Abstract
To investigate the possible role of p53 in the progression of paraquat-induced apoptosis, the authors used two cell lines that were wild-type p53–expressing human lung epithelial-like cell line (L132) and a p53-deficient human promyelocytic leukemia cell line (U937) and explored the linkage between p53, DNA damage, and apoptosis. Following paraquat exposure to L132 cells, the percentage of S-phase cells decreased significantly and the expression of p53 protein increased, suggesting that entry into S phase from G1 phase was blocked. U937 cells showed complete resistance to paraquat, although paraquat-evoked initial single-stranded DNA breaks was shown equally in either L132 or U937 cells, as assessed by single-cell gel electrophoresis. U937 and L132 cells die normally with similar kinetics when exposed to tumor necrosis factor in the presence of cycloheximide, indicating that their capacity to undergo p53-independent mechanisms of inducing apoptosis has an equal rate. These results suggest that paraquat-induced DNA damage caused G1 arrest and apoptosis only in L132 cells, and that p53 protein accumulation is required for the induction of apoptosis by paraquat.
Paraquat dichloride (methyl viologen), 1,1′-dimethyl-4,4′-bipyridinium dichloride, is a widely used herbicide that is highly toxic to humans and most mammalian species. It is accepted that the toxicity of paraquat is based on an oxygen free radical mechanism (Bus, Aust, and Gibson 1974; Hassan and Fridovich 1978). The mechanism of cellular damage after exposure to paraquat involves the P-450 reductase–dependent formation of reactive oxygen intermediates (ROIs) and subsequent interactions with lipids, proteins, and nucleic acids. Cycling of the paraquat molecule between its ionic and radical forms appears to induce conversion of molecular oxygen to the superoxide anion and to other ROIs. Treatment of cells with stimuli known to elicit excessive ROI formation (Buttke and Sandstrom 1994) as well as the depletion of cellular antioxidants (Jacobson 1996) often seems to produce apoptosis. Also, it has been demonstrated that paraquat induces apoptosis in 32D mouse myeloid cells (Fabisiak et al. 1997) and N9 mouse microglial cells (Vogt et al. 1998).
The wild-type tumor suppresser protein, p53, is known to be a critical regulator of the cellular response to DNA damage (Kastan et al. 1991). Normally, cells show low levels of p53 expression due to the extremely short half-life of this protein. Irradiation or exposure to DNA-damaging agents such as mitomycin A causes post-transcriptional stabilization of p53, which prolongs the half-life of the protein and results in its accumulation. Up-regulation of p53 either blocks cellular proliferation by causing G1 cell-cycle arrest and/or induces apoptosis. p53 has been proposed to inhibit progression through G1 into S phase via transactivation of p21WAF/CIP1, which inhibits G1 cyclin-dependent kinases, allowing time for DNA repair (El-Deiry et al. 1993; Harper et al. 1993).
Although the precise molecular mechanisms by which paraquat mediates apoptosis have not been elucidated, there are links suggested between p53, DNA damage, and the production of ROIs (Lotem et al. 1996; Polyak et al. 1997). In this study, we therefore explored the role of p53 during paraquat-induced DNA damage and apoptosis by using a wild-type p53–expressing human lung epithelial-like cell line (L132) and a p53-deficient human promyelocytic leukemia cell line (U937).
MATERIALS AND METHODS
Cell Culture
We used two cell lines that were wild-type p53–expressing fatal human lung epithelial-like cell line (L132) and a p53-deficient human promyelocytic leukemia cell line (U937). U937 cells lack detectable amounts of normal p53 mRNA because of a point mutation causing aberrant splicing (Sugimoto et al. 1992). L132 and U937 cells were respectively cultured in Eagle’s minimal essential medium (MEM) and RPMI-1640 medium (containing 10% heat inactivated fetal bovine serum, 100 U/ml of penicillin, 0.1 mg/ml of streptomycin, and 0.25 μg/ml of amphotericin B) at 37°C in a 5% CO2 atmosphere. All experiments were initiated at 48 h after plating, cells were treated by direct addition to the culture medium with 0 to 20 mM paraquat for 8 h. Paraquat was freshly prepared as concentrated stocks in phosphate-buffered saline (PBS). In another experiment, cells were exposed to 4 μM cycloheximide (CHX) for 2 h, and then treated with 50 IU/ml of recombinant human tumor necrosis factor (TNF) for 8 h to assess the p53-independent apoptotic capacity. TNF and CHX were administered from stock solutions prepared in PBS and stored −70°C. TNF was purchased from Genzyme (Cambridge, MA).
