Glutathione-S-transferase π inhibits As₂O₃-induced apoptosis in lymphoma cells: involvement of hydrogen peroxide catabolism

Arsenic trioxide (As₂O₃) is an effective agent for the treatment of relapsed and refractory acute promyelocytic leukemia (APL) and acts by inducing apoptosis and partial differentiation.^1-3  Although As₂O₃ induces apoptosis in other forms of malignant tumors, different mechanisms of action have been identified. In some leukemia and lymphoma cells, As₂O₃, at therapeutic concentrations (1-2 μM), induces apoptosis through radical oxygen species–mediated pathways.^4-8  Moreover, in leukemia cells, the H₂O₂ scavenging systems (glutathione peroxidase and catalase) and arsenic detoxification systems (glutathione-S-transferase and glutathione) are key factors in controlling cell sensitivity to As₂O₃-induced apoptosis.⁴  However, in solid tumors, As₂O₃, at high concentrations (greater than 10 μM), is required to induce apoptosis that involves the Jun N-terminal kinase (JNK)–mediated pathway.^9-13 

Glutathione-S-transferase π (GSTP1), a member of the As₂O₃ detoxification pathway, is increased in several arsenic-resistant cell lines.^14-18  As₂O₃-sensitive NB4 cells have low levels of GSTP1 compared with other leukemia cell lines.⁴  These observations suggest that GSTP1 may be directly involved in As₂O₃-induced apoptosis. Jurkat cells are less sensitive to arsenic-induced apoptosis than Raji cells,¹⁹  and GSTP1 is highly expressed in Jurkat cells but undetectable in Raji cells.²⁰  In the present report, the role of GSTP1 in As₂O₃-induced apoptosis was studied in these 2 lymphoma cell lines. Raji cells were stably transfected with GSTP1, and the effects of GSTP1 on apoptosis, H₂O₂ accumulation, intracellular accumulation of arsenic, and JNK activation were compared in GSTP1-expressing and nonexpressing cells. It was found here that the overexpression of GSTP1 in Raji cells decreased the ability of therapeutic concentrations of As₂O₃ to induce apoptosis. As₂O₃-treated GSTP1-expressing Raji cells were also found to accumulate less intracellular H₂O₂ and As₂O₃ than Raji cells not expressing GSTP1, demonstrating 2 mechanisms for GSTP1 mediation of As₂O₃-induced apoptosis.

Arsenic (III) oxide (As₂O₃, at least 99% pure) and anticatalase monoclonal antibody were purchased from Sigma (St Louis, MO); G418 sulfate was purchased from Fisher Scientific (Pittsburgh, PA); anti-GST P1-1, anti-GST M1-1, and anti-GST A1-1 polyclonal antibodies were purchased from Calbiochem (La Jolla, CA); and antiphospho-SAPK/JNK (Thr183/Tyr185) monoclonal antibody (anti–p-JNK), anti-SAPK/JNK polyclonal antibody, antiphospho-p38 mitogen-activated protein (MAP) kinase (Thr180/Thr182) polyclonal antibody, and anti-p38 MAP kinase polyclonal antibody were purchased from Cell Signaling Technology (Beverly, MA). Antiglutathione peroxidase was obtained from Abcam (Cambridge, MA). Carrier-free [⁷³As]-arsenite (⁷³As^v) was purchased from Los Alamos Meson Production Facility (Los Alamos, NM). [⁷³As]-arsenate (⁷³As^III) was prepared from ⁷³As^v by reduction with metabisulfite/thiosulfate reagent.³⁴  Yields of ⁷³As^III in this reaction typically exceeded 95%, as determined by thin-layer chromatography (TLC).²¹ 

Raji and Jurkat human lymphoma cell lines were obtained from the American Type Culture Collection (ATCC, Rockville, MD) and were cultured in RPMI 1640 medium adjusted to contain 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 10 mM HEPES (N-2-hydroxyethylenepiperazine-N′-2-ethanesulfonic acid), 1.0 mM sodium pyruvate, 100 U/mL penicillin, 100 μg/mL streptomycin (Invitrogen, Carlsbad, CA), and 10% fetal bovine serum (JRH BioScience, Lenexa, KS) in a humidified atmosphere of 95% air and 5% CO₂ at 37°C.

