- Email: [email protected]
Activation of Na+, K+, Cl−-Cotransport Mediates Intracellular Ca2+ Increase and Apoptosis Induced by Pinacidil in HepG2 Human Hepatoblastoma Cells
Biochemical and Biophysical Research Communications 281, 511–519 (2001) doi:10.1006/bbrc.2001.4371, available online at http://www.idealibrary.com on
Activation of Na ⫹, K ⫹, Cl ⫺-Cotransport Mediates Intracellular Ca 2⫹ Increase and Apoptosis Induced by Pinacidil in HepG2 Human Hepatoblastoma Cells Jung-Ae Kim,* Young Shin Kang,† and Yong Soo Lee† ,1 *College of Pharmacy, Yeungnam University, Kyongsan 712-749, Korea; and †Department of Physiology, College of Medicine, Kwandong University, Kangnung 210-701, Korea
Received January 23, 2001
The role of Na ⴙ, K ⴙ, Cl ⴚ-cotransport (NKCC) in apoptosis of HepG2 human hepatoblastoma cells was investigated. Pinacidil (Pin), an activator of ATPsensitive K ⴙ (K ATP) channels, induced apoptosis in a dose- and time-dependent manner in HepG2 cells. Pin increased intracellular K ⴙ concentration ([K ⴙ] i). Bumetanide and furosemide, NKCC inhibitors, significantly inhibited the Pin-induced increased [K ⴙ] i and apoptosis, whereas K ATP inhibitors (glibenclamide and tolbutamide) had no effects. The Pin-induced [K ⴙ] i increase was significantly prevented by reducing extracellular Cl ⴚ concentration, and Pin also increased intracellular Na ⴙ concentration ([Na ⴙ] i), further indicating that these effects of Pin may be due to NKCC activation. In addition, Pin induced a rapid and sustained increase in intracellular Ca 2ⴙ concentration ([Ca 2ⴙ] i), which was completely prevented by the NKCC inhibitors. Treatment with EGTA or BAPTA/AM markedly inhibited the Pin-induced apoptosis. Inhibitors of Na ⴙ, Ca 2ⴙ-exchanger, bepridil, and benzamil significantly prevented both [Ca 2ⴙ] i increase and apoptosis induced by Pin. Taken together, these results suggest that Pin increases [Na ⴙ] i through NKCC activation, which leads to stimulation of reverse-mode of Na ⴙ, Ca 2ⴙ exchanger, resulting in [Ca 2ⴙ] i increase, and in turn, apoptosis. These results further suggest that NKCC may be a good target for induction of apoptosis in human hepatoma cells. © 2001 Academic Press Key Words: pinacidil; apoptosis; Na ⴙ, K ⴙ, Cl ⴚ-cotransport; intracellular Ca 2ⴙ; Na ⴙ-Ca 2ⴙ exchanger; HepG2 cells.
Apoptosis, a highly organized cell death process is characterized by early and prominent condensation of nuclear chromatin, loss of plasma membrane phospho1 To whom correspondence should be addressed. Fax: ⫹82 33 6411074. E-mail: [email protected].
lipid asymmetry, activation of proteases and endonucleases, enzymatic cleavage of the DNA into oligonucleosomal fragments, and segmentation of the cells into membrane-bound apoptotic bodies (1). Apoptosis has been recognized to play an important role in maintenance of tissue homeostasis by selective elimination of excessive cells (2). Genetic changes resulting in loss of apoptosis or derangement of apoptosis-signaling pathways are likely to be critical components of carcinogenesis (3, 4). In addition, apoptosis induction of cancer cells appears to be useful for cancer treatment (5) including chemotherapy (6) and radiation therapy (7). However, signaling pathways for the induction of apoptosis are not completely understood. Na ⫹, K ⫹, Cl ⫺-cotransport (NKCC) proteins are expressed in nearly every animal cell type (8). Two isoforms of the NKCC protein (⬃120 –130 kDa, unglycosylated) are currently known. NKCC1 is found in nearly all cell types, whereas NKCC2 is found exclusively in the kidney (9). Although the cellular functions of the NKCC are not clearly identified, the NKCC appears to maintain intracellular Cl ⫺ concentration [Cl ⫺] i at levels above the predicted electrochemical equilibrium (8). This high [Cl ⫺] i is important for promoting net salt transport in epithelial tissues (10) and for setting synaptic potentials in neuronal cells (11). The NKCC has been suggested to control cell volume by mediating the net influx of osmotically active ions (12). The NKCC activity may also be involved in the regulation of cell cycle (13). However, there is no published data on evidence that the NKCC regulates apoptosis. In the results of this study we identified for the first time that pinacidil (Pin), an activator of ATP-sensitive K ⫹ (K ATP) channels (14), has a novel activity to stimulate the NKCC without activation of the K ATP channels in HepG2 human hepatoblastoma cells. Thus the purposes of this study were to investigate (i) whether Pin
511
0006-291X/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
induces apoptosis in the HepG2 cells, (ii) whether the NKCC activity is involved in the mechanism of apoptosis induced by Pin, and (iii) whether intracellular Ca 2⫹ acts as a mediator between the Pin-induced NKCC activation and apoptosis. MATERIALS AND METHODS Materials. The powders Eagle minimum essential medium (MEM) and Earle’s basal salt solution, trypsin solution (EBSS), ethylene glycol-bis-(aminoethyl ether)N,N,N⬘,N⬘-tetraacetic acid (EGTA), ouabain, bumetanide, furosemide, 3,4,5-trimethoxybenzoic acid-8-(diethylamino)-octyl ester (TMB-8), dantrolene, sodium pyruvate, probenecid, propidium iodide (PI), ribonuclease A and all salt powders were obtained from Sigma Chemical Co. (St. Louis, MO). Pinacidil (Pin), glibenclamide (Glib), tolbutamide (Tolb), bepridil and benzamil were from RBI (Natick, MA). 1-(2,5Carboxyoxazol-2-yl-6-aminobenzfuran-5-oxyl)-2-(2⬘-aminomethylphenoxy)-ethane-N,N,N,N-tetraacetoxylmethyl ester (Fura-2/AM), bis-(o-aminophenoxy)-ethane-N,N,N,N-tetraacetic acid/acetoxymethyl ester (BAPTA/AM), potassium-binding benzofuran isophthalate aceoxylmethyl ester (PBFI/AM) and sodiumbinding benzofuran isophthalate aceoxylmethyl ester (SBFI/AM) were from Molecular Probes, Inc. (Eugene, OR). Fetal bovine serum (FBS) and antibiotics (penicillin and streptomycin mixture) were purchased from GIBCO (Grand Island, NY). Pin, Fura-2/AM, BAPTA/AM, PBFI/AM and SBFI/AM were prepared as stock solutions in dimethyl sulfoxide (DMSO), then diluted with aqueous medium to the final desired concentrations. The stock solutions of drugs were sterilized by filtration through 0.2 m disc filters (Gelman Sciences: Ann Arbor, MI). Cell lines and cell culture. The HepG2 human hepatoblastoma cell line was purchased from American Type Culture Collection (Rockville, MA). HepG2 cells were grown at 37°C in a humidified incubator under 5% CO 2/95% air in a MEM supplemented with 10% FBS, 200 IU/ml penicillin, 200 g/ml of streptomycin and 1 mM sodium pyruvate. Culture medium was replaced every other day. After attaining confluence the cells were subcultured following trypsinization with 0.25% trypsin-EDTA solution. DNA isolation and electrophoresis. DNA isolation was done according to Hockenbery et al. (15). HepG2 cells were collected by centrifugation (200g, 10⬘), washed twice in phosphate buffered saline (PBS) (pH 7.4) and resuspended at a density of 4 ⫻ 10 6 cells/400 l in hypotonic lysing buffer (5 mM Tris, 20 mM EDTA, pH 7.4) containing 0.5% Triton X-100 for 30⬘ at 4°C. The lysates were centrifuged at 13,000g, for 15⬘ at 4°C. Fragmented DNA was extracted from the supernatant with phenol-chloroform-isoamylalcohol, precipitated by addition of 2 volume of absolute ethanol and 0.1 volume of 3 mM sodium acetate, and treated with RNAse A (500 U/ml) at 37°C for 3 h. The pattern of DNA fragmentation was visualized by electrophoresis in 1.8% agarose gel containing ethidium bromide and photographed under UV light. Flow cytometric analysis of apoptosis. For flow cytometric analysis, HepG2 cells were collected and washed twice with PBS buffer (pH 7.4). After fixing in 80% ethanol for 30⬘, cells were washed twice, and resuspended in PBS buffer (pH 7.4) containing 0.1% Triton X-100, 5 g/ml PI and 50 g/ml ribonuclease A for DNA staining. Cells were then analyzed by a FACScan (BIO-RAD, Hercules, CA). At least 20,000 events were evaluated. All histograms were analyzed using WinBryte software (BIO-RAD, Hercules, CA) to determine percentage of nuclei with hypodiploid content indicative of apoptosis (16). The normal lipid organization of the plasma membrane is altered soon after apoptosis is initiated. Thus, annexin-V binding was also employed as an indicator of apoptosis (17) to demonstrate the loss of