Determination of Cell Viability
The capacity to produce water-soluble formazan from WST-1 (Wako, Tokyo, Japan) was measured by the MTT tetrazolium salt assay (Mosmann 1983). Briefly, cells were incubated in 96-well plates (3 ×104 cells/well) in the presence of paraquat (Tokyo Chemical, Tokyo, Japan) or TNF/CHX for the times indicated and then were incubated with 5 mM WST-1 for 1 h at 37°C. The formazan product of WST-1 reduction by mitochondrial enzymes was measured by microplate reader at 450 nm.
Cell Cycle/Proliferation Assay
DNA synthesis was assessed by incorporation of 5-bromodeoxyuridine (BrdU; Becton Dickinson, Bridgewater Lane, NJ) and flow cytometric analysis (Hoy, Seamer, and Schimke 1989). Briefly, after incubation in the absence or presence of paraquat for the times indicated, cells were washed in PBS and incubated with 10 μM BrdU for 30 min. After that, the cells were fixed in 70% ethanol for 30 min at −20°C, and resuspended in 2 N HCl/0.5% Triton X-100 for 30 min at 25°C. After neutralization with 0.1 M Na2B4O7 (pH 8.5), the cells were resuspended in PBS containing 0.5% Tween 20/1% bovine serum albumin (BSA) and incubated with anti–BrdU-fluorescein isothiocyanate (Becton Dickinson) for 30 min at 25°C. After washing, the cells were resuspended in 5 μg/ml propidium iodide (Molecular Probes, Eugene, OR) in PBS. Up to 15,000 cells were studied using Profile Acquisition software and an Epics Profile flow cytometer (Coulter, Hialeah, FL). Detection of fluorescence was accomplished using 525 nm and 575 nm bandpass optical filters.
Quantification of DNA Fragmentation
Cytoplasmic histone-associated DNA fragments (mononucleosomes and oligonucleosomes) were quantified by photometric enzyme immunoassay with a cell death detection enzyme-linked immunosorbent assay (ELISA) (Boehringer Mannheim, Mannheim, Germany). Cells were washed in PBS and placed on ice for 30 min in lysis buffer (10 mM K-Na-phosphate buffer, pH 6.8 to 7.4, 138 mM NaCl, 2.7 mM KCl, 1% BSA, 1 mM EDTA, 0.2% Tween 20, and 0.1% Kathon CG). The cell lysate was centrifuged at 20,000 ×g for 10 min to remove nuclei containing high-molecular-weight unfragmentated DNA, and the supernatant was diluted to 1:10 with lysis buffer and incubated for 1 h at 25°C in a 96-well plate coated with anti-histone mouse monoclonal antibody. After washing, the plate was incubated with a peroxidase-conjugated anti-DNA mouse monoclonal antibody for 90 min at 25°C and washed again. The amount of peroxidase in the immune complexes corresponded to the nucleosome level in the cytoplasmic fraction and was determined by addition of a substrate (2,2′-azino-di-[3- ethylbenzthiazoline sulfonate]) to the plate and detection at 405 nm/490 nm.
TUNEL Assay
Apoptotic nuclear change were detected by the TUNEL assay (deoxynucleotidyltransferase [TdT]-mediated dUTP nick and labeling) using an in situ fluorescein apoptosis detection kit (Oncor, Gaithersburg, MD). Briefly, 1 ×106 cells were fixed in freshly prepared 4% neutral buffered formalin for 10 min at room temperature. After drying on a microscope slide, the cells were stained according to the manufacturer’s protocol. Positive controls were treated with DNase 1 (1 mg/ml; Oncor) for 10 min at room temperature. For negative controls, TdT was omitted from the reaction mixture. Examination was done using a confocal microscope GB200 (Olympus, Tokyo, Japan), with a krypton/argon laser (488 nm and 568 nm) being used for fluorescence excitation of fluorescein isothiocyanate and propidium iodide.
Single-Cell Gel Electrophoresis (Comet) Assay
Single-stranded DNA breaks were detected by a microgel electrophoreasis technique (Trevigen’s Comet Assay; Trevigen, Gaithersburg, MD). In brief, cells embedded in low-melting-point agarose gel were placed on a microscope slide, lysed with 2.5 M sodium chloride, 100 mM EDTA, 10 mM Tris, 1% sodium lauryl sarcosinate, and 0.01% Triton X-100 for 30 min at 4°C, and then immersed in an alkaline solution (pH >13) for 30 min at room temperature to remove the cellular cytoplasm and DNA-associated proteins. The liberated DNA was electrophoresed under alkaline conditions and stained with SYBR Green fluorescent dye, after which the comet tail length (CTL) was viewed under a confocal fluorescent microscope (Rojas, Lopez, and Valverde 1999). A total of 50 digitized image data points (25 from each of two duplicate slides) were measured.