Raji cells were transfected by electroporation with a pcDNA3.1 plasmid with or without a GSTP1 expression sequence (kindly provided by Dr Ze'ev Ronai, Mount Sinai School of Medicine, New York). Briefly, 10⁶ Raji cells in 1 mL mixed with 20 μg pcDNA3.1/GSTP1 plasmid were transferred to sterile electroporation cuvettes (Bio-Rad, Hercules, CA) and were electroporated in a GenePulser (Bio-Rad) with the voltage set at 300 V and the capacitor at 250 μF. After transfection, the cells were incubated in fresh medium containing 1 mg/mL G418 for 4 weeks. Subsequently, cell clones resistant to G418 were isolated and screened by limited dilution. Two GSTP1-expressing and 2 GSTP1-nonexpressing cell clones were selected for further study. Transfected Raji cell clones were routinely cultured in medium containing 1 mg/mL G418 and then were cultured for at least 24 hours without G418 before an experiment was initiated.

Cells were seeded at 1.5 to 2 × 10⁵ cells/mL and were cultured in medium described with or without the indicated concentrations of test compounds for the times indicated. Cell viability was estimated by trypan blue dye exclusion, and cell numbers were determined by hemocytometer.

Apoptotic cells were detected by Annexin V assay. In general, 10⁶ cells were washed twice with phosphate-buffered saline (PBS), then labeled by Annexin V–fluorescein isothiocyanate (FITC) and propidium iodide (PI) in binding buffer according to the instructions in the Annexin V–FITC Apoptosis Detection Kit provided by the manufacturer (Oncogene, Cambridge, MA). Fluorescence signals of FITC and PI were detected by FL1 (FITC detector) at 518 nm and FL2 at 620 nm, respectively, on a FACScan (Becton Dickinson, San Jose, CA). The log of Annexin V–FITC fluorescence was displayed on the x-axis, and the log of PI fluorescence was displayed on the y-axis. Data were analyzed using the CELLQuest (Becton Dickinson) software. For each analysis, 10 000 events were recorded.

Intracellular H₂O₂ level was detected as previously reported by using 5,6-carboxy-2′,7′-dichlorofluorescein-diacetate (DCFH-DA; Molecular Probes, Eugene, OR).⁴  Briefly, 2 hours before ending the indicated treatment, 5 μM DCFH-DA was added to the medium and was continuously incubated for 2 hours at 37°C; then the fluorescence intensity was measured by FACScan (Becton Dickinson).

GSTP1-1 activity was measured as previously described using 1-chloro-2, 4-dinitrobenzene (CDNB) as a high-affinity substrate and compared with that of other GST isoforms.²²  Briefly, 5 × 10⁷ cells were washed twice with cold PBS, resuspended in 300 μL of 100 mM potassium phosphate buffer (pH 6.8), and sonicated for 10 seconds at 4°C. After centrifugation at 17 000 g for 30 minutes, 50 μL cell lysate were mixed with 850 μL of 0.1 mM EDTA (ethylenediaminetetraacetic acid) (pH 6.5), 50 μL of 20 mM glutathione, and 50 μL of 20 mM CDNB. Absorbance of the mixture was continuously recorded for 2 minutes at 340 nm on a spectrophotometer (Ultrospec 2000; Pharmacia Biotec, Uppsala, Sweden).

Intracellular glutathione (GSH) contents were measured using the Glutathione Assay Kit (Calbiochem, San Diego, CA). In brief, 5 × 10⁶ cells were homogenized in 5% metaphosphoric acid using a Teflon pestle (Racine, WI). Particulate matter was separated by centrifugation at 4000g. Supernatant was used for GSH measurement according to the manufacturer's instruction. The GSH content was expressed as nmol/10⁶ cells.

Cells were centrifuged, washed with cold PBS, and lysed on ice for 30 minutes in RIPA buffer (1 × PBS, 1% nonidet P-40 [NP40], 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS]) containing protease and phosphatase inhibitors (1 mM phenylmethylsulfonyl fluoride, 1 mM Na₃VO₄, 5 mM NaF, 1 × protease inhibitor [Boehringer Mannheim GmbH, Mannheim, Germany]). Protein concentrations were determined with a Bio-Rad protein assay kit (Bio-Rad). Twenty-five micrograms to 70 μg total protein was electrophoresed on 8% to 12% SDS polyacrylamide gels, then transferred to a nitrocellulose membrane (Amersham, Piscataway, NJ). After incubating with 5% nonfat milk (Nestle Carnation) for an hour, the membrane was incubated with the indicated primary antibody overnight at 4°C, washed with Tris-buffered saline (pH 6.8) with Tween-20 (TBS-T) 3 times, incubated with secondary antibody for 1 hour at room temperature, and washed with TBS-T 3 times. The membrane was analyzed by autoradiography using a chemiluminescence kit (Amersham Life Science, Buckinghamshire, United Kingdom).

