Apoptosis, a programmed process of cell suicide, has been proposed as the most plausible mechanism for the chemopreventive activities of selenocompounds. In our study, we found that Se-methylselenocysteine (MSC) induced apoptosis through caspase activation in human promyelocytic leukemia (HL-60) cells. Measurements of cytotoxicity, DNA fragmentation and apoptotic morphology revealed that MSC was more efficient at inducing apoptosis than selenite, but was less toxic. Moreover, MSC increased both the apoptotic cleavage of poly(ADP-ribose) polymerase (PARP) and caspase-3 activity, whereas selenite did not. We next examined whether caspases and serine proteases are required for the apoptotic induction by MSC. A general caspase inhibitor, z-VAD-fmk, dramatically decreased cytotoxicity in MSC-treated HL-60 cells and several other apoptotic features, such as, caspase-3 activation, the apoptotic DNA ladder, TUNEL-positive staining and the DNA double-strand break. Interestingly, a general serine protease inhibitor, AAPV-cmk, also effectively inhibited MSC-mediated cytotoxicity and apoptosis. These results demonstrate that MSC is a selenocompound that efficiently induces apoptosis in leukemia cells and that proteolytic machinery, in particular caspase-3, is necessary for MSC-induced apoptosis. On the other hand, selenite-induced cell death could be derived from necrosis rather than apoptosis, since selenite did not significantly induce several apoptotic phenomena, including the activation of caspase-3.
Selenium is an essential trace element with several important biological functions. Recently, studies on selenium have been focused on its chemopreventive activity (1). Although several mechanisms have been proposed for the chemopreventive effects of selenocompounds (2–5), their inhibitory effect on cell proliferation, with a preference for tumor cells versus non-transformed cells, is considered to be the most feasible mechanism (6). Therefore, it has been suggested that selenium-induced apoptosis in cancer cells is related to its chemopreventive activity. Several groups have shown that selenocompounds induce apoptosis in cell culture systems (7–9).
Various forms of selenium compounds have been reported to be effective at chemoprevention, and selenate, selenite, selenium dioxide and selenomethionine have been shown to be effective in dimethylbenz[a]pyrene-induced mammary tumor models (10). It has also been reported that methylated selenocompounds including Se-methylselenocysteine (MSC) have anticarcinogenic activities in the same dimethylbenz[a] pyrene model (11). A recent study of methylseleninic acid, which is a simplified version of MSC, without an amino acid moiety, proposed that a monomethylated selenium metabolite is important for cancer chemoprevention (12). Several organic selenocompounds have shown considerable chemopreventive activities but fewer side effects than selenite (13–17). These organoselenium compounds were effective at inhibiting chemically induced tumorigenesis and the formation of DNA adducts (18). However, the mechanisms underlying the inhibitory actions of selenocompounds upon chemically induced tumor formation remains unclear. Recently, apoptosis has been recognized as an important determinant of the response of leukemia cells to chemotherapeutic agents (19). Therefore, the manipulation of apoptosis can probably provide novel strategies for cancer chemoprevention or chemotherapy in general, either by selectively activating apoptosis in malignant cells or by suppressing apoptosis in normal cells.
Our previous studies have shown that selenium compounds are potent inducers of apoptosis and exert a cytotoxic effect in a dose-dependent manner in HL-60 cells (9). However, the intracellular responses to treatment with selenocompounds and their mechanisms have not been clearly elucidated. In the present study, we examined whether cell death induced by selenocompounds is mediated by apoptosis and if the proteolytic machinery, including caspases and serine proteases, is involved in this process. HL-60 cells are used as a valid model system for testing antileukemic and general antitumor compounds (20). Therefore, we investigated the mechanisms underlying apoptosis induced by treatment with MSC and selenite in HL-60 cells. Selenite was used as a negative control in comparisons with MSC.
Materials and methods
Chemicals and cell culture
Sodium selenite was obtained from Sigma Chemical Co. (St Louis, MO, USA). Se-methylselenocysteine (MSC) was chemically synthesized as previously described (13,21). [3H]-NAD+ was purchased from NEN Life Science Products (Boston, MA, USA), and [3H]thymidine was obtained from Amersham Pharmacia Biotech (Buckinghamshire, UK). The general caspase inhibitor, N-benzyloxycarbonyl-Val-Ala-Asp-fluoromethylketone (z-VAD-fmk), and the serine protease inhibitor, Ala-Ala-Pro-Val-chloromethylketone (AAPV-cmk) were purchased from Enzyme System Products (Livermore, CA, USA).
