

Spongistatin 1 furthermore showed effects on the clonogenic survival of MCF-7 cells. Stimulation with spongistatin 1 (500 pmol/L) over 3 h and culturing for 7 days reduced the growth of colonies ∼50% compared with the untreated cells, whereas a stimulation with a higher dose of spongistatin 1 (1 nmol/L) eliminates completely the formation of colonies (Fig. 1C, top). Exposure of cells to staurosporine (500 nmol/L) does not inhibit the survival of MCF-7 cells. Although Taxol induces apoptosis in MCF-7 cells, in concentration corresponding the apoptosis rate (1 nmol/L spongistatin 1; Fig. 1C, bottom) and proliferation inhibition of spongistatin 1 (data not shown), it decreased MCF-7 colony formation by only 40% and 30%.

As shown by Western blot analysis (Fig. 2A), treatment with spongistatin 1 does not provoke an upregulation of Bax protein. However, Bax undergoes a NH₂-terminal conformational change resulting in cytochrome c release, which was analyzed using specific antibodies against the normally occluded NH₂ terminus. Fluorescence-activated cell sorting analysis of MCF-7 cells revealed an activation of Bax as early as 8 h, evidenced by a shift in FL-1 channel (Fig. 2A). As a consequence of Bax activation, spongistatin 1 triggers the release of cytochrome c, Smac/DIABLO, and Omi/HtrA2 from the mitochondria to the cytosol after 8 to 16 h (Fig. 2B).

Recently, we showed that spongistatin 1 induces apoptosis in leukemia cells via degradation of the inhibitor of apoptotic protein family member XIAP (19). However, spongistatin 1 does not degrade XIAP in MCF-7 (Fig. 2C, top), indicating an alternative apoptotic mechanism in these cells. To this end, activation of initiator and effector caspases was elucidated (Fig. 2C, bottom) and revealed a rather weak activation of the initiator caspase-8 and caspase-9 16 h after stimulation and only a slight activation of effector caspase-2, caspase-6, and caspase-7 after 32 h. In accordance, downregulation of Apaf-1, which forms with cytochrome c, an activation platform for caspase-9, as well as downregulation of Smac/DIABLO and Omi/HtrA2, do not protect cells against DNA fragmentation (Fig. 3A). Finally, the actual effect of caspases was examined using the broad-range caspase inhibitor Q-VD-OPh (Fig. 3B). Presence of Q-VD-OPh led to only a moderate reduction in the apoptosis rate of spongistatin 1 (21.4%), whereas a more pronounced reduction was seen when cells were treated with Taxol (69.5%). To show that the effect of spongistatin 1 is not limited to MCF-7 cells, this experiment was repeated with similar results in two additional cell lines derived from different tumors, the melanoma cell line SK-Mel-5 and the pancreatic cell line Panc-1. Moreover, MCF-7 cells reconstituted with caspase-3 acquire no greater sensitivity to spongistatin 1 compared with caspase-3–deficient cells, whereas the stimulation with staurosporine induces a stronger DNA fragmentation. This indicates a marginal effect of caspase-3 in the apoptotic signaling of spongistatin in MCF-7 cells (Fig. 3C). In summary, further apoptotic factors in addition to caspases may play a role in the spongistatin 1–induced cell death in MCF-7 cells.

Spongistatin 1 led to a translocation of AIF and EndoG from the mitochondria to the nucleus as shown by Western blot analysis and confocal microscopy (Fig. 4A and B). To elucidate the effect of these factors on the spongistatin 1–induced cell death, the expression of AIF and EndoG was silenced by siRNA. Interestingly, downregulation of AIF does not lead to a significant reduction in DNA fragmentation, whereas silencing of EndoG by siRNA induces a marked decrease in the apoptosis rate. Cells cotransfected with AIF and EndoG siRNA, however, show an increased reduction in DNA fragmentation showing the functional role of these factors collaborating in the apoptotic signaling induced by spongistatin 1 (Fig. 4C).

Next, upstream effectors in spongistatin 1–induced apoptosis should be identified. Spongistatin 1 has been reported to depolymerize the microtubule network by interacting with tubulin (17). MCF-7 cells express readily detectable levels of Bim. The working hypothesis was that spongistatin 1 releases Bim from the microtubule-associated dynein motor complex. Spongistatin decreases the level of β-tubulin in the unsoluble fraction and soluble fraction (Fig. 5A). Thereby Bim is released from β-tubulin shown by a diminished Bim level in the unsoluble fraction and soluble fraction on treatment with spongistatin 1.

