Abstract
Some, but not all, of a series of novel pyrrolo-1,5-benzoxazepines (PBOXs) induce apoptosis as shown by cell shrinkage, chromatin condensation, and DNA fragmentation in three human cell lines, HL-60 promyelocytic, Jurkat T lymphoma, and Hut-78 s.c. lymphoma cells. This chemical selectivity, together with the lack of apoptotic activity against rat Leydig cells, argues against a general cell poisoning effect. PBOX-6, a potent member of the series, caused activation of a member of the caspase-3 family of proteases. In addition, the caspase-3-like inhibitor z-DEVD-fmk, but not the caspase-1-like inhibitor z-YVAD-fmk prevented PBOX-6-induced apoptosis, suggesting that caspase 3-like proteases are involved in the mechanism by which PBOX compounds induce apoptosis. The release of cytochromec into the cytosol in HL-60 cells in response to PBOX-6 suggests that this cellular response may be important in the mechanism by which PBOX-6 induces apoptosis. However, reactive oxygen intermediates do not play a key role in PBOX-6-induced apoptosis because neither the free radical scavenger TEMPO nor the antioxidantN-acetylcysteine had any effect on PBOX-6-induced apoptosis. The apoptotic induction seems independent of the mitochondrial peripheral-type benzodiazepine receptor (PBR) that binds these pyrrolobenzoxazepines with high affinity, due to the lack of correlation between their affinities for the receptor and their apoptotic potencies, their high apoptotic activity in PBR-deficient cells such as Jurkats, and their lack of apoptotic induction in PBR-rich rat Leydig cells. These PBOXs also can overcome nuclear factor-κB-mediated resistance to apoptosis. This suggests an important potential use of these compounds in drug-resistant cancers.
Recently, a series of pyrrolo-1,5-benzoxazepines, classed herein as PBOX compounds (Fig. 1), has been synthesized. These compounds are high-affinity ligands for the peripheral benzodiazepine receptor (PBR) and have been used as novel probes to study the physiological role of the PBR. This receptor has been implicated in controlling cell growth, apoptosis, and steroidogenesis, among other functions, but its true physiological function still remains unresolved (Zisterer and Williams, 1997). In a recent study, three compounds from the series were found to inhibit the proliferation of rat C6 glioma and human 1321N1 astrocytoma cells. The antiproliferative effect was found to be mediated by arrest in the G1 phase of the cell cycle (Zisterer et al., 1998).
In this study, while examining the effect of PBOX compounds on the proliferation of human promyelocytic leukemia HL-60 cells, we observed that some PBOX compounds, along with other more commonly used PBR ligands, PK 11195 and Ro5-4864, could induce apoptosis in these cells. Apoptosis is a cell suicide mechanism invoked in disparate situations to remove redundant, damaged, or infected cells. Apoptosis is classically defined by a characteristic set of morphological changes in the cell, including membrane blebbing, cell shrinkage, chromatin condensation, DNA fragmentation, and the packaging of what remains in membrane-enclosed vesicles (Kerr et al., 1972). In this study, we examined the mechanism by which these pyrrolobenzoxazepines could induce apoptosis.
Several studies have demonstrated activation of caspases, a family of cysteine proteases, in different pathways of apoptosis (Polverino and Patterson, 1997). Caspases are clustered into three groups according to their specificities and their biological functions: group I (caspases 1, 4, and 5), group II (caspases 2, 3, and 7), and group III (caspases 6, 8, 9, and 10) (Thornberry et al., 1997). We show that caspase 3-like proteases play an important role in the mechanism of the induction of apoptosis by these PBOX compounds. Cytochrome c is a mitochondrial protein that induces apoptosis when accumulated in the cytosol in response to diverse stress inducers and that can then go on to activate caspase-3 (Kluck et al., 1997). In some cell lines, however, such as multiple myeloma cells, there are at least two pathways that lead to apoptosis, one involving and one not involving cytochrome c release (Chauhan et al., 1997). In HL-60 cells, we demonstrate that apoptotic cell death by a selected PBOX compound, PBOX-6, is correlated with release of cytochrome c into the cytosol. Several observations suggest an involvement of ROI in the signal transduction pathways leading to apoptosis. This mode of cell death is sometimes associated with increases in intracellular reactive oxygen intermediate (ROI) levels and addition of exogenous antioxidants such as N-acetylcysteine can inhibit apoptosis (Buttke and Sandstrom, 1994). In this study, we determined whether PBOX compound-induced apoptosis was affected by the presence of antioxidants.
