Abstract
The ability of the protein kinase C down-regulator bryostatin 1 to potentiate 1-β-d-arabinofuranosylcytosine (ara-C)-induced apoptosis was examined in human leukemia cells (U937) over-expressing the antiapoptotic protein Bcl-xL. Coadministration of bryostatin 1 with ara-C resulted in enhanced cytosolic release of cytochrome c and Smac/DIABLO, procaspase-3 and -9 activation, loss of mitochondrial membrane potential (Δψm), poly(ADP-ribosyl)phosphorylase degradation, apoptosis, and loss of clonogenic survival in U937/Bcl-xL cells, although effects were not as marked as in empty-vector control cells. Whereas the broad caspase inhibitor ZVAD-fluoromethyl ketone blocked ara-C/bryostatin 1-mediated caspase activation, loss of Δψm,and apoptosis in U937 cells, it failed to diminish cytochrome c release. In contrast, ectopic expression of Bcl-xL blocked cytochromec redistribution as well as all other events involved in ara-C/bryostatin 1-mediated apoptosis. The ability of ectopic expression of cytokine response modifier A to attenuate, albeit partially, bryostatin 1-mediated potentiation of ara-C-related apoptosis suggested a contributory role for activation of the extrinsic pathway in this phenomenon. Finally, the F0F1ATPase inhibitor oligomycin effectively blocked cytochromec release as well as loss of Δψm and apoptosis in U937/Bcl-xL cells. Together, these findings support the concept that bryostatin 1 potentiates ara-C lethality in human leukemia cells ectopically expressing Bcl-xL by diminishing the capacity of this antiapoptotic protein to antagonize cytochrome c release. In addition, they raise the possibility that activation of caspase cascades operating independently of Bcl-xL-associated mitochondrial actions may also contribute to enhanced lethality.
Apoptosis in both normal and neoplastic cells is controlled by a family of homologous proteins related to the prototypical member, Bcl-2, over-expression of which was originally discovered in the cells of patients with lymphoma containing the 14:18 translocation (Reed, 1994). This family consists of proteins that promote cell death (e.g., Bax, Bag, Bak, Bad, Bcl-xs, etc.), as well as those that oppose this process (e.g., Bcl-2, Bcl-xL, A1, and Mcl-1, among others) (Chao and Korsmeyer, 1998). The antiapoptotic protein Bcl-x has been shown to consist of two alternatively spliced forms; i.e., a long form, Bcl-xL, which protects cells from various noxious stimuli, and a short form, Bcl-xs, which promotes cell death (Boise et al., 1993). Although both Bcl-2 and Bcl-xL can reduce lethality in leukemic cells exposed to cytotoxic drugs (Chinnaiyan et al., 1996), the observation that these proteins exhibit different cytoprotective patterns, depending upon the cytotoxic agent (Simonian et al., 1997), raises the possibility that disparate modes of action may be involved.
The mechanism by which Bcl-2 and Bcl-xL protect cells from apoptosis is not known with certainty. One possibility is that these proteins prevent loss of the mitochondrial membrane potential (ΔΨm), which under some conditions may represent the central executioner of apoptosis (Petit et al., 1998). In addition, Bcl-xL may interact with apaf-1 (Hu et al., 1998), a critical component of a multiprotein complex referred to as the apoptosome, which also contains procaspase-9, dATP, and cytochrome c (Jiang and Wang, 2000). After oligomerization, the apoptosome induces cleavage of procaspase-9 and activation of the downstream apoptotic caspase cascade (Zou et al., 1999). However, a growing body of evidence suggests that Bcl-2 and Bcl-xL interact in an as yet to be determined way with the proapoptotic protein Bax to modify the structure and/or function of mitochondrial membrane channels, resulting in cytosolic release of several proapoptotic compounds, including cytochromec, AIF, and Smac/DIABLO (Zou et al., 1999; Jiang and Wang, 2000; Srinivasula et al., 2001). The possibility that more than one mechanism of action may be responsible for the antiapoptotic actions of Bcl-2 family members also cannot be excluded.