phospholipid asymmetry and the presence of phosphatidylserine on the outer layer of the plasma membrane. It was analyzed using a commercial kit (Boehringer Mannheim Biochemicals, Mannheim, Germany). Cells were washed in cold PBS, and resuspended in binding buffer. A portion of cell suspension (500 l) was exposed to Annexin-V-FLUOS. The cells were gently vortexed, incubated at room temperature for 20⬘ in the dark, and then analyzed by FACScan within 1 h of staining. Intracellular Ca 2⫹ measurement. Aliquots of the HepG2 cells were washed in EBSS. Then, 5 M Fura-2/AM was added, and the cells were incubated for 30⬘ at 37°C. Unloaded Fura-2/AM was removed by centrifugation at 150g for 3⬘. Cells were resuspended at a density of 2 ⫻ 10 6 cells/ml in Krebs-Ringer buffer containing 125 mM NaCl, 5 mM KCl, 1.3 mM CaCl 2, 1.2 mM KH 2PO 4, 1.2 mM MgSO 4, 5 mM NaHCO 3, 25 mM Hepes, 6 mM glucose and 2.5 mM probenecid (pH 7.4). Fura-2/AM-loaded cells were maintained at 25°C for 90⬘ before fluorescence measurement. For each experiment, 0.5 ml aliquot of Fura-2/AM-loaded cells was equilibrated to 37°C in a stirred quartz cuvette. Fluorescence emission (510 nm) was monitored with the excitation wavelength cycling between 340 and 380 nm using a Hitachi F4500 fluorescence spectrophotometer. At the end of an experiment, fluorescence maximum and minimum values at each excitation wavelength were obtained by lysing cells with 20 g/ml digitonin (maximum) and then adding 10 mM EGTA (minimum). With the maximum and minimum values, the 340:380 nm fluorescence ratios were converted into free Ca 2⫹ concentrations using a software, F-4500 Intracellular Cation Measurement System, provided by Hitachi. Intracellular K ⫹ and Na ⫹ measurement. Intracellular K ⫹ and Na ⫹ levels were monitored with the K ⫹- and Na ⫹-sensitive fluorescent dyes PBFI/AM and SBFI/AM, respectively (18). Cells were washed, and resuspended at a density of 4 ⫻ 10 5 cells/ml in KrebsRinger buffer. The cells were loaded with 5 M PBFI/AM or SBFI/AM in Krebs-Ringer buffer containing 0.02% pluronic F-127, a nonionic surfactant, for 2 h at 37°C. Unloaded dye was removed by centrifugation at 150g for 3⬘. The dual-wavelength excitation method for measurement of PBFI and SBFI fluorescence was used. Fluorescence was monitored at 500 nm with excitation wavelengths of 340 and 380 nm in a stirred quartz cuvette. In the results relative changes in intracellular K ⫹ and Na ⫹ concentrations were reported as the 340:380 fluorescence ratios. Data analysis. All experiments were performed four times. Data were expressed as mean ⫾ standard error of the mean (SEM) and were analyzed using one way analysis of variance (ANOVA) and Student-Newman-Keul’s test for individual comparisons. P values less than 0.05 are considered statistically significant.
RESULTS Induction of apoptotic cell death by Pin in HepG2 cells. Pin induced DNA fragmentation in a dosedependent manner studied by agarose gel electrophoresis as shown in Fig. 1A. This effect of Pin was prominent at the concentration of 1 mM. Pin also induced loss of phospholipid asymmetry, resulting in appearance of phosphatidylserine on the outer layer of the plasma membrane detected by annexin-V binding, as depicted in Fig. 1B. Furthermore, Pin induced apoptosis in a time-related manner tested by flow cytometry by determining hypodiploid DNA content stained with PI as shown in Fig. 1C. Taken together, these results indicate that Pin induced apoptotic cell death in the HepG2 cells.
512
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Pin induces apoptotic cell death in HepG2 human hepatoblastoma cells. In the experiments of (A) cells were treated for 48 h with or without each concentration of Pin. DNA was isolated from the cells and analyzed by 1.8% agarose gel electrophoresis. Lane M represents DNA marker. In the experiments of (B) cells were treated for 36 h with or without Pin. Cells were stained with Annexin-V-FLUOS and analyzed by flow cytometry. In the experiments of (C) the cells were incubated with Pin (1 mM) for each designated time. The number of apoptotic cells was measured by flow cytometry as described in text. The region to the left of the G 0/G 1 peak, designated A 0, was defined as cells undergoing apoptosis-associated DNA degradation. In bar graphs the data represent the mean values of four replications with bars indicating SEM. *P ⬍ 0.05 compared to control.
No involvement of K ATP channels in the Pin-induced apoptosis. Since the only known pharmacological action of Pin is to activate K ATP channels (14), we determined the role of the K ⫹ channels in the Pin-induced apoptosis of HepG2 cells. To this end, we firstly examined whether Pin decreases [K ⫹] i as a consequence of K ⫹ channel activation assessed by the K ⫹-sensitive fluorescent dye PBFI/AM. Unexpectedly, Pin induced a rapid and sustained increase in [K ⫹] i in a dosedependent manner as shown in Fig. 2A. The Pininduced increase in [K ⫹] i was not altered by specific inhibitors of K ATP channels (100 M Glib, 100 M Tolb). Secondly, we investigated the effects of these inhibitors on the Pin-induced apoptosis. They did not significantly affect the apoptosis induced by Pin (1 mM) as shown in Fig. 2B. These results strongly suggest that K ATP channels may not play a role in the mechanism of the Pin-induced apoptosis in the HepG2 cells.
achieved by activation of Na ⫹, K ⫹-ATPase and/or NKCC, we examined the effects of selective inhibitors of these pathways on the [K ⫹] i increase and apoptosis induced by Pin (1 mM). Treatment with ouabain (100 M), a selective inhibitor of Na ⫹, K ⫹-ATPase did not alter both [K ⫹] i increase and apoptosis by Pin as shown in Figs. 3A and 3B, respectively, suggesting that activation of Na ⫹, K ⫹-ATPase may not mediate these actions of Pin. However, NKCC inhibitors (10 M bumetanide, 50 M furosemide) markedly suppressed these effects of Pin as illustrated in Figs. 3A and 3B. In addition, the Pin-induced [K ⫹] i increase was profoundly inhibited by using extracellular Cl ⫺-free buffer (Fig. 3A), which further demonstrates that Pin activates the NKCC. Taken together, these results suggest that Pin may activate the NKCC, and in turn, induce apoptosis in the HepG2 cells. To our knowledge this is the first report that revealed the role for NKCC in induction of apoptosis.
NKCC mediates the Pin-induced [K ⫹] i increase and apoptosis. Next, we investigated the possible sources of the increased [K ⫹] i by Pin. Since K ⫹ influx can be
Role for intracellular Ca 2⫹ in the Pin-induced apoptosis. Since intracellular Ca 2⫹ signal appears to act as a common mediator of apoptosis (19), we investi-
513
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Effects of inhibitors of K ATP channels on [K ⫹] i (A) and apoptosis (B) induced by Pin in HepG2 human hepatoblastoma cells. The data (A) show changes in [K ⫹] i as a function of time, which was measured by using K ⫹-sensitive fluorescent dye PBFI/AM. The arrows show the time points for addition of Pin (1 mM). Glib (100 M) and Tolb (100 M), specific inhibitors of K ATP channels, were added 10 min before Pin treatment. In the experiments of (B) the cells were incubated with Pin (1 mM) for 48 h. Drugs (100 M Glib, 100 M Tolb) were added 30 min before Pin treatment. The number of apoptotic cells was measured by flow cytometry. In bar graphs the data represent the mean values of four replications with bars indicating SEM. *P ⬍ 0.05 compared to control.
gated whether intracellular Ca 2⫹ plays a role in the Pin-induced apoptosis. Intracellular Ca 2⫹ concentration ([Ca 2⫹] i) was measured by using Fura-2 fluorescence technique. As shown in Fig. 4A, Pin (1 mM) induced a rapid and sustained increase in [Ca 2⫹] i. To determine the source of the Pin-induced intracellular Ca 2⫹ increase, we measured [Ca 2⫹] i using a nominal Ca 2⫹-free medium containing 100 M EGTA. This experimental protocol can effectively reduce extracellular free Ca 2⫹ concentration, and thus, blunt available Ca 2⫹ influx. Under these conditions cellular Ca 2⫹ response to Pin (1 mM) was markedly reduced as illustrated in Fig. 4A. However, treatment with intracellular Ca 2⫹ release blockers (20 M TMB-8, 50 M dantrolene), did not significantly alter the Ca 2⫹-increasing effect of Pin. These results indicate that the increased [Ca 2⫹] i by Pin may be predominantly due to Ca 2⫹ influx. As shown in Fig. 4B, the Pin-induced apoptosis was significantly suppressed either by 0.5 M BAPTA/AM, an intracellular Ca 2⫹ chelator, or by 1 mM EGTA, an extracellular Ca 2⫹ chelator. However, intracellular Ca 2⫹ release blockers, TMB-8 (20 M) and dantrolene (50 M) did not significantly alter the apoptosis induced by Pin. These results suggest that Ca 2⫹ influx may mediate the Pin-induced apoptosis.
