Role of phospholipase A2 activation and calcium in CYP2E1-dependent toxicity in HepG2 cells.

Previous studies suggested a role for calcium in CYP2E1-dependent toxicity. The possible role of phospholipase A2 (PLA2) activation in this toxicity was investigated. HepG2 cells that overexpress CYP2E1 (E47 cells) exposed to arachidonic acid (AA) +Fe-NTA showed higher toxicity than control HepG2 cells not expressing CYP2E1 (C34 cells). This toxicity was inhibited by the PLA2 inhibitors aristolochic acid, quinacrine, and PTK. PLA2 activity assessed by release of preloaded 3H-AA after treatment with AA+Fe was higher in the CYP2E1 expressing HepG2 cells. This 3H-AA release was inhibited by PLA2 inhibitors, a -tocopherol, and by depleting Ca 2+ from the cells (intracellular + extracellular sources), but not by removal of extracellular calcium alone. Toxicity was preceded by an increase in intracellular calcium caused by influx from the extracellular space and this was prevented by PLA2 inhibitors. PLA2 inhibitors also blocked mitochondrial damage in the CYP2E1 expressing HepG2 cells exposed to AA+Fe. Ca2+ depletion and removal of extracellular calcium inhibited toxicity at early time periods, although a delayed toxicity was evident at later times in Ca2+-free media. This later toxicity was also inhibited by PLA2 inhibitors. Analogous to PLA2 activity, Ca2+ depletion but not removal of extracellular calcium alone prevented the activation of calpain activity by AA+Fe. These results suggest that release of stored calcium by AA+Fe, induced by lipid peroxidation, can initially activate calpain and PLA2 activity, that PLA2 activation is critical for a subsequent increased influx of extracellular Ca2+, and that the combination of increased PLA2 and calpain activity, increased calcium and oxidative stress cause mitochondrial damage, that ultimately produces the rapid toxicity of AA+Fe in CYP2E1-expressing HepG2 cells. activation of PLA2 AA+Fe in the CYP2E1-expressing cells. As non-CYP2E1 dependent C34 and E47 cells were preloaded which results in cell death after more prolonged incubation. Given the importance of CYP2E1 in activating hepatotoxins such as acetaminophen, CCl4, benzene, nitrosamines, and in producing ROS, and contributing to alcohol-induced liver injury, these studies suggest that the control of intracellular calcium release and inhibition of PLA2 activity may represent strategies to block early events of CYP2E1-dependent cytotoxicity.


INTRODUCTION
Phospholipase A 2 1 (PLA2) comprises a set of extracellular and intracellular enzymes that catalyze the hydrolysis of the sn-2 fatty acyl bond of phospholipids to yield free fatty acids and lysophospholipids. The extracellular (secreted) PLA2s (sPLA2) require millimolar calcium concentrations for catalytic activity, and do not manifest significant fatty acid selectivity in vitro. In mammalian cells, as many as five different sPLA2 exist: groups I, IIA, IIC, V and X. The intracellular PLA2s are further divided into groups IV (cytosolic Ca 2+ dependent PLA2 or cPLA2) and VI (intracellular Ca 2+ independent PLA2 or iPLA2) based on the Ca 2+ requirements needed for basal activity. cPLA2 requires µM Ca 2+ for membrane translocation but not for catalysis, and possesses a preference for phospholipids containing AA. iPLA2 exhibits no substrate specificity for AA-containing phospholipids and no Ca 2+ requirement for activity.
Coexpression of different forms of PLA2 has been found within the same cell or tissue (1,2). The physiological functions of individual PLA2s in the hepatocyte are currently unknown (3).
Phospholipase A 2 activities of the mitochondrial and cytosolic fraction in the kidney were significantly enhanced after 1h ischemia followed by 1h reperfusion (15). Liver microsomes isolated from rats chronically treated with ethanol showed increased PLA2 and CYP2E1 activity, decreased content of AA and increased content of conjugated dienes (16).