Human promyelocytic leukemia (HL-60) cells were obtained from the American Type Culture Collection. Cells were maintained below 5×105 cells/ml in suspension at 37°C in a humidified atmosphere of 5% CO2 – 95% air in RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum (FBS), 50 μg/ml gentamycin, 18 mM sodium bicarbonate and 10 mM N-(2-hydroxy)piperazine-N-(2-ethanesulfonic acid). In the experiment, to determine the effects of protease inhibitors, cells were pretreated with 50 μM each inhibitor and 0.1% DMSO (vehicle) for 1 h prior to treatment with selenocompounds. Cells were counted using a haemacytometer.
Cytotoxicity was measured by WST-1 and the lactate dehydrogenase (LDH) release assay, as described previously (9). The tetrazolium salt WST-1 assay (4-[3-(4-lodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate), purchased from Boehringer Mannheim (Germany), was performed according to the manufacturer's instructions. In brief, HL-60 cells, at 1×105 cells/ml per 200 μl culture medium, and various amounts of MSC and selenite were placed in a 96-well microtiter plate. After a 24 h incubation, 20 μl ready-to-use WST-1 solution was added and incubation continued for 2 h at 37°C. Absorbance was read at 450 nm in a microplate reader. Cytotoxicity (%) was calculated using the following formula.
Lactate dehydrogenase (LDH) release assay was performed using a Cytotox96 Assay kit (Promega, Madison, WI, USA). This assay quantitatively measures the activity of LDH, a stable cytosolic enzyme that is released upon cell lysis. After cells were treated with selenocompounds, as explained in the WST-1 method, the assay was performed according to the manufacturer's instructions. Cytotoxicity (%) was calculated using the following formula.
In all cases, test compounds were added to wells containing medium alone as blanks, to determine whether these compounds interfere with the two assays.
Quantification of DNA fragmentation
DNA fragmentation assay was carried out as previously described (22) with some modifications. Log-phase HL-60 cells were labeled with 1 μCi/ml [3H]thymidine for 24 h and washed twice with media. 1×105 labeled cells in 200 μl culture medium were incubated with MSC and selenite for 18 h, and the culture medium removed by centrifugation and saved. Cells were lysed with lysis buffer containing 10 mM Tris–HCl pH 8.0, 20 mM EDTA and 0.5% Triton X-100 for 30 min on ice, and then centrifuged at 10 000 g for 30 min at 4°C to separate the intact chromatin (pellet) from the cytoplasmic fragmented DNA (supernatant). The supernatant was then mixed with the saved culture medium and the pellets were resuspended in 0.2 ml Soluene P-350 (Packard Instrument Co. (Moridan, CT, USA)). Radioactivities of the pellets and supernatant samples were measured using a Beckman LS3801 liquid scintillation counter. Fragmentation (%) was calculated from the fraction of supernatant d.p.m. to total d.p.m.
Identification of apoptotic cell morphologies
To analyze the appropriate features of apoptotic cells, terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick-end labeling (TUNEL) assays were carried out with an Apoptosis Detection System (Promega (Madison, WI, USA)) as previously described (23) and according to the manufacturer's instructions. Cells were washed with PBS, and resuspended at a concentration of 2×106 cells/ml in PBS. A 50 μl aliquot of cell suspension was dropped onto a pre-cleaned PolyPrep slide coated with poly-l-lysine (Sigma Chemical Co (St Louis, MO, USA). Samples were fixed in ice-cold 10% TCA for 15 min, and 70% ethanol, 90% ethanol and absolute ethanol for 3 min, and permeabilized by immersion in 0.2% Triton X-100 in PBS for 5 min on ice. Thereafter the slide was incubated for 10 min at room temperature in equilibration buffer (200 mM potassium cacodylate, pH 6.6, 25 mM Tris–HCl, pH 6.6, 0.2 mM DTT, 0.25 mg/ml BSA and 2.5 mM CoCl2) and the enzymatic reaction performed for 60 min at 37°C in reaction buffer, which was the equilibration buffer supplemented with 10 μM dATP, 1 mM Tris–HCl (pH 7.5), 0.1 mM EDTA, 5 μM fluorescin-12-dUTP and 25 μU TdT (based on final concentrations). After washing with 2× SSC buffer for 15 min and three times for 5 min with PBS, the slide was mounted with PBS containing 10% glycerol and 0.02% sodium azide. Observation of apoptosis-related morphologies was carried out using a Microphot-FXA fluorescence microscope (Nikon).