Because all these PBOX compounds reportedly bind to the PBR, we investigated whether this receptor was involved in the mechanism by which these compounds cause apoptosis. We determined whether there was any correlation between the affinity of these compounds for the PBR in HL-60 cells and the potency with which they induce apoptosis.
Nuclear factor-κB (NF-κB) is a member of the Rel family of transcription factors. NF-κB has been implicated as both a promoter and inhibitor of cell death, depending on the cell type and apoptotic stimulus. The activation of NF-κB is initiated by a variety of stress stimuli, such as tumor necrosis factor (TNF), ceramide, and several chemotherapeutic drugs, which themselves all induce apoptosis (Baeuerle and Henkel, 1994). In addition, some cell lines such as the T-cell lymphoma Hut-78 cells constitutively express high levels of the activated form of NF-κB that render them resistant to apoptosis induced by agents such as TNF and ceramide (Giri and Aggarwal, 1998). The mechanism whereby NF-κB protects against apoptosis is presently unclear. The observation that apoptotic cell death by TNF and other apoptotic agents, which activate NF-κB, is enhanced by the protein synthesis inhibitor cycloheximide, suggests that the activation of NF-κB probably acts by transcriptionally up-regulating genes encoding proteins involved in protection against cell death. An antiapoptotic role of NF-κB also has been suggested from the observation that mice that lack the NF-κB gene die early in embyronic development from massive apoptosis within the liver (Beg and Baltimore, 1996). In this study, we examine whether NF-κB is involved in the mechanism by which the PBOX-compounds induce apoptosis. Finally, we discuss the potential use of these compounds as novel anticancer drugs.
Experimental Procedures
Materials.
HL-60, Jurkat T cells, and Hut-78 cells (all obtained from the European Cell Culture Collection, Salisbury, UK) were grown in suspension culture in RPMI 1640 supplemented with 10% fetal calf serum, gentamycin (0.1 mg/ml), and l-glutamate (final concentration, 2 mM), all obtained from Sigma (Poole, Dorset, UK). R2C rat Leydig cells, obtained from the American Type Culture Collection (Rockville, MD) were grown in modified Waymouth's medium with 10% horse serum as previously described (Garnier et al., 1994). Poly(dI-dC) was from Pharmacia Biosystems (Milton Keynes, UK), T4 polynucleotide kinase and oligonucleotide containing the consensus sequence (5′-GG GAC TTT CC-3′), corresponding to the κ light chain enhancer motif, were purchased from Promega (Southhampton, UK). [γ-32P]ATP (3000 Ci/mmol), [3H]PK 11195 (85.8 Ci/mmol), and enhanced chemiluminescence reagent were from Amersham Pharmacia Biotech (Aylesbury, UK). The benzodiazepine Ro5-4864 [7-chloro-5-(4-chlorophenyl)-1,3-dihydro-1-methyl-2-H-1,4-benzodiazepin-2-one] was obtained from Fluka Chemie AG (Buchs, Switzerland). PK 11195 [1-(2-chlorophenyl)-N-methyl-N-(1-methylpropyl)-3-isoquinoline carboxamide] was a gift from Dr. Alan Doble (Pharmuka Laboratories, Gennevilliers, France). The pyrrolobenzoxazepines 7-[(dimethylcarbamoyl)oxy]-6-phenylpyrrolo- [2,1-d][1,5]benzoxazepine (PBOX-1), 7-[(dimethylcarbamoyl)oxy]-6-p-tolylpyrrolo[2,1-d]-[1,5]benzoxazepine (PBOX-2), 4-acetoxy-5-phenylnaphtho[2,3-b]pyrrolo[1,2-d][1,4]-oxazepine (PBOX-3), 7-acetoxy-6-(1-naphthyl)pyrrolo[2,1-d][1,5]-benzoxazepine (PBOX-4), 4-[(dimethylcarbamoyl)oxy]5-phenylnaphtho[2,3-b]-pyrrolo[1,2-d][1,4]oxazepine (PBOX-5), 7-[(dimethylcarbamoyl)oxy]-6-(1-naphthyl)pyrrolo-[2,1-d][1,5]-benzoxazepine (PBOX-6), and 7-[(methylcarbamoyl)oxy]-6-(1-naphthyl)pyrrolo[2,1-d][1,5]-benzoxazepine (PBOX-7) were synthesized by a strategy described previously (Campiani et al., 1996). The RapiDiff kit was obtained from Diagnostic Developments (Burscough, Lancashire, UK). The caspase 3-like fluorogenic substrate [Ac-DEVD-AMC was obtained from Alexis (Nottingham, UK). The caspase 3-like protease inhibitor z-DEVD-fmk and the caspase 1-like protease inhibitor z-YVAD-fmk were supplied by Calbiochem-Novabiochem (Nottingham, UK). Anti-cytochrome cwas a mouse monoclonal antibody obtained from PharMingen (San Diego, CA). Anti-pro-caspase 3 was a monoclonal antibody from Transduction Laboratories (Lexington, KY). All other reagents were supplied by Sigma.