In previous communications, we have reported that the macrocyclic lactone bryostatin 1, which initially activates and subsequently down-regulates the Ca2+- and lipid-dependent serine/threonine kinase protein kinase C (PKC) (Grant et al., 1992), potentiates apoptosis induced by the antimetabolite 1-β-d-arabinofuranosylcytosine (ara-C) in human myeloid leukemia cells in a dose- and sequence-dependent manner (Jarvis et al., 1994; Grant et al., 1996a). In addition, we have also observed that bryostatin 1 potentiates ara-C-mediated apoptosis in human promyelocytic leukemia cells (HL-60) that ectopically express Bcl-2 (Wang et al., 1997). The present study had two major goals. First, given evidence that Bcl-2 and Bcl-xL can exert disparate cytoprotective effects (Simonian et al., 1997), it would be important to determine whether bryostatin 1 is able to potentiate ara-C-induced lethality in leukemic cells over-expressing Bcl-xL, analogous to findings involving cells ectopically expressing Bcl-2 (Wang et al., 1997). Second, whereas potentiation of ara-C-induced apoptosis by bryostatin 1 has been described, very little information is available concerning early events, particularly those related to mitochondrial dysfunction, that might account for this phenomenon. To address these issues, we have examined interactions between bryostatin 1 and ara-C in human leukemia cells (U937) that ectopically express Bcl-xL. Our results indicate that, as in the case of cells ectopically expressing Bcl-2, bryostatin 1 potentiates, albeit partially, ara-C-induced apoptosis and inhibition of clonogenicity in Bcl-xL-over-expressing leukemic cells. In addition, the present findings suggest that bryostatin 1 acts, at least in part, by interfering with the ability of Bcl-xL to block the release of cytochromec in response to ara-C exposure.
Materials and Methods
Cells.
The human monocytic leukemic cell line U937, isolated from the peripheral blood of a patient with diffuse histiocytic lymphoma, was maintained as previously described (Grant et al., 1992). Cells were cultured in RPMI 1640 medium supplemented with sodium pyruvate, minimal essential vitamins,l-glutamate, penicillin and streptomycin, and 10% heat-inactivated fetal calf serum. They were maintained in a 37°C, 5% CO2, fully humidified incubator, passed twice weekly, and prepared for experimental procedures when in log-phase growth.
For the generation of stable transfectants, the cDNA encoding Bcl- xL or CrmA (obtained from Dr. Kapil Bhalla, H. Lee Moffitt Cancer Center, Tampa, FL) was cloned into an expression vector (pcDNA 3.1) containing a G418 resistance marker. U937 cells were transfected by electroporation as described previously (Wang et al., 1999). Single cells were obtained by limiting dilution and expanded under selection pressure in medium containing 400 μg/ml G418. Clones were tested for the expression of the constructs by Western blotting, and those exhibiting the greatest degree of over-expression relative to controls were selected for use in the indicated experiments. For all studies, cells containing empty vectors were used as controls.
Drugs and Chemicals.
1-β-d-Arabinofuranosylcytosine hydrochloride (ara-C) was purchased from Sigma-Aldrich (St. Louis, MO) and maintained as a dry powder at −20°C. It was reformulated in PBS before use. Bryostatin 1 was provided by the Cancer Treatment and Evaluation Program, National Institutes of Health (Bethesda, MD) and stored desiccated at −20°C. It was formulated in sterile dimethyl sulfoxide and subsequently diluted in RPMI 1640 medium so that the final concentration of dimethyl sulfoxide was in all cases <0.05%. 3,3-Dihexyloxacarbocyanine (DiOC6) and carbamoyl cyanidem-chlorophenylhydrazone were purchased from Molecular Probes (Eugene, OR). Caspase inhibitors were purchased from Bio-Rad Laboratories (Hercules, CA). Oligomycin was purchased from Sigma.
Assessment of Apoptosis.
After drug exposures, cytocentrifuge preparations were stained with White-Giemsa and viewed by microscopy to evaluate features of cellular differentiation as well as apoptosis as previously described (Grant et al., 1992). For the latter studies, the percentage of apoptotic cells was determined by evaluating >500 cells/condition in triplicate. We have previously reported that the incidence of apoptosis as determined by these morphological criteria correlates very closely with the degree of low molecular weight DNA fragmentation assayed quantitatively by spectrofluorometry, and qualitatively with the amount of internucleosomal DNA fragmentation determined by agarose gel electrophoresis (Jarvis et al., 1994).
Agarose Gel Electrophoresis.
For qualitative assessment of internucleosomal DNA fragmentation, DNA was extracted from cell lysates after the appropriate drug treatment and subjected to agarose gel electrophoresis as previously described (Jarvis et al., 1994).
Caspase Activity.
The activities of caspase-3 and -9 were determined using commercially available kits (ApoAlert; CLONTECH, Palo Alto, CA) according to the manufacturer's specifications. The caspase-3 kit employs a colorimetric assay to monitor cleavages of DEVD-pNA substrate. Liberated pNA is monitored colorimetrically by absorbance at 405 nm. The caspase-9 kit uses the substrate LEHD-AMC, which is cleaved by caspase-9. When LEHD-AMC is cleaved, the released AMC molecule displays green fluorescence. By comparing the fluorescence of apoptotic samples versus untreated control, caspase activity can be quantified.
Western Analysis.