Quantification of p53
The p53 protein level was quantified by immunoprecipitation followed by Western blot analysis (Messmer et al. 1994). Briefly, 1 ×107 cells were scraped off the culture plate, washed in PBS, and lysed for 30 min at 4°C in 1 ml lysis buffer (50 mM Tris, 150 mM NaCl, 5 mM EDTA, 0.5% Nonidet-P40, 1 mM phenylmethylsulfonyl fluoride, pH 7.5). The lysed cells were sonicated for 10 s using a Branson sonifier (100% duty cycle, output control 1). After centrifugation for 5 min at 13,000 ×g, nonspecific absorbants were removed from the resulting supernatant by incubation with 40 μl of 50% (ν/ν) protein A-Sepharose for 30 min at 4°C, followed by centrifugation for 15 min at 13,000 ×g. Then p53 was immunoprecipitated overnight at 4°C by adding 200 μl of anti–p53 mouse immunoglobulin (Ig)G2a monoclonal antibody (Bp53–11) (Progen, Heidelberg, Germany) and 50 μl of 50% protein A-Sepharose. Immune complexes were precipitated at 13,000 ×g for 60 s and washed 3 times with 500 μl of 5% sucrose, 1% Nonidet-P40, 0.5 M NaCl, 50 mM Tris, and 5 mM EDTA (pH 7.4). Finally, the samples were resuspended in 40 μl of 125 mM Tris, 2% sodium dodecyl sulfate (SDS), 10% glycerin, 1 mM dithiothreitol, and 0.002% bromophenol blue (pH 6.9) and boiled for 5 min. Proteins were separated on 10% SDS-polyacrylamide gels and blotted onto nitrocellulose sheets using the semidry blot system (BIO-RAD, 5.5 mA/cm2, 45 min; 25 mM Tris, 192 mM glycine buffer system). The sheets were washed twice with TBS (140 mM NaCl, 50 mM Tris, pH 7.2) containing 0.1% Tween-20 before blocking nonspecific binding with TBS containing 2% BSA overnight at 4°C. Then an anti-p53 antibody was added (Bp53-11; 1:10 in TBS containing 0.2% BSA) and incubation was done for 2 h at 25°C. p53 was then detected using a horseradish peroxidase–conjugated secondary antibody and a enhanced chemiluminescent (ECL) substrate (Amersham International).
Assessment of DNA Staining by Ethidium Bromide
Flow cytometric analysis of the early phase of apoptosis was performed by evaluation of the cellular light scatter profile and of nuclear DNA staining with ethidium bromide (Ferlini et al. 1996). Cells were washed in PBS and incubated with 4 μg/ml ethidium bromide (Molecular Probes) for 15 min at room temperature. Up to 15,000 cells were analyzed by flow cytometry and detection of fluorescence was accomplished using a 575-nm bandpass optical filter.
Statistical Analysis
All data are presented as the mean ±SD. Significant differences among group means were evaluated by one-way analysis of variance. If the analysis gave a significant result (p < .05), individual treatment means were further compared by Scheffé’s test.
RESULTS
Effect of Paraquat and TNF/CHX on Cell Viability
The MTT assay was used to assess the cytotoxicity of paraquat and TNF/CHX. Although treatment with TNF in the presence of the de novo protein synthesis inhibitor CHX induced cell death with similar kinetics in both L132 (Figure 1A ) and p53-deficient U937 (Figure 1B ) cells, U937 cells displayed a dramatic resistance to the effects of paraquat (Figure 1B ). Incubation of L132 cells with paraquat resulted in dose-dependent cell death (Figure 1A ). A progressive decrease of viability was observed after L132 cell exposure to 10 and 20 mM paraquat for 4 h or more.