Activation of the reverse mode of Na ⫹, Ca 2⫹exchanger mediates the Pin-induced [Ca 2⫹] i increase and apoptosis. Although the Pin-induced [Ca 2⫹] i increase was found to be due to Ca 2⫹ influx (Fig. 4A), The exact pathway responsible for Ca 2⫹ influx is not known. Thus to determine the exact Ca 2⫹ influx pathway activated by Pin, we investigated the effects of various pharmacological agents which can inhibit the potential pathways of Ca 2⫹ influx, on the Pin-induced [Ca 2⫹] i increase. Neither inhibitors of voltage-sensitive Ca 2⫹ channels (100 M verapamil, 100 M nifedipine), nor an inhibitor of nonselective cation channels (100 M FA) did significantly alter the Ca 2⫹ influx induced by Pin (1 mM) as depicted in Fig. 5A. In addition, these agents did not significantly influence the Pin-induced apoptosis as shown in Fig. 5B. These results suggest that Ca 2⫹ influx as well as apoptosis induced by Pin may be mediated not through these ion channels, but through other mechanisms. Interestingly, inhibitors of Na ⫹ , Ca 2⫹ exchanger (50 M bepridil, 50 M benzamil) completely suppressed both Ca 2⫹ influx and apoptosis induced by Pin (1 mM) as illustrated in Figs. 5A and 5B, respectively. These results indicate that the reverse mode of the Na ⫹ , Ca 2⫹ exchanger may be activated by Pin
514
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Roles of NKCC in the Pin-induced [K ⫹] i increase and apoptosis in HepG2 human hepatoblastoma cells. The data (A) show changes in [K ⫹] i as a function of time, which was measured by using K ⫹-sensitive fluorescent dye PBFI/AM. The arrows show the time points for addition of Pin (1 mM). Ouabain (100 M), an inhibitor of Na ⫹, K ⫹-ATPase, and NKCC inhibitors (10 M bumetanide, 50 M furosemide) were added 10 min before Pin treatment. For extracellular Cl ⫺ ([Cl ⫺] ex)-free buffer solution, external Cl ⫺ was replaced with gluconate. In the experiments of (B) the cells were incubated with Pin (1 mM) for 48 h. Drugs (100 M ouabain, 10 M bumetanide, 50 M furosemide) were added 30 min before Pin treatment. The number of apoptotic cells was measured by flow cytometry. In bar graphs, the data represent the mean values of four replications with bars indicating SEM. *P ⬍ 0.05 compared to control. # P ⬍ 0.05 compared to Pin alone.
and that Ca 2⫹ influx through the exchanger may lead to apoptosis. Role of NKCC in the activation of the reverse mode of Na ⫹, Ca 2⫹-exchanger induced by Pin. To determine whether the activation of the reverse mode of Na⫹/Ca 2⫹ exchanger is secondarily achieved by increased intracellular Na ⫹ concentration ([Na⫹] i), we monitored [Na ⫹] i using a Na ⫹-sensitive fluorescent dye SBFI/AM. As shown in Fig. 6A, Pin increased [Na⫹] i. This [Na ⫹] i increase was markedly suppressed by NKCC inhibitors (10 M bumetanide, 50 M furosemide). Moreover, inhibitors of Na ⫹, Ca 2⫹ exchanger (50 M bepridil, 50 M benzamil) profoundly augmented the [Na⫹] i increase by Pin. Next, we tested whether activation of NKCC mediates the [Ca 2⫹] i-increasing effect of Pin. As shown in Fig. 6B, treatment with the NKCC inhibitors (10 M bumetanide, 50 M furosemide) completely prevented the [Ca 2⫹] i increase by Pin, implying that activation of NKCC may be necessary for the [Ca 2⫹] i-increasing effect of Pin. Taken together, these results suggest that Pin induces Ca 2⫹ influx through activation of the reverse mode of Na ⫹/Ca 2⫹ exchanger which is achieved by increased [Na ⫹] i attributed to activation of NKCC. DISCUSSION In this study we demonstrated for the first time that Pin induces apoptosis and that NKCC is involved in
the mechanism of apoptosis in HepG2 human hepatoma cells. The induction of apoptosis by Pin was demonstrated by using three independent methods, detection of DNA fragmentation through agarose gel electrophoresis (Fig. 1A), detection of phosphatidylserine translocation through annexin-V binding assay (Fig. 1B), and measurement of hypodiploid DNA contents through flow cytometry (Fig. 1C). The following results strongly suggest the essential role for the NKCC in the mechanism of apoptosis by Pin. Pin induced increases in [K ⫹] i as well as [Na ⫹] i, which was markedly suppressed by bumetamide and furosemide, a highly selective and less selective inhibitors of the NKCC, respectively (8) (Figs. 3A and 6A). The Pininduced increased [K ⫹] i was significantly reduced by [Cl ⫺] ex-free buffer (Fig. 3A). The Pin-induced apoptosis was completely suppressed by the NKCC inhibitors (Fig. 3B). The results that Pin induced [K ⫹] i increase rather than decrease (Fig. 2A), imply no involvement in the action mechanism of Pin of activation of K ATP channels which is the only known pharmacological action of Pin (14). No significant effects of selective inhibitors of the K ATP channels (Glib and Tolb) on the Pin-induced increased [K ⫹] i and apoptosis (Fig. 2) consistently support the notion that the K ATP channels may not play a role in these activities of Pin. Recently, Malhi et al. have suggested that K ATP channels exist in HepG2
515
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. Role of Ca 2⫹ influx in the Pin-induced apoptosis in HepG2 human hepatoblastoma cells. The data (A) represent intracellular Ca 2⫹ changes with time assessed by Fura-2 fluorescence technique. The arrows show the time points for addition of Pin (1 mM). EGTA (100 M) in a nominal Ca 2⫹-free medium, TMB-8 (20 M) and dantrolene (50 M) were added 10 min before Pin application. Quantitative changes (right bar graphs) were expressed as percent increases in [Ca 2⫹] i induced by the drug compared to control condition in which the cells were incubated with a drug-free vehicle. In the experiments of (B) the cells were incubated with Pin (1 mM) for 48 h. Drugs (0.5 M BAPTA/AM, 500 M EGTA, 20 M TMB-8, 50 M dantrolene) were added 30 min before Pin treatment. The number of apoptotic cells was measured by flow cytometry. In bar graphs (A and B) the data represent the mean values of four replications with bars indicating SEM. *P ⬍ 0.05 compared to control. # P ⬍ 0.05 compared to Pin alone.
cells, detected by using method of polymerase chain reaction (PCR) (20). However, the results of this study suggest that the total K ⫹ conductance due to the activation of the K ATP channels by Pin may be too small to reduce [K ⫹] i. In addition to activation of NKCC, [K ⫹] i increase can be possibly achieved by other mechanisms, i.e., inhibition of background K ⫹ channels and/or activation of Na ⫹, K ⫹-ATPase. The contribution of these pathways to the Pin-induced increase in [K ⫹] i was investigated. Treatment with ouabain, a specific inhibitor of Na ⫹, K ⫹-ATPase, did not alter both [K ⫹] i increase and apoptosis induced by Pin (Figs. 3A and 3B), suggesting that Na ⫹, K ⫹-ATPase may not be involved in these actions of Pin. Since the increased [K ⫹] i by Pin was not completely blocked by NKCC inhibitors (Fig. 3A), the residual portion may be due to inhibition of the background K ⫹ channels. Although the contribution of this channel activity to apoptosis can not be completely excluded, it is evident that the NKCC plays a major role in the Pin-induced apoptosis, since it was completely suppressed by the NKCC inhibitors (Fig. 3B). Accumulating evidence implies that intracellular Ca 2⫹ is commonly involved in the mechanism of apoptosis (19). One of the targets for elevated intracellular Ca 2⫹ is the activation of the Ca 2⫹-dependent protein
kinases and phosphatases that has been seen during apoptosis (21, 22). Direct activation of the Ca 2⫹dependent proteinase may represent another target for intracellular Ca 2⫹ action in apoptosis (23). Ca 2⫹/Mg 2⫹dependent endonuclease whose activation results in DNA fragmentation, the most characteristic biochemical feature of apoptosis (24), and Ca 2⫹-dependent transglutaminase which is highly activated in apoptotic cells (25), also appear to be a target for Ca 2⫹ action (26, 27). In HepG2 cells we have found that a sustained Ca 2⫹ influx mediates apoptotic cell death induced by various substances including tamoxifen (28) and tertbutyl hydroperoxide (29). Therefore, we investigated a possible role of Ca 2⫹ signal in the Pin-induced apoptosis. The results showed that Pin increased [Ca 2⫹] i through Ca 2⫹ influx, since the increased [Ca 2⫹] i by Pin was significantly prevented by using nominal Ca 2⫹-free medium containing 100 M EGTA, but not altered by internal Ca 2⫹ release blockers, TMB-8 and dantrolene (Fig. 4A). Consistently, the Pin-induced apoptosis was significantly prevented by EGTA as well as BAPTA/ AM, an intracellular Ca 2⫹ chelator, but not affected by these inhibitors of internal Ca 2⫹ release (Fig. 4B). Interestingly, the inhibitors of NKCC, bumetanide and furosemide completely suppressed the Pin-induced [Ca 2⫹] i increase (Fig. 6B). Thus these results suggest
516
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 5. Involvement of Na ⫹, Ca 2⫹-exchanger in the Pin-induced [Ca 2⫹] i increase and apoptosis in HepG2 human hepatoblastoma cells. Experimental procedures and data presentations are the same as Fig. 4. In the experiments inhibitors of Na ⫹, Ca 2⫹-exchanger (50 M bepridil, 50 M benzamil), voltage-sensitive Ca 2⫹ channel blockers (100 M verapamil, 100 M nifedipine) and an inhibitor of non-selective cation channels (100 M FA) were used. In bar graphs the data represent the mean values of four replications with bars indicating SEM. *P ⬍ 0.05 compared to control. # P ⬍ 0.05 compared to Pin alone.
