Activation of c-Jun NH₂-terminal Kinase and Subsequent CPP32/Yama during Topoisomerase Inhibitor β-Lapachone-induced Apoptosis through an Oxidation-dependent Pathway¹

β-Lap³ (3,4-dihydro-2,2-dimethyl-2H-naphtho[1,2-b]pyran-5,6-dione) is a naturally occurring plant quinone obtained from the lapacho tree (Tabebuia avellanedae), which is native to South America (1). β-Lap possesses a variety of pharmacological effects, including antibacterial, antifungal, and antitrypanocidal activities (2, 3, 4). In addition, it prolongs the survival of mice infected with Rauscher leukemia virus, probably through inhibition of reverse transcriptase (5, 6). β-Lap is also a potent DNA repair inhibitor that sensitizes tumor cells to DNA-damaging agents (7, 8). Several lines of evidence suggest that β-Lap can directly target Topo I (9) and Topo II (10) and inhibit their activity, which results in cytotoxicity. However, the inhibitory mechanism of β-Lap is distinct from that of other typical Topo inhibitors, such as CPT. More recent studies have shown that β-Lap potentially induces apoptotic cell death in human leukemia and prostate cancer cell lines in a p53-independent pathway (11, 12). The increase of ROS by β-Lap was mainly attributed to apoptosis in human leukemic HL-60 cells (13). The selective action of β-Lap in inducing apoptosis in human leukemia and prostate cancer cells has implicated its potential clinical utility against both cancers. However, the signaling transduction pathway or pathways that lead to apoptosis program in response to β-Lap in such cell system are not known completely.

Response to numerous types of extracellular signals is mediated by MAPKs, which are members of a serine/threonine kinase family (14, 15). A well-defined MAPK subfamily consists of JNKs, also called stress-activated protein kinases, which are responsible for a variety of stresses (16). Activation of JNKs requires dual phosphorylation at conserved threonine and tyrosine residues by MAPK kinase 4 or JNK kinase, which in turn can be phosphorylated by upstream kinase MEKK1 (17). JNK activity can be induced by diverse stimuli, such as growth factors, cytokines, certain protein synthesis inhibitors, UV irradiation, heat shock, H₂O₂, and osmotic shock (18, 19, 20, 21, 22, 23). Recent emerging evidence suggest that sustained activation of JNK is a requisite for the initiation of apoptosis. The overexpression of MEKK, the JNK kinase kinase, had a lethal effect on fibroblasts (24). In addition, overexpression of activated JNK caused cell death in transfected Jurkat cells. In contrast, expression of a dominant-negative mutant of MAPK kinase 1 or JNK prevented the UV-C- and γ-irradiation-induced cell death (25). Various well-known chemotherapeutic drugs, such as Adriamycin, vinblastine, VP-16, and CPT, are also capable of activating JNK. These drugs are critical in triggering apoptosis program in different cell lines (26, 27). These data suggest that the JNK kinase cascade may dominantly participate in apoptosis.

ICE and related cysteine-proteases, such as CED-3, CPP32/Yama, Ich-1/Nedd2, Ich-2/ICErel-II/TX, or Mch2, are thought to be downstream regulators of apoptosis (28, 29). Overexpression of these proteases leads to apoptosis of various cell types (30). Although some of these proteases seem to constitute a multiple protease cascade, the molecular mechanism of the cascade triggering is poorly understood. Interestingly, a recent study indicated that JNK induction by anticancer drugs can lead to the activation of CED-3-like protease and, ultimately, cell death (27). This finding provided a relevance to delineate the signaling flow from upstream JNK to downstream executor cysteine protease in apoptotic cell death program.