Measurement of caspase-3 activity
Caspase-3 activity was measured according to the manufacturer's protocol (Promega, Madison, WI, USA). Treated and untreated HL-60 cells were lysed with chilled cell lysis buffer. After micro-centrifugation (14 000 r.p.m. for 20 min at 4°C), 50 μg total protein from the clear supernatant was mixed with 32 μl caspase assay buffer, 2 μl DMSO, 10 μl 100 mM DTT and 2 μl 10 mM Asp-Glu-Val-Asp-p-nitroanilide (DEVD-pNA). After incubation at 37°C for 4 h, samples were read at 405 nm.
Detection of DNA ladder formation
2×106 cells in 60 mm dishes were pretreated with protease inhibitors for 1 h, and treated with 50 μM MSC for 24 h. Treated cells were washed twice with ice-cold PBS and resuspended in lysis buffer containing 10 mM Tris–HCl (pH 8.0), 20 mM EDTA and 0.5% Triton X-100, and incubated for 30 min on ice. After centrifugation at 10 000 g for 30 min at 4°C, DNA was extracted five times with phenol and once with chloroform, precipitated with 0.1 vol 3 M sodium acetate and 2.5 vol ethanol, and stored at –20°C overnight. DNA was pelleted by centrifugation at 10 000 g for 5 min at 4°C, rinsed with 70% ethanol and then resuspended in TE buffer (pH 8.0) containing 30 μg/ml RNase. Contaminated RNA was removed by incubation for 6 h at 37°C. DNA was loaded into a 1.8% agarose gel at 100 V for 2 h, stained with 50 μg/ml ethidium bromide and visualized under UV light.
Control and treated HL-60 cells (2×106 cells of each) were collected, washed twice with ice-cold PBS, and lysed with 50 mM Tris–HCl (pH 8.0), 150 mM NaCl, 1% Triton X-100, 100 μg/ml phenylmethylsulfonyl fluoride (PMSF) and 1 μg/ml aprotinin. After incubation on ice for 30 min, the lysate was clarified by centrifugation at 14 000 r.p.m. for 10 min at 4°C. An aliquot of total protein (40 μg/lane) was separated by SDS–PAGE and transferred to a PVDF membrane (Millipore, Bedford, MA, USA). The membrane was stained with 0.1% Ponceau S (Sigma) solution to confirm equal loading and transfer. Immunoblotting was performed using anti-PARP (rabbit polyclonal; Boehringer Mannheim, Germany), anti-caspase-3 antibody (mouse monoclonal; Santa Cruz Biotech, Santa Cruz, CA, USA) and the secondary antibody of HRP-linked anti-rabbit (Pharmacia Biotech, Amersham, Buckinghamshire, UK) or mouse IgG (Pierce Chemical Co., Rockford, IL, USA). The immune complexes were detected using an enhanced chemiluminescence (ECL) system (Amersham Pharmacia Biotech, Buckinghamshire, UK).