that intracellular Ca 2⫹ may mediate apoptosis induction by Pin and that the Pin-induced activation of the NKCC may be necessarily involved in the [Ca 2⫹] i increase. Extracellular free Ca 2⫹ ions may enter into the cells by the following two mechanisms: (i) activation of plasma membrane Ca 2⫹ channels, and (ii) activation of reverse mode of Na ⫹/Ca 2⫹ exchange mechanism. In this study voltage-dependent Ca 2⫹ channel blockers, verapamil and nifedipine did not significantly alter the Pin-induced [Ca 2⫹] i increase and apoptosis (Figs. 5A and 5B), suggesting no involvement of these ion channels in the activity of Pin in the HepG2 cells. Recently, Ca 2⫹-permeable nonselective cation channels have been reported to exist in the HepG2 cells (30). Since these channels have mediated Ca 2⫹ influx in many different types of cells (31, 32), as well as in HepG2 cells (28), we determined the possible role for this channel in the Pin-induced Ca 2⫹ influx using FA, a known inhibitor of the channels. No significant effects of FA on the Pin-induced Ca 2⫹ influx and apoptosis (Figs. 5A and 5B) further indicate that these types of plasma membrane Ca 2⫹ channels may not act as the Ca 2⫹ influx pathway activated by Pin. On the other hand, bepridil and benzamil, known inhibitors of Na ⫹ /Ca 2⫹ exchanger significantly suppressed both Ca 2⫹ influx and apoptosis induced by Pin (Figs. 5A and 5B), implying that the activation of reverse mode of Na ⫹ /Ca 2⫹ exchanger may be critical for these actions of Pin. Na ⫹ /Ca 2⫹ exchanger nor-
mally acts to extrude Ca 2⫹ ions when intracellular Ca 2⫹ rises above certain levels (33). On the contrary, Ca 2⫹ ions enter into the cells under conditions that favor the reverse mode of operation of the Na ⫹ /Ca 2⫹ exchanger (33). Reverse operation of the Na ⫹ /Ca 2⫹ exchanger during anoxia has been reported to be a critical mechanism of Ca 2⫹ influx and subsequent neuronal cell injury (34). The activation of reverse mode of Na ⫹ /Ca 2⫹ exchanger may be a second event following to the increased [Na ⫹ ] i through the activated NKCC by Pin. This possibility was investigated, and the results showed that Pin indeed increased [Na ⫹ ] i immediately which was markedly reduced by bumetanide and furosemide (Fig. 6A). Furthermore, enhancement of the Pin-induced [Na ⫹ ] i increase by bepridil and benzamil (Fig. 6A) strongly support the idea that the reverse mode of Na ⫹ /Ca 2⫹ exchanger is activated by increased [Na ⫹ ] i through the Pin-induced stimulation of NKCC. Currently, we do not know the exact mechanism by which Pin activates NKCC in the HepG2 cells. Although it may be the most plausible explanation that Pin directly activates the NKCC, indirect regulation of the activity of this cotransport may also be possible. The NKCC activity seems to be regulated by phosphorylation and dephosphorylation of the protein via protein kinases and phosphatases, respectively (8). Thus Pin may regulate the activity of the cotransport via these mechanisms. However, the possibility remains to be determined in the future studies.
517
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 6. NKCC mediates activation of reverse mode of Na ⫹, Ca 2⫹-exchanger and [Ca 2⫹] i increase induced by Pin in HepG2 human hepatoblastoma cells. The data (A) show changes in [Na ⫹] i as a function of time, which was measured by using Na ⫹-sensitive fluorescent dye SBFI/AM. The arrows show the time points for addition of Pin (1 mM). NKCC inhibitors (10 M bumetanide, 50 M furosemide), and of Na ⫹, Ca 2⫹-exchanger (50 M bepridil, 50 M benzamil) were added 10 min before Pin treatment. The data (B) represent intracellular Ca 2⫹ changes with time assessed by Fura-2 fluorescence technique. Results were expressed as percent increases in [Ca 2⫹] i induced by the drug compared to control condition in which the cells were incubated with a drug-free vehicle. Bumetanide (10 M) and furosemide (50 M) were added 10 min before Pin (1 mM) application.
In conclusion, Pin induced apoptosis through activation of intracellular Ca 2⫹ signal in HepG2 human hepatoblastoma cells. Pin may activate NKCC, which results in [Na ⫹ ] i increase, and in turn, leads to the stimulation of the reverse mode of Na ⫹ , Ca 2⫹ exchanger, resulting in [Ca 2⫹ ] i increase. These results further suggest that NKCC may be a good target for the induction of apoptosis of human hepatoma cells. ACKNOWLEDGMENT This work was supported by Korea Research Foundation Grant (KRF-2000-D0341).
REFERENCES 1. Kidd, V. J. (1998) Annu. Rev. Physiol. 60, 533–573. 2. Nayfield, S. G., Karp, J. E., Ford, L. G., Dorr, F. A., and Kramer, B. S. (1991) J. Natl. Cancer Inst. 83, 1450 –1459. 3. Schulte-Hermann, R., Bursch, W., Low-Baselli, A., Wagner, A., and Grasl-Kraupp, B. (1997) Cell. Biol. Toxicol. 13, 339 –348. 4. Kastan, M. B., Canman, C. E., and Leonard, C. J. (1995) Cancer Metastasis Rev. 14, 3–15. 5. Kornblau, S. M. (1998) Leukemia 12, S41–S46. 6. Kamesaki, H. (1998) Int. J. Hematol. 68, 29 – 43. 7. Crompton, N. E. (1998) Acta Oncol. 37, 1– 49.
8. Russell, J. M. (2000) Physiol. Rev. 80, 211–276. 9. Delpire, E., Rauchman, M. I., Beier, D. R., Hebert, S. C., and Gullans, S. R. (1994) J. Biol. Chem. 269, 25677–25683. 10. Greger, R. (1996) Pflugers Arch. 432, 579 –588. 11. Alvarez-Leefmans, F. J., Gamino, S. M., Giraldez, F., and Nogueron, I. (1988) J. Physiol. 406, 225–246. 12. Hara, E., Reinach, P. S., Wen, Q., Iserovich, P., and Fischbarg, J. (1999) J. Membr. Biol. 171, 75– 85. 13. Leung, S., O’Donnell, M. E., Martinez, A., and Palfrey, H. C. (1994) J. Biol. Chem. 269, 10581–10589. 14. Edwards, G., and Weston, A. H. (1995) Cardiovasc. Drugs Ther. 9, 185–193. 15. Hockenbery, D., Nunez, G., Milliman, C., Schreiber, R. D., and Korsmeyer, S. J. (1990) Nature 348, 334 –336. 16. Bombeli, T., Karsan, A., Tait, J. F., and Harlan, J.M. (1997) Blood 89, 2429 –2442. 17. Vermes, I., Haanen, C., Steffens-Nakken, H., and Reutelingsperger, C. (1995) J. Immunol. Methods 184, 39 –51. 18. Ito, S., Ohta, T., Kadota, H., Kitamura, N., and Nakazato, Y., (1997) J. Neurosci. Methods 75, 21–27. 19. McConkey, D. J., and Orrenius, S. (1996) J. Leukoc. Biol. 59, 775–783. 20. Malhi, H., Irani, A. N., Rajvanshi, P., Suadicani, S. O., Spray, D. C., McDonald, T. V., and Gupta, S. (2000) J. Biol. Chem. 275, 26050 –26057. 21. Bonnefoy-Berard, N., Genestier, L., Flacher, M., and Revillard, J. P. (1994) Eur. J. Immunol. 24, 325–329.
518
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
22. Shibasaki, F., Kondo, E., Akagi, E., and McKeon, F. (1997) Nature 386, 728 –731. 23. Squier, M. K. T., and Cohen, J. J. (1997) J. Immunol. 158, 3690 –3697. 24. Wyllie, A. H., Morris, R. G., Smith, A. L., and Dunlop, D. (1984) J. Pathol. 142, 67. 25. Fesus, L., Thomazy, V., and Falus, A. (1987) FEBS Lett. 224, 104 –108. 26. Cohen, J. J., and Duke, R. C. (1984) J. Immunol. 132, 38 – 42. 27. Melino, G., Annicchiarico-Petruzzeli, M., Piredda, L., Candi, E., Gentile, V., Davies, P. J., and Piacentini, M. (1994) Mol. Cell. Biol. 14, 6584 – 6596.
28. Kim, J. A., Kang, Y. S., Lee, S. H., Lee, E. H., and Lee, Y. S. (1999) Cancer Lett. 147, 115–123. 29. Kim, J. A., Kang, Y. S., Lee, S. H., and Lee, Y. S. (2000) Free Rad. Res. 33, 267–277. 30. Chen, W. H., Yeh, T. H., Tsai, M. C., Chen, D. S., and Wang, T. H. (1997) J. Form. Med. Assoc. 96, 503–510. 31. Demaurex, N., Monod, A., Lew, D. P., and Krause, K. H. (1994) Biochem. J. 297, 595– 601. 32. Ling, B. N., and O’Neill, W. C. (1992) Am. J. Physiol. 263, H1827–H1838. 33. DiPolo, R., and Beauge, L. (1987) Hypertension 10, 115–119. 34. Stys, P. K., Waxman, S. G., and Ransom, B. R. (1991) Ann. Neurol. 30, 375–380.