Apoptosis and DNA Fragmentation Assays.
Cells were seeded at a density of 3 × 105 cells/ml and following treatment with the indicated compound, an aliquot (100 μl) was cytocentrifuged onto glass slides precoated with poly(l-lysine). They were then stained with the RapiDiff kit (eosin/methylene blue) under conditions described by the manufacturer. The degree of apoptosis and necrosis was determined by counting ∼300 cells under a light microscope. At least three fields of view per slide, with an average of ∼100 cells per field, were counted and the percentage of apoptosis and necrosis was determined. Apoptotic cells were characterized by cell shrinkage, membrane blebbing, and nuclear condensation and fragmentation, whereas necrotic cells were identified by cell swelling and loss of cell membrane. DNA isolation and fragmentation assays were performed as previously described (Martin et al., 1995).
Flourogenic Assay of Caspase 3-Like Proteases.
Cells (5 × 106 cells) were harvested by centrifugation, washed in ice-cold PBS, and the pellets resuspended in 200 μl of harvesting buffer [20 mM HEPES, pH 7.5, containing 10% (w/v) sucrose, 0.1% (w/v) 3-[(3-cholamidopropyl)dimethylammino]propanesulfonate, 2 mM dithiothreitol, 0.1% (v/v) Nonidet NP40, 1 mM sodium EDTA, and 1 mM phenylmethylsulfonyl fluoride] supplemented with protease inhibitors (1 μg/ml pepstatin A and 1 μg/ml leupeptin). Following incubation on ice for 10 min, samples were passed up and down 10 times through a 21-gauge needle. Following a further incubation on ice for 10 min, the homogenates were centrifuged at 20,000g for 20 min and the resulting supernatants used to measure caspase 3-like protease activity. This activity was determined by a fluorometric assay with the substrate Ac-DEVD-AMC, which is cleaved by caspase 3-like proteases to release the fluorescent leaving group amino-4-methyl coumarin (AMC). Enzyme extracts (50 μg of protein) were incubated with 100 mM HEPES, pH 7.5, containing 10% (w/v) sucrose, 0.1% (w/v) 3-[(3-cholamidopropyl)dimethylammino]propanesulfonate, 10 mM dithiothreitol, and 20 μM substrate in a total reaction volume of 3 ml. Following incubation for 60 min at 25°C, fluoresence was monitored continuously with a spectrofluorimeter (excitation wavelength 380 nm, emission wavelength 460 nm). The amount of AMC released was determined by comparison with a standard curve generated with known amounts of AMC.
Measurement of Cytochrome c and Pro-Caspase 3 by Western Blot.
Cells (15 × 106) were harvested by centrifugation at 1800g for 10 min at 4°C. After being washed once with ice-cold PBS, the cell pellet was suspended in 100 μl of ice-cold buffer A for assay of cytochromec (20 mM HEPES, pH 7.5, containing 10 mM KCl, 1.5 mM MgCl2, 1 mM sodium EDTA, 1 mM sodium EGTA, 1 mM dithiotreitol, and 0.1 mM phenylmethylsulfonyl fluoride) supplemented with protease inhibitors (5 μg/ml pepstatin A, 10 μg/ml leupeptin, and 2 μg/ml of aprotinin). For the measurement of pro-caspase 3, the cell pellet was resuspended in 200 μl of ice-cold buffer B [PBS containing 1% (v/v) Nonidet NP40, 0.5% (w/v) sodium deoxycholate, 0.1% (w/v) SDS, 3% (v/v) aprotinin, 0.1 mM sodium orthovanadate, and 0.1 mM phenylmethylsulfonyl fluoride). In both cases, cells were left to sit on ice for 15 min and then centrifuged at 20,000g for 20 min. The resulting supernatants were stored at −70°C until measurement of cytochrome c or the disappearance of pro-caspase 3. Protein determination was measured with the Bradford assay (Bradford, 1976). Equal amounts of protein were resolved by SDS-polyacrylamide gel electrophoresis (PAGE) in 15% gels and transferred onto nitrocellulose. Membranes were blocked with PBS/5% (w/v) dry milk and probed with anti-cytochrome c antibody or anti-pro-caspase 3. Blots were washed, incubated with goat anti-mouse IgG peroxidase conjugate, and developed by enhanced chemiluminescence according to the manufacturer's recommendations.