Here, we examined whether the JNK signaling pathway is involved in β-Lap-induced apoptosis in human leukemic HL-60 cells. Our data showed that β-Lap treatment induced sustained activation of JNK but not other MAPK subfamilies, i.e., ERK1/2 and p38, during the cell death process. β-Lap was also found to induce a large amount of ROS but not the other Topo inhibitors, such as VP-16, CPT, and GL331. Antioxidant ascorbic acid can effectively inhibit the production of ROS, JNK activation, and ultimate apoptosis induced by β-Lap; ascorbic acid failed to prevent other Topo inhibitor-induced JNK activation and cell death. Furthermore, expression of a MEKK1-DN or treatment with JNK-specific antisense oligonucleotide can abolish JNK activity and subsequent CPP32/Yama activity. These results indicate that, unlike other Topo inhibitors, β-Lap induces JNK/CPP32-associated apoptosis mediated by an oxidation-dependent fashion.

HL-60 cells, a human promyelocytic leukemia cell line, were obtained from American Type Culture Collection (Rockville, MD). Cells were maintained in a humidified 5% CO₂ atmosphere and cultured in RPMI supplemented with 10% FCS, 2 mm l-glutamine, 100 units/ml penicillin, and streptomycin. β-Lap was prepared according to the procedures described by Schaffner-Sabba et al. (1). β-Lap was dissolved in ice-cold absolute alcohol as a stock solution at 10 mm concentration and stored in aliquots at −20°C. DCFH-DA and HE were obtained from Molecular Probes. Ascorbic acid, CPT, etoposide, and propidium iodide were obtained from Sigma Chemical Co.

Intracellular hydrogen peroxide production was monitored by fluorescence microscopy using DCFH-DA. Briefly, cells (2 × 10⁵) were pretreated with 25 μm Vit C for 1 h and then coincubated with 50 μm DCFH-DA in the absence or presence of anticancer drugs at 37°C for 1 h. After incubation, cells were washed and resuspended in ice-cold PBS and placed on coverglass at dark for fluorescence microscopy (Nikon) analysis.

The fluorochrome HE was used to measure superoxide anion generation as described (31). Briefly, after drugs treatment, cells were incubated at 37°C for 15 min in the presence of 2 μm HE, followed by immediate analysis of fluorochrome incorporation in a FACScan flow cytometry (Becton Dickinson).

Cells were harvested and washed with PBS, and hypodiploid cells were analyzed by flow cytometer as described previously (13).

HL-60 cells were treated with various anticancer drugs or combined with antioxidants for 4 h (indicated in Fig. 1 C). After that, treated cells were harvested and washed with PBS, and DNA fragmentation was analyzed by agarose gel electrophoresis as described previously (13).

Transfection was created by electroporation (model T800; BTX, San Diego, CA) of HL-60 cells with the glucocorticoid-inducible pSRa-MEKK(K432M) vector (a gift from Dr. Michael Karin of the Department of Pharmacology, School of Medicine, University of California, San Diego, La Jolla, CA). Briefly, cells were suspended in 1 ml of HEPES-buffered saline containing plasmid DNA and then received electric treatment as follows: electric amplitude, 900 V; pulse width, 99 ms. After 10 min on ice, the cells were transferred to fresh complete medium and cultured for 24 h before addition of hygromycin. To avoid problems with clonal variation, the transfected cells were selected for hygromycin for 4 weeks, and all of the clones were pooled.

Cell lysis and immune complex kinase assays were performed as described (32). HL-60 cells were treated with different drugs, washed twice with ice cold PBS, and lysed in buffer containing 20 mm HEPES (pH 7.4), 50 mm β-glycerophosphate, 1% Triton X-100, 10% glycerol, 2 mm EGTA, 1 mm DTT, 10 mm sodium fluoride, 1 mm sodium orthovanadate, 1 μg/ml leupeptin, 1 μg/ml aprotinin, and 1 mm phenylmethylsulfonyl fluoride. The soluble extracts were prepared by centrifugation at 14,500 rpm for 15 min at 4°C. Following normalization of protein concentration, equal amounts of protein were incubated with protein A-Sepharose and anti-JNK1 (1 μg; C17; Santa Cruz Biotechnology, Santa Cruz, CA), anti-ERK1 (1 μg; C16; Santa Cruz Biotechnology), or anti-p38 (1 μg; N20; Santa Cruz Biotechnology) for 3 h at 4°C. The immune complexes were washed twice with lysis buffer and then once with kinase assay buffer[20 mm MOPS (pH 7.2), 2 mm EGTA, 20 mm MgCl₂, 1 mm DTT, and 0.1% Triton X-100], following which they were resuspended in 20 μl of kinase assay buffer containing 5 μCi of [γ-³²P]ATP, 30 μm cold ATP, and 2 μg of substrate and incubated for 20 min at 30°C. Reactions were terminated by the addition of the SDS sample buffer and boiling for 5 min. The phosphorylated proteins were resolved by SDS-PAGE and visualized by autoradiography. GST-c-jun(1/79) was used as a substrate for JNK1, myelin basic protein was used for assaying ERK1, and ATF-2 was used as a substrate for p38.