Simultaneous detection of DNA double- and single-strand breaks
To assess DNA strand breaks mediated by treatment with selenocompounds, a filter elution assay was performed, as previously described (24). Exponentially growing HL-60 cells were exposed to 0.1 μCi/ml [3H]thymidine for 24 h and treated with selenocompounds. The filter elution protocol was based on the gravity flow method. Briefly, 1×106 of treated cells were loaded per column, each equipped with a polycarbonate filter (2 μm pore size; Millipore). The neutral eluate fraction (N) was collected by applying the neutral lysis buffer (2% SDS, 25 mM EDTA and 0.1 M glycine, pH 10) to each column, and weighed to estimate the elution volume. Thereafter alkaline elution buffer (0.1% SDS, 20 mM EDTA and 0.8% (v/v) tetraethylammonium hydroxide, pH 12.3) was added to each column to collect the alkaline eluate (A) and the elution volume was estimated by weight. DNA retained on the filter (F) was retrieved and an aliquot of each fraction or the entire filter was counted in a liquid scintillation counter (Beckman LS 3801). The elution parameters were defined as
Effects of MSC or selenite on cytotoxicity and DNA fragmentation
To analyze the chemopreventive activity of MSC and selenite, we first examined their effects on cytotoxicity and apoptosis in HL-60 cells. The cytotoxicity assay involved two different assays. As was the case for the MTT assay, the WST-1 assay was based on the measurement of succinate-tetrazolium reductase activity in viable cells; therefore, it represents the number of metabolically active cells in the culture (25). On the other hand, LDH release assay measures the activity of LDH, which is released upon cell lysis (26). Treatment with MSC and selenite increased cytotoxicity in a dose-dependent manner (Figure 1). Moreover, cytotoxicity was increased more severely by selenite than MSC. The concentrations of MSC and selenite required to induce 50% cytotoxicity were approximately 85 (WST-1) or 165 μM (LDH release) and 15 (WST-1) or 18 μM (LDH release), respectively. This result demonstrates that MSC treatment caused extensive loss of metabolic activity rather than cell lysis, while selenite influenced both.
MSC- and selenite-mediated apoptosis was monitored by quantifying DNA fragmentation. Interestingly, this assay showed that MSC was more efficient at inducing apoptosis, despite its lower cytotoxicity (Figure 1A). Although selenite also increased DNA fragmentation, its cytotoxicity was more marked (Figure 1B). In addition, a similar pattern of DNA fragmentation was observed in the gel-electrophoretic DNA ladder assay (data not shown). On the basis of these observations, especially apoptosis induction, we decided upon concentrations of MSC and selenite, which were used throughout the subsequent experimentation. The concentrations of MSC and selenite were determined as 50 and 20 μM, respectively, at which levels MSC and selenite caused a similar level of DNA fragmentation (47% for 50 μM MSC and 41.5% for 20 μM selenite).
Effect of MSC or selenite on apoptotic morphologies
The apoptotic features of MSC- and selenite-treated cells were further examined by TUNEL staining (Figure 2). Treatment with 50 μM MSC increased the number of TUNEL-positive apoptotic cells with typical morphologies, including chromatin condensation, nuclear collapse, membrane blebbing and apoptotic body formation. Treatment with 20 μM selenite also increased the TUNEL-positive DNA fragmentation and chromatin condensation, but did not induce membrane blebbing and apoptotic body formation. These results demonstrate that MSC is a more potent apoptosis-inducing compound than selenite, as selenite did not induce the full range of apoptotic features.
Treatment of MSC activates caspase-3
Caspases are known to execute apoptosis in a variety of systems (27,28). However, the role of caspases in selenium-treated apoptotic cells has not been established. Since the above results imply that caspases might be involved in the apoptotic process of MSC- or selenite-treated cells, we next examined the activation of caspase-3, one of the important effector caspases. As shown in Figure 3A, cells treated with MSC increased the apoptotic cleavage of PARP, an endogenous substrate of caspase-3, in a dose-dependent manner. Activated caspase-3 (17 and 20 kDa fragments) was increased in accord with the appearance of the apoptotic fragment of PARP (85 kDa) and a decreased level of 32 kDa procaspase-3 (Figure 3A, upper panel). The activation of caspase-3 by MSC was also confirmed directly by measuring the activity of caspase-3, using the specific substrate DEVD-pNA (Figure 3A, lower panel). In a time-course study, the 85 kDa fragment of PARP and activated caspase-3 began to increase 8 h after MSC treatment (Figure 3B, upper panel). Caspase-3 activity was also elevated by MSC treatment in a time-dependent manner (Figure 3B, lower panel). However, when cells were treated with selenite, neither PARP processing nor caspase-3 activation was observed (Figure 3). Based on these results, we conclude that caspase-3 plays an important role in MSC-induced apoptosis, and that the two selenocompounds, MSC and selenite, have different apoptosis inducing mechanisms.
