Adaptive concentrations of hydrogen peroxide suppress cell death by blocking the activation of SAPK/JNK pathway

Although oxidative stress is generally considered to be cytotoxic, it can be beneficial to cell viability under certain circumstances. For example, when cells are first exposed to low H₂O₂ concentrations, they develop a resistance to subsequent challenges with high concentrations of the same agent that would otherwise be lethal (Farr and Kogoma, 1991; Davies et al., 1995; Wiese et al., 1995; Lee and Um, 1999). As this adaptive response requires a time lag (Wiese et al., 1995; Lee and Um, 1999), it is widely accepted that this protective effect of H₂O₂ is induced by an accumulation of H₂O₂-responsive proteins. Given that H₂O₂ can elevate the cellular levels and activities of H₂O₂-degrading enzymes such as glutathione peroxidase and catalase (Shull et al., 1991; Lu et al., 1993; Lee and Um, 1999), it is believed that cells adapt to H₂O₂, at least in part, by enhancing their capacity to degrade H₂O₂.

In recent years, evidence has been accumulating that this adaptation may not always depend on H₂O₂-degrading enzymes. It has been shown that H₂O₂ induced the synthesis of 20-25 new proteins in vascular endothelial and fibroblast cells (Lu et al., 1993; Wiese et al., 1995). Although the nature of these proteins is largely unknown, the induction of numerous proteins by H₂O₂ implies a system of H₂O₂ adaptation involving multiple mechanisms. Consistent with this possibility, H₂O₂ pretreatment was found to confer on human U937 leukemia cells a cross-resistance to other lethal stimuli such as serum withdrawal and C₂-ceramide (Lee and Um, 1999). Interestingly, it was observed that C₂-ceramide, in the absence of such pretreatment, killed U937 cells in a manner independent of H₂O₂. A lethal concentration of C₂-ceramide did not elevate the cellular levels of H₂O₂ in U937 cells, and C₂-ceramide-induced cell death was not suppressed by the addition of antioxidants (Lee and Um, 1999). Therefore, the H₂O₂-induced resistance of U937 cells to C₂-ceramide cannot be explained by the enhanced cellular capacity to degrade H₂O₂. Moreover, H₂O₂ was able to adapt U937 cells without signs of enhanced antioxidant capacity, such as an increase in protein levels and activities of the H₂O₂-degrading enzymes or an enhancement in the cellular capacity to degrade H₂O₂ (Lee and Um, 1999). A similar observation was also reported using hamster fibroblast cells (Wiese et al., 1995). Taken together, these observations suggests that H₂O₂ can impart cells with a survival advantage in a manner independent of H₂O₂-degrading activity. In the case of U937 cells, this alternative mechanism was selectively induced when the cells were exposed to 0.05 mM H₂O₂, whereas an increase in H₂O₂-degrading activity required 0.25 mM H₂O₂ (Lee and Um, 1999). Therefore, the former mechanism appears to be more sensitive to H₂O₂ than the latter, at least in U937 cells. On the basis of these observations, it was suggested that relatively low H₂O₂ concentrations protect the cells by blocking the lethal signaling triggered by subsequent stimuli. It was the purpose of this study to investigate this hypothesis using U937 cells. A prerequisite of this investigation was to define the cell death pathways induced by H₂O₂, serum withdrawal and C₂-ceramide, because H₂O₂ adaptation controls the degree of lethality of all of these stimuli. The data presented in this report indicates that stress-activated protein kinase/c-Jun N-terminal kinase (SAPK/JNK), a member of the mitogen-activated protein kinase (MAPK) family, acts as a common mediator of U937 cell death induced by high H₂O₂ concentrations, serum withdrawal and C₂-ceramide. Moreover, evidence is provided showing that low H₂O₂ concentrations induce a protective response against all these lethal stimuli by specifically blocking their ability to activate SAPK and its upstream kinases. The mechanism of this interesting phenomenon is discussed.