Total RNA was isolated from 10⁶ cells with an RNA isolation kit (Gentra, Minneapolis, MN). Twenty micrograms RNA was size fractionated on a 1.2% agarose–2.2 M formaldehyde gel, transferred to hybrid-N+ membrane (Amersham) in 20 × standard sodium citrate (SSC) solution, and UV cross-linked (Stratalinker; Stratagene, La Jolla, CA). Whole-length complementary DNA (cDNA) of GSTP1 was used as the probe. Probes were labeled with ³²P-dCTP by random priming to a specific activity of 0.5 to approximately 1 × 10⁹ cpm/ng. Membranes were prehybridized for 4 hours at 42°C in 50% formamide, 6 × sodium chloride, sodium phosphate, EDTA (SSPE), 5 × Denhardt reagent, and 0.2 mg/mL salmon sperm DNA and then were hybridized with a radiolabeled probe. Membranes were washed twice in 6 × SSC containing 0.1% SDS, followed by a stringent wash with 0.2 × SSC containing 0.1% SDS at 65°C.

Cells (10⁶) were incubated in complete medium with 1 μM ⁷³AsIII–carrier at 37°C in a humidified atmosphere containing 5% CO₂ for 24 hours. Cell were centrifuged at 500 rpm for 5 minutes and washed twice with PBS. Radioactivity in the sediment and in the medium plus washes was determined with the aid of a gamma counter.

Data were analyzed for statistical significance using the Student t test (Microsoft Excel; Microsoft, Seattle, WA). Differences were considered significant at P less than .05.

Apoptosis induction and growth inhibition abilities of As₂O₃ in Jurkat and Raji cells were compared after As₂O₃ treatment at therapeutic concentrations of 1 to 2 μM. As shown in Figure 1A, As₂O₃ treatment for 3 days caused more growth inhibition in Raji cells (inhibitory concentration 50% [IC₅₀], approximately 0.9 μM) than in Jurkat cells (IC₅₀, approximately 2 μM). In addition, the viability of Raji, but not of Jurkat, cells was decreased by As₂O₃ treatment in concentration- and time-dependent manners. As₂O₃ (2 μM) treatment for 3 days reduced the viability of Raji cells by 35%, whereas the viability of Jurkat cells was greater than 92% before and after As₂O₃ treatment. Percentages of apoptotic cells induced by 2 μMAs₂O₃ in Raji and Jurkat cells was 34% and 9%, respectively. The extent of apoptosis induction as measured by Annexin V and PI staining and by PARP cleavage was consistent with the decrease in cell viability (Figure 1B-C). PARP, a substrate of caspase protease in an apoptosis-signaling pathway, was cleaved by 2 μMAs₂O₃ treatment in Raji, but not in Jurkat, cells.

We used Western and Northern blot analyses to determine GSTP1-1, GSTA1-1, and GSTM1-1 protein levels and GSTP1 mRNA levels in Jurkat and Raji cells. Jurkat cells expressed higher levels of GSTP1-1 protein and GSTP1 mRNA than did Raji cells (Figure 2). GSTM1-1 protein expression was almost the same in both cell types and appeared not to contribute to the total GST activity using CDNB as a substrate.²²  GSTA1-1 protein was not detectable in either cell line (data not shown). GSH level was the same in both cell lines (data not shown). Jurkat cells had higher levels of GSTP1-1 activity than Raji cells (Figure 2C), suggesting that the basal GSTP1 level might determine cell sensitivity to As₂O₃-induced apoptosis.

To investigate the role of GSTP1 in the regulation of cell sensitivity to As₂O₃, Raji cells were stably transfected with a plasmid with or without a GSTP1 expression sequence. Two Raji cell clones with GSTP1 expression (RG19 and RG20) and 2 Raji clones without GSTP1 expression (RV5 and RV7) were selected for further study. RG19 and RG20 cells had high GSTP1-1 activity—up to 120 nmol CDNB/min per milligram protein—which is similar to that in Jurkat cells, whereas RV5 and RV7 cells had low GSTP1-1 activity (less than 20 nmol CDNB/min per milligram protein) (Figure 3). Western blot analysis demonstrated that RG19 and RG20, but not RV5 and RV7, expressed GSTP1 mRNA and protein (Figure 3). Moreover, these 4 clones contained similar amounts of GSTM1-1, GSTA1-1, catalase, and glutathione peroxidase (GPx) based on Western blot analysis (Figure 3). As in parental Raji cells (Figure 1), As₂O₃ (2 μM) treatment induced apoptosis (23%) and reduced cell viability in RV7 cells (Figure 4). The apoptosis induction ability of As₂O₃ was partially blocked in GSTP1-transfected RG20 cells (Figure 4). These data suggest that GSTP1 might be one of the factors inhibiting As₂O₃-induced apoptosis.