Electrophoretic Mobility Shift Assays.
Cells (1 × 106 cells/3 ml) were treated either with TNF-α or the indicated compounds for various incubation periods and nuclear extracts were prepared as previously described (Boland et al., 1997). Protein determinations were made with the Bradford assay (Bradford, 1976), with BSA as standard. Nuclear NF-κB was assessed by the electrophoretic mobility shift assay with a 22-base pair oligonucleotide containing the human κ-light chain enhancer motif, which had been previously end-labeled [γ-32P]ATP (Boland et al., 1997). Typically, 2 to 4 μg of nuclear extract protein was incubated with radiolabeled oligonucleotide (10,000 cpm) at room temperature for 30 min under conditions as described previously (Boland et al., 1997). NF-κB complexes were resolved on 5% acrylamide gels and identified following autoradiography.
Radioligand Binding Assays.
HL-60 cells were harvested by centrifugation at 600g for 5 min and washing in PBS. The resulting cell pellet was homogenized in 50 mM Tris-HCl buffer, pH 7.4 (2 ml), with an Ultraturrax homogenizer (10 s) and then passed five times through a 21-gauge needle. Cell homogenates (50 μg of protein) were incubated with 0.5 to 50 nM [3H]PK 11195 in 50 mM Tris-HCl buffer, pH 7.4 (incubation buffer), in a total volume of 0.5 ml on ice. Total and nonspecific/nonsaturable binding in each case was determined in the absence and presence of 10 μM unlabeled PK 11195, respectively. All samples were incubated in triplicate for 60 min. The incubation mixtures were then filtered and counted as previously described (O'Beirne and Williams, 1988). When testing the potency of a compound to inhibit [3H]PK 11195 binding, samples were incubated with 2 nM [3H]PK 11195 and various concentrations (0.1 nM to 1 μM) of compound and subsequently treated exactly as described above. The resulting Ki values were then generated by the use of the computer programs EBDA and LIGAND (Munson and Rodbard, 1980).
Results
Pyrrolobenzoxazepines Induce Apoptosis in HL-60 Cells.
Some pyrrolobenzoxazepines were found to induce apoptosis in HL-60 cells. The characteristic morphological effects of apoptosis, i.e., shrinkage of cells, extensive membrane blebbing, condensation of chromatin, and DNA fragmentation, were observed in these cells (Fig.2). To make a direct comparison of potencies, all the PBR ligands were tested in the same experiment at a single concentration (10 μM). Of the PBOX compounds tested, PBOX-3, -4, -5, -6, and -7 were found to be the most potent apoptotic inducers (Fig. 3A). After treatment of HL-60 cells for 16 h with a final concentration of 10 μM drug, the cells exhibited between 25 and 40% apoptosis. The degree of necrosis observed under the same conditions was negligible. At the same time, other members of the PBOX series such as PBOX-1 and -2 elicited no effect on cell viability even at the highest concentration tested (50 μM, limits of solubility), suggesting a structure-activity relationship. Some of the more widely known PBR ligands, PK 11195 and Ro5-4864, also induced apoptosis in HL-60 cells, albeit at a higher concentration (100 μM) (Fig. 3B). These compounds did not induce apoptosis at 10 μM (Fig. 3B). Many of the subsequent experiments described in this study were performed with PBOX-6 as a representative apoptotic PBOX. Incubation with PBOX-6 for 16 h at 50 μM (limits of solubility) did not induce apoptosis in a rat R2C Leydig cell line (data not shown), demonstrating that these novel apoptotic agents do not elicit a general toxic effect.