U937 cells were obtained from the American Type Culture Collection (Rockville, MD). All the antibodies used in this study were raised against human antigens. The anti-MKK4/SEK1 (anti-MAPK kinase 4) and anti-MKK7/SKK4 (anti-MAPK kinase 7) antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA), and the anti-SAPK was obtained from PharMingen/Transduction Laboratories (San Diego, CA). All other antibodies were purchased from New England Biolabs (Beverly, MA). New England Biolabs and Upstate Biotechnology (Lake Placid, NY) supplied the recombinant c-Jun and SAPK proteins. The phosphorylated heat and acid-stable protein (PHAS)-1 was purchased from Stratagene (La Jolla, CA).

U937 cells were cultured in a RPMI 1640 medium supplemented with 10% heat-inactivated FBS and gentamicin (50 μg/ml). Cells at a concentration of 3×10⁵/ml were exposed to H₂O₂ (1 mM), C₂-ceramide (0.06 mM), or washed in PBS and then cultured in a serum-depleted medium. Where indicated, the cells were pretreated with 0.05 mM H₂O₂ for set periods before receiving the lethal treatments. For DNA transfection, the dominant negative mutants of MKK4 (Sanchez et al., 1994) and MKK7 (Moriguchi et al., 1997) were cloned into the pcDNA vector and delivered into the U937 cells by electroporation. The transfected cells were selected by using 1 mg/ml of G418 sulfate, after which they received the indicated treatments.

The treated and untreated control cells received propidium iodide (PI) (5 μg/ml) followed by flow cytometry analysis to simultaneously monitor the PI uptake (FL-2 channel) and cell size (forward light scatter). The cells that displayed both a normal size and a low permeability to PI were understood to be viable cells, as previously defined (Mangan et al., 1991). All other populations were understood to be dead.

The cells were lysed in a solution containing 70 mM β-glycerophosphate (pH 7.2), 0.1 mM sodium orthovanadate, 2 mM MgCl₂, 1 mM EDTA, 1 mM DTT, 0.5% Triton X-100 and protease inhibitors (2 μg/ml aprotinin, 2 μg/ml leupeptin and 100 μg/ml PMSF). After the removal of the cell debris by centrifugation at 13,000 g for 15 minutes, equal amounts of proteins (50 μg) were separated by 12% SDS-PAGE. The proteins were then electrotransferred to Immobilon membranes (Millipore, Bedford, MA), which were subsequently blotted using the indicated antibodies and visualized by chemiluminescence (ECL; Amersham, Arlington Heights, IL).

The cells were lysed in a Hepes buffer (50 mM, pH 7.4) containing 100 mM NaCl, 10% glycerol, 1% NP-40, 1 mM EDTA, 20 mM β-glycerophosphate, 1 mM NaF, 1 mM p-nitrophenyl phosphate, 1 mM sodium orthovanadate and the protease inhibitors listed above. The lysates were clarified by centrifugation at 13,000 g for 15 minutes. Immunoprecipitation was performed by using 400 μg of the lysate proteins and the indicated antibodies. The precipitates were resolved in 20 μl of a kinase buffer containing 20 mM Hepes (pH 7.4), 10 mM MgCl₂, 20 mM β-glycerophosphate, 10 mM NaF, 1 mM DTT, 0.5 mM sodium orthovanadate, 50 μM ATP, and 10 μCi [γ-³²P]ATP. The kinase reactions were initiated by adding 2 μg of the specified substrates to the solution. After a certain incubation time, the reaction was stopped by adding a boiled sample buffer, and the proteins were then separated by 12% SDS-PAGE. The gels were dried, and a PhosphoImager using Tina 2.0 software visualized the radioactive bands.