Effect of protease inhibitors on MSC- or selenite-mediated cytotoxicity
To elucidate the roles of the proteolytic machinery in selenium-induced apoptosis, we used two kinds of protease inhibitors, one was z-VAD-fmk, a general caspase inhibitor and the other AAPV-cmk, a general serine protease inhibitor. As shown in Figure 4, pre-treatment with 50 μM z-VAD-fmk recovered the cytotoxicity of MSC-treated cells to the level of the control group (asterisk). Moreover, 50 μM AAPV-cmk also reduced the cytotoxicity to half that of the MSC-treated cells. As expected, z-VAD-fmk and AAPV-cmk had negligible effects on the cytotoxicity of selenite-treated cells. These results indicate that the elevated cytotoxicity, caused by treatment with MSC, is clearly related to the increased activity of caspases and serine proteases, and eventually results in cell death.
Protease inhibitors suppress apoptotic features induced by MSC
MSC-mediated apoptosis was further analyzed by performing experiments with caspase and serine protease inhibitors. As expected, z-VAD-fmk effectively reduced the cleavage of PARP and the processing of caspase-3 (Figure 5A). Due to treatment with z-VAD-fmk, the 85 kDa apoptotic fragment of PARP and the fragments of activated caspase-3 (20 and 17 kDa) decreased. In this experiment, a general serine protease inhibitor, AAPV-cmk also efficiently inhibited caspase-3 activation and PARP cleavage. Interestingly, the inhibitory effect of AAPV-cmk was similar to that of z-VAD-fmk. In addition, DNA ladder formation and TUNEL-positive cells, which were induced by MSC treatment, were effectively decreased by pre-treatment with both z-VAD-fmk and AAPV-cmk (Figure 5B and C). These results confirm previous observations that caspase-3 is crucial for inducing apoptosis by MSC treatment, and further suggest the possibility that an unknown member of the serine protease family is involved in MSC-mediated apoptosis.
DNA strand breakage
In order to investigate the effect of selenocompounds on total DNA damage, DNA single- and double-strand breaks were measured by filter elution assay. As previously reported (7), our experiment using HL-60 cells also showed that MSC treatment increased only DNA double-strand break for 24 h, and that selenite induced both DNA single- and double-strand breaks, which were more abrupt and severe than those of MSC-treated cells (data not shown). To evaluate whether DNA strand breaks induced by MSC and selenite result from the apoptotic machinery, the effects of z-VAD-fmk and AAPV-cmk were investigated. As shown in Figure 6A, z-VAD-fmk significantly reduced DNA double-strand breaks in MSC-treated cells, and AAPV-cmk was also effective. On the other hand, they exerted no significant effect on DNA single- and double-strand breaks induced by selenite (Figure 6B), which suggests that DNA double-strand breakage, induced by MSC treatment, is closely related to caspase activity.
Two different forms of selenocompounds were used in these experiments: MSC as an organic form and selenite as an inorganic form. MSC has been regarded as a potent chemopreventive agent on the basis of its organic nature, lowered toxicity and substantial anticarcinogenic properties (13,29), when compared with selenite, which has considerable toxicity. Our study demonstrates that MSC is a more efficient selenocompound for the induction of apoptosis in HL-60 cells than selenite, after considering their respective apoptotic characteristics and cytotoxic effects. Human promyelocytic leukemia cell line, HL-60, readily undergoes apoptosis by various cell death stimuli, especially treatment with anticancer agents (30). Furthermore, HL-60 cells have been used as a model system for testing antileukemic or general antitumor compounds, notably in studies about the mechanisms underlying apoptosis (31–33). Thus, we used HL-60 cells to explore the hypothesis that apoptosis is a plausible mechanism of chemoprevention by selenocompounds, while some of the earlier studies involved investigations in mammary epithelial tumor cells (4,34).