Amounts of intracellular H₂O₂ were determined in Jurkat, Raji, and GSTP1-expressing Raji cells. The basal H₂O₂ level in Jurkat cells was lower than it was in Raji cells. As₂O₃ (2 μM) treatment increased H₂O₂ levels in Raji cells, but not in Jurkat cells, after 3 days of treatment (Figure 5A). The effect of GSTP1 on intracellular H₂O₂ levels in Raji clones grown in medium with or without 2 μM As₂O₃ for 48 and 72 hours was tested. RG20 and RG19 cells had lower intracellular H₂O₂ basal levels than RV5 and RV7 cells. H₂O₂ levels in RV5 and RV7 cells exposed to 2 μM As₂O₃ obviously increased at 48 hours and then continuously accumulated, up to as much as 5-fold, at 72 hours compared with untreated cells (Figure 5B), similar to the increases in parental Raji cells. However, H₂O₂ levels in RG20 and RG19 cells were only slightly increased at 48 hours and were 3-fold lower than those in RV5 and RV7 cells at 72 hours after the addition of 2 μM As₂O₃ (Figure 5B). Intracellular GSH levels were tested before and after As₂O₃ treatment in RV5 and RG19 cells. As₂O₃ treatment increased intracellular GSH content in RV5 cells, but not in RG19 cells (Figure 5C). After exogenous H₂O₂ was added to the growth medium, the elimination rate of H₂O₂ in RV5 and RG19 cells was determined. H₂O₂ was eliminated more rapidly from RG19 cells than from RV5 cells (Figure 5D). Mean levels of intracellular H₂O₂ 2 hours after the addition of 500 μMH₂O₂ to the growth medium were 79.6 in RV5 cells but only 39.5 in RG19 cells. Under the same conditions, the levels of GSH were much lower after H₂O₂ treatment in RG19 cells (Figure 5E). These data suggest that As₂O₃ treatment results in the accumulation of H₂O₂ by inhibiting GSH-involved peroxidase activity, whereas GSTP1-1 may compromise an alternative GSH-involved peroxidase pathway not inhibited by As₂O₃ treatment.

To test the role of GSTP1 on the cellular accumulation of As₂O₃, intracellular arsenic uptake was measured using ⁷³AsIII–arsenite. Similar ⁷³As uptake was found in GSTP1-expressing or -nonexpressing Raji cells after incubation for 1 hour (data not shown). Percentages of intracellular ⁷³As in RV5 and RV7 cells were 1.5- to approximately 2-fold higher than in RG20 and RG19 cells after incubation for 24 hours (Figure 6). These data indicate that overexpression of GSTP1 decreases As retention in the cells.

P-JNK is not detectable by Western blot analysis in any of the untreated cells. The phosphorylated form of JNK was not detected in either RV7 or RG20 cells after treatment with 1 to approximately 10 μMAs₂O₃ for 72 hours, although apoptotic cells were detected (data not shown; Figure 4). However, high concentrations of As₂O₃ (approximately 50-80 μM) treatment for 2 hours activated the phosphorylation of JNK in RV7 cell clones and, to a lesser extent, in RG20 cells (Figure 7B). Moreover, phosphorylation of p38 was markedly increased after treatment with high concentrations of As₂O₃ (approximately 50-80 μM), and there was no detectable difference in p-p38 levels between GSTP1-expressing and -nonexpressing Raji cells (Figure 7). These data suggest that JNK activation represents a stress response and would not contribute to the apoptosis observed in Raji cells after treatment with As₂O₃ at therapeutic concentrations.

Jurkat cells were less sensitive to As₂O₃-induced apoptosis than Raji cells, and Jurkat cells express higher levels of GSTP1 mRNA and activity than Raji cells, which are devoid of GSTP1 expression and activity (Figures 1, 2). Because other potential factors, such as GSTM1, catalase, GPx, and GSH levels, were equally expressed in both cell lines (Figure 3) and GSTA1 was absent, it appears that GSTP1 might be the factor that mediates the observed different cell sensitivities to As₂O₃. Stable transfection with GSTP1 in Raji cells decreased As₂O₃-induced apoptosis, suggesting that GSTP1 is indeed a potent inhibitor of As₂O₃-induced apoptosis (Figure 4).