SAPK has emerged as a mediator of the cell death induced by H₂O₂ (Verheij et al., 1996) and many other cytotoxic agents (Cross et al., 2000). As with other members of the MAPK family, such as the extracellular signal-regulated kinase (ERK) and the p38 MAPK, SAPK is activated by its phosphorylation (Whitmarsh and Davis, 1996). Therefore, to confirm the ability of H₂O₂ to activate SAPK under the experimental conditions of this study, U937 cells were exposed to lethal H₂O₂ concentrations (1 mM). After various incubation times, the SAPK phosphorylation levels were monitored by western blot analysis using an antibody specific to the phosphorylated forms of SAPK. This H₂O₂ treatment resulted in an increase in the phosphorylation levels of two SAPK isoforms, p54 and p46 (Fig. 1A). This effect was most clear when the cells were exposed to H₂O₂ for 15-30 minutes. Under these conditions, the SAPK protein levels did not change significantly, suggesting that the enhanced SAPK phosphorylation levels reflected its activation. This was confirmed by an in vitro kinase assay using c-Jun as a substrate (Fig. 1B). Although it was reported that H₂O₂ can activate ERK and p38 MAPK under other experimental conditions (Guyton et al., 1996; Wesselborg et al., 1997), the phosphorylation levels of these MAPKs were not significantly enhanced after up to 30 minutes of H₂O₂ treatment (Fig. 1A). Similar results were obtained when the ERK and p38 MAPK activities were analyzed using PHAS-1 as a substrate (Wang et al., 2000; Zhang et al., 2000) (Fig. 1B). All these results suggest that 1 mM H₂O₂ specifically, or at least most efficiently, activates SAPK in U937 cells.

SAPK can be phosphorylated/activated by two upstream MAPK kinases, MKK4 (Sanchez et al., 1994) and MKK7 (Moriguchi et al., 1997; Tournier et al., 1997). To determine which one was involved in the H₂O₂-induced activation of SAPK, the activities of MKK4 and MKK7 were analyzed using SAPK as a substrate. Treatment with H₂O₂ resulted in increased activity of both MKK4 and MKK7 (Fig. 1B), which occurred more rapidly than SAPK activation, being evident as early as 5 minutes after H₂O₂ treatment. Therefore, both MKK4 and MKK7 appeared to act as mediators of the H₂O₂-induced SAPK activation. To confirm this, the U937 cells were stably transfected with a pcDNA vector containing the dominant negative mutant of either MKK4 (U937/MKK4) or MKK7 (U937/MKK7). Cells receiving the empty vector (U937/pcDNA) were used as a control. As in the case of untransfected U937 cells, 1 mM H₂O₂ efficiently activated SAPK in U937/pcDNA cells (Fig. 1C). However, this activation was greatly reduced in both U937/MKK4 and U937/MKK7 cells. Therefore, it was clear that both MKK4 and MKK7 mediate H₂O₂-induced SAPK activation.

The expression of MKK4 mutants has been reported to suppress H₂O₂-induced death in U937 cells (Verheij et al., 1996). To investigate whether this can also be achieved using the MKK7 mutant, the transfectants were exposed to various H₂O₂ concentrations, and their viability was compared using flow cytometry. Unlike the untransfected U937 cells that were marginally susceptible to 0.1 mM H₂O₂ (Lee and Um, 1999), the exposure of U937/pcDNA cells to the same H₂O₂ concentration for 48 hours resulted in a cell death of more than 60% (Fig. 1D). Therefore, the susceptibility of U937 cells to H₂O₂ appeared to be enhanced during the transfection procedure, as reported previously (Kim et al., 2001). Increasing the concentration of H₂O₂ to 0.25 mM killed almost all of the U937/pcDNA cells. However, the cell death was dramatically reduced when the U937/MKK7 cells were exposed to the same H₂O₂ concentrations. A direct comparison of U937/MKK7 and U937/MKK4 cells revealed that the MKK7 mutant was slightly more protective than the MKK4 mutant. These results suggested that both MKK4 and MKK7 act as major mediators of H₂O₂-induced death in U937 cells.

The role of the SAPK pathways in the cell death induced by serum withdrawal and C₂-ceramide was investigated. Both of these stimuli activated SAPK and MKK7 (Fig. 2A), but interestingly MKK4 was not stimulated (Fig. 2B). Activation of SAPK, induced by either serum withdrawal or C₂-ceramide, was consistently reduced only by mutant MKK7 expression (Fig. 2A), but not by mutant MKK4 (Fig. 2B). Moreover, although the MKK7 mutant efficiently reduced the cell death induced by serum withdrawal and C₂-ceramide, such protection was not detected using the MKK4 mutant (Fig. 2C). This finding contrasts with a previous report showing that the expression of MKK4 mutant rescued U937 cells from C₂-ceramide (Verheij et al., 1996). Although the reason for this discrepancy is not clear, our data suggests that serum withdrawal and C₂-ceramide selectively utilize MKK7 to activate SAPK and kill U937 cells, at least under the experimental conditions of this study. Considering these results and the above findings together, the route leading to SAPK activation appears to vary in a single cell type depending on the types of stimuli. This has also been reported in other studies (Moriguchi et al., 1997).