Oxidative stress-associated activation of PLA2 has been proposed to be a critical factor in cytotoxicity.
PLA2 activation results not only in the degradation of membrane phospholipids but also in the accumulation of unsaturated free fatty acids and lysophospholipids which by themselves can be injurious (17). On the contrary, PLA2 has been shown to be protective in some settings of lipid peroxidation and hypoxia/reoxygenation (18)(19)(20)(21). There is mounting evidence that membrane lipid peroxidation stimulates phospholipid hydrolysis via Ca 2+ dependent PLA2 activity. It is well known that Ca 2+ mobilization and phospholipid degradation are closely interrelated. Membrane lipid peroxidation alters Ca 2+ homeostasis, and this appears to be an important initial signal for PLA2 activity. Peroxides are known to increase intracellular free Ca 2+ concentrations through either membrane disruption and physical disruption of ionic homeostasis, or by signaling release of intracellular calcium (22). In other cell systems, the mechanism of oxidant-induced AA release appears to be independent of Ca 2+ , resulting either from the activation of an iPLA2 (8,23,24), or the inhibition of AA esterification into phospholipids (25,26).
Many cytochrome P450s especially CYP2E1 can produce superoxide anion and hydrogen peroxide during their catalytic cycle (27,28). CYP2E1 is of interest because of its ability to metabolize and activate many substrates to reactive intermediates and to its role in alcohol liver injury. CYP2E1 overexpression in HepG2 cells (E47 cell line), in the absence of added toxin, is associated with increased cellular production of ROS and lipid peroxidation (29,30) and increased cytotoxicity of arachidonic acid (31), GSH depletion (29,30), iron (32) and ethanol (33). These cells when exposed to Fe-NTA showed an elevation of intracellular calcium and reactive oxygen species that occurs before the onset of cellular toxicity, events blocked by antioxidants. Control HepG2 cells not expressing P450 (C34 cells) showed very low toxicity and a small increase in intracellular calcium under the same conditions (34). Activation of calpain was observed together with cell death, both processes blocked by calpain inhibitors (34). An increase in intracellular calcium seems to be a very early and critical event in the toxicity caused by CYP2E1dependent oxidative stress, leading to the activation of Ca 2+ -dependent processes such as calpain activity.
The objectives of this work were: i) to investigate the possible activation of PLA2 activity in HepG2 cells  Cytotoxicity measurements: 5 x 10 4 cells were plated onto 24 well plates and after the corresponding treatment, the medium was removed, and cell viability was evaluated by the MTT test (38) as previously described (39). Another index of cytotoxicity used was the leakage of lactate dehydrogenase (40). LDH activity was assessed in a medium containing 50 mM KPi, pH 7.4, 170 µM NADH, and 562 µM pyruvate.
Cytotoxicity was expressed as % LDH release: (100 x (LDHout/LDH out+LDH in)). Lipid peroxidation assay: The production of thiobarbituric acid-reactive substances (TBARS) was assayed as previously described (38). The protein concentration of the cell suspension was determined using a protein assay kit based on the Lowry assay (BioRad DC kit). The concentration of MDA was calculated from a standard curve prepared using malonaldehyde bisdimethylacetal (42).

Measurement of calpain activity in situ:
The activity of calpain in intact cells was assessed with the use of the cell permeable fluorogenic substrate SUCC-LLVY-AMC (45) as previously described (34). The tested compounds at the concentrations used did not interfere with the quantification of AMC fluorescence.
Flow cytometry analysis of the mitochondrial membrane potential: The mitochondrial transmembrane potential was analyzed from the accumulation of rhodamine 123, a membrane-permeable cationic fluorescent dye. Cells were plated onto 6 well plates, and at the end of the treatment the medium was replaced with fetal bovine serum-free MEM exps containing 5 µg/ml rhodamine 123, and incubated at 37 °C for 1 h. The cells were then harvested by trypsinization, washed with PBS, and resuspended in 1 mL of fetal bovine serum-free MEM exps . The intensity of fluorescence from rhodamine 123 was determined using a BD FACSCalibur Flow Cytometer (San Jose, CA) as previously described (38).