The rationale of JNK1 antisense oligonucleotide design and treatment is based on the report by Seimiya et al. (27). The JNK1-specific antisense (5′-GTCACGCTTGCTTCTGCTCATGAT-3′) and sense (5′-ATCATGAGCAGAAGCAAGCGTGAC-3′) phosphorothioates were synthesized and purified by high-performance liquid chromatography (Genset Co.). These sequences represent amino acids −1 to +7 of JNK1. The oligonucleotides were dissolved in distilled and sterilized water and added into culture medium. After treatment with the oligonucleotides for 16 h, cells were analyzed the JNK1 activity and JNK1 protein level.

Western blot was measured by the method described previously (13). Briefly, cell lysates were prepared, electrotransferred, and then immunoblotted with anti-JNK1, anti-ERK1, anti-p38, anti-PARP, and anti-MEKK1 (C22) antibody (Santa Cruz Biotechnology). Detection was performed with Western blotting reagent ECL (Amersham), and chemiluminescence was exposed by the filters of Kodak X-Omat films.

CPP32 protease activity was measured by the method described previously (30). In brief, cytosolic extracts were prepared by repeated cycles of freezing and thawing in 300 μl of extraction buffer[12.5 mm Tris (pH 7.0), 1 mm DTT, 0.125 mm EDTA, 5% glycerol, 1 mm phenylmethylsulfonyl fluoride, 1 μg/ml leupeptin, and 1 μg/ml aprotinin]. Cell lysates (100 μg) were then diluted with 1 ml of assay buffer[50 mm Tris, 1 mm EDTA, and 10 mm EGTA (pH 7.0)] and incubated at 37°C for 30 min in dark with 10 μm fluorescence substrate, Ac-DEVD-AMC. The fluorescence of the cleaved substrate was determined using a spectrofluorometer (Hitachi F-3000) set at an excitation wavelength of 380 nm and an emission wavelength of 460 nm.

To verify whether β-Lap induces ROS generation, we used DCFH-DA fluorescent dye to detect intracellular peroxides level and HE dye to detect superoxide anion (O₂⁻). Under examination by fluorescence microscope, we found that ∼94+ 1.01% of β-Lap (1 μm)-treated leukemia cells displayed evident DCFH fluorescence after 30 min treatment (Fig. 1,A). No such fluorescence was observed in the control or other Topo inhibitor-treated cells. Ascorbic acid (Vit C) treatment completely prevented the DCFH fluorescence induced by β-Lap. We further examined the O2⁻ level using HE detection and quantitated by flow cytometry. Fig. 1,B shows that ∼19.77+ 0.86% HL-60 cells produced significant amount of O₂⁻ following β-Lap (1 μm) treatment. Similarly, we did not detect the production of O₂⁻ in the control or VP-16-, CPT-, or GL331-treated HL-60 cells (Fig. 1 B). We also observed that Vit C almost blocked the elevation of O₂⁻ by β-Lap.

We then examined whether ROS generation was involved in β-Lap-mediated apoptosis in HL-60 cells. As Fig. 1 C shows, Vit C treatment effectively prevented β-Lap-induced internucleosomal DNA fragmentation, a hallmark of apoptosis, whereas Vit C failed to inhibit CPT-, VP-16-, or GL331-induced DNA fragmentation. The inhibitory effect of Vit C on β-Lap-induced apoptosis was also seen in morphological characteristics by Hoechst 33258 staining (data not shown). Our previous study has indicate that β-Lap induced ∼70% apoptosis of HL-60 cells at 1 μm for 24 h (13).