519
Activation of Na ⫹, K ⫹, Cl ⫺-Cotransport Mediates Intracellular Ca 2⫹ Increase and Apoptosis Induced by Pinacidil in HepG2 Human Hepatoblastoma Cells Jung-Ae Kim,* Young Shin Kang,† and Yong Soo Lee† ,1 *College of Pharmacy, Yeungnam University, Kyongsan 712-749, Korea; and †Department of Physiology, College of Medicine, Kwandong University, Kangnung 210-701, Korea
Received January 23, 2001
The role of Na ⴙ, K ⴙ, Cl ⴚ-cotransport (NKCC) in apoptosis of HepG2 human hepatoblastoma cells was investigated. Pinacidil (Pin), an activator of ATPsensitive K ⴙ (K ATP) channels, induced apoptosis in a dose- and time-dependent manner in HepG2 cells. Pin increased intracellular K ⴙ concentration ([K ⴙ] i). Bumetanide and furosemide, NKCC inhibitors, significantly inhibited the Pin-induced increased [K ⴙ] i and apoptosis, whereas K ATP inhibitors (glibenclamide and tolbutamide) had no effects. The Pin-induced [K ⴙ] i increase was significantly prevented by reducing extracellular Cl ⴚ concentration, and Pin also increased intracellular Na ⴙ concentration ([Na ⴙ] i), further indicating that these effects of Pin may be due to NKCC activation. In addition, Pin induced a rapid and sustained increase in intracellular Ca 2ⴙ concentration ([Ca 2ⴙ] i), which was completely prevented by the NKCC inhibitors. Treatment with EGTA or BAPTA/AM markedly inhibited the Pin-induced apoptosis. Inhibitors of Na ⴙ, Ca 2ⴙ-exchanger, bepridil, and benzamil significantly prevented both [Ca 2ⴙ] i increase and apoptosis induced by Pin. Taken together, these results suggest that Pin increases [Na ⴙ] i through NKCC activation, which leads to stimulation of reverse-mode of Na ⴙ, Ca 2ⴙ exchanger, resulting in [Ca 2ⴙ] i increase, and in turn, apoptosis. These results further suggest that NKCC may be a good target for induction of apoptosis in human hepatoma cells. © 2001 Academic Press Key Words: pinacidil; apoptosis; Na ⴙ, K ⴙ, Cl ⴚ-cotransport; intracellular Ca 2ⴙ; Na ⴙ-Ca 2ⴙ exchanger; HepG2 cells.
Apoptosis, a highly organized cell death process is characterized by early and prominent condensation of nuclear chromatin, loss of plasma membrane phospho1 To whom correspondence should be addressed. Fax: ⫹82 33 6411074. E-mail: [email protected].
lipid asymmetry, activation of proteases and endonucleases, enzymatic cleavage of the DNA into oligonucleosomal fragments, and segmentation of the cells into membrane-bound apoptotic bodies (1). Apoptosis has been recognized to play an important role in maintenance of tissue homeostasis by selective elimination of excessive cells (2). Genetic changes resulting in loss of apoptosis or derangement of apoptosis-signaling pathways are likely to be critical components of carcinogenesis (3, 4). In addition, apoptosis induction of cancer cells appears to be useful for cancer treatment (5) including chemotherapy (6) and radiation therapy (7). However, signaling pathways for the induction of apoptosis are not completely understood. Na ⫹, K ⫹, Cl ⫺-cotransport (NKCC) proteins are expressed in nearly every animal cell type (8). Two isoforms of the NKCC protein (⬃120 –130 kDa, unglycosylated) are currently known. NKCC1 is found in nearly all cell types, whereas NKCC2 is found exclusively in the kidney (9). Although the cellular functions of the NKCC are not clearly identified, the NKCC appears to maintain intracellular Cl ⫺ concentration [Cl ⫺] i at levels above the predicted electrochemical equilibrium (8). This high [Cl ⫺] i is important for promoting net salt transport in epithelial tissues (10) and for setting synaptic potentials in neuronal cells (11). The NKCC has been suggested to control cell volume by mediating the net influx of osmotically active ions (12). The NKCC activity may also be involved in the regulation of cell cycle (13). However, there is no published data on evidence that the NKCC regulates apoptosis. In the results of this study we identified for the first time that pinacidil (Pin), an activator of ATP-sensitive K ⫹ (K ATP) channels (14), has a novel activity to stimulate the NKCC without activation of the K ATP channels in HepG2 human hepatoblastoma cells. Thus the purposes of this study were to investigate (i) whether Pin
511
0006-291X/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
induces apoptosis in the HepG2 cells, (ii) whether the NKCC activity is involved in the mechanism of apoptosis induced by Pin, and (iii) whether intracellular Ca 2⫹ acts as a mediator between the Pin-induced NKCC activation and apoptosis. MATERIALS AND METHODS Materials. The powders Eagle minimum essential medium (MEM) and Earle’s basal salt solution, trypsin solution (EBSS), ethylene glycol-bis-(aminoethyl ether)N,N,N⬘,N⬘-tetraacetic acid (EGTA), ouabain, bumetanide, furosemide, 3,4,5-trimethoxybenzoic acid-8-(diethylamino)-octyl ester (TMB-8), dantrolene, sodium pyruvate, probenecid, propidium iodide (PI), ribonuclease A and all salt powders were obtained from Sigma Chemical Co. (St. Louis, MO). Pinacidil (Pin), glibenclamide (Glib), tolbutamide (Tolb), bepridil and benzamil were from RBI (Natick, MA). 1-(2,5Carboxyoxazol-2-yl-6-aminobenzfuran-5-oxyl)-2-(2⬘-aminomethylphenoxy)-ethane-N,N,N,N-tetraacetoxylmethyl ester (Fura-2/AM), bis-(o-aminophenoxy)-ethane-N,N,N,N-tetraacetic acid/acetoxymethyl ester (BAPTA/AM), potassium-binding benzofuran isophthalate aceoxylmethyl ester (PBFI/AM) and sodiumbinding benzofuran isophthalate aceoxylmethyl ester (SBFI/AM) were from Molecular Probes, Inc. (Eugene, OR). Fetal bovine serum (FBS) and antibiotics (penicillin and streptomycin mixture) were purchased from GIBCO (Grand Island, NY). Pin, Fura-2/AM, BAPTA/AM, PBFI/AM and SBFI/AM were prepared as stock solutions in dimethyl sulfoxide (DMSO), then diluted with aqueous medium to the final desired concentrations. The stock solutions of drugs were sterilized by filtration through 0.2 m disc filters (Gelman Sciences: Ann Arbor, MI). Cell lines and cell culture. The HepG2 human hepatoblastoma cell line was purchased from American Type Culture Collection (Rockville, MA). HepG2 cells were grown at 37°C in a humidified incubator under 5% CO 2/95% air in a MEM supplemented with 10% FBS, 200 IU/ml penicillin, 200 g/ml of streptomycin and 1 mM sodium pyruvate. Culture medium was replaced every other day. After attaining confluence the cells were subcultured following trypsinization with 0.25% trypsin-EDTA solution. DNA isolation and electrophoresis. DNA isolation was done according to Hockenbery et al. (15). HepG2 cells were collected by centrifugation (200g, 10⬘), washed twice in phosphate buffered saline (PBS) (pH 7.4) and resuspended at a density of 4 ⫻ 10 6 cells/400 l in hypotonic lysing buffer (5 mM Tris, 20 mM EDTA, pH 7.4) containing 0.5% Triton X-100 for 30⬘ at 4°C. The lysates were centrifuged at 13,000g, for 15⬘ at 4°C. Fragmented DNA was extracted from the supernatant with phenol-chloroform-isoamylalcohol, precipitated by addition of 2 volume of absolute ethanol and 0.1 volume of 3 mM sodium acetate, and treated with RNAse A (500 U/ml) at 37°C for 3 h. The pattern of DNA fragmentation was visualized by electrophoresis in 1.8% agarose gel containing ethidium bromide and photographed under UV light. Flow cytometric analysis of apoptosis. For flow cytometric analysis, HepG2 cells were collected and washed twice with PBS buffer (pH 7.4). After fixing in 80% ethanol for 30⬘, cells were washed twice, and resuspended in PBS buffer (pH 7.4) containing 0.1% Triton X-100, 5 g/ml PI and 50 g/ml ribonuclease A for DNA staining. Cells were then analyzed by a FACScan (BIO-RAD, Hercules, CA). At least 20,000 events were evaluated. All histograms were analyzed using WinBryte software (BIO-RAD, Hercules, CA) to determine percentage of nuclei with hypodiploid content indicative of apoptosis (16). The normal lipid organization of the plasma membrane is altered soon after apoptosis is initiated. Thus, annexin-V binding was also employed as an indicator of apoptosis (17) to demonstrate the loss of
phospholipid asymmetry and the presence of phosphatidylserine on the outer layer of the plasma membrane. It was analyzed using a commercial kit (Boehringer Mannheim Biochemicals, Mannheim, Germany). Cells were washed in cold PBS, and resuspended in binding buffer. A portion of cell suspension (500 l) was exposed to Annexin-V-FLUOS. The cells were gently vortexed, incubated at room temperature for 20⬘ in the dark, and then analyzed by FACScan within 1 h of staining. Intracellular Ca 2⫹ measurement. Aliquots of the HepG2 cells were washed in EBSS. Then, 5 M Fura-2/AM was added, and the cells were incubated for 30⬘ at 37°C. Unloaded Fura-2/AM was removed by centrifugation at 150g for 3⬘. Cells were resuspended at a density of 2 ⫻ 10 6 cells/ml in Krebs-Ringer buffer containing 125 mM NaCl, 5 mM KCl, 1.3 mM CaCl 2, 1.2 mM KH 2PO 4, 1.2 mM MgSO 4, 5 mM NaHCO 3, 25 mM Hepes, 6 mM glucose and 2.5 mM probenecid (pH 7.4). Fura-2/AM-loaded cells were maintained at 25°C for 90⬘ before fluorescence measurement. For each experiment, 0.5 ml aliquot of Fura-2/AM-loaded cells was equilibrated to 37°C in a stirred quartz cuvette. Fluorescence emission (510 nm) was monitored with the excitation wavelength cycling between 340 and 380 nm using a Hitachi F4500 fluorescence spectrophotometer. At the end of an experiment, fluorescence maximum and minimum values at each excitation wavelength were obtained by lysing cells with 20 g/ml digitonin (maximum) and then adding 10 mM EGTA (minimum). With the maximum and minimum values, the 340:380 nm fluorescence ratios were converted into free Ca 2⫹ concentrations using a software, F-4500 Intracellular Cation Measurement System, provided by Hitachi. Intracellular K ⫹ and Na ⫹ measurement. Intracellular K ⫹ and Na ⫹ levels were monitored with the K ⫹- and Na ⫹-sensitive fluorescent dyes PBFI/AM and SBFI/AM, respectively (18). Cells were washed, and resuspended at a density of 4 ⫻ 10 5 cells/ml in KrebsRinger buffer. The cells were loaded with 5 M PBFI/AM or SBFI/AM in Krebs-Ringer buffer containing 0.02% pluronic F-127, a nonionic surfactant, for 2 h at 37°C. Unloaded dye was removed by centrifugation at 150g for 3⬘. The dual-wavelength excitation method for measurement of PBFI and SBFI fluorescence was used. Fluorescence was monitored at 500 nm with excitation wavelengths of 340 and 380 nm in a stirred quartz cuvette. In the results relative changes in intracellular K ⫹ and Na ⫹ concentrations were reported as the 340:380 fluorescence ratios. Data analysis. All experiments were performed four times. Data were expressed as mean ⫾ standard error of the mean (SEM) and were analyzed using one way analysis of variance (ANOVA) and Student-Newman-Keul’s test for individual comparisons. P values less than 0.05 are considered statistically significant.