Statistics: Data are expressed as mean ± standard error of the mean from 1 to 5 independent experiments run in duplicate. One-way analysis of variance (ANOVA) with subsequent post hoc comparisons by Scheffe was performed. A p < 0.05 was considered as statistically significant.

CYP2E1-dependent toxicity and protection by PLA2 inhibitors.
As previously described (38), addition of Fe+AA to CYP2E1-expressing HepG2 cells caused a rapid loss of cell viability (Fig 1A-E, 0 concentration of inhibitor). This mode of cell death was shown to be necrosis based upon assays of trypan blue uptake, release of lactic dehydrogenase and propidium iodide staining, and no DNA ladder formation (38). The effect of phospholipase A 2 inhibitors on this toxicity of Fe+AA was tested. Aristolochic acid and quinacrine (two general non-specific PLA2 inhibitors) partially prevented the toxicity of Fe+AA in these cells, with optimal concentrations (maximal inhibition with lowest toxicity by itself) of 200 µM and 100 µM, respectively) ( Fig 1A and 1B respectively). PTK, a specific inhibitor of cPLA2 and iPLA2, also showed significant protection against the toxicity, with 20 µM optimal concentration ( Fig 1C). BPB (selective inhibitor of sPLA2) and BEL (selective inhibitor of iPLA2) did not show significant protection against the toxicity of Fe+AA (Fig 1D and E, respectively). Control HepG2 cells not expressing CYP2E1 did not show significant toxicity after exposure to Fe+AA and/or aristolochic acid ( Fig 1F), confirming the role of CYP2E1 in this Fe+AA toxicity model.

CYP2E1-dependent toxicity and activation of PLA2.
As several PLA2 inhibitors partially protected the CYP2E1-expressing HepG2 cells from the toxicity of Fe+AA, we evaluated the activation of PLA2 in this system, by measuring in situ the release of 3 H-AA as described in Materials and Methods. Cells were first loaded with 2 µM AA + 0.2 µCi/ml 3 H-AA for 12h and after several washing steps, the percentage of release of 3 H-AA was measured at different time points after exposure to Fe-NTA. E47 cells treated with Fe+AA showed a time-dependent increased release of 3 H-AA (denoting increased PLA2 activity) with respect to cells which were treated only with AA (Fig 2A).
The non CYP2E1-expressing cells also showed an increased release of 3 H-AA especially after 3h of exposure to Fe-NTA, but the release was lower than that measured in the cells expressing CYP2E1 (Fig   2A). These results suggest an increased activation of PLA2 by AA+Fe in the CYP2E1-expressing cells. As a non-CYP2E1 dependent control, C34 and E47 cells were preloaded with 2 µM AA + 0.2 µCi/ml 3 H-AA and then exposed to a calcium ionophore, A23187 (a potent PLA2 activator), for up to 3h. Both cell lines showed comparable release of 3 H-AA by the calcium ionophore at all time points evaluated ( Fig 2B).
To evaluate simultaneously the kinetics of activation of PLA2 and the toxicity by Fe+AA, the release of 3 H-AA and of LDH was measured in the same experimental sample, as described in Materials and Methods.
3 H-AA release and LDH release proceeded simultaneously, both reactions starting after 1.25h of exposure of AA-loaded CYP2E1-expressing cells to Fe-NTA (Fig 3). Interestingly, a 5-fold increase in AA release (2% to 10% at 3h) was associated with a corresponding 5-fold increase in LDH release (5% to 25% at 3h).
If the tested PLA2 inhibitors prevented toxicity by effectively inhibiting PLA2 activity, then they should also block the AA+Fe-activated release of 3 H-AA. Table 1 shows that aristolochic acid, quinacrine and PTK added at their optimal protective concentrations (Fig 1) significantly inhibited the release of 3 H-AA triggered by Fe-NTA in AA-loaded cells expressing CYP2E1. BPB and BEL, inhibitors which did not block the toxicity of Fe+AA, also did not block the release of 3 H-AA. α-Tocopherol, which was shown to completely block the toxicity of AA+Fe (38), completely blocked the Fe+AA-activated release of 3 H-AA.
