Nitric oxide is produced in high amounts in asthmatic lungs and has an important role as a regulator of lung inflammation. In this study, we found that NO prevents the survival-prolonging effect of GM-CSF in eosinophils by inducing apoptosis consistently with our previous reports
[11, 12]. We focused here to the cascade of events leading to NO-induced eosinophil apoptosis particularly concentrating on the role of mitochondria, ROS and JNK. We showed that NO-induced eosinophil apoptosis is dependent on early ROS production, JNK and late mPT. In addition, we found that NO induced an early partial mPT and mPT-dependent JNK activation but those events seemed not mediate NO-induced apoptosis detected at later time points.
Mitochondrial permeability transition pore is a channel in the inner mitochondrial membrane that is composed of several proteins in a complex manner. The molecular structure of the channel has not yet been resolved despite of numerous attempts and it has been postulated that the structure may vary depending on the cell type and/or the trigger. Inhibitor of mPT, bongkrekic acid, acts as a ligand to adenine nucleotide translocator (ANT), which is either a component or an important regulator of mPT
[28, 29]. ANT has been shown to act as a critical target in mitochondrial membrane permeabilization induced by NO and peroxynitrite
, giving ground for usage of bongkrekic acid as an inhibitor of mPT in this study. Here, we also showed that bongkrekic acid inhibits mPT in eosinophils.
Our results demonstrate that NO has a marked and varied effect on mPT at long time-scale. We found that NO induces partial mPT at 1 h in eosinophils that does not lead to early permanent loss of ΔΨm. This strongly suggests that early NO-induced mPT is transient or flickering. Previously, flickering mPT has been demonstrated in healthy intact cells
[21, 31] and it has been postulated to act as a fast calcium release mechanism
[32, 33] participating thereby in diverse range of Ca2+ -mediated cellular activities. Flickering mPT may also be involved in cell protection during minor stress
 or it may act as an early signal for oxidative stress-induced apoptosis
. In accordance with our results, NO was shown to induce and modulate mPT in a reversible manner in isolated mitochondria
. We found that the early mPT is not significant for initiation of NO-induced apoptosis because addition of mPT inhibitor bongkrekic acid at 16 h after SNAP and GM-CSF still efficiently prevented NO-induced apoptosis. This seems to be in contrast to the findings of Ma et al.
 who showed that early flickering mPT and associated superoxide flashes induced by selenite are early signals of apoptosis. They showed that manipulation of selenite-induced flickering mPT and associated superoxide flashes by knockdown or overexpression of mPT component cyclophilin D decreased or increased selenite-induced apoptosis, respectively
. However, this manipulation also affects the irreversible mPT and there is no conclusive evidence that flickering mPT and the associated superoxide flashes are necessary for apoptosis. Nevertheless, because treatment with NO ends up in mPT-mediated apoptosis it is likely that flickering mPT is an important point where anti-apoptotic and pro-apoptotic signals converge and fate of the cell is determined. Most probably, if a certain threshold is achieved flickering mPT is turned into permanent mPT and cell undergoes apoptosis or necrosis.
We showed that the early partial mPT led to strong activation of JNK at 2 h. A smaller activation of JNK was observed at later time-points (20–30 h). Consistently, a previous study showed that apoptosis-inducing N-methyl-4-phenylpyridinium (MPP+) induced early and late phases of JNK activation in a mammalian cell line and the early phase was preventable by bongkrekic acid
. They also found that the late JNK activation was independent of mPT, which remains unsolved in our study. However, in contrast to our study, Casarino et al. did not show whether the early mPT-dependent JNK activation or late phase of JNK activation are relevant for MPP+ −induced apoptosis. To our knowledge, this is the first study to demonstrate mPT-mediated JNK activation not related to apoptosis. According to our study, one physiological function of flickering mPT may, therefore, be initiation of JNK signalling in response to oxidative stress probably aiming to cell rescue. Plenty of evidence from studies conducted by others supports the conclusion that the early mPT and the following JNK activation are a protective response. First, several groups have demonstrated that rapid, strong and transient JNK activation is a stress response resulting in cell survival signalling while delayed and sustained JNK activation is related to apoptosis
[23, 24]. Second, flickering mPT was shown to be involved in cell protection during minor stress
. Additionally, Beltran et al. have demonstrated that long exposure to NO by DETA-NONOate initially stimulates a protective response by inhibiting complex IV in the mitochondrial respiratory chain. This led to maintenance of ΔΨm by hydrolysis of glycolytic ATP instead of the respiratory chain and increased cell viability. The protective response induced by NO turned into apoptotic response by an unknown mechanism that was speculated to involve accumulation of oxidative damage
. In eosinophils ΔΨm is maintained exceptionally by hydrolysis of glycolytic ATP rather than respiratory chain in contrast to most eukaryotic cells
 indicating that this mechanism is already functional in eosinophils. Evidence of the survival-increasing potential of NO in eosinophils has been shown by several groups
[39–41]. Hebestreit and co-workers showed that NO stimulates eosinophil survival in the presence of apoptosis-inducing Fas at 24 h time-point
. Further time-points were not studied to see whether the survival signalling would have turned into apoptotic signalling. In our experiments GM-CSF produced maximal survival of eosinophils at 10 pM concentration making it impossible to show that NO activates survival machinery in eosinophils at early time-points. This evidence supports the conclusion that NO-induced early mPT and the following JNK activation are a protective response of the cell.