As Bim was originally described as a Bcl-2–interacting protein capable of triggering the mitochondrial pathway either by directly activating Bax or by binding prosurvival Bcl-2 proteins, Bim released by spongistatin might block Bcl-2 and/or activate Bax via binding (12). Immunoprecipitation experiments (Fig. 5B, top) showed that Bim interaction with Bax was reduced by spongistatin 1 treatment. Spongistatin does not alter Bim interaction with Bcl-2 or Bcl-x_L. Because Mcl-1 is another well-known binding partner for Bim (28), immunoprecipitation experiments by either precipitating Bim or Mcl-1 were done. In both settings, spongistatin 1 was able to disrupt the Mcl-1/Bim complex and this effect is not due to a degradation of Mcl-1 and Bim (Fig. 5B, bottom). Most importantly, silencing of Bim by siRNA led to a marked reduction in DNA fragmentation on stimulation with spongistatin 1 (Fig. 5C), showing that Bim functions as a major proapoptotic factor in the spongistatin 1–induced cell death.

Bim is a proapoptotic factor activated by spongistatin 1, which acts upstream of mitochondria as it triggers the translocation of mitochondrial AIF and EndoG to the nucleus (Fig. 6A). In MCF-7 cells transfected with Bim siRNA, no translocation of AIF and EndoG to the nucleus was detectable. Finally, to prove the effect of Bim and caspase-independent players such as EndoG on spongistatin 1–induced apoptosis, cells were cotransfected with Bim and EndoG siRNA and a marked reduction in DNA fragmentation equal to the level of the control cells in response to spongistatin 1 exposure could be seen (Fig. 6B).

Our study presents the marine natural compound spongistatin 1 as a powerful apoptosis-inducing agent in MCF-7 cells. As MCF-7 cells are quite insensitive to many chemotherapeutic agents, this cell line is a good model to study mechanisms to combat chemoresistant cells (20). Both spongistatin 1 and Taxol induce apoptosis in MCF-7 cells; however, spongistatin 1 leads to apoptosis at a concentration 1,000-fold lower than Taxol and showed more pronounced long term effects on the clonogenic survival of MCF-7 cells than Taxol in concentrations corresponding to the apoptosis rate and proliferation inhibition of spongistatin 1. Thus, spongistatin 1 can be regarded as a powerful apoptosis-inducing and growth-inhibiting agent in breast carcinoma cells.

Spongistatin 1 was recently reported to induce caspase-dependent apoptosis in Jurkat T cells by degradation of XIAP (19). This target is not affected by spongistatin 1 in MCF-7 cells. Investigating the underlying apoptotic signaling pathways in MCF-7 cells, we were able to define two major factors. First, the spongistatin 1–induced cell death in MCF-7 cells is mainly caspase-independent and involves the apoptotic proteins AIF and EndoG. Secondly, we identified Bim as a proapoptotic factor upstream of mitochondria executing a central role in the apoptosis signaling pathway besides caspase activation by spongistatin 1.

The release of mitochondrial intermembrane space proteins to the cytosol is a key event during apoptosis (29, 30). For instance, cytochrome c is required for the initiation of the apoptosome and activation of caspases, whereas Smac/DIABLO and Omi/HtrA2 are believed to enhance caspase activation through the neutralization of the inhibitors of apoptosis proteins. Spongistatin released cytochrome c, Smac/DIABLO, and Omi/HtrA2 from mitochondria to the cytosol early on; however, caspase activation occurred late and only slightly (Fig. 2C). Furthermore, presence of the pan-caspase inhibitor Q-VD-OPh led to a just moderate reduction in DNA fragmentation (Fig. 3B) and knockdown experiments of Smac/DIABLO and Omi/HtrA2 (Fig. 3A) suggest that caspases play only an inferior role in the apoptosis signaling pathway induced by spongistatin 1.

Based on these data, we focused on proapoptotic factors working in addition to caspases such as AIF (6) and EndoG (5, 31). AIF is a mitochondrial flavoprotein first identified and characterized in the laboratory of Guido Kroemer; its main function during apoptosis is to translocate to the nucleus and initiate large-scale (50-kb) DNA fragmentation. In vitro studies using recombinant AIF showed (6, 32) that AIF is not able to cleave DNA by itself but recruits or activates endonucleases to facilitate DNA fragmentation and chromatin condensation (33), building up a so-called “degradeosome” (34). On apoptotic stimuli including various chemotherapeutic compounds, EndoG, like AIF, translocates from the mitochondria to the nucleus mediating DNA fragmentation in a caspase-independent manner.

Niikura et al. proposed in a recent study (35) that both AIF and EndoG are required in the caspase-independent cell death signaling pathway, whereas one of these two factors alone is not able to induce apoptosis. In contrast to these findings, Arnoult et al. showed that AIF and EndoG define a caspase-dependent mitochondria-initiated apoptotic DNA degradation pathway (36, 37). Hence, it remains to be examined whether caspases are required for the release of AIF from mitochondria and for DNA fragmentation and whether EndoG could be the endonuclease interacting and cooperating with AIF.