PBOX-6 caused a dose- (Fig. 4A) and time-dependent (Fig. 4B) induction of apoptosis in HL-60 cells. The morphological signs of apoptosis became apparent at 6 h and increased linearly up to 16 h. Apoptosis was negligible below a final concentration of 5 μM drug. Similar results for time- and dose-dependence were obtained with PBOX-7 (data not shown). PBOX-6-induced apoptosis also resulted in DNA fragmentation. When the DNA from HL-60 cells incubated with PBOX-6 for 24 h was analyzed by agarose gel electrophoresis, it generated a characteristic “ladder pattern” of discontinuous DNA fragments (Fig. 4C).
Apoptosis Induced by PBOX-6 Results in Activation of Caspase 3-Like Proteases.
Several studies have demonstrated activation of caspases in different pathways of apoptosis (Polverino and Patterson, 1997). To directly address the involvement of caspase 3-like proteases in PBOX-6-mediated apoptosis, we studied caspase 3-like activity in HL-60 cells following PBOX-6 treatment. Cytosolic extracts from HL-60 cells treated with PBOX-6 were incubated with the fluorogenic caspase 3-like substrate DEVD-AMC. As shown in Fig.5A, treatment of cells with PBOX-6 caused a dose-dependent activation of caspase 3-like proteases. This protease activity became evident at 6 h and increased linearly up to 16 h (Fig. 5B). This dose-dependent and time-dependent activation of caspase 3-like proteases directly parallels the observed morphological effects of apoptosis induced by PBOX-6, as determined from cytospinning and staining of cells. This result was confirmed by Western blotting, demonstrating that PBOX-6 induces, in a dose-dependent manner, the processing of pro-caspase 3, as monitored by the disappearance of the 32-kDa form of the enzyme (Fig.6), and this correlates with the appearance of the morphological signs of apoptosis.
Inhibition of Caspase 3-Like Proteases Prevents PBOX-6-Mediated Apoptosis.
Caspases are specifically inhibited in vitro and in vivo by cell-permeable tetrapeptides designed to mimic cleavage sites of their respective substrates (Nicholson et al.,1995). Pretreatment of HL-60 cells for 1 h with a caspase 3-like protease inhibitor, z-DEVD-fmk, followed by treatment for a further 8 h with PBOX-6, inhibited both the appearance of the morphological signs of apoptosis (Fig. 5C), and the activity of caspase 3-like proteases (Fig. 5D). No protective effect against apoptosis was observed when cells were pretreated with the caspase 1-like inhibitor, z-YVAD-fmk (results not shown). This would suggest that activity of caspase 3-like proteases is an essential part of the mechanism by which PBOX-6 induces apoptosis in HL-60 cells.
PBOX-6-Induced Apoptosis Causes Cytochrome c Release into Cytosol in HL-60 Cells.
Previous studies have shown that accumulation of cytochrome c in the cytosol occurs in response to multiple apoptotic stimuli and that this released cytochrome c in turn activates caspase 3, thus playing an important part in inducing apoptosis (Kluck et al., 1997). The effect of PBOX-6 treatment of HL-60 cells on accumulation of cytochromec in the cytosol was analyzed. Treatment of HL-60 cells with PBOX-6 under conditions of induction of apoptosis (Fig.7) caused an accumulation of cytochromec in the cytosol, suggesting that its release may be important in the mechanism by which PBOX-6 induces apoptosis.
PBOX-6-Induced Apoptosis Is Not Triggered by Oxidative Stress.
Several observations suggest an involvement of ROIs in the signal transduction pathways leading to apoptosis, and that they lie upstream of cytochrome c release and caspase 3 activation. (Jacobson, 1996). To determine whether the induction of apoptosis in HL-60 cells by PBOX-6 involved the production of ROIs, these cells were pretreated with either a commonly used antioxidant NAC (5 mM) or the free radical scavenger 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO; 1 μM) for 30 min before incubation with PBOX-6 for a further 8 h (Fig.8). Both NAC and TEMPO, at the concentrations used, have previously been shown to prevent the formation of ROIs in HL-60 cells (Kakeya et al., 1998). Neither of these compounds was found to protect against PBOX-6-induced apoptosis, suggesting that the mechanism by which this compound causes apoptosis in HL-60 cells does not involve the production of ROIs.
Pyrrolobenzoxazepine-Induced Apoptosis Is Independent of PBR.