RESULTS Induction of apoptotic cell death by Pin in HepG2 cells. Pin induced DNA fragmentation in a dosedependent manner studied by agarose gel electrophoresis as shown in Fig. 1A. This effect of Pin was prominent at the concentration of 1 mM. Pin also induced loss of phospholipid asymmetry, resulting in appearance of phosphatidylserine on the outer layer of the plasma membrane detected by annexin-V binding, as depicted in Fig. 1B. Furthermore, Pin induced apoptosis in a time-related manner tested by flow cytometry by determining hypodiploid DNA content stained with PI as shown in Fig. 1C. Taken together, these results indicate that Pin induced apoptotic cell death in the HepG2 cells.
512
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Pin induces apoptotic cell death in HepG2 human hepatoblastoma cells. In the experiments of (A) cells were treated for 48 h with or without each concentration of Pin. DNA was isolated from the cells and analyzed by 1.8% agarose gel electrophoresis. Lane M represents DNA marker. In the experiments of (B) cells were treated for 36 h with or without Pin. Cells were stained with Annexin-V-FLUOS and analyzed by flow cytometry. In the experiments of (C) the cells were incubated with Pin (1 mM) for each designated time. The number of apoptotic cells was measured by flow cytometry as described in text. The region to the left of the G 0/G 1 peak, designated A 0, was defined as cells undergoing apoptosis-associated DNA degradation. In bar graphs the data represent the mean values of four replications with bars indicating SEM. *P ⬍ 0.05 compared to control.
No involvement of K ATP channels in the Pin-induced apoptosis. Since the only known pharmacological action of Pin is to activate K ATP channels (14), we determined the role of the K ⫹ channels in the Pin-induced apoptosis of HepG2 cells. To this end, we firstly examined whether Pin decreases [K ⫹] i as a consequence of K ⫹ channel activation assessed by the K ⫹-sensitive fluorescent dye PBFI/AM. Unexpectedly, Pin induced a rapid and sustained increase in [K ⫹] i in a dosedependent manner as shown in Fig. 2A. The Pininduced increase in [K ⫹] i was not altered by specific inhibitors of K ATP channels (100 M Glib, 100 M Tolb). Secondly, we investigated the effects of these inhibitors on the Pin-induced apoptosis. They did not significantly affect the apoptosis induced by Pin (1 mM) as shown in Fig. 2B. These results strongly suggest that K ATP channels may not play a role in the mechanism of the Pin-induced apoptosis in the HepG2 cells.
achieved by activation of Na ⫹, K ⫹-ATPase and/or NKCC, we examined the effects of selective inhibitors of these pathways on the [K ⫹] i increase and apoptosis induced by Pin (1 mM). Treatment with ouabain (100 M), a selective inhibitor of Na ⫹, K ⫹-ATPase did not alter both [K ⫹] i increase and apoptosis by Pin as shown in Figs. 3A and 3B, respectively, suggesting that activation of Na ⫹, K ⫹-ATPase may not mediate these actions of Pin. However, NKCC inhibitors (10 M bumetanide, 50 M furosemide) markedly suppressed these effects of Pin as illustrated in Figs. 3A and 3B. In addition, the Pin-induced [K ⫹] i increase was profoundly inhibited by using extracellular Cl ⫺-free buffer (Fig. 3A), which further demonstrates that Pin activates the NKCC. Taken together, these results suggest that Pin may activate the NKCC, and in turn, induce apoptosis in the HepG2 cells. To our knowledge this is the first report that revealed the role for NKCC in induction of apoptosis.
NKCC mediates the Pin-induced [K ⫹] i increase and apoptosis. Next, we investigated the possible sources of the increased [K ⫹] i by Pin. Since K ⫹ influx can be
Role for intracellular Ca 2⫹ in the Pin-induced apoptosis. Since intracellular Ca 2⫹ signal appears to act as a common mediator of apoptosis (19), we investi-
513
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Effects of inhibitors of K ATP channels on [K ⫹] i (A) and apoptosis (B) induced by Pin in HepG2 human hepatoblastoma cells. The data (A) show changes in [K ⫹] i as a function of time, which was measured by using K ⫹-sensitive fluorescent dye PBFI/AM. The arrows show the time points for addition of Pin (1 mM). Glib (100 M) and Tolb (100 M), specific inhibitors of K ATP channels, were added 10 min before Pin treatment. In the experiments of (B) the cells were incubated with Pin (1 mM) for 48 h. Drugs (100 M Glib, 100 M Tolb) were added 30 min before Pin treatment. The number of apoptotic cells was measured by flow cytometry. In bar graphs the data represent the mean values of four replications with bars indicating SEM. *P ⬍ 0.05 compared to control.
gated whether intracellular Ca 2⫹ plays a role in the Pin-induced apoptosis. Intracellular Ca 2⫹ concentration ([Ca 2⫹] i) was measured by using Fura-2 fluorescence technique. As shown in Fig. 4A, Pin (1 mM) induced a rapid and sustained increase in [Ca 2⫹] i. To determine the source of the Pin-induced intracellular Ca 2⫹ increase, we measured [Ca 2⫹] i using a nominal Ca 2⫹-free medium containing 100 M EGTA. This experimental protocol can effectively reduce extracellular free Ca 2⫹ concentration, and thus, blunt available Ca 2⫹ influx. Under these conditions cellular Ca 2⫹ response to Pin (1 mM) was markedly reduced as illustrated in Fig. 4A. However, treatment with intracellular Ca 2⫹ release blockers (20 M TMB-8, 50 M dantrolene), did not significantly alter the Ca 2⫹-increasing effect of Pin. These results indicate that the increased [Ca 2⫹] i by Pin may be predominantly due to Ca 2⫹ influx. As shown in Fig. 4B, the Pin-induced apoptosis was significantly suppressed either by 0.5 M BAPTA/AM, an intracellular Ca 2⫹ chelator, or by 1 mM EGTA, an extracellular Ca 2⫹ chelator. However, intracellular Ca 2⫹ release blockers, TMB-8 (20 M) and dantrolene (50 M) did not significantly alter the apoptosis induced by Pin. These results suggest that Ca 2⫹ influx may mediate the Pin-induced apoptosis.