ROS/RNS are known inducers of JNK in eosinophils
, which suggests that ROS formation may be involved in the early mPT-dependent activation of JNK. Studies by Zorov et al.
 have shown a relationship between ROS and mPT which raises an interesting possibility for the mechanism of mPT-induced JNK activation. They showed that mitochondrial ROS accumulation leads to mPT, which was followed by mitochondrial ROS burst that can be released to the cytosol
. Also Ma et al. demonstrated mPT-dependent superoxide flashes in response to oxidative stress
. However, whether this mechanism explains mPT-mediated early JNK activation in NO-treated eosinophils remains to be clarified.
Stimulation of GM-CSF-treated eosinophils with NO resulted in permanent loss of ΔΨm and apoptosis at 40 h. Lost ΔΨm is often, but not always, a consequence of permanent mPT
. Prevention of NO-mediated apoptosis by late addition of bongkrekic acid, however, gives further evidence that the threshold for permanent mPT is crossed at time-point beyond 16 h. Previously, NO was shown to induce permanent mPT in isolated mitochondria and thymocytes
. In eosinophils, mPT mediated dexamethasone-induced apoptosis
JNK has been previously reported to mediate spontaneous apoptosis and apoptosis induced by several drugs and nitric oxide in eosinophils
[11, 42, 45]. In concordance with our previous results, we found here that NO activates JNK and JNK has a role in NO-induced apoptosis
. Similarly to the results with JNK peptide inhibitor 1 (L-JNKI1) in the previous study
, two additions of JNK inhibitor SP600125 (at 30 min before and 16 h after SNAP) was required to suppress the pro-apoptotic effect of NO. Either addition alone had no statistically significant effect on SNAP-induced apoptosis. The result implies that the later phase of JNK activation is required for apoptosis but the later phase may initiate before 16 h. This is possible because we did not study pJNK levels in time-points between 2 h and 20 h. With the limited amounts of cells available for these studies and the sensitivity of the current assays, it was not possible to determine the exact time, when the later phase of JNK activation was initiated. The result also suggests that the initial 10 μM concentration of SP600125 was not adequate to inhibit JNK during long time-scale of activation in contrast to 1 μM concentration of JNK inhibitor VIII that slightly but statistically significantly suppressed the pro-apoptotic effect of SNAP. The mechanism by which JNK participates in NO-induced eosinophil apoptosis remains unclear. In previous studies with other cell types and stimulants JNK has promoted apoptosis by inducing mPT
[25, 26]. However, this seems not to be the mechanism in NO-induced apoptosis in eosinophils since inhibition of JNK had no effect on the SNAP-induced loss of ΔΨm. It is also possible that JNK has a pro-apoptotic mechanism that is independent on mPT. For example, JNK-mediated transcription has been shown to enhance expression of Fas-ligand
 indicating that JNK activation may stimulate extrinsic Fas pathway of apoptosis in parallel to intrinsic mitochondrial pathway. Alternatively, JNK-dependent mechanisms may mediate the pathway from mPT to DNA fragmentation.
Under normal physiological condition, few per cent of the oxygen consumed by mitochondria is converted to superoxide. Cellular stress often leads to further increase in superoxide production. Inhibition of the components of the mitochondrial respiratory chain has been demonstrated as one mechanism by which NO increases superoxide production
. Reaction between nitric oxide and superoxide leads to formation of peroxynitrite. By using SOD mimetic AEOL 10150, we showed that the pro-apoptotic effect of SNAP on eosinophils in the presence of GM-CSF is partly dependent on superoxide and/or peroxynitrite production. However, addition of SOD mimetic at 16 h time-point was no longer effective in reversing the pro-apoptotic effect of SNAP suggesting that early formation of superoxide and/or peroxynitrite is critical for SNAP-induced apoptosis. Early NO-induced ROS-production is in concordance with the results of other studies
[48, 49]. Previously in other cell types the peak of ROS production by NO has been demonstrated to occur before caspase activation and the following apoptosis
[49, 50]. However, the importance of this early ROS peak for apoptosis has been unclear. Our study shows that early increase of ROS is a critical event mediating NO-induced eosinophil apoptosis. NO-induced apoptosis seems to be initiated relatively late suggesting importance of accumulation of oxidative damage. Interestingly, because both apoptosis-related formation of ROS and activation of JNK seem to initiate before 16 h and ROS/RNS are known inducers of JNK
, ROS may also participate in activating the later apoptosis-related JNK. Inhibition of apoptosis by AEOL was only partial suggesting that NO-induced formation of ROS is not the only key event for initiation of apoptosis. NO may also have direct apoptosis-stimulating effects on eosinophils. Another possibility that may explain the result is inefficiency of the used SOD-mimetic in dismutating all superoxide and peroxynitrite production. According to our results with inhibitors of NADPH oxidase, this enzyme is not the major source of superoxide in NO-treated eosinophils. This leaves mitochondrial electron transport chain as the most likely source of superoxide.
Nitric oxide is abundant in the lungs of asthmatics and thereby most likely affects eosinophil functions in a physiological situation. In this study, in addition to the mechanism of NO-induced apoptosis, we showed some interesting early events in NO-stimulated eosinophils that may take place even if the threshold for irreversible mPT and apoptosis is not crossed. In fact, levels of NO in the exhaled air has been shown to correlate to eosinophilic inflammation
[8, 9] which makes it tempting to speculate that NO might only induce flickering mPT and JNK activation ending up in a protective response but not apoptosis in a physiological situation.