During apoptosis, cytochrome c has been reported to exit the mitochondria through the mitochondrial porin channel (also called the voltage-dependent anion channel or VDAC) (Shimizu et al., 1999). Because VDAC is associated with the PBR in the mitochondrial outer membrane (McEnery et al., 1992), and because all these PBOX compounds can bind with high affinity to the PBR, we examined whether this receptor was involved in the mechanism by which these compounds cause apoptosis. HL-60 cell homogenates displayed saturable, high-affinity binding of [3H]PK 11195, a selective ligand for the PBR, yielding Kd andBmax values of 17.7 ± 5.0 nM and 9.5 ± 1.5 pmol/mg protein, respectively. All of the PBOX compounds were shown to inhibit [3H]PK 11195 binding to HL-60 homogenates with Kivalues between 1 and 7 nM (Table 1), yet micromolar concentrations of these compounds were required to induce apoptosis. All the ligands totally inhibit [3H]PK 11195 binding at high concentrations, showing them to be fully competitive inhibitors. In addition, although PBOX-1 and PBOX-2 bind to the PBR withKi values in the nanomolar range, these drugs did not induce apoptosis. Furthermore, the effect of PBOX compounds on a human T Jurkat cell line was examined, a cell line previously shown to the lack the PBR (Carayon et al., 1996), an observation confirmed by us by ligand-binding studies (data not shown). Figure 9 shows that apoptosis is induced in Jurkat cells treated with PBOX compounds, with similar potency to that observed in treated HL-60 cells. Finally, PBOX-6 was unable to induce apoptosis in the rat R2C Leydig cells, which have been previously shown to be PBR rich (Garnier et al., 1994). These results suggest that the mechanism by which these PBOX compounds induce apoptosis does not involve their interaction with the PBR.
Pyrrolobenzoxazepines Can Overcome NF-κB-Mediated Resistance to Apoptosis.
NF-κB activation has been implicated in induction of resistance of tumor cells to apoptosis (Giri and Aggarwal, 1998). Pretreatment of HL-60 cells with TNF for 30 min, at a concentration which activated NF-κB but did not induce apoptosis, followed by treatment with PBOX-6 for 16 h, afforded no protection against PBOX-6-induced apoptosis (Fig. 10A). We then examined a human Sezary lymphoma cell line, Hut-78, which constitutively expresses NF-κB, and as such reportedly causes these cells to be resistant to apoptosis (Giri and Aggarwal, 1998). The five PBOX drugs tested induced apoptosis in Hut-78 cells with similar potency to that observed in HL-60 cells (Fig. 10B). In addition, PBOX-6 had no effect on NF-κB in these cells (Fig. 10D). These results would imply that activation of NF-κB does not necessarily cause resistance to apoptosis by all apoptotic-inducing drugs.
The activation of the transcription factor NF-κB is initiated by a variety of stress stimuli such as ceramide and H2O2, which themselves cause apoptosis (Baeuerle and Henkel, 1994). We tested whether PBOX-6 itself could affect NF-κB expression in HL-60 cells. We used TNF treatment as a positive control for activation of NF-κB in these cells. Although TNF, in agreement with other reports, was shown to activate NF-κB in HL-60 cells, as was demonstrated by the detection of protein-DNA complexes in nuclear extracts, PBOX-6 failed to affect NF-κB expression (Fig. 10C). These results would suggest that PBOX-6-induced apoptosis uses a NF-κB-independent mechanism.
Discussion
In this study, we describe how a new class of apoptotic agents, PBOXs, induce cell death in a number of human leukemia and lymphoma cell lines. The morphological characteristics associated with apoptosis with previously defined criteria (Kerr et al., 1972) such as cell shrinkage, chromatin condensation, membrane blebbing, and DNA fragmentation was observed. Of all the compounds screened, PBOX-3, -4, -5, -6, and -7 were found be most potent, whereas PBOX-1 and -2 had no effect. This chemical selectivity, together with the lack of apoptotic activity against rat Leydig cells, argues against a general cell poisoning effect.