Western blot analysis was performed essentially as described previously (Wang et al., 1999). In brief, for each sample, 25 μg of protein per lane were separated by SDS-PAGE and electroblotted to nitrocellulose(Schleicher & Schuell, Keene, NH). Subsequently, after incubation in PBS-Tween 20 (0.05%) supplemented with 5% nonfat dry milk for 1 h at 22°C, the blots were incubated for 2 h at 22°C in fresh blocking solution with an appropriate dilution of primary antibodies as follows: Bcl-2 1:1000, Bcl- xL 1:1000, and Bax 1:1000 (Santa Cruz Biotechnology, Santa Cruz, CA); cytochrome c 1:1000, caspase-3, -8, and -9 1:1000, and apaf-1 1:1000 (BD PharMingen, San Diego, CA); and PARP 1:2500 and Smac/DIABLO 1:500 (BIOMOL Research Laboratories, Plymouth Meeting, PA). Blots were washed three times for 5 min in PBS-Tween 20 and then incubated with a 1:2000 dilution of horseradish peroxidase-conjugated secondary antibody for 1 h at 22°C. Blots were again washed three times for 5 min in PBS-Tween 20 and then developed by enhanced chemiluminescence (Amersham Biosciences, Braunschweig, Germany).
apaf-1, Bcl-2, and Bcl-xL Immunoprecitation.
After treatment, cells were washed twice with cold PBS and lysed in 0.2% Nonidet P-40 isotonic lysis buffer containing protease inhibitors for 30 min. Nuclei and unlysed cellular debris were removed by centrifugation at 15000g for 10 min. Supernatant protein (200 μg/condition) was incubated with antibody (1 μg) for 4 h at 4°C. Immunoprecipitates were then captured utilizing Dynabeads (Dynal A.S., Oslo, Norway) for another 4 h according to the manufacturer's instructions. The beads were washed three to six times with 0.5 ml of lysis buffer, and the bound proteins were eluted by boiling in 25 μl of loading buffer(0.5 M Tris-HCl, pH 6.8, 4% SDS, 20% glycerol, 0.005% bromophenol blue, 50 mM dithiothreitol). Electrophoresis of samples was carried out on 12% SDS-PAGE gels as described above. Antibodies included mouse monoclonal antihuman Bcl-2, rabbit polyclonal anti-Bcl-xL, and rabbit polyclonal antihuman apaf-1 (rabbit polyclonal; BD PharMingen). Secondary horseradish peroxidase-conjugated goat antibodies to mouse, and rabbit Ig were used. After washing twice in PBS-Tween 20, the proteins were visualized by chemiluminescence reagent (NEN Life Science Products, Boston, MA).
Cytochrome c and Smac/DIABLO Release Assay.
After treatment, cells were harvested by centrifugation at 600g for 10 min at 4°C. Cell pellets were washed once with ice-cold PBS and resuspended in 5 volumes of buffer A (20 mM HEPES-KOH, pH 7.5, 10 mM KCl, 1.5 mM MgCl2, 1 mM sodium EDTA, 1 mM sodium EGTA, 1 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride in 250 mM sucrose). After chilling for 30 min on ice, the cells were disrupted by 20 passages through a syringe fitted with a 25-gauge needle. The homogenate was centrifuged twice to remove unbroken cells and nuclei (750g for 30 min at 4°C). The S-100 fractions (supernatants) were then obtained by centrifugation at 100,000g for 30 min at 4°C. All steps were performed on ice or at 4°C. In some cases, an alternative protocol was used in which cytosolic fractions were obtained by selective plasma membrane permeabilization with digitonin (Single et al., 1998). Briefly, 2 × 106 cells were lysed by including for 1 to 2 min in lysis buffer (75 mM NaCl, 8 mM Na2HPO4, 1 mM NaH2PO4, 1 mM EDTA, and 350 μg/ml digitonin). The lysates were centrifuged at 12,000gfor 1 min, and the supernatant was collected and added to an equal volume of 2× sample buffer. The protein samples were quantified, separated by 15% SDS-PAGE, and subjected to immunoblot analysis as described above.
Assessment of Mitochondrial Membrane Potential (Δψm).
Mitochondrial membrane potential was monitored utilizing 3,3-dihexyloxacarbocyanine iodide (DiOC6). For each condition, 4 ×105 cells were incubated for 15 min at 37°C in 1 ml of 40 nM DiOC6 and subsequently analyzed using a BD Biosciences (San Jose, CA) FACScan cytofluorometer with excitation and emission settings of 488 and 525 nm, respectively. Control experiments documenting the loss of Δψm were performed by exposing cells to 5 μM carbamoyl cyanide m-chlorophenylhydrazone (15 min, 37°C), an uncoupling agent that abolishes the mitochondrial membrane potential. Results were expressed as the percentage of cells exhibiting a reduced Δψm, manifested by a reduction in the uptake of DiOC6 relative to untreated control cells.
Clonogenic Assay.