At the beginning of the present study, we examined the different natures of MSC and selenite as apoptosis or cytotoxicity inducers. As shown in Figure 1, treatment with 50 μM MSC showed more apoptosis and less cytotoxicity than 20 μM selenite. In microscopic observations (Figure 2), common morphological features of apoptosis were detected in MSC-treated cells. TUNEL-positive cells were increased, chromatin condensation and nuclear collapse were observed, and terminal apoptotic cells were evident. In contrast, selenite-treated cells exhibited only TUNEL-positive DNA fragmentation and chromatin condensation, but the other apoptotic features like membrane blebbing and apoptotic body formation were lacking. Further studies on the apoptotic processing of PARP and the activation of caspase-3 were performed by treating with MSC and selenite, to clarify the apoptotic process. During apoptosis, PARP is cleaved sequentially by a specific caspase (35). Among the identified caspases, caspase-3 is known to be an effector caspase, whose activation or processing triggers apoptotic cells to the `point of no return' (27,28). The appearance of the apoptotic 85 kDa fragment of PARP occurred in parallel with the activation of caspase-3 by treatment with MSC (Figure 3). This result indicates that MSC triggers the activation of the proteolytic machinery during apoptosis, especially that of the caspase family. In contrast, selenite did not affect the proteolytic activities related to PARP cleavage and caspase-3 activation. While the above findings give valuable insight into the understanding of the MSC-induced apoptotic pathway, its role on apoptotic progression has not been fully elucidated.
Subsequent experiments were designed to determine the role of caspases and serine proteases during apoptosis in cells treated with MSC, using their specific inhibitors. The results showed that the cytotoxicity, morphological changes, DNA ladder formation, caspase-3 activation, PARP cleavage and DNA double-strand break induced by MSC were significantly inhibited by pre-treatment with 50 μM z-VAD-fmk and 50 μM AAPV-cmk (Figures 4, 5 and 6A). These results imply that serine proteases as well as caspases play an important role in the cellular apoptosis induced by MSC. On the other hand, these inhibitors had no effect on the cytotoxicity and DNA strand breaks induced by selenite (Figures 4 and 6B). These relationships could be explained on the basis that cell death induced by selenite has another pathway, unrelated to the apoptosis mediated by the caspase system. It has been established that serine proteases are involved in the apoptotic process, but their role in apoptosis is not understood clearly. Several reports (36–38) have demonstrated that serine proteases are mainly concerned with DNA fragmentation, especially by the cleavage of nuclear structural proteins, processing of endonuclease or direct fragmentation of DNA. Furthermore, it has also been suggested that cytosolic serine proteases play a crucial role in apoptosis and is closely related to caspase-3 activation (39). On the basis of the above suggestions, serine proteases probably play a dual role in the apoptosis induced by MSC: one involves contribution to DNA fragmentation, and the other the processing of caspase-3. Further studies are required to elucidate which serine protease is involved, and the relation between serine proteases and caspases.
We previously observed that low concentrations of selenite may induce necrosis as well as apoptosis, while high concentrations induce more necrosis than apoptosis (9). Our present results on DNA strand breaks, cytotoxicity and apoptosis strengthen the possibility that selenite causes extensive DNA damage and that this damage leads the cell to necrosis rather than apoptosis. On the other hand, DNA double-strand breaks induced by MSC are related to the apoptotic DNA fragmentation formed during apoptosis. It has been postulated that selenite undergoes reductive metabolism to hydrogen selenide, and that MSC is metabolized to methylselenol, by passing hydrogen selenide (40). Hydrogen selenide induces DNA single-strand breaks, in association with the production of reactive oxygen species, which has been demonstrated by several workers (8,11,41). Furthermore, it has also been shown that selenite-induced oxidative stress and apoptosis are remarkably attenuated by antioxidants (8). It will be interesting to study further whether the apoptosis induced by MSC can be prevented by antioxidants.
In summary, we demonstrate that MSC is more efficient at inducing apoptosis and less toxic than selenite. Treatment with MSC causes DNA double-strand breakage, apoptotic DNA fragmentation, caspase-3 activation and TUNEL-positive apoptotic morphologies in HL-60 cells. Notably, we found that the novel pathway involving caspase is pivotal in MSC-mediated apoptosis. Our results strengthen the possibility that MSC can be used to prevent or cure cancer as either a chemopreventive or a chemotherapeutic agent. Further studies are required to differentiate between the chemopreventive effects on normal cells and cancer cells in vitro.