Multiple lines of evidence indicate that apoptosis can be triggered by the activation of a set of death effector cysteine proteases called caspases with specificity for Asp-X bonds. Some experimental observations would however suggest that a caspase-independent mechanism for commitment to cell death also exists. For example, overexpression of a proapoptotic protein such as Bax in mammalian cells can induce DNA condensation and membrane alterations leading to apoptosis, without any caspase activation (Xiang et al., 1996). In this study, we determined whether caspase-3 like proteases were involved in cell death induced by PBOX-6. PBOX-6 caused a dose- and time-dependent activation of caspase 3-like proteases that directly correlated with the observed morphological effects of apoptosis induced by this compound. The caspase 3-like protease inhibitor z-DEVD-fmk prevented both the activity of caspase 3-like proteases and the appearance of the morphological signs of apoptosis, whereas the caspase 1-like protease inhibitor z-YVAD-fmk had no effect. This would suggest the involvement of caspase 3-like proteases in the mechanism by which PBOX-6 induces apoptosis in HL-60 cells.
Cytochrome c is a mitochondrial protein that induces apoptosis when accumulated in the cytosol in response to diverse stress inducers (Kluck et al., 1997; Yang et al., 1997). This protein also has been shown to cause apoptosis when added to cell free extracts (Liu et al., 1996). In some cell lines however, such as multiple myeloma cells, there are at least two different pathways that lead to apoptosis, one involving and one not involving cytochrome c release from mitochondria (Chauhan et al., 1997). These researchers studied the role of cytochrome c in dexamethasone-, anti-Fas mAb-, and ionizing radiation-induced apoptosis, and they demonstrated that although ionizing radiation-induced apoptosis is associated with an increase in cytosolic cytochrome c levels, apoptosis induced by the two other agents had no detectable effect on cytochromec release. In addition, there are many reports that during apoptosis, accumulation of cytochrome c in the cytosol results in the activation of caspase 3-like proteases (Chu et al., 1997; Kluck et al., 1997), although pathways leading to caspase 3 activation without cytochrome c release also have been described (Chauhan et al., 1997). In this study, PBOX-6-induced apoptosis in HL-60 cells was associated with an accumulation of cytochrome c in the cytosol. This result indicates that release of cytochrome c from the mitochondria may be important for triggering apoptosis in response to PBOX-6.
Apoptosis is sometimes associated with increases in intracellular ROI levels and addition of exogenous antioxidants such as NAC can inhibit apoptosis (Buttke and Sandstrom, 1994). The specific molecular mechanisms involved, however, remain to be elucidated. In this study, it has been shown that PBOX-6-induced apoptosis in HL-60 cells was unaffected by the presence of either NAC or the presence of the spin trap and free radical scavenger TEMPO. This would suggest that PBOX-6-induced apoptosis is not mediated by ROIs. This is in agreement with recent reports that have indicated that ROIs are not necessarily a requirement for apoptosis. For example, programmed cell death induced by the Fas ligand or by staurosporine do not appear to require the generation of ROIs, and are not inhibited by the use of antioxidants (Jacobson and Raff, 1995).
Recently, there has been some suggestion that the PBR may be involved in apoptosis. A group of workers has shown that PK 11195, a prototypic ligand of the PBR, facilitates the induction of apoptosis by a variety of stimuli in a number of cell types, including thymocytes and the T-cell leukemia CEM cells (Hirsch et al., 1998). However, PK 11195 by itself had no apoptotic effect. In addition, it has been recently reported that during apoptosis, cytochrome c can exit the mitochondria through VDAC (Shimizu et al., 1999), which is itself associated with the PBR in the mitochondrial outer membrane (McEnery et al., 1992). In this study, we describe how PK 11195 and the PBOX compounds induce apoptosis by themselves, independently of other apoptosis-inducing stimuli. It is unlikely however that the PBR is involved in the mechanism by which these PBOX compounds induce apoptosis. Much higher concentrations of the compounds were required to induce apoptosis than were necessary to saturate the receptor. Furthermore, some of the compounds, e.g., PBOX-1 and PBOX-2 did not induce apoptosis, yet all of these compounds bind to the receptor with similar affinity. In addition, PBOX-6 could not induce apoptosis in the PBR-rich rat Leydig R2C cell line. Finally, we have shown that some PBOX compounds induce apoptosis in Jurkat cells that have previously been shown to be devoid of the PBR (Carayon et al., 1996), a result that we have confirmed. These studies demonstrate that the apoptotic effects of the PBR ligands are incompatible with PBR involvement.
