Immunostimulated glial cells simultaneously produce nitric oxide (NO) and superoxide anion (O₂^·-) via expression and activation of inducible nitric oxide (iNOS) (Simmons and Murphy, 1992). NO and O₂^·- rapidly react in a 1:1 stoichiometry to form a strong oxidant ONOO^- (Beckman et al., 1990). ONOO^- has been found to play a critical role in tissue injury and appears to mediate many of the cytotoxic effects (Beckman et al., 1990). Previously, we showed that immunostimulated astrocytes themselves became highly vulnerable to glucose deprivation (GD) (Choi and Kim, 1998; Choi et al., 2001). This enhanced vulnerability was attributed to the accumulation of ONOO^- due to the disruption of intracellular antioxidant system (i.e., decreased level of reduced glutathione) in glucose-deprived immunostimulated astrocytes (Choi and Kim, 1998a; 1998b; Choi et al., 2000; 2001). Expectedly, the ONOO^--releasing reagent 3-morpholinosydnonimine (SIN-1) also augmented the death of glucose-deprived astrocytes. Interestingly, we further found that the augmented death of immunostimulated or SIN-1-treated astrocytes under GD was dependent on the proliferative state of the astrocytes (Choi and Kim, 1998b).

Activation of iNOS was reported to induce cytotoxicity in part by disturbing cell cycle progression (Gu et al., 2000; Kotamraju et al., 2007). Apoptotic death of some cells can be blocked by coordinating the progression through the cell cycle and/or by preventing the inappropriate activation of cell cycle signalling pathways (Park et al., 1997; Zhu et al., 2007). Because proliferating immunostimulated astrocytes are much more vulnerable to GD, it is reasonable to think that the augmented death could be prevented by inhibition of cell cycle progression.

In the present study, therefore, we examined the possible cytoprotective effect of the DNA topoisomerase II inhibitor etoposide on peroxynitrite toxicity in proliferating astrocytes. Interestingly, the present results demonstrates for the first time that etoposide markedly blocks the augmented toxicity of ONOO^- in SIN-1-treated/glucose-deprived astrocytes by directly scavenging peroxynitrite, not by inhibiting cell cycle progression.

As we previously reported (Choi and Kim, 1998a; Choi et al., 2002a; 2002b), SIN-1 potentiated the LDH release (i.e., a cell death marker) in glucose-deprived astrocytes (Fig. 1). Presence of etoposide (10µM) 16 h before (i.e., pre-treatment) and during GD (i.e., co-treatment) completely blocked the potentiated cell death (Fig. 1). Similarly, cotreatment of etoposide alone also blocked the increased cell death. However, FACS analysis using propidium iodide showed that co-treatment of etoposide alone did not arrest the cell cycle progression (Table 1). It implies that the cytoprotective effect of etoposide may not caused by its inhibition of cell cycle progression.

In brain, one of many well-known actions of astrocytes is to produce antioxidants, which scavenge reactive oxygen species and buffer oxidative stress. Astrocytes are known to be the main source of GSH which acts as a free radical scavenger and is important in recycling other antioxidants (Makar et al., 1994; Ju et al., 2000). Thus, astrocytes appear to be resistant to the actions of the strong oxidant ONOO^- (Barker et al., 1996). In the present study, the level of GSH in astrocytes was synergistically reduced by SIN-1 and GD (Fig. 2A). Co-treatment of etoposide prevented the synergistic decrease of GSH (Fig. 2A). Pretreatment of Lbuthionine-S,R-sulfoximine (BSO, a relatively specific inhibitor of GSH biosynthesis) for 16 h depleted the intracellular GSH (Fig. 2B). Depletion of GSH itself did not induce the cell death, but increased the release of LDH in astrocytes exposed to SIN-1 or SIN-1/GD (Fig. 2C). Although etoposide did not block the GSH depletion by BSO, it inhibited the increased peroxynitrite toxicity by BSO (Fig. 2C).

The ONOO^- plays a critical role in the enhanced vulnerability of astrocytes to GD (Ju et al., 2000; Choi et al., 2002a; 2002b). To quantify the ONOO^- scavenging effect of etoposide, we used the dihydrorhodamine 123. We found that etoposide effectively inhibited ONOO^- (SIN-1)-induced oxidation of dihydrorhodamine 123 to rhodamine 123 in the presence of catalase (Fig. 3A). However, etoposide was little or less efficacious in scavenging NO (Fig. 3B) and O₂^·- (Fig. 3C).

Previously, we reported that GD increased the intracellular concentration of H₂O₂ in immunostimulated astrocytes, leading to astroglial cell death (Choi et al., 2004). Our present study showed that both GSH depletion and death in astrocytes simultaneously exposed to GD and H₂O₂ were not reversed by etoposide (Fig. 4A, B, respecively). We further found that etoposide did not effectively scavenge H₂O₂ (Fig. 4C).

Many researchers have used etoposide for experimentally inducing apoptosis in tumor cells. Studies on etoposide-induced apoptotic cell death include the processing of stabilized cleavage complexes into frank DNA strand breaks, sensing of the DNA damage, leading to activation of stress-asociated signaling pathways and cell cycle arrest (Kubota et al., 1990; Ryan et al., 1991; Tsao et al., 1992). Etoposide can also activate a pre-existing group of enzymes and enzyme precursors, typified by the cysteine-dependent aspartate-directed protease (caspases) (Henderson et al., 2005). The present study, however, demonstrates that the anti-tumoric DNA topoisomerase II inhibitor etoposide can reduce the peroxynitrite toxicity in glucose-deprived astrocytes by directly scavenging peroxynitrite, not by inhibiting cell cycle progression. Furtherore, etoposide interacts with ONOO^- most likely due to the high reactivity of ONOO^-, but not with other products NO, O₂^·- and H₂O₂.

The mammalian cell cycle maintains internal checkpoints at which cells are arrested when previous events have not been completed in response to external signals (Hartwell and Weinert, 1989). The cell cycle checkpoints are thought to exist at the G1/S and G2/M transitions (Agarwal et al., 1995). Etoposide has been used as a potent and reversible blocker of the cell cycle that appears to act at the G2/M transition. Previously, we found that the G1/S transition blockers mimosine and ciclopirox also blocked the peroxynitrite toxicity. Like etoposide, mimosine reduced the augmented death of glucose-deprived immunostimulated astrocytes by directly scavenging peroxynitrite rather than suppressing the cell cycle progression (Choi et al., 2002a). Unlike mimosine, however, ciclopirox blocked the peroxynitrite toxicity in glucose-deprived SIN-1-treated astrocytes by attenuating peroxynitrite-induced mitochondrial dysfunction, not by scavenging peroxynitrite (Choi et al., 2002b; Lee et al., 2005). Ciclopirox neither prevented the depletion of reduced glutathione in glucose-deprived SIN-1-treated astrocytes, nor scavenged oxidants such as peroxynitrite, nitric oxide, superoxide anion, hydrogen peroxide and hydroxyl radical (Choi et al., 2002b). Because of those various mechanisms of action, therefore, such cell cycle inhibitors should be carefully selected for the studies on oxidative stress-related cell deaths.

Table 1. Changes of cell cycle distribution by etoposide. Astrocytes were exposed to 200 mM SIN-1 and/or glucose deprivation with or without pre-(16 h) and co-treatment, or just co-treatment of etoposide