A previously described method was employed (Jarvis et al., 1994). Briefly, after drug treatment, cells were washed three times in drug-free RPMI 1640 medium containing 10% fetal bovine serum, and cell counts were determined using a model ZBI Coulter counter (Beckman Coulter, Inc., Fullerton, CA). After normalization, cells were plated in 18-mm 12-well plates (Costar, Cambridge, MA). Each well contained 1 ml of supplemented RPMI 1640 medium, 20% fetal bovine serum, 0.3% Bacto agar (Difco, Detroit, MI), and 500 cells/condition. The plates were placed in a fully humidified 37°C, 5% CO2 incubator for 13 days, after which colonies, consisting of groups of ≥50 cells, were scored with the aid of an Olympus (Tokyo, Japan) model CK inverted microscope. The total number of colonies for each condition was calculated by multiplying the cell number at the end of the initial incubation period by the cloning efficiency. Values for each condition were expressed as a percentage relative to untreated controls.
Statistical Analysis.
The significance of differences between experimental conditions was determined using the Student'st test for unpaired observations.
Results
U937/Bcl-xL cells, which show approximately a 4-fold increase in expression of Bcl-xL relative to empty-vector controls (U937/neo), but equivalent levels of Bcl-2 and bax (Wang et al., 1999), were exposed to various concentrations of ara-C for 6 h, after which various features of apoptosis were monitored. Analogous to findings by Kim et al. (1997), who reported that ectopic expression of Bcl-xL protected human promyelocytic leukemic cells (HL-60) from ara-C-induced apoptosis, U937/Bcl-xL cells were clearly less susceptible to 1 to 100 μM ara-C-induced DNA fragmentation compared with their empty-vector counterparts (Fig. 1A). In accord with these findings, U937/Bcl-xL was substantially protected from ara-C-induced loss of mitochondrial membrane potential (Δψm; Fig. 1B), cytochromec release, procaspase-3 and PARP degradation (Fig. 1C), and DEVD-pNA cleavage, reflecting caspase-3 activity (Fig. 1D). Thus, ectopic expression of Bcl-xL in human monocytic leukemia cells (U937) exerted cytoprotective effects with respect to ara-C-induced mitochondrial damage, caspase activation, and apoptosis that were similar to those observed in promyelocytic leukemia cells (Kim et al., 1997).
Recently, debate has arisen regarding the role of mitochondrial damage in the apoptotic process (Finkel, 2001). To examine this issue, the effects of the broad caspase inhibitor ZVAD-fmk, the caspase-8 inhibitor IETD-fmk, and Bcl-xLover-expression were compared with respect to ara-C-induced cytochromec release, loss of Δψm, caspase activation, and apoptosis (Fig. 2). When ara-C-treated cells (6 h) were coincubated with 25 μM IETD, which substantially blocked activation of the extrinsic, caspase-8-dependent pathway (e.g., by TNF + cycloheximide; data not shown), no diminution was observed in either cytochrome c release, procaspase-3 activation (Fig. 2A), loss of Δψm (Fig. 2B), or apoptosis (Fig. 2C). These findings argue against a major role for the receptor-mediated pathway in ara-C lethality. In contrast, coadministration of the broad caspase inhibitor ZVAD-fmk, a potent inhibitor of caspase-3, with ara-C substantially reduced procaspase-3 activation (Fig. 2A), mitochondrial discharge (Fig. 2B), and apoptosis (Fig. 2C). It did not, however, diminish cytochrome crelease. This indicates that ara-C-induced loss of Δψm represents a secondary, caspase-dependent event. In contrast to these observations, ectopic expression of Bcl-xL completely blocked ara-C-induced cytochrome c release, procaspase-3 degradation (Fig. 2A), loss of Δψm (Fig. 2B), and apoptosis (Fig.2C). Together, these findings suggest that Bcl-xL-mediated protection from ara-C lethality in U937 cells occurs upstream of cytochrome c release and that ara-C-induced loss of Δψm represents a secondary, caspase-dependent phenomenon. The finding that mitochondrial membrane potential loss after ara-C treatment is a caspase-dependent event is consistent with the results of earlier studies involving other cytotoxic agents (Bossy-Wetzel et al., 1998).