For the inhibitors tested, controls for comparison were carried out using the solvent in which the inhibitor was dissolved (ethanol for PTK and α-tocopherol, DMSO for BPB and BEL, and water for quinacrine and aristolochic acid) ( Experiments were carried out to check for the possible inhibitory effect of quinacrine, aristolochic acid and PTK on CYP2E1 activity and lipid peroxidation. CYP2E1 activity was measured in microsomes obtained from acetone-treated rats. The tested PLA2 inhibitors did not block CYP2E1 activity at the optimal concentration used, while a known CYP2E1 inhibitor, 50 µM 4-methyl pyrazole blocked CYP2E1 activity by 70% (data not shown). Since α-tocopherol blocked the AA+Fe toxicity and activation of PLA2, it was important to evaluate whether aristolochic acid, quinacrine or PTK had an antioxidant action similar to that of α-tocopherol. NADPH + (Fe)ADP-dependent lipid peroxidation, assessed as TBARS generation in microsomes obtained from control rats, was not blocked by the inhibitors tested, although 50 µM trolox significantly blocked (96%) TBARS generation (data not shown). The effect of the inhibitors on the TBARS content of homogenates from cells expressing CYP2E1 and exposed to Fe+AA was also tested to evaluate whether they exerted an antioxidant action in intact cells. As shown in Table 2, AA+Fe increased the production of TBARS 5 fold, an increase totally prevented by α-tocopherol. The PLA2 inhibitors did not have any effect on the content of TBARS of the Fe+AA-treated cells.

Activation of PLA2 and calcium.
Experiments were performed in order to test the Ca 2+ -dependence of the activation of PLA2 detected in the Fe+AA-treated E47 cells, since BEL, an inhibitor of the calcium-independent PLA2, did not protect against the AA+Fe toxicity or the 3 H-AA release. Cells were preloaded with 2 µM AA + 0.2 µCi/ml 3 H-AA, and after several washing steps the medium was replaced with PLA2 assay buffer containing 1 mM or 0 mM CaCl 2 . Buffer or 25 µM Fe-NTA was added, and the release of 3 H-AA was measured at different time points. As shown in Fig 4A, there was no statistical difference in 3 H-AA release between Ca 2+ containing or Ca 2+ -deficient medium in Fe+AA-treated cells, although after substracting out the release of 3 H-AA in the AA alone controls, a 42% decrease was found for the minus calcium medium.
Nevertheless, substantial AA release occurs in the Ca 2+ -deficient medium. In order to minimize adventitious calcium contamination in the buffer, PBS was treated with 5% (w/v) Chelex-100 for 1h (46).
The Chelex treatment did not significantly modify the release of 3 H-AA triggered by Fe-NTA in 0 mM Ca 2+ medium, suggesting that the effect was not due to contaminating extracellular calcium (data not shown). The release of 3 H-AA in Fe+AA treated cells in 0 mM Ca 2+ medium was efficiently blocked by aristolochic acid and α-tocopherol (Fig 4B), results similar to that observed in 1 mM Ca 2+ (Table 1).
Although  Table 3). Removal of extracellular Ca 2+ , and removal of extracellular calcium +intracellular calcium stores strongly reduced the toxicity of Fe+AA in the CYP2E1 expressing cells (Table 3).
An increase in intracellular calcium can be detected after exposure of these cells to Fe+AA in MEM but not in SMEM. This increase occurs prior to the developing toxicity and is inhibited by α-tocopherol, but not by calpeptin (34). Fig 5 shows

Activation of PLA2 and mitochondrial damage.