AIF and EndoG collaborate in the spongistatin 1–induced cell death as combined silencing of these genes with siRNA results in a marked reduction in DNA fragmentation, whereas gene silencing of AIF and EndoG individually did not rescue apoptosis or only by 45%, respectively. Thus, our results correspond with the findings of Niikura et al. (35) and support the notion that AIF alone is not able to induce DNA fragmentation but enhances the activity of endonucleases. We hypothesize that EndoG is an endonuclease interacting and cooperating with AIF in spongistatin 1–induced apoptosis.

The release of mitochondrial proteins, especially AIF and EndoG, is largely regulated by members of the Bcl-2 protein family. Because the BH3-only protein Bim executes a highly proapoptotic function by antagonizing all the prosurvival Bcl-2 family proteins (10), Bim is considered to be a central factor regulating the release of AIF and EndoG from the mitochondria. A functional link between these two factors and Bim was presented by Liou et al. (38). In two additional studies, it is shown that BH3-only proteins, especially tBid and Bim, are able to induce the translocation of EndoG from mitochondria to the nucleus (5, 31). As we could not detect any cleavage of Bid to the active form tBid (data not shown), our study focused on the involvement of Bim. Puthalakath et al. showed that Bim is normally bound to the dynein light chain (LC8) of microtubules and thereby sequestered from other Bcl-2 family members (11). Apoptotic stimuli are thought to disrupt this interaction, freeing Bim to translocate to the mitochondria and releasing proapoptotic factors to the cytosol. Accordingly, we could show that, in healthy cells, Bim is associated with β-tubulin and the microtubular complex, respectively. Spongistatin 1 depolymerizes microtubules, thereby freeing Bim from sequestration by β-tubulin and microtubules. Originally, Bim was described as a Bcl-2–interacting protein capable of initiating the mitochondrial pathway either by directly activating Bax-like proteins or by binding to prosurvival Bcl-2 family members (12). Recently, Weber et al. (39) showed that the Bim_S apoptosis-inducing potential is correlated with mitochondrial localization but not the ability to bind to Bcl-2. Nevertheless, the essential activity of Bim_S is assumed to be the activation of Bax. Contrary to this thesis, the results of Willis et al. (40) show that Bim is able to induce apoptosis without binding Bax. Our experiments showed that, even in untreated cells, Bim is already associated with Bcl-2, Bcl-x_L, and Bax and no enhanced translocation of Bim to Bcl-2 or Bcl-x_L occurs. Bax binding to Bim, however, is reduced in spongistatin-treated cells. In addition, the antiapoptotic Bcl-2 family member Mcl-1 can bind BH3-only proteins such as Bim, thereby functioning as a reservoir for those proapoptotic proteins (41). It has been shown that Bim has higher affinity for Mcl-1 than Bcl-2, suggesting Bim to counteract Mcl-1 more actively than Bcl-2 (28, 42). In fact, our data obtained from immunoprecipitation experiments suggest that spongistatin 1 is able to disrupt the Mcl-1/Bim complex, thereby abolishing the sequestration of the potent proapoptotic protein Bim. Based on two facts, Bim could be a major proapoptotic regulator targeted by spongistatin 1 upstream of mitochondria. First, silencing of Bim by siRNA rescued cells from apoptosis and led to a diminished release of mitochondrial proteins. Second, AIF and EndoG translocation by spongistatin 1 was slightly inhibited in cells transfected with Bim siRNA. Along this line, Bim could also act as a general regulator of AIF and EndoG trafficking and exposure to spongistatin could induce signaling events that render these proteins capable of inducing apoptosis. Cotransfection experiments with Bim siRNA and EndoG siRNA, however, indicated a direct functional link between Bim and the caspase-independent factor EndoG, showing the involvement of Bim in caspase-independent apoptotic pathways.

In summary, we propose the mechanism of spongistatin 1–induced cell death as illustrated in Fig. 6C. The natural marine compound spongistatin 1 potently induces apoptosis and inhibits long-term survival of the epithelial breast cancer cells MCF-7. Spongistatin 1 proves to be both a tool to discover novel aspects in apoptotic signaling, including the involvement of Bim in caspase-independent cell death, and a promising new anticancer agent.

No potential conflicts of interest were disclosed.

We thank Dr. Schulze-Osthoff (University of Münster) for supplying the used MCF-7 cells and caspase-3 reconstituted MCF-7 cells and Dr. Stefan Zahler for help in confocal microscopy analysis.