Several recent articles have shown that activation of the transcription factor NF-κB is linked to apoptosis (Beg and Baltimore, 1996; Giri and Aggarwal, 1998). There have been some suggestions that this factor, once activated, plays an antiapoptotic role, most likely by inducing expression of gene products such as cIAP2 (cellular inhibitor for apoptosis) (Chu et al., 1997) that inhibit the apoptotic pathway. However, a general role for NF-κB as a transcription factor that prevents apoptosis is far from established. The activation of NF-κB is initiated by a variety of stress stimuli, such as TNF, ceramide, and daunorubicin, which themselves cause apoptosis (Boland et al., 1997). In this case NF-κB, activation may then cause cell death. Thus, the role of NF-κB as a promoter or inhibitor of cell death may depend on both the cell type and the apoptosis-inducing stimulus.
In this study, we have shown that although PBOX-6 induced apoptosis in both HL-60 and Hut-78 cells, this compound did not affect NF-κB levels. Hut-78 cells constitutively express NF-κB and as such have been reported to be resistant to a range of stress stimuli, including TNF, lipopolysacharide, H2O2, ceramide, and okadaic acid (Giri and Aggarwal, 1998). In this study, the PBOX compounds induced apoptosis in both HL-60 and Hut-78 cells, with similar potency. Furthermore, pretreatment of HL-60 cells with TNF at a concentration that activates NF-κB afforded no protection against PBOX-6-induced apoptosis. It can thus be concluded that PBOX-6-induced apoptosis most likely uses a NF-κB-independent mechanism. Furthermore, this study argues against a general role for NF-κB as a transcription factor that prevents cell death.
Several anticancer drugs such camptothecin (Piret and Piette, 1996), etoposide (Perez et al., 1997), and the anthracycline antibiotics daunorubicin and doxorubicin (Das and White, 1997) activate NF-κB in addition to inducing cell death. Daunorubicin is widely used in cancer chemotherapy and although its mechanism of antitumor action has not been fully elucidated, it ultimately induces apoptosis in cells. The concomitant activation of NF-κB may counteract the therapeutic effects of these chemotherapeutic compounds. Therefore, anticancer drugs that do not activate NF-κB may result in more effective anticancer treatments.
It is evident that apoptosis can be induced by a variety of drugs with diverse chemical structure and different mechanisms of action. Apoptosis-inducing agents include a wide range of anticancer drugs, including inhibitors of the mitotic spindle apparatus such as vinca alkaloids, inhibitors of DNA synthesis such as aphidicolin, and drugs such as campothecin that cause protein-associated DNA strand breaks mediated by DNA topoisomerase I (Sen and D'Incalci, 1992). All these drugs have been shown to induce apoptosis in cancerous cells derived from the hemopoietic system such as HL-60 cells (Sen and D'Incalci, 1992). We propose that these PBOXs may be potential anticancer drugs. The observation that these compounds do not activate NF-κB also may result in them being more effective anticancer agents. The effect of these PBOX compounds on tumors in in vivo animal models are warranted to evaluate their anticancer potential.
In conclusion, we have described a series of novel apoptotic agents and indicated the potential of these compounds for use as anticancer drugs. Although we have elucidated some of the mechanisms by which these compounds induce apoptosis, more work is required to piece together the exact events occurring upstream of cytochrome c release and caspase 3-like protease activation.
Acknowledgments
We thank Marion Boland, Adrienne Gorman, and Luke O'Neill for useful advice.
Footnotes
-
Send reprint requests to: Dr. Daniela M. Zisterer, Biochemistry Department, Trinity College Dublin, Ireland. E-mail:dzistrer{at}tcd.ie
-
↵1 This study was supported by BioResearch Ireland, National Pharmaceutical Biotechnology Center.
- Abbreviations:
- PBR
- peripheral-type benzodiazepine receptor
- ROI
- reactive oxygen intermediate
- NAC
- N-acetylcysteine
- NF-κB
- nuclear factor-κB
- TNF
- tumor necrosis factor
- AMC
- amino-4-methyl coumarin
- PAGE
- polyacrylamide gel electrophoresis
- VDAC
- voltage-dependent anion channel
- TEMPO
- 2,2,6,6-tetramethyl-1-piperidinyloxy
- PBOX
- pyrrolo-1,5-benzoxazepine
- fmk
- fluoromethyl ketone
- Received July 9, 1999.
- Accepted November 8, 1999.
- The American Society for Pharmacology and Experimental Therapeutics