It has been found that As₂O₃-induced apoptosis at therapeutic concentrations is associated with the up-regulation of H₂O₂.^4-8  As₂O₃ treatment significantly increased H₂O₂ levels in parental Raji cells but not in Jurkat cells (Figure 5). Raji cells transfected with GSTP1 had reduced levels of H₂O₂ production compared with vector-transfected cells. Moreover, given that vector-transfected cells contained relatively higher levels of H₂O₂ than GSTP1-transfected cells and that H₂O₂-scavenging enzymes, such as catalase, glutathione peroxidase, and GSTA1-1 were not changed in these cells (Figure 3), it appears that GSTP1 may function as a peroxidase to diminish intracellular H₂O₂. GSTA1, but not GSTP1, has been reported to have selenium-independent glutathione peroxidase activity.^23-26  Studies have reported that GSTP1 plays an important role in the detoxification of carcinogens and the prevention of DNA damage but not in H₂O₂ scavenging.^27-29  However, the observations that less H₂O₂ accumulation (Figure 5D) and more GSH depletion (Figure 5E) in Raji cells expressing GSTP1 treated with H₂O₂ support the hypothesis that GSTP1 can function as a glutathione-dependent peroxidase. The nature of GSTP1 must be further studied by chemical methods.

GSTP1 has also been reported to be involved in the detoxification of arsenic by an efflux system.^15,17,30  Increased GSTP1 expression levels and activity were observed in arsenic-tolerant and -resistant cells.^14-18  It is possible that GSTP1 facilitates the efflux of arsenite in the cells expressing GSTP1; in turn, GPx is not inhibited, and H₂O₂ does not accumulate after As₂O₃ treatment. The retention of As₂O₃ in Raji clones expressing GSTP1 was less than it was in Raji clones not expressing GSTP1 (Figure 6). That H₂O₂ did not significantly increase in GSTP1-expressing Raji cells after As₂O₃ treatment supports this possibility.

It has been reported that high concentrations of As₂O₃ treatment activated JNK and p38, members of stress-activated signal transduction pathways, and resulted in apoptosis in several leukemia and lymphoma cell lines.^11,12,31,32  Basal levels of the phosphorylated form of JNK and p38 were not detectable in Raji cell clones with or without GSTP1 expression. Although a significant apoptotic effect was observed after treatment with 2 μM As₂O₃ in vector-transfected Raji cells, activation of JNK and p38 was still not detectable in these cells, suggesting that JNK and p38 activation might not contribute to As₂O₃-induced apoptosis at therapeutic concentrations. Furthermore, as reported by other groups,^10-13  the phosphorylated forms of JNK and p38 were significantly increased by As₂O₃ treatment at higher concentrations (Figure 7B), suggesting that JNK and p38 activation might mediate a stress response. Recently, it was found that GSTP1 is an inhibitor of JNK activity through direct protein–protein interaction.³³  Forced expression of GSTP1 in Raji cells decreased JNK phosphorylation after exposure to high concentrations of As₂O₃, which is consistent with previous reports that GSTP1 inhibited JNK activation after stress treatments (high-dose H₂O₂ and As₂O₃). The phosphorylated form of p38 was equally increased after exposure to high concentrations of As₂O₃ treatment in Raji cells expressing or not expressing GSTP1 (Figure 7). These data suggest that under stress condition after high As₂O₃ treatments, GSTP1 may function as an inhibitor of JNK rather than of p38 kinase. However, at therapeutic As₂O₃ concentrations, caspase activation (Figure 1) but not JNK-mediated pathway(s) is correlated with apoptosis induction in lymphoma cells.

In conclusion, the data presented here indicate that GSTP1 might be an important factor in determining the cell sensitivity to As₂O₃-induced apoptosis in lymphoma. GSTP1 may inhibit arsenic-induced apoptosis through at least 2 mechanisms: the detoxification mechanism, which decreases arsenic intracellular retention, and the peroxidase mechanism, which catabolizes H₂O₂. Each mechanism results in a decrease in intracellular H₂O₂ level and in the inhibition of apoptosis observed in Raji cells transfected with GSTP1 and Jurkat cells after low-concentration As₂O₃ treatment.