Alkaline Comet Assay
Figure 2 shows the typical appearance of comets produced by alkaline comet assay. Untreated control L132 (Figure 2A ) and U937 (Figure 2C ) cells fluorescence were largely confined to the nucleus, because undamaged DNA is supercoiled and thus does not migrate very far. The mean CTL of control L132 and U937 cells ranged from 2.5 to 8.0 μm (3.1 ±4.5 μm) and from 2.0 to 9.5 μm (5.0 ±4.9 μm), respectively. L132 (Figure 2B ) and U937 (Figure 2D ) cells exposed to paraquat, showed extensive migration of DNA out of the nucleus to form a large comet tail. After 2 h of incubation in the presence of 10 mM paraquat, the mean CTL of L132 cells increased to 25.4 ±5.1 μm (range: 2.0–40 μm; p < .01) and the mean CTL of U937 cells also increased to 21.2 ±4.4 μm (range: 2.5–36 μm; p < .01). In cells with total DNA damage, alkaline treatment unwinds the DNA and releases fragments that migrate during electrophoresis.
DNA Fragmentation
We measured in situ nick-end labeling and cytosolic oligonucleosome-bound DNA to investigate whether the decrease of viability caused by paraquat and TNF/CHX was related to apoptosis. The time course of DNA fragmentation was assessed by ELISA for detection of oligonucleosome-bound DNA. L132 cells exposed to increasing concentrations of paraquat showed a significant increase of DNA fragmentation, which first became detectable after incubation with 10 and 20 mM paraquat for 6 h (Figure 3A ). In contrast, significant DNA fragmentation was not observed in U937 cells after 8 h of exposure to10 and 20 mM paraquat (Figure 3B ).
TNF plus CHX was capable of inducing significant DNA fragmentation both in the L132 and U937 cultured cells. After 4 h, significant DNA fragmentation was observed (Figure 3A and B ).
Confocal microscopy of L132 cells after exposure to paraquat for 8 h showed apoptotic changes (i.e., chromatin condensation and nuclear fragmentation), and there were TUNEL-positive cells (Figure 4B ). In contrast, no apoptotic cells and no 3′-OH ends produced by DNA fragmentation were detected when U937 cells were treated with paraquat (Figure 4D ).
Forward Scatter Plus DNA Staining with Ethidium Bromide
To detect the early phase of apoptosis after paraquat exposure, we evaluated the changes of forward scatter and DNA staining with ethidium bromide by flowcytometry (Figure 5). Four cell subsets could be recognized by assessing a combination of forward scatter and DNA staining according to the method of Ferlini et al. (1996). Viable cells (region 1) had no detectable DNA staining and unchanged forward scatter. Conversely, cells with altered forward scatter and strong DNA staining (region 4) were late apoptotic cells. Cells in regions 2 and 3 were considered to be in the early phase of apoptosis. After exposure of L132 cells to paraquat, there was a progressive shift of cells from regions 1 to 4, indicating the progression of apoptosis. In contrast, DNA staining and forward scatter were not altered at 8 h after exposure of U937 cells to paraquat.
Changes of Cell-Cycle Progression
We examined the effect of paraquat exposure on the cell cycle using unsynchronized L132 and U937 cells (Figure 6). In the absence of paraquat, the percentage of G0/G1 and S phase cells was 48.7% ±6.9% and 32.3% ±4.2% for L132 versus 43.3% ±2.5% and 38.0% ±2.7% for U937, respectively. After exposure to paraquat, L132 cells showed an increase of G1 cells and a decrease of S-phase cells, indicating cell cycle block at G1. The percentage of cells in G0/G1 phase increased to 66.3% ±2.9% (p < .01), and that of cells in S phase decreased to 14.0% ±2.2% (p < .01) after exposure to 10 mM paraquat for 5 h. In contrast, paraquat did not affect the cell cycle progression of U937 cells. The percentage of G0/G1- and S-phase cells after exposure to 10 mM paraquat for 5 h was 45.1% ±3.6% and 38.2% ±3.4% for U937, respectively.
p53 Protein Level
To investigate whether paraquat-related apoptosis involved p53 protein, we used immunoblot analysis with an antibody directed against p53 (Figure 7). L132 cells showed a time-dependent increase of p53 protein levels after exposure to 10 mM paraquat. In contrast, p53 accumulation was not observed up to 6 h after exposure of U937 cells to paraquat.
DISCUSSION
To investigate the possible role of p53 in the progression of paraquat-induced apoptosis, we used two cell lines that were wild-type p53 expressing (L132) and p53 deficient (U937). We showed that paraquat-induced DNA damage caused G1 arrest and apoptosis only in L132 cells, and that p53 protein accumulation was required for the induction of apoptosis by paraquat.