Mitochondrial damage is a very early effect leading to necrosis in several experimental models (48). In the CYP2E1-expressing HepG2 cells treated with Fe+AA, mitochondrial damage manifested as a decrease in mitochondrial membrane potential was a very early event that preceded the permeabilization of the plasma membrane (49). The effect of PLA2 inhibitors on this AA+Fe-induced mitochondrial damage was assessed. As shown in Fig 6, Fe+AA increased the number of cells in the M1 low membrane potential zone (low rhodamine 123 fluorescence). The PLA2 inhibitors aristolochic acid and PTK significantly prevented the decrease of mitochondrial membrane potential observed in the Fe+AA-treated E47 cells.
A control experiment was performed in order to test for a possible direct effect of PLA2 inhibitors on the mitochondrial permeability transition. Mitochondria were isolated, energized and loaded with Ca 2+ . After exposing mitochondria to 10 µM AA, the permeability transition-dependent swelling was measured on the basis of the absorbance changes at 540 nm, following the protocol described in (50). As was shown by Scorrano et al., (2001) (50), AA-dependent swelling was inhibited (60-80%) by 2 µM cyclosporin A, a specific inhibitor of the permeability transition (Fig 7). Stock solutions of cyclosporin A and AA were prepared in ethanol; control experiments showed that the vehicle alone did not induce any mitochondrial swelling and that the vehicle alone did not prevent the AA-induced swelling. As discussed in (51), inhibition and persistence of inhibition by cyclosporin A was dependent on the inducer of the mitochondrial permeability transition, ranging from complete to approximately 50%. AA-dependent swelling was not significantly inhibited by 20 µM PTK, 100 µM quinacrine, or 100 µM aristolochic acid, suggesting that PLA2 inhibitors do not directly affect the permeability transition (Fig 7). A PLA2 inhibitor, trifluoperazine, was reported to inhibit the mitochondrial permeability transition, however, this inhibition was related to a change of surface charge of the mitochondrial membrane and not inhibition of PLA2 (51). Thus, the ability of these inhibitors to prevent the decrease of the mitochondrial membrane potential by Fe+AA in the CYP2E1-expressing HepG2 cells is a reflection of their ability to inhibit PLA2 activity.

CYP2E1-dependent activation of calpain and calcium.
It has been previously shown that calpain is activated in the Fe+AA-treated cells expressing CYP2E1, and that toxicity is partially inhibited by 2 calpain inhibitors, calpeptin and E64d (34). Experiments were performed in order to evaluate if Fe+AA-activated calpain activity requires extracellular calcium, or if intracellular calcium stores can play a role in the activation analogous to what was observed for PLA2 activation. Fe+AA-activated calpain activity assessed in buffer containing 1 mM CaCl 2 was significantly blocked by calpeptin and completely prevented by α-tocopherol (Table 4). When the experiment was performed in buffer containing 0 mM CaCl 2 , Fe+AA again activated calpain activity; the net increase by Fe-NTA was 65% that in Ca 2+ -containing medium. This activation was also significantly inhibited by calpeptin and α-tocopherol (Table 4). These results suggest that the activation of calpain in the Fe+AAtreated cells does not require extracellular calcium, although it is still possible that calcium from intracellular stores may play a role in the activation. Thus, AA-loaded cells were depleted from calcium in intracellular stores with A23187+EGTA as described above, and Fe-NTA-activated calpain activity was assessed in these cells. Fe-NTA did not activate calpain activity in the Ca 2+ -depleted cells when incubated in Ca 2+ -deficient medium.

PLA2 activation and toxicity of Fe+AA in Ca 2+ -deficient medium.
PLA2 was activated by Fe+AA in the CYP2E1-expressing HepG2 cells in Ca 2+ -deficient medium (Fig   3). Toxicity by Fe+AA in Ca 2+ -deficient medium is much less than in Ca 2+ -containing medium (Table 3) especially at shorter incubation times e.g. <6h after addition of Fe-NTA, but significant toxicity is observed after 12h of exposure (34). This toxicity in SMEM at 12h incubation was not prevented by calpeptin (49). Is the activation of PLA2 in Ca 2+ -deficient medium related to the delayed toxicity of Fe+AA in these conditions? To test this, inhibitors of PLA2 were preincubated for 1h in AA-loaded cells, in SMEM, followed by the addition of Fe-NTA for 12h. As shown in Fig 8, Fe+AA caused an approximate 50% loss of cell viability in control incubations. The PLA2 inhibitors significantly blocked this toxicity.