Activation of the reverse mode of Na ⫹, Ca 2⫹exchanger mediates the Pin-induced [Ca 2⫹] i increase and apoptosis. Although the Pin-induced [Ca 2⫹] i increase was found to be due to Ca 2⫹ influx (Fig. 4A), The exact pathway responsible for Ca 2⫹ influx is not known. Thus to determine the exact Ca 2⫹ influx pathway activated by Pin, we investigated the effects of various pharmacological agents which can inhibit the potential pathways of Ca 2⫹ influx, on the Pin-induced [Ca 2⫹] i increase. Neither inhibitors of voltage-sensitive Ca 2⫹ channels (100 M verapamil, 100 M nifedipine), nor an inhibitor of nonselective cation channels (100 M FA) did significantly alter the Ca 2⫹ influx induced by Pin (1 mM) as depicted in Fig. 5A. In addition, these agents did not significantly influence the Pin-induced apoptosis as shown in Fig. 5B. These results suggest that Ca 2⫹ influx as well as apoptosis induced by Pin may be mediated not through these ion channels, but through other mechanisms. Interestingly, inhibitors of Na ⫹ , Ca 2⫹ exchanger (50 M bepridil, 50 M benzamil) completely suppressed both Ca 2⫹ influx and apoptosis induced by Pin (1 mM) as illustrated in Figs. 5A and 5B, respectively. These results indicate that the reverse mode of the Na ⫹ , Ca 2⫹ exchanger may be activated by Pin
514
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Roles of NKCC in the Pin-induced [K ⫹] i increase and apoptosis in HepG2 human hepatoblastoma cells. The data (A) show changes in [K ⫹] i as a function of time, which was measured by using K ⫹-sensitive fluorescent dye PBFI/AM. The arrows show the time points for addition of Pin (1 mM). Ouabain (100 M), an inhibitor of Na ⫹, K ⫹-ATPase, and NKCC inhibitors (10 M bumetanide, 50 M furosemide) were added 10 min before Pin treatment. For extracellular Cl ⫺ ([Cl ⫺] ex)-free buffer solution, external Cl ⫺ was replaced with gluconate. In the experiments of (B) the cells were incubated with Pin (1 mM) for 48 h. Drugs (100 M ouabain, 10 M bumetanide, 50 M furosemide) were added 30 min before Pin treatment. The number of apoptotic cells was measured by flow cytometry. In bar graphs, the data represent the mean values of four replications with bars indicating SEM. *P ⬍ 0.05 compared to control. # P ⬍ 0.05 compared to Pin alone.
and that Ca 2⫹ influx through the exchanger may lead to apoptosis. Role of NKCC in the activation of the reverse mode of Na ⫹, Ca 2⫹-exchanger induced by Pin. To determine whether the activation of the reverse mode of Na⫹/Ca 2⫹ exchanger is secondarily achieved by increased intracellular Na ⫹ concentration ([Na⫹] i), we monitored [Na ⫹] i using a Na ⫹-sensitive fluorescent dye SBFI/AM. As shown in Fig. 6A, Pin increased [Na⫹] i. This [Na ⫹] i increase was markedly suppressed by NKCC inhibitors (10 M bumetanide, 50 M furosemide). Moreover, inhibitors of Na ⫹, Ca 2⫹ exchanger (50 M bepridil, 50 M benzamil) profoundly augmented the [Na⫹] i increase by Pin. Next, we tested whether activation of NKCC mediates the [Ca 2⫹] i-increasing effect of Pin. As shown in Fig. 6B, treatment with the NKCC inhibitors (10 M bumetanide, 50 M furosemide) completely prevented the [Ca 2⫹] i increase by Pin, implying that activation of NKCC may be necessary for the [Ca 2⫹] i-increasing effect of Pin. Taken together, these results suggest that Pin induces Ca 2⫹ influx through activation of the reverse mode of Na ⫹/Ca 2⫹ exchanger which is achieved by increased [Na ⫹] i attributed to activation of NKCC. DISCUSSION In this study we demonstrated for the first time that Pin induces apoptosis and that NKCC is involved in
the mechanism of apoptosis in HepG2 human hepatoma cells. The induction of apoptosis by Pin was demonstrated by using three independent methods, detection of DNA fragmentation through agarose gel electrophoresis (Fig. 1A), detection of phosphatidylserine translocation through annexin-V binding assay (Fig. 1B), and measurement of hypodiploid DNA contents through flow cytometry (Fig. 1C). The following results strongly suggest the essential role for the NKCC in the mechanism of apoptosis by Pin. Pin induced increases in [K ⫹] i as well as [Na ⫹] i, which was markedly suppressed by bumetamide and furosemide, a highly selective and less selective inhibitors of the NKCC, respectively (8) (Figs. 3A and 6A). The Pininduced increased [K ⫹] i was significantly reduced by [Cl ⫺] ex-free buffer (Fig. 3A). The Pin-induced apoptosis was completely suppressed by the NKCC inhibitors (Fig. 3B). The results that Pin induced [K ⫹] i increase rather than decrease (Fig. 2A), imply no involvement in the action mechanism of Pin of activation of K ATP channels which is the only known pharmacological action of Pin (14). No significant effects of selective inhibitors of the K ATP channels (Glib and Tolb) on the Pin-induced increased [K ⫹] i and apoptosis (Fig. 2) consistently support the notion that the K ATP channels may not play a role in these activities of Pin. Recently, Malhi et al. have suggested that K ATP channels exist in HepG2
515
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. Role of Ca 2⫹ influx in the Pin-induced apoptosis in HepG2 human hepatoblastoma cells. The data (A) represent intracellular Ca 2⫹ changes with time assessed by Fura-2 fluorescence technique. The arrows show the time points for addition of Pin (1 mM). EGTA (100 M) in a nominal Ca 2⫹-free medium, TMB-8 (20 M) and dantrolene (50 M) were added 10 min before Pin application. Quantitative changes (right bar graphs) were expressed as percent increases in [Ca 2⫹] i induced by the drug compared to control condition in which the cells were incubated with a drug-free vehicle. In the experiments of (B) the cells were incubated with Pin (1 mM) for 48 h. Drugs (0.5 M BAPTA/AM, 500 M EGTA, 20 M TMB-8, 50 M dantrolene) were added 30 min before Pin treatment. The number of apoptotic cells was measured by flow cytometry. In bar graphs (A and B) the data represent the mean values of four replications with bars indicating SEM. *P ⬍ 0.05 compared to control. # P ⬍ 0.05 compared to Pin alone.
cells, detected by using method of polymerase chain reaction (PCR) (20). However, the results of this study suggest that the total K ⫹ conductance due to the activation of the K ATP channels by Pin may be too small to reduce [K ⫹] i. In addition to activation of NKCC, [K ⫹] i increase can be possibly achieved by other mechanisms, i.e., inhibition of background K ⫹ channels and/or activation of Na ⫹, K ⫹-ATPase. The contribution of these pathways to the Pin-induced increase in [K ⫹] i was investigated. Treatment with ouabain, a specific inhibitor of Na ⫹, K ⫹-ATPase, did not alter both [K ⫹] i increase and apoptosis induced by Pin (Figs. 3A and 3B), suggesting that Na ⫹, K ⫹-ATPase may not be involved in these actions of Pin. Since the increased [K ⫹] i by Pin was not completely blocked by NKCC inhibitors (Fig. 3A), the residual portion may be due to inhibition of the background K ⫹ channels. Although the contribution of this channel activity to apoptosis can not be completely excluded, it is evident that the NKCC plays a major role in the Pin-induced apoptosis, since it was completely suppressed by the NKCC inhibitors (Fig. 3B). Accumulating evidence implies that intracellular Ca 2⫹ is commonly involved in the mechanism of apoptosis (19). One of the targets for elevated intracellular Ca 2⫹ is the activation of the Ca 2⫹-dependent protein
kinases and phosphatases that has been seen during apoptosis (21, 22). Direct activation of the Ca 2⫹dependent proteinase may represent another target for intracellular Ca 2⫹ action in apoptosis (23). Ca 2⫹/Mg 2⫹dependent endonuclease whose activation results in DNA fragmentation, the most characteristic biochemical feature of apoptosis (24), and Ca 2⫹-dependent transglutaminase which is highly activated in apoptotic cells (25), also appear to be a target for Ca 2⫹ action (26, 27). In HepG2 cells we have found that a sustained Ca 2⫹ influx mediates apoptotic cell death induced by various substances including tamoxifen (28) and tertbutyl hydroperoxide (29). Therefore, we investigated a possible role of Ca 2⫹ signal in the Pin-induced apoptosis. The results showed that Pin increased [Ca 2⫹] i through Ca 2⫹ influx, since the increased [Ca 2⫹] i by Pin was significantly prevented by using nominal Ca 2⫹-free medium containing 100 M EGTA, but not altered by internal Ca 2⫹ release blockers, TMB-8 and dantrolene (Fig. 4A). Consistently, the Pin-induced apoptosis was significantly prevented by EGTA as well as BAPTA/ AM, an intracellular Ca 2⫹ chelator, but not affected by these inhibitors of internal Ca 2⫹ release (Fig. 4B). Interestingly, the inhibitors of NKCC, bumetanide and furosemide completely suppressed the Pin-induced [Ca 2⫹] i increase (Fig. 6B). Thus these results suggest
516
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 5. Involvement of Na ⫹, Ca 2⫹-exchanger in the Pin-induced [Ca 2⫹] i increase and apoptosis in HepG2 human hepatoblastoma cells. Experimental procedures and data presentations are the same as Fig. 4. In the experiments inhibitors of Na ⫹, Ca 2⫹-exchanger (50 M bepridil, 50 M benzamil), voltage-sensitive Ca 2⫹ channel blockers (100 M verapamil, 100 M nifedipine) and an inhibitor of non-selective cation channels (100 M FA) were used. In bar graphs the data represent the mean values of four replications with bars indicating SEM. *P ⬍ 0.05 compared to control. # P ⬍ 0.05 compared to Pin alone.