Because the SAPK pathway was shown to be crucial for the lethal actions of H₂O₂, serum withdrawal, and C₂-ceramide, it was of interest to investigate whether H₂O₂ pretreatment protects the cells from those stimuli by interfering with their ability to activate the SAPK pathway. To address this question, U937 cells were incubated in the presence or absence of 0.05 mM H₂O₂ for 24 hours, an optimal condition for inducing the antioxidant-independent adaptation (Lee and Um, 1999). The pretreated and untreated control cells were then challenged with 1 mM H₂O₂ for 30 minutes, and analyzed for the levels of SAPK phopshorylation/activity. Pretreatment with H₂O₂ itself did not significantly influence the levels of SAPK phopshorylation/activity as analyzed from 5 minutes (data not shown) up to 24 hours. However, it did dramatically reduce the challenge-induced enhancement of SAPK phosphorylation/activity (Fig. 3A). By contrast, H₂O₂ pretreatment did not significantly influence phosphorylation levels of ERK and p38 MAPK, regardless of the challenge (data not shown). This data suggested that H₂O₂-induced SAPK activation was specifically downregulated in the cells that experienced sublethal concentrations of H₂O₂. To determine whether the pretreated cells also have a defect in MKK4 and MKK7 activation, the ability of H₂O₂ to activate MKK4 and MKK7 in the pretreated and untreated cells was compared. As shown in Fig. 3A, the H₂O₂-induced activation of both MKK4 and MKK7 was dramatically reduced by H₂O₂ pretreatment. Therefore, the cells that have previously experienced low concentrations of H₂O₂ lose their ability to efficiently activate the MKK4 and MKK7 pathways in response to subsequent high H₂O₂ concentrations.

To investigate whether H₂O₂ pretreatment similarly influences the SAPK activation induced by other stimuli, the pretreated and untreated cells received a challenge with either serum withdrawal or C₂-ceramide. As shown in Fig. 3B, H₂O₂ pretreatment efficiently reduced the enhancement of SAPK phosphorylation/activity induced by both serum withdrawal and C₂-ceramide. Moreover, the ability of these stimuli to activate MKK7 was also attenuated by pretreatment. These findings indicate that H₂O₂ pretreatment interferes with the SAPK pathway activation induced by diverse stimuli.

H₂O₂ requires a time lag to induce cellular resistance to death stimuli (Wiese et al., 1995; Lee and Um, 1999). In the case of U937 cells, pretreatment with 0.05 mM H₂O₂ for 4 hours was not sufficient for inducing such resistance. This resistance was initially detected 8 hours after the pretreatment, and was further enhanced as the pretreatment was extended up to 24 hours (Lee and Um, 1999). To investigate whether a similar time lag was also required to inhibit SAPK activation, U937 cells were pre-exposed to 0.05 mM H₂O₂ for various time periods, challenged with 1 mM H₂O₂ for 15 minutes and then analyzed for MKK4 and MKK7 activity. Alternatively, SAPK activity was analyzed 30 minutes after the challenge. Pretreatment itself did not significantly influence the activities of all those kinases under any of the tested conditions. Interestingly, the challenge-induced activation of MKK4, MKK7 and SAPK were not significantly suppressed in the cells that had been pretreated for 4 hours (Fig. 4). Pretreatment for 8 hours resulted in a dramatic suppression of MKK7, but not MKK4, activation. Therefore, the MKK4 and MKK7 pathways appear to be differentially regulated by H₂O₂ pretreatment. Pretreatment for 8 hours also displayed some suppressive effect on SAPK activation, suggesting that the selective MKK7 suppression can influence SAPK activation. When the cells were pretreated for 24 hours, the challenge-induced activation of all of MKK4, MKK7, and SAPK were almost completely blocked. These results revealed that the time course required for the inhibition of SAPK activation co-relates exceedingly well with that observed for the induction of cellular resistance to H₂O₂. This strongly supports the suggestion that the suppression of SAPK activation is a mechanism whereby H₂O₂ induces this protection.