In view of evidence that subsequent exposure of ara-C-pretreated cells to bryostatin 1 increases apoptosis (Grant et al., 1996) and that bryostatin 1 has been shown to reverse, at least in part, resistance of Bcl-2-over-expressing HL-60 cells to ara-C (Wang et al., 1997), an attempt was made to determine whether and to what extent similar effects might occur in Bcl-xL-over-expressing U937 cells (Fig. 3). To this end, empty-vector and U937/Bcl-xL cells were exposed to varying concentrations of ara-C for 6 h, washed, and incubated for an additional 24 h in the presence or absence of 10 nM bryostatin 1, which by itself minimally induced cell death (e.g., <8%; data not shown), after which apoptosis was monitored (Fig. 3A). Several findings were apparent. First, subsequent exposure to bryostatin 1 increased ara-C-related apoptosis in both cell lines. Second, ectopic expression of Bcl-xL clearly protected cells from the lethal effects of combined treatment with ara-C and bryostatin 1. However, when ara-C-pretreated U937/Bcl-xL cells were subsequently exposed to a marginally toxic concentration of bryostatin 1 (e.g., 10 nM), the extent of cell death was not significantly different from that observed in empty-vector cells exposed to ara-C alone (P ≥ 0.05 in each case). Because induction of apoptosis may not correlate with loss of clonogenic survival (Yin and Schimke, 1995), colony-forming assays were performed in parallel (Fig. 3B). Again, Bcl-xL over-expression protected clonogenic cells from the lethal consequences of ara-C alone or ara-C/bryostatin 1 exposure. However, sequential exposure of Bcl-xL-over-expressing cells to ara-C followed by bryostatin 1 resulted in loss of clonogenicity equivalent to that observed in empty-vector control cells treated with ara-C alone (P ≥ 0.05). Thus, whereas Bcl-xLover-expression protected U937 cells from combined treatment with ara-C and bryostatin 1, subsequent exposure of ara-C-pretreated cells to bryostatin 1 clearly increased the extent of death in cells otherwise resistant to this nucleoside analog.
Subsequent studies were carried out to assess further the impact of combined exposure to ara-C and bryostatin 1 on mitochondrial damage and caspase activation in Bcl-xL-over-expressing cells (Fig. 4). As shown in Fig. 4A, ectopic expression of Bcl-xL protected cells from ara-C-induced loss of Δψm. Although over-expression of Bcl-xL also protected cells from loss of Δψm after exposure to the ara-C/bryostatin 1 combination, mitochondrial discharge was not significantly different from that observed in wild-type cells exposed to ara-C alone (P > 0.05). As shown in Fig. 4B, exposure of empty-vector control cells to ara-C in combination with bryostatin 1 resulted in an increase in cytosolic release of both cytochrome c and the recently described proapoptotic protein Smac/DIABLO (Chai et al., 2000). Furthermore, whereas ectopic expression of Bcl-xL substantially attenuated ara-C-mediated cytochrome c and Smac/DIABLO redistribution, addition of bryostatin 1 clearly increased release of these proteins. Combined treatment of Bcl-xL-over-expressing cells with ara-C and bryostatin 1 also resulted in an increase in cleavage/activation of procaspase-9, -8, and -3, as well as PARP degradation (Fig. 4B, bottom).
To gain insight into the hierarchy of events accompanying ara-C- and bryostatin 1-induced apoptosis in Bcl-xL-over-expressing cells, parallel studies were conducted in the presence of the pan-caspase inhibitor ZVAD-fmk. As shown in Fig. 5A, ZVAD-fmk substantially blocked apoptosis in both wild-type and Bcl-xL-over-expressing cells after exposure to ara-C and bryostatin 1. A parallel attenuation of the loss of Δψm was observed (Fig. 5B), analogous to results obtained in untransfected cells exposed to ara-C alone (Fig.2). However, in marked contrast to results obtained in the case of apoptosis and loss of Δψm, potentiation of ara-C-mediated cytochrome c and Smac/DIABLO release by bryostatin 1 was not diminished by ZVAD-fmk in either empty-vector or Bcl-xL-over-expressing cells (Fig. 5C). ZVAD-fmk did block ara-C/bryostatin 1-induced cleavage of procaspase-3 in both cell lines. Together, these findings suggest that potentiation of ara-C-induced release of cytochrome c and Smac/DIABLO by bryostatin 1 in Bcl-xL-over-expressing cells represents a primary, caspase-independent process, whereas loss of Δψm and induction of procaspase-3 cleavage represent secondary, caspase-dependent events.
Concordant results were obtained when caspase-3 (i.e., DEVD-pNA cleavage) and caspase-9 (i.e., LEHD-pNA cleavage) activities were monitored (Fig. 6). Thus, administration of bryostatin 1 increased caspase-3 and -9 activity in both empty-vector and U937/Bcl-xL cells, although it is apparent that caspase activation was attenuated in the Bcl-xL-over-expressing cell line. Nevertheless, activity of caspase-3 and -9 in U937/Bcl-xL cells after exposure to ara-C and bryostatin 1 did not differ significantly from that observed in empty-vector controls treated with ara-C alone (P ≥ 0.05 in each case). Furthermore, ZVAD-fmk blocked caspase-3 and -9 activity in both wild-type and U937/Bcl-xL cells, consistent with its effects on apoptosis (Fig. 4A).