Oxidative stress resulting from the toxic effects of ROIs plays an important role in the pathogenesis of a variety of diseases and important biological processes (Boobis, Fawthrop, and Davies 1989). It is well established that paraquat undergoes repeated cycles of reduction and reoxidation, producing large amounts of ROIs that cause oxidative stress and cell damage (Farrington et al. 1973; Talcott, Shu, and Wei 1979; Kitazawa et al. 1991). The formation of paraquat radicals by NADPH-dependent cy-tochrome c reductase in the microsomes (Talcott, Shu, and Wei 1979) and xanthine oxidase–dependent paraquat reduction in the cytosol (Kitazawa et al. 1991) are the initial steps in paraquat toxicity.
Mechanisms proposed for ROI-mediated damage include interactions with protein thiol groups (Boobis, Fawthrop, and Davies 1989) and iron-sulfur proteins, or direct DNA damage (Peter et al. 1992; Dusinska et al. 1998) and lipid peroxidation (Dusinska et al. 1998; Bus, Aust, and Gibson 1975). Piotrowski et al. (1996) showed that lipid peroxides are not present in the lungs, the critical organ affected by paraquat toxicity, at 48 h after paraquat treatment of rats. In contrast, DNA damage is responsible for paraquat-induced toxicity in Ehrlich ascites cells (Peter et al. 1992) and in rat pneumocytes (Dusinska et al. 1998). The present study also demonstrated that paraquat could induce single-stranded DNA breaks after 2 h of treatment, as assessed by single cell gel electrophoresis. The fact that total DNA damage occurred before cell death indicates that such damage was indeed related to the toxicity of paraquat.
Paraquat-induced apoptosis was first demonstrated in mouse 32D cells (Fabisiak et al. 1997). It was found that overexpression of Bcl-2 afforded significant protection against paraquat-induced apoptosis and that phospholipase A2–independent peroxidation of specific membrane phospholipids is a critical event leading to apoptosis (Fabisiak et al. 1998). Moreover, splenocytes from Bcl-2–deficient mice were more susceptible to oxidants after incubation with paraquat than cells from wild-type mice, suggesting that paraquat-induced oxidative stress is involved in triggering apoptosis (Hockman et al. 1998). The involvement of ROIs in the occurrence of paraquat-induced apoptosis has also been reported using an in vivo model of the intrahippocampal injection of paraquat (Melchiorri et al. 1998). On the other hand, Vogt et al. (1998) showed that exposure of microglial cells to paraquat causes a rapid increase of APO-1/Fas ligand mRNA, suggesting that enhanced expression of APO-1/Fas might be involved in paraquat-induced apoptosis. Thus, the mechanism through which paraquat produces apoptotic cell death is not known. We hypothesized that paraquat-mediated DNA damage might cause the protein product of tumor suppresser gene p53 to accumulate at high levels, leading to growth arrest and apoptosis.
The differing sensitivity of L132 and U937 cells to paraquat seemed to be correlated with the expression of p53 protein, because both cell lines exhibited similar initial DNA damage after 2 h of treatment with 10 mM paraquat. Moreover, an equal rate of apoptosis induced by TNF plus CHX between these two cell lines indicates that their capacity to undergo p53-independent apoptosis is similar. p53 is known to be a cell-cycle regulator, arresting cells in the G1 phase, besides being a potent inducer of apoptosis (Lowe et al. 1993). We found that only L132 cells with the normal p53 gene exhibited a rapid increase in the level of p53 protein after exposure to paraquat, along with accumulation in G1 phase and induction of apoptosis, suggesting that p53 protein is an essential participant in the induction of paraquat toxicity. The changes of the cell cycle following paraquat exposure observed in this study were similar to the results reported by Tomita (1996). He found marked cell-cycle delay in HeLa cells treated with paraquat, suggesting that G1-phase arrest may have developed.
DNA-damaging agents such as either ionizing or ultraviolet (UV) radiation, and some chemicals, could induce p21 in wild-type p53–expressing cells, but not in cells with mutant p53 (El-Deiry et al. 1993). It is clear, however, that oxidative stress-induced apoptosis can occur in p53-deficient cells as well as in cells with mutations of the p53 gene. Therefore, p53 induction may not contribute to oxidative stress-induced apoptosis in all types of cells, but may be involved in genotoxic stress-mediated apoptosis.
In conclusion, our results indicate that paraquat-mediated oxidative stress is able to induce apoptosis, by acting in the G1 phase of the cell cycle and increasing the p53 protein level.