To explore whether the apoptosis-related signaling pathway(s) are activated in HL-60 cells in response to β-Lap, we examined JNK, ERK1/2, and p38 activities by using immunocomplex kinase assay. Fig. 2,A shows that, following 1 μm β-Lap treatment, JNK activity was detectable and increased at 15 min and was sustained up to 2 h. Western blot analysis showed that these JNK activations were not due to enhanced expressions of JNK protein (Fig. 2,A, bottom). However, under the same dose, we did not detect any significant p38 or ERK1/2 activation during the apoptosis process (Fig. 2, B and C). A kinetic study on DNA fragmentation shows that an initial DNA fragmentation occurred at 2 h after 1 μm β-Lap treatment (Fig. 2 D). These observations indicated that JNK activation preceded the occurrence of DNA fragmentation (apoptosis).

A dose-response experiment for these kinases was also studied. Treatment of HL-60 cells with 0.5, 1, and 2 μm β-Lap induced 1.61-, 7.5-, and 7.1-fold increase of JNK activity, respectively (Fig. 3,A). We did not observe any change in ERK1/2 or p38 activity in HL-60 cells even when the dose was increased up to 2 μm (Fig. 3, B and C). This suggests that JNK but not ERK1/2 or p38 is persistently activated during β-Lap-induced apoptosis.

To verify the role of JNK in β-Lap-induced apoptosis, we transfected and expressed a MEKK1-DN plasmid[pSRα-MEKK1 (K432M)], which was under control by a glucocortcoid-inducible promoter, in HL-60 cells. We subsequently treated the MEKK1-DN-transfected HL-60 (HL-60/MEKK-DN) and parental HL-60 cells with 1 μm β-Lap and then examined the JNK activity and apoptosis. Under the presence of Dex, the β-Lap-stimulated JNK activation was inhibited down to the basal level in HL-60/MEKK-DN cells (Fig. 4,A, top). To confirm the expression of the MEKK1-DN protein, the transfectants were analyzed their MEKK1 protein level under the presence or absence of Dex by Western blotting with an antibody specific to MEKK1. Indeed, HL60/MEKK-DN cells expressed a truncated form of MEKK1 (MEKK▵) protein in the presence of Dex. (Fig. 4 A, bottom). These results indicate that the MEKK1-DN expression vector is functional in suppressing JNK activity.

In the presence of Dex, HL-60/MEKK-DN cells became resistant to DNA fragmentation (apoptosis) induced by β-Lap (Fig. 4,B). In contrast, in the absence of Dex, HL-60/MEKK-DN cells were equally sensitive to β-Lap compared with parental HL-60 cells either with or without Dex (Fig. 4,B). Consistent with other reports (9), we also observed that HL-60/MEKK-DN cells, under the presence of Dex, were resistant to CPT-, VP-16-, and GL331-induced apoptosis (data not shown). Given direct evidence to prove the importance of JNK in β-Lap-induced apoptosis, we treated HL-60 cells with JNK1-specific antisense oligonucleotide. HL-60 cells were treated with 25 μm JNK1-specific antisense oligonucleotide for 16 h before the addition of 1 μm β-Lap for another 4 h. Upon treatment with JNK1 antisense oligonucleotide for up to 16 h, cellular amounts of JNK1 protein decreased in HL-60 cells. Control sense oligonucleotide marginally affected the JNK1 contents (data not shown). Under this condition, the β-Lap induction of apoptosis could obviously be prevented by the JNK1-specific antisense oligonucleotide phosphorothioate but not by its sense oligonucleotide (Fig. 5,A). The antisense oligonucleotide treatment was nontoxic to the HL-60 cells because antisense-treated cells retained membrane integrity and normal proliferation rate for up to 24 h (data not shown). This observation strongly suggests that the reduction of JNK activity by antisense oligonucleotide is not an artifact of nonspecifically toxic to cells. Immunocomplex kinase assay shows again that β-Lap-elicited JNK activity could be blocked by JNK1 antisense oligonucleotide but not by its sense oligonucleotide (Fig. 5 B).