DISCUSSION
Previous work (38,52,53) has shown that the toxicity of Fe+AA in CYP2E1-overexpressing HepG2 cells is principally mediated by oxidative stress. Toxicity is dependent on CYP2E1 expression, as HepG2 cells that do not express P450 activity show low levels of cytotoxicity by Fe+AA (38,53, and Fig 1). CYP2E1mediated generation of ROS in an intracellular environment containing an active metal catalyst (Fe) and an oxidizable substrate (AA) is a key component of oxidative stress and toxicity in this system. Previous studies indicated that there was an early toxicity phase which required extracellular Ca 2+ and a later toxicity phase which occurred in Ca 2+ -deficient medium (34). One Ca 2+ -activated hydrolase, calpain, played a role in the early but not later toxicity phase (49). The current study was carried out to assess the role of another Ca 2+ -activated hydrolase, PLA2, in the toxicity.
The CYP2E1-expressing HepG2 cells treated with Fe+AA showed increased PLA2 activity, measured as release of pre-labeled 3 H-AA. Release of prelabeled 3 H-AA is an appropriate measure of PLA2 activity because: i) prelabeled 3 H-AA was mainly (86%) incorporated into phospholipids, in the sn-2 position predominantly, thus presenting an adequate substrate for PLA2 catalysis; ii) released radioactivity was shown to be mainly 3 H-AA (79%); iii) release of 3 H-AA was inhibited by the PLA2 inhibitors quinacrine, aristolochic acid, and PTK. The non CYP2E1-expressing HepG2 (C34) cells exposed to Fe+AA showed lower activation of PLA2, indicating that the higher activity of PLA2 in E47 cells is a CYP2E1-dependent effect. A possible lower content of PLA2 capable of being activated by Ca 2+ is not a cause for the low activation of PLA2 in C34 cells, as a calcium ionophore activated PLA2 to the same extent in both HepG2 cell lines.
Increased AA release in AA+Fe-treated cells expressing CYP2E1 is likely a reflection of increased PLA2 activity, and not simply an inhibition of AA reacylation due to ATP depletion because: i) cPLA2 activity in cell lysates of Fe+AA-treated cells, measured using a synthetic substrate in a reaction system not dependent on ATP, significantly increased, and ii) CYP2E1-expresing HepG2 cells exposed to AA+Fe in a medium lacking extracellular Ca 2+ , showed increased AA release (Fig 4), although ATP levels did not drop under these conditions (49).
PLA2 activity and cytotoxicity appear to be causally related because inhibiting PLA2 activity with quinacrine, aristolochic acid and PTK significantly blocks toxicity, and PLA2 activation and LDH release are simultaneous early events. Interestingly, in contrast to calpain, PLA2 activation is an early event that is linked to cytotoxicity in cells exposed to Fe-NTA in a buffer lacking Ca 2+ , since significant release of 3 H-AA occurred in Ca 2+ -deficient medium, and the above PLA2 inhibitors also blocked the AA+Fe-induced toxicity. Thus, PLA2 plays a role in the AA+Fe early toxicity which occurs in Ca 2+ -containing medium, and in the later toxicity which occurs in Ca 2+ -deficient medium. Low activation of PLA2 in C34 cells treated with Fe+AA in buffer containing 1 mM Ca 2+ was associated with lower toxicity. To our knowledge, this is the first report of a critical role for PLA2 in a CYP2E1-dependent model of toxicity.
The specific isoform involved is probably cPLA2, since selective inhibitors of sPLA2 (BPB) and iPLA2 (BEL) were ineffective in blocking toxicity and PLA2 activity, while a selective inhibitor of i plus cPLA2  Fig 4), suggesting an oxidant-dependent process was involved in the PLA2 activation.