To examine further what role, if any, activation of the receptor-mediated apoptotic cascade might play in ara-C/bryostatin 1-induced lethality, the effects of ectopic expression of Crm A were monitored in relation to responses to these agents. As shown in Fig.7, enforced expression of Crm A substantially protected cells from treatment with TNF/cycloheximide, and the effects on apoptosis and loss of Δψmin ara-C/bryostatin 1-exposed cells were also diminished, albeit partially. These findings are consistent with results indicating that combined treatment of cells with ara-C and bryostatin 1 induced degradation/activation of procaspase-8 (Fig. 4C). It should be noted that Crm A did not protect cells from lethality induced by ara-C alone. These findings suggest that activation of the extrinsic, receptor-mediated apoptotic cascade contributes, at least in part, to apoptosis induced by coadministration of ara-C and bryostatin 1.
In view of reports that cytoprotective actions of Bcl-xL may involve binding to apaf-1 (Hu et al., 1998) and the observation that bryostatin 1 increased caspase-9 activation in wild-type and Bcl-xL-over-expressing cells (Fig. 4), an attempt was made to determine whether bryostatin 1 might reduce binding of Bcl-xL to apaf-1. To this end, apaf-1 was immunoprecipitated from U937/Bcl-xL cells, and Western analysis was performed to assess the amount of Bcl-xL coimmunoprecipitating with it, as previously described (Fang et al., 1998). However, coimmunoprecipitating Bcl-xL was not detected in cells exposed to either ara-C administered alone or ara-C in combination with bryostatin 1 (data not shown). In addition, Western analysis revealed the absence of changes in expression of apaf-1 in immunoprecipitates for all experimental conditions (data not shown). Together, these findings argue against the possibility that bryostatin 1 increases ara-C-mediated caspase-9 activation and apoptosis in Bcl-xL-over-expressing cells by interfering with putatively disruptive Bcl-xL/apaf-1 interactions.
Lastly, to explore further the role of cytochrome c release in ara-C/bryostatin 1-induced apoptosis in U937/Bcl-xL cells, studies were carried out in the presence of oligomycin, an inhibitor of the F0F1 ATPase that has previously been shown to block cytochrome c redistribution (Goldstein et al., 2000). As shown in Fig.8A, cotreatment of U937/Bcl-xL cells with the combination of ara-C and bryostatin 1 (8 h) resulted in an increase in cytochromec release, and this effect, in contrast to lack of responses to caspase inhibitors, was blocked by coadministration of oligomycin. Similarly, oligomycin blocked ara-C/bryostatin 1-induced procaspase-3 degradation. Furthermore, oligomycin also prevented the increase in apoptosis and loss of Δψm observed in U937/Bcl-xL cells exposed to the combination of ara-C and bryostatin 1 (Fig. 8B). Taken together with the preceding findings, these observations support the notion that the ability of bryostatin 1 to enhance ara-C lethality in U937/Bcl-xL cells involves, at least in part, interference with the capacity of Bcl-xL to block cytochrome c redistribution.
Discussion
The present findings demonstrate that the PKC down-regulator bryostatin 1 potentiates ara-C-induced apoptosis in human U937 leukemia cells over-expressing the antiapoptotic protein Bcl-xL, analogous to results previously obtained in HL-60 cells over-expressing Bcl-2 (Wang et al., 1997). As in the latter study, cells over-expressing Bcl-xLremained less sensitive to the combination of ara-C and bryostatin 1 than their empty-vector counterparts. Nevertheless, the extent of apoptosis in Bcl-xL-over-expressing cells exposed to both agents was equivalent to that observed in control cells treated with ara-C alone. Thus, from a functional standpoint, a subtoxic concentration of bryostatin 1 was able to restore, at least in part, ara-C susceptibility to cells otherwise protected by ectopic expression of Bcl-xL. Although antiapoptotic proteins such as Bcl-2 and Bcl-xL share a number of features, including structural homologies to pore-forming proteins, the capacity to interact with proapoptotic family members (e.g., Bax), and the ability to block mitochondrial damage and caspase activation (Chang et al., 1997; Basanez et al., 2001; Vander Heiden et al., 2001), a number of differences have been identified, including disparate protective effects toward various cytotoxic agents (Simonian et al., 1997). In addition, recent studies have suggested that levels of Bcl-xL may represent a response determinant in leukemia (Schaich et al., 2001). Whatever their ultimate role in conferring drug resistance turns out to be, the present and earlier findings (Wang et al., 1997) raise the possibility that a common mechanism underlies the ability of bryostatin 1 to circumvent, at least to an extent, ara-C resistance conferred by increased expression of Bcl-2 and Bcl-xL.