These results showed that interfering with JNK signaling can prevent β-Lap-induced cell death, an indication that JNK kinase cascade is required for cell death signaling.

The results obtained thus far in this study led us to propose that ROS generation may initiate JNK activation. To resolve this issue, we examined the effect of Vit C on JNK activity in HL-60 cells treated with β-Lap or other drugs. As Fig. 6,A shows, β-Lap induced JNK activation was suppressed by the addition of Vit C. However, Vit C did not affect CPT-, VP-16-, or GL331-stimulated JNK activation. Quantitation of sub-G₁ cells by flow cytometer also revealed that Vit C prevented β-Lap-mediated apoptosis but failed against other Topo inhibitors-induced apoptosis (Fig. 6 B). These data suggested that β-Lap initially induces the generation of ROS that can activate JNK, which, in turn, triggered the apoptosis program.

Because CPP32/Yama protease plays an important role in various drug-induced apoptoses, we examined the role of CPP32/Yama in the β-Lap-mediated apoptotic process. To address this issue, we determined the cleavage of PARP protein and CPP32/Yama activity by using Western blotting and spectrofluorometry, respectively. When HL-60 cells were treated with 1 μm β-Lap or CPT, cleavage of PARP protein was obviously detected after exposure to both drugs. However, only β-Lap-mediated PARP cleavage was prevented by Vit C (Fig. 7,A). The anti-PARP antibodies used in this study recognized an epitope at the NH₂ terminus of the PARP polypeptides; thus the NH₂-terminal M_r 30,000 fragment can be detected by immunoblotting. Consistent with the immunoblotting, we found that β-Lap-induced increase of CPP32/Yama activity, which is determined by using a fluorogenic tetrapeptide substrate Ac-DEVD-AMC, was completely inhibited by Vit C. Vit C treatment did not affect CPT-induced CPP32/Yama activation (Fig. 7,B). However, when JNK activity was suppressed by expression of MEKK-DN, both β-Lap- and CPT-induced CPP32/Yama activation were effectively diminished (Fig. 7 C). These data suggest that CPP32/Yama protease acts as a downstream executor of JNK signaling triggered by β-Lap.

In this study, the capability of inducing ROS generation by β-Lap is clearly demonstrated in human leukemia HL-60 cells. In contrast, we did not detect any significant ROS generation in same cell system by treatment with other commonly known Topo inhibitors. This indicates that the mechanism of action of β-Lap is somehow different from those of other Topo inhibitors. According to previous studies, the inhibitory mechanism of β-Lap on Topo I and II is possibly due to directly chemical modification of these enzymes. This is characteristically different from other Topo poisons, which inhibit Topo in an ATP-dependent manner (10). Other findings raised by Li et al. (12) showed that the cytotoxic effect of Topo poisons is always associated with G₂-M arrest. On the one hand, cell death induced by β-Lap in human prostate PC-3 cancer cells was associated with increases in the fractions of G₁, sub-G₁, and cells with higher DNA ploidy. The structure of β-Lap is relatively similar to that of menadione. Frydman et al. (10) did not rule out the possibility that free radical formation is involved in the proposed chemical modification of the Topo-DNA complex. They also found that β-Lap possesses a thiol reactivity under in vitro conditions. These findings seemed to indicate that the characteristic of ROS production is likely one of major different mechanisms between β-Lap and other commonly used Topo poisons. It is also relevant that Topo may be one of targets attacked by β-Lap-induced ROS, which ultimately leads to cell death. Whether the inhibition of Topo by β-lap is through ROS modification remains unknown and need further investigation.

Antioxidant Vit C treatment effectively blocked β-Lap-induced apoptosis but failed to prevent cell death induced by other Topo poisons. Another antioxidant, N-acetyl-l-cysteine, also strongly antagonized the apoptosis induced by β-Lap (data not shown). ROS generation may specifically be involved in β-Lap- but not in other Topo poison-mediated apoptosis programs. Supportive of this finding, many drugs, such as 1-β-d-arabinofuranosylcytosine, ceramide, Adriamycin, and methylprednisolone, induced apoptosis, which is accompanied with an increase of intracellular oxidants and elevated level of lipid peroxidation (22, 26, 33, 34).