that intracellular Ca 2⫹ may mediate apoptosis induction by Pin and that the Pin-induced activation of the NKCC may be necessarily involved in the [Ca 2⫹] i increase. Extracellular free Ca 2⫹ ions may enter into the cells by the following two mechanisms: (i) activation of plasma membrane Ca 2⫹ channels, and (ii) activation of reverse mode of Na ⫹/Ca 2⫹ exchange mechanism. In this study voltage-dependent Ca 2⫹ channel blockers, verapamil and nifedipine did not significantly alter the Pin-induced [Ca 2⫹] i increase and apoptosis (Figs. 5A and 5B), suggesting no involvement of these ion channels in the activity of Pin in the HepG2 cells. Recently, Ca 2⫹-permeable nonselective cation channels have been reported to exist in the HepG2 cells (30). Since these channels have mediated Ca 2⫹ influx in many different types of cells (31, 32), as well as in HepG2 cells (28), we determined the possible role for this channel in the Pin-induced Ca 2⫹ influx using FA, a known inhibitor of the channels. No significant effects of FA on the Pin-induced Ca 2⫹ influx and apoptosis (Figs. 5A and 5B) further indicate that these types of plasma membrane Ca 2⫹ channels may not act as the Ca 2⫹ influx pathway activated by Pin. On the other hand, bepridil and benzamil, known inhibitors of Na ⫹ /Ca 2⫹ exchanger significantly suppressed both Ca 2⫹ influx and apoptosis induced by Pin (Figs. 5A and 5B), implying that the activation of reverse mode of Na ⫹ /Ca 2⫹ exchanger may be critical for these actions of Pin. Na ⫹ /Ca 2⫹ exchanger nor-
mally acts to extrude Ca 2⫹ ions when intracellular Ca 2⫹ rises above certain levels (33). On the contrary, Ca 2⫹ ions enter into the cells under conditions that favor the reverse mode of operation of the Na ⫹ /Ca 2⫹ exchanger (33). Reverse operation of the Na ⫹ /Ca 2⫹ exchanger during anoxia has been reported to be a critical mechanism of Ca 2⫹ influx and subsequent neuronal cell injury (34). The activation of reverse mode of Na ⫹ /Ca 2⫹ exchanger may be a second event following to the increased [Na ⫹ ] i through the activated NKCC by Pin. This possibility was investigated, and the results showed that Pin indeed increased [Na ⫹ ] i immediately which was markedly reduced by bumetanide and furosemide (Fig. 6A). Furthermore, enhancement of the Pin-induced [Na ⫹ ] i increase by bepridil and benzamil (Fig. 6A) strongly support the idea that the reverse mode of Na ⫹ /Ca 2⫹ exchanger is activated by increased [Na ⫹ ] i through the Pin-induced stimulation of NKCC. Currently, we do not know the exact mechanism by which Pin activates NKCC in the HepG2 cells. Although it may be the most plausible explanation that Pin directly activates the NKCC, indirect regulation of the activity of this cotransport may also be possible. The NKCC activity seems to be regulated by phosphorylation and dephosphorylation of the protein via protein kinases and phosphatases, respectively (8). Thus Pin may regulate the activity of the cotransport via these mechanisms. However, the possibility remains to be determined in the future studies.
517
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 6. NKCC mediates activation of reverse mode of Na ⫹, Ca 2⫹-exchanger and [Ca 2⫹] i increase induced by Pin in HepG2 human hepatoblastoma cells. The data (A) show changes in [Na ⫹] i as a function of time, which was measured by using Na ⫹-sensitive fluorescent dye SBFI/AM. The arrows show the time points for addition of Pin (1 mM). NKCC inhibitors (10 M bumetanide, 50 M furosemide), and of Na ⫹, Ca 2⫹-exchanger (50 M bepridil, 50 M benzamil) were added 10 min before Pin treatment. The data (B) represent intracellular Ca 2⫹ changes with time assessed by Fura-2 fluorescence technique. Results were expressed as percent increases in [Ca 2⫹] i induced by the drug compared to control condition in which the cells were incubated with a drug-free vehicle. Bumetanide (10 M) and furosemide (50 M) were added 10 min before Pin (1 mM) application.
In conclusion, Pin induced apoptosis through activation of intracellular Ca 2⫹ signal in HepG2 human hepatoblastoma cells. Pin may activate NKCC, which results in [Na ⫹ ] i increase, and in turn, leads to the stimulation of the reverse mode of Na ⫹ , Ca 2⫹ exchanger, resulting in [Ca 2⫹ ] i increase. These results further suggest that NKCC may be a good target for the induction of apoptosis of human hepatoma cells. ACKNOWLEDGMENT This work was supported by Korea Research Foundation Grant (KRF-2000-D0341).
REFERENCES 1. Kidd, V. J. (1998) Annu. Rev. Physiol. 60, 533–573. 2. Nayfield, S. G., Karp, J. E., Ford, L. G., Dorr, F. A., and Kramer, B. S. (1991) J. Natl. Cancer Inst. 83, 1450 –1459. 3. Schulte-Hermann, R., Bursch, W., Low-Baselli, A., Wagner, A., and Grasl-Kraupp, B. (1997) Cell. Biol. Toxicol. 13, 339 –348. 4. Kastan, M. B., Canman, C. E., and Leonard, C. J. (1995) Cancer Metastasis Rev. 14, 3–15. 5. Kornblau, S. M. (1998) Leukemia 12, S41–S46. 6. Kamesaki, H. (1998) Int. J. Hematol. 68, 29 – 43. 7. Crompton, N. E. (1998) Acta Oncol. 37, 1– 49.
8. Russell, J. M. (2000) Physiol. Rev. 80, 211–276. 9. Delpire, E., Rauchman, M. I., Beier, D. R., Hebert, S. C., and Gullans, S. R. (1994) J. Biol. Chem. 269, 25677–25683. 10. Greger, R. (1996) Pflugers Arch. 432, 579 –588. 11. Alvarez-Leefmans, F. J., Gamino, S. M., Giraldez, F., and Nogueron, I. (1988) J. Physiol. 406, 225–246. 12. Hara, E., Reinach, P. S., Wen, Q., Iserovich, P., and Fischbarg, J. (1999) J. Membr. Biol. 171, 75– 85. 13. Leung, S., O’Donnell, M. E., Martinez, A., and Palfrey, H. C. (1994) J. Biol. Chem. 269, 10581–10589. 14. Edwards, G., and Weston, A. H. (1995) Cardiovasc. Drugs Ther. 9, 185–193. 15. Hockenbery, D., Nunez, G., Milliman, C., Schreiber, R. D., and Korsmeyer, S. J. (1990) Nature 348, 334 –336. 16. Bombeli, T., Karsan, A., Tait, J. F., and Harlan, J.M. (1997) Blood 89, 2429 –2442. 17. Vermes, I., Haanen, C., Steffens-Nakken, H., and Reutelingsperger, C. (1995) J. Immunol. Methods 184, 39 –51. 18. Ito, S., Ohta, T., Kadota, H., Kitamura, N., and Nakazato, Y., (1997) J. Neurosci. Methods 75, 21–27. 19. McConkey, D. J., and Orrenius, S. (1996) J. Leukoc. Biol. 59, 775–783. 20. Malhi, H., Irani, A. N., Rajvanshi, P., Suadicani, S. O., Spray, D. C., McDonald, T. V., and Gupta, S. (2000) J. Biol. Chem. 275, 26050 –26057. 21. Bonnefoy-Berard, N., Genestier, L., Flacher, M., and Revillard, J. P. (1994) Eur. J. Immunol. 24, 325–329.
518
Vol. 281, No. 2, 2001
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
22. Shibasaki, F., Kondo, E., Akagi, E., and McKeon, F. (1997) Nature 386, 728 –731. 23. Squier, M. K. T., and Cohen, J. J. (1997) J. Immunol. 158, 3690 –3697. 24. Wyllie, A. H., Morris, R. G., Smith, A. L., and Dunlop, D. (1984) J. Pathol. 142, 67. 25. Fesus, L., Thomazy, V., and Falus, A. (1987) FEBS Lett. 224, 104 –108. 26. Cohen, J. J., and Duke, R. C. (1984) J. Immunol. 132, 38 – 42. 27. Melino, G., Annicchiarico-Petruzzeli, M., Piredda, L., Candi, E., Gentile, V., Davies, P. J., and Piacentini, M. (1994) Mol. Cell. Biol. 14, 6584 – 6596.
28. Kim, J. A., Kang, Y. S., Lee, S. H., Lee, E. H., and Lee, Y. S. (1999) Cancer Lett. 147, 115–123. 29. Kim, J. A., Kang, Y. S., Lee, S. H., and Lee, Y. S. (2000) Free Rad. Res. 33, 267–277. 30. Chen, W. H., Yeh, T. H., Tsai, M. C., Chen, D. S., and Wang, T. H. (1997) J. Form. Med. Assoc. 96, 503–510. 31. Demaurex, N., Monod, A., Lew, D. P., and Krause, K. H. (1994) Biochem. J. 297, 595– 601. 32. Ling, B. N., and O’Neill, W. C. (1992) Am. J. Physiol. 263, H1827–H1838. 33. DiPolo, R., and Beauge, L. (1987) Hypertension 10, 115–119. 34. Stys, P. K., Waxman, S. G., and Ransom, B. R. (1991) Ann. Neurol. 30, 375–380.
519