Whereas the ability of pharmacologic PKC inhibitors to promote apoptosis is well described (Grant et al., 1994; Grant and Jarvis,1996b; Sordet et al., 1999), the basis for this phenomenon remains obscure. Similarly, the mode of action by which bryostatin 1, which potently down-regulates PKC activity (Grant et al., 1992; Jarvis et al., 1994), lowers the threshold for drug-induced lethality (Grant et al., 1994; Wang et al., 1997, 1999) is currently unknown, as is the mechanism by which it opposes the cytoprotective effects of antiapoptotic proteins such as Bcl-xL. Several plausible possibilities exist. For example, bryostatin 1 could oppose the capacity of Bcl-xL to block the release of proapoptogenic proteins (e.g., cytochrome c, AIF, Smac) (Yang et al., 1997; Kroemer and Reed, 2000). Alternatively, it could interfere with the ability of Bcl-xL to prevent the mitochondrial permeability transition characterized by the loss of ΔΨm, which under some circumstances may represent the central initiating apoptotic event (Zamzami et al., 1995). Bryostatin 1 could also activate the extrinsic, receptor-mediated apoptotic cascade, which is in large part independent of regulation by Bcl-2/Bcl-xL (Gross et al., 1999). Finally, bryostatin could act downstream of mitochondrial injury; e.g., by interfering with direct inhibitory interactions between Bcl-xL and components of the apoptotic caspase cascade. The ability of Bcl-xL to form a complex with apaf-1 and thereby inhibit its actions has been reported (Hu et al., 1998), although the significance of this phenomenon has been called into question (Gross et al., 2000).
The results of this study support the concept that bryostatin 1 acts, at least in part, by opposing the ability of Bcl-xL to block the release of proapoptotic proteins (e.g., cytochrome c), or at some point upstream of that level. For example, over-expression of Bcl-xL limited ara-C-induced loss of ΔΨm, a phenomenon that has previously been described in HL-60 cells (Wang et al., 1997). However, although subsequent exposure of ara-C-pretreated U937/Bcl-xL cells with bryostatin 1 potentiated the loss of ΔΨm, this phenomenon, as well as the increase in apoptosis, was abrogated by the caspase inhibitor ZVAD-fmk. This suggests that the capacity of bryostatin 1 to increase mitochondrial depolarization in Bcl-xL-over-expressing cells represents a secondary event. In marked contrast, caspase inhibition failed to block enhanced cytochrome c release in Bcl-xL over-expressors after treatment with ara-C followed by bryostatin 1. This indicates that bryostatin 1 overcomes, albeit partially, the ability of Bcl-xL to prevent ara-C-induced cytochrome c redistribution, an event that lies upstream of caspase activation. Moreover, the ability of oligomycin to block both cytochrome c release and its downstream consequences; i.e., caspase activation and apoptosis, further supports a critical role for enhanced mitochondrial release of this protein in ara-C/bryostatin 1-mediated lethality in Bcl-xL-over-expressing cells. It is noteworthy that combined treatment of cells with ara-C and bryostatin 1 resulted in enhanced activation of Smac/DIABLO, an antiapoptotic protein that promotes cell death by antagonizing inhibitor of apoptosis proteins and caspase activation (Srinivasula et al., 2001). Recent studies have suggested that cytochrome c and Smac/DIABLO release may be independently regulated and stimulus-dependent (Chauhan et al., 2001). Based upon the present results, the possibility that the enhanced lethality of the bryostatin 1/ara-C combination in Bcl-xL-over-expressing cells involves an increase in Smac/DIABLO release cannot be excluded.
An alternative explanation for the present findings is that bryostatin 1 might activate the extrinsic, receptor-related apoptotic pathway, which is largely independent of mitochondrial injury, and generally insensitive to inhibition by Bcl-2 or Bcl-xL(Gross et al., 1999). The finding that ectopic expression of CrmA, a potent inhibitor of caspase-8 (Zhou et al., 1997) or administration of the caspase-8 inhibitor IETD-fmk was relatively ineffective in opposing apoptosis induced by ara-C alone compared with that induced by TNFα argues against a major role for the extrinsic pathway in ara-C lethality. In this regard, cytochrome c release, by activating caspase-9 and -3, can induce secondary caspase-8 activation, which then feeds back to trigger Bid cleavage, resulting in further mitochondrial damage (Sun et al., 1999). Thus, engagement of the receptor-related apoptotic pathway can amplify mitochondrial dysfunction induced by various noxious stimuli, including cytotoxic drugs (Slee et al., 1999). It is noteworthy that in contrast to results obtained in cells exposed to ara-C alone, ectopic expression of CrmA did provide partial protection against apoptosis and loss of ΔΨm induced by the combination of ara-C and bryostatin 1. Thus, the concept that activation of the extrinsic pathway contributes to ara-C/bryostatin 1-mediated lethality appears plausible. Finally, the possibility that bryostatin 1/ara-C triggers release of mitochondrial factors such as AIF that activate a caspase-independent pathway of apoptosis cannot be excluded.