Recent studies have suggested that β-Lap is capable of accepting a single electron to form semiquinone radical. Quinone radicals undergo oxidation in the presence of quinone reductase, such as NADH, NADPH, or SOD, to produce H₂O₂ and O₂⁻ in mitochondria or microsomes (35, 36). A decrease of mitochondrial membrane potential was correlated with an elevation of ROS in HL-60 cells following treatment with β-Lap.⁴ These observations strongly indicate that β-Lap may interfere with the mitochondrial function (i.e., electron transport chain) through its thiol reactivity and subsequently lead to the generation of ROS.

The role of JNK activation in apoptosis has been extensively discussed (37). However, this study is the first to demonstrate that the new drug β-Lap can stimulate a rapid and pronounced activation of JNK during the onset of apoptosis. Our data further showed that ROS was fully attributed to β-Lap-induced JNK activation, as evidenced by the blocking effect of Vit C on JNK activity. Although ROS has been reported to activate ERK1/2 or p38 kinase under certain experimental conditions (38), we found in our study that only JNK, not other MAPKs, was stimulated by β-Lap-generated ROS. Because the activities of these three MAPK subfamilies are tightly regulated by their respective upstream kinases (39), an initial molecule responsible for the JNK-related signaling may have been highly reactive to β-Lap-generated ROS. A similar finding by Lo et al. (40) showed that ROS acted as a mediator for cytokine-stimulated JNK activation in bovine chondrocytes. These and our data suggest that overstimulation of JNK activity by excessive production of ROS may have to do with pathological conditions in different cell types in response to exogenous stimuli. Expression of MEKK1-DN significantly abolished β-Lap and other Topo poison-induced JNK activation and apoptosis. This finding is consistent with other previous studies, which suggested that JNK activation plays a prominent role in apoptosis process induced by a diverse type of drugs (41). On the other hand, many studies (42, 43) have recently shown that JNK activation is involved in tumor necrosis factor receptor-associated protein-mediated antiapoptotic signals in lymphocytes upon TNF treatment. These contrary findings suggest that JNK exerts multifaceted roles in regulation of cellular survival and death. ICE/CED-3-like proteases, such as ICE and CPP32, are functionally involved in apoptosis of many cell types (44). Consistent with other studies, we demonstrated that CPP32 is activated during β-Lap-induced apoptosis. Again, expression of MEKK1-DN or treatment with Vit C effectively abolished the β-Lap-induced CPP32 activity in HL-60 cells. Only MEKK1-DN overexpression, not Vit C, was able to antagonize the Topo 1 inhibitor CPT-induced CPP32 activity. This implies a commonly downstream role of CPP32 in JNK signaling. However, a recent study showed that caspase can induce apoptosis by activating MEKK1, which, in turn, activates more caspase activity, comprising a positive feedback loop (45). Whether this feedback regulation mechanism is involved in β-Lap-induced caspase activity and apoptosis remain elusive. Here, we found that the CPP32-specific tetrapeptide inhibitor, Ac-DEVD-CHO, only partially reduced β-Lap-induced apoptosis (data not shown). Thus, it is likely that other ICE-like proteases may also be involved in such cell death processes. Studies by Davis (46) showed that ICE-like cysteine protease does not have a proline-directed serine/threonine residue that is preferentially phosphorylated by ERK/JNK family. Therefore, it seems unlikely that JNK directly phosphorylated and activated the protease. There must be additional transducers that connected the two events, JNK activation and the protease activation. These transducers must be elucidated in a future study.

In conclusion, here we present the first evidence demonstrating that β-Lap induces a novel ROS-dependent apoptosis program. In this study, we delineated the β-Lap-induced apoptosis signaling pathway in which ROS initially generated and, in turn, activated JNK that lead to triggering the CPP32 protease and facilitated apoptosis of HL-60 cells (Fig. 8). However, other Topo inhibitors activated JNK, CPP32, and apoptosis (i.e., not through ROS generation).