This study showed that H₂O₂ can induce an adaptive response by specifically suppressing the activation of SAPK and its upstream kinases. As H₂O₂ also exhibits such an effect against serum withdrawal and C₂-ceramide, the suppression of the SAPK pathways appears to be a mechanism whereby H₂O₂-adapted cells withstand diverse stimuli. Given that the H₂O₂ concentration employed for the pretreatment does not enhance cellular antioxidant capacity (Lee and Um, 1999), the suppression of the SAPK pathways is believed to be due to a blockage of the pathways rather than a degradation of the challenged H₂O₂. This belief is further supported by the suppression of SAPK pathways induced by C₂-ceramide, which acts on U937 cells in a H₂O₂-independent manner (Lee and Um, 1999). Although cellular adaptation to oxidative stress has long been observed, to our knowledge this is the first case demonstrating that the response can modulate a specific signaling pathway. Our findings also suggest that oxidative stress can regulate the SAPK pathway in two opposing manners depending on its intensity. Although relatively high-intensity stress rapidly activates the pathway, lower-intensity stress induces a blockage of the pathway. This latter mechanism may have applications in certain types of cancer cells that constitutively generate reactive oxygen intermediates to levels higher than their normal counterparts, and display a resistance to oxidative cytolysis and chemotherapy (O’Donnell-Tormey et al., 1985; Yagoda, 1989; Toyokuni, 1995).

The time lag required for inducing the SAPK-suppressing response implies that H₂O₂ induces macromolecules that can inhibit the activation of the SAPK pathways. Given that the MKK4- and MKK7-suppressing responses are induced by different kinetics, multiple factors may be involved in suppressing the SAPK pathways. Although the identity of such factors is currently unclear, several anti-death proteins such as Bcl-2, Bcl-X_L and heat-shock proteins have been eliminated as potential candidates (Lee and Um, 1999). Because SAPK is activated by its phosphorylation, we have investigated the possibility that H₂O₂ pretreatment (0.05 mM) induces phosphatases such as MKP-1 and MKP-2 that can dephosphorylate SAPK (Chu et al., 1996; Sanchez-Perez et al., 2000). However, no evidence from western blot analysis was found (data not shown). Although it has been reported that the overexpression of p21^{WAF1/CIP1/Sdi1}, an inhibitor of cyclin-dependent kinases, can attenuate UV-induced SAPK activation (Shim et al., 1996), the adaptive concentrations (0.05 mM) of H₂O₂ in this study did not significantly alter the cellular levels of p21 (data not shown). Exposure of the U937 cells to 0.05 mM H₂O₂ consistently showed no significant influence on their cycling distribution. Therefore, the adaptation induced by 0.05 mM H₂O₂ does not appear to depend on the p21 level or the cellular cycling status. The authors have recently reported that nuclear factor κB (NF-κB) is a mediator of the U937 cell adaptation induced by 0.05 mM H₂O₂ (Kim et al., 2001). However, the promotion of H₂O₂-induced NF-κB activation by the overexpression of I-κB kinase α (IKKα) (Kim et al., 2001) did not enhance the SAPK-suppressing activity of H₂O₂ (data not shown). This observation suggests that NF-κB is not involved in the SAPK-suppressing pathway induced by H₂O₂. Therefore, the antioxidant-independent adaptation appears to be induced by at least two different mechanisms: one mediated by NF-κB and the other that suppresses SAPK activation. The authors are currently attempting to establish experimental conditions for proteomics to enable them to identify a cellular factor involved in these phenomena.

This work was supported by the Korea Science and Engineering Foundation in the program years of 1999-2000 and also by a Nuclear R&D Program (2001) from the Ministry of Science and Technology in Korea.