The observation that Bcl-2 and related proteins block apoptosis in cells microinjected with cytochrome c (Zhivotovsky et al., 1998) raises the possibility that such antiapoptotic proteins could act downstream of mitochondrial injury; i.e., at the level of caspase activation. In accord with this notion, Hu et al. (1998) reported that Bcl-xL formed a complex with apaf-1, and in so doing, limited its capacity to trigger procaspase-9 cleavage/activation (Hu et al., 1998). Thus, it is possible that bryostatin 1 acts by opposing the putative Bcl-xL/apaf-1 interaction. However, other investigators have called this interaction into question, and have suggested that Bcl-xL/apaf-1 associations may be artifactual (Moriishi et al., 1999). In this context, we were unable to detect apaf-1 coimmunoprecipitating with Bcl-xL in cells treated with either ara-C or with the combination of ara-C and bryostatin 1. Such findings argue against the possibility that bryostatin 1 potentiates ara-C-induced apoptosis in Bcl-xL-over-expressing cells by disrupting Bcl-xL/apaf-1 interactions.
Previous studies have demonstrated that a discordance may exist between the ability of Bcl-2 to protect cells from drug-induced apoptosis and its capacity to restore clonogenic potential (Yin and Schimke, 1995). It is noteworthy that subsequent exposure of ara-C-treated Bcl-xL-over-expressors to bryostatin 1 resulted in a significant reduction in leukemic cell colony formation. Although this effect was less pronounced in cells ectopically expressing Bcl-xL, inhibition of clonogenicity was essentially equivalent to that observed in parental cells treated with ara-C alone. Thus, from a functional standpoint, administration of a marginally toxic concentration of bryostatin 1 restored the ara-C sensitivity of Bcl-xL-over-expressors to wild-type levels as far as self-renewal capacity was concerned. These results are similar to those of previous studies in which the PKC and checkpoint kinase 1 inhibitor UCN-01 (7-hydroxystaurosporine) was combined with ara-C in human leukemic cells over-expressing Bcl-2 (Tang et al., 2000). However, in contrast to the present findings, ectopic expression of Bcl-2 failed to exert any protective effects toward clonogenic leukemic cells exposed sequentially to ara-C followed by UCN-01 (Tang et al., 2000). Whether these divergent results reflect intrinsic differences between the actions of UCN-01 and bryostatin 1 or disparate protective of Bcl-2 versus Bcl-xL(Simonian et al., 1997), remains to be determined.
In summary, the present studies demonstrate that the ability of bryostatin 1 to potentiate ara-C-induced apoptosis in human leukemia cells (U937) over-expressing Bcl-2 can be extended to include cells ectopically expressing the antiapoptotic protein Bcl-xL. Furthermore, this process appears to involve a diminution in the capacity of Bcl-xL to block ara-C-related cytochrome c (and possibly Smac/DIABLO) release or upstream events, rather than potentiation of the loss of ΔΨm or a disruption of Bcl-xL/apaf-1 interactions. In addition, activation of the extrinsic apoptotic pathway by bryostatin 1/ara-C may also contribute to the increased lethality of this drug combination. Lastly, administration of bryostatin 1 enhances the inhibitory effects of ara -C toward the clonogenic growth of Bcl-xL-over-expressing leukemia cells. Given the possibility that Bcl-xL may be a determinant of leukemic cell responsiveness to chemotherapeutic drugs (Schaich et al., 2001), these findings could have implications for the clinical development of regimens combining ara-C with bryostatin 1 in the treatment of acute leukemia.
Footnotes
-
This work was supported by awards CA63753 and CA83705 from the National Institutes of Health and Award LSA 6630-01 from the Leukemia and Lymphoma Society of America. Portions of this work have been presented in preliminary form at the meeting of the American Association for Cancer Research, New Orleans, LA, March 31–April 4, 2001.
- Abbreviations:
- Δψm
- mitochondrial membrane potential
- apaf-1
- apoptosis-activating factor-1
- AIF
- apoptosis-inducing factor
- PKC
- protein kinase C
- ara-C
- 1-β-d-arabinofuranosylcytosine
- PBS
- phosphate-buffered saline
- DiOC6
- 3,3-dihexyloxacarbocyanine iodide
- pNA
- p-nitroanilide
- AMC
- 7-amino-4-methylcoumarin
- PAGE
- polyacrylamide gel electrophoresis
- PARP
- poly(ADP-ribosyl)phosphorylase
- fmk
- fluoromethyl ketone
- TNF
- tumor necrosis factor
- UCN-01
- 7-hydroxystaurosporine
- CrmA
- cytokine response modifier A
- Received October 9, 2001.
- Accepted January 7, 2002.
- The American Society for Pharmacology and Experimental Therapeutics