Control experiments indicated that the PLA2 inhibitors did not show an antioxidant activity under the condition utilized, and the PLA2 inhibitors did not block the generation of TBARS triggered by AA+Fe-NTA (Table 2).
A previous study showed that inhibition of calpain by calpeptin did not block the AA+Fe-induced influx of Ca 2+ from the extracellular medium (49). However, the PLA2 inhibitors did block this increase in intracellular calcium (Fig 5). PLA2-mediated increase of intracellular calcium may represent a critical step in the mechanism of the early toxicity phase. A recent report shows that the Ca 2+ influx in PC12 cells exposed to polychlorinated biphenyls was mediated through PLA2 activation (63). The mechanism for this PLA2-mediated increase in intracellular calcium may involve a direct effect on the plasma membrane by depletion of phospholipids or by free fatty acids and lysophospholipids that can increase membrane permeability to Ca 2+ (64), or an indirect effect mediated by store operated channel activation caused by P450 metabolites of free AA. Such activation of Ca 2+ channels by coupled cPLA2 plus P450 activity has been proposed as a possible mechanism for capacitative calcium entry (65). The fact that SKF 96365 (a store operated channel inhibitor) did not inhibit toxicity of Fe+AA in E47 cells (34) may indicate that direct effects on the plasma membrane are more important.
While PLA2 (Fig 4) and lipid peroxidation (34) are activated to similar or slightly different extents in AAloaded cells exposed to Fe-NTA in 0 or 1 mM Ca 2+ medium, toxicity was much lower in the buffer lacking Ca 2+ , and was delayed to longer incubation times e.g. 12h instead of 3h. Increased PLA2 activity and lipid peroxidation in E47 cells treated with Fe-NTA in Ca 2+ -deficient medium is not sufficient to produce rapid toxicity probably because of the lack of Ca 2+ incorporation from the extracellular medium.
The results obtained suggest that lipid peroxidation of cellular membranes is not sufficient per se for rapid toxicity, but that Ca 2+ incorporation from the extracellular medium is, and peroxidation-induced PLA2 activity (but not lipid peroxidation alone) is a key mediator of this increase.
Toxicity of AA + Fe in SMEM develops at longer time periods (i.e. 12h). Toxicity is still blocked by αtocopherol, but it is not associated with measurable increase in intracellular Ca 2+ (49). PLA2 activity (Fig   4) and the later toxicity (Fig 8) in AA-loaded E47 cells exposed to Fe-NTA in a medium without Ca 2+ were blocked by PLA2 inhibitors. This suggests that apart from increasing influx of Ca 2+ , PLA2 activity is acting on other cellular targets that result in toxicity without requiring influx of extracellular Ca 2+ . Possible targets are lysosomes and mitochondria. Destabilization of the lysosomal membrane and leakage of lysosomal contents were associated with activation of PLA2 in H 2 O 2 -treated J774 cells (66).
Phospholipids are important constituents of the mitochondrial membrane and are essential for activity of certain mitochondrial enzymes (64). Ischemia-reperfusion induced a PLA2-dependent mitochondrial damage in renal mitochondria (67). Structural alteration of the inner mitochondrial membrane was associated with activation of PLA2 (68). A decrease in the mitochondrial membrane potential by AA+Fe was shown to occur in both Ca 2+ -containing and Ca 2+ -deficient medium, and the decline preceded the early and late toxicity phases induced by AA+Fe (49).
Calpain activation by Fe+AA in E47 cells is Ca 2+ -dependent because it was completely blocked in Ca 2+depleted cells incubated in 0 mM Ca 2+ (Table 4). AA-loaded cells exposed to Fe-NTA in 1 or 0 mM CaCl 2 exhibited increases of calpain activity ( partially blocked the toxicity and the decline in mitochondrial membrane potential (49). However, calpain, unlike PLA2, did not appear to play a role in the late toxicity phase in Ca 2+ -deficient medium since calpeptin had no effect on the toxicity or the decline in membrane potential (49).