The Effects of COX and nNOS Inhibitors in the Hypoxia-Ischaemia Model of Brain Injury: Roles of Arachidonic Acid and L-Arginine Metabolism

Abstract

Ischaemic brain injury and neurodegenerative pathologies, including perinatal hypoxia-ischaemia (HI), involve a number of inflammatory and excitotoxic pathways associated with, and contributing to, apoptotic and necrotic cell death. A number of inflammatory enzymes are known to be involved including: cyclooxygenase (COX) metabolism of arachidonic acid and nitric oxide synthase (NOS) production of the free radical nitric oxide. Consistent with published data, we found, both COX-2 and inducible (i)NOS immunoreactive cells, located primarily within the peri-infarct regions following HI, and identified these as non-neuronal in origin due to the lack of co-localization with the neuronal marker NeuN. This suggests that modulation of COX, and in particular COX-2, or NOS could have beneficial effects after HI.

The non-selective COX inhibitor, aspirin, and the COX-2 selective antagonist, nimesulide, have previously been shown to afford significant protection in animal models of stroke. However, neither compound has been assessed using an HI model of brain injury, nor is there clear evidence as to the underlying protective mechanisms of these drugs. We initially hypothesised that both aspirin and nimesulide would provide significant protection against HI-induced neuronal injury. We also hypothesised that nimesulide would exert greater protection due to having a higher affinity for COX-2. Prophylactic treatment with aspirin (4mg/kg) reduced the infarct volume to 23.5±11.9mm3 (P<0.01) following HI-induced injury compared to saline treated controls (94.3±12.8mm3). However, contrary to our hypothesis, nimesulide treatment (12mg/kg) failed to afford any significant neuroprotection when compared with vehicle-treated Polyvinylpyrrolidone (PVP) controls. These results suggest that inhibition of COX-2 alone is unlikely to play a significant role in protection against cell death following HI.

The neuroprotective effects of ARL 17477, a selective nNOS inhibitor, were also assessed following HI. Previous work by our group had shown ARL 17477 to be neuroprotective and that this protection was mediated, in part, via increased arginase activity (18±3.0μg urea/mg compared to 12±2.0μg urea/mg for HI+saline). Expanding on these findings, we found that ARL 17477 significantly reduced total NOS activity (500±2.0pM L-citrulline/min/mg compared to 760±4.0pM L-citrulline/min/mg for HI+saline); and significantly protected all the mitochondrial complexes against HI-induced damage. These data further confirm the involvement of NOS activity in the regulation of cell death and impairment of mitochondrial enzyme kinetics. This also suggests that the arginase enzyme, which has had very little attention in terms of modulating cell death/neuronal injury, may play a significant role that warrants further investigation.

To determine the underlying neuroprotective properties associated with aspirin treatment compared with treatment with nimesulide, we investigated the associated changes in NOS, arginase, and COX activities. Both aspirin (0.15±0.67pg/min/mg) and nimesulide (0.13±0.04pg/min/mg) treatment groups showed a significant decrease in COX activity compared to their respective vehicle controls (HI+saline 0.53±0.15pg/min/mg; and HI+PVP 0.57±0.22pg/min/mg). Arginase was unaltered by HI or either of the drug treatments. HI+saline resulted in elevated nitrite levels (73.19±4.8μM/mg) compared with non-intervention control (30.9±4.0μM/mg), aspirin treatment partially mitigated the increase in nitrite (59.3±2.6μM/mg). In the ipsilateral hemisphere HI+saline led to increased 15-HETE (3480±511.0pg/mg) compared with non-intervention control (1822±143.8pg/mg) which was ameliorated by aspirin (2406±219.6pg/mg) However, within the contralateral hemisphere aspirin elevated 15-HETE levels (3500±200.7pg/mg) compared with non-intervention controls (2156±194.1pg/mg) while HI+saline did not. There is evidence that 15-HETE and NO mediate part of the anti-inflammatory action associated with aspirin treatment by decreasing inflammatory cell infiltration and inhibiting the production of pro-inflammatory eicosanoids. This provides an additional mechanism of action for the neuroprotection seen with aspirin treatment. Nimesulide treatment led to decreased levels of nitrite (39.1±3.9μM/mg compared with HI+PVP (56.6 ± 5.3μM/mg) and decreased iNOS activity (292.1±89.82pM L-citrulline/min/mg) compared with HI+PVP treatment (593.2±57.51pM L-citrulline/min/mg). Despite these decreases nimesulide was not neuroprotective, suggesting drugs that target multiple and broad targets are more likely to be successful.

We have found that prophylactic treatment with low-dose aspirin is beneficial in the HI model of brain injury and could be clinically useful as a neuroprotective agent. However, contrary to our hypothesis and to previously published data, nimesulide was not found to be neuroprotective. These results suggest that COX-1 inhibition, rather than COX-2, may offer benefit in protection against HI-injury. However, further studies are required to confirm and substantiate these findings. Based on our findings and the clinical failure of so many compounds, it appears that no single target offers significant protection against ischaemic cell death and that combination therapy maybe a better option. There is recent pre-clinical data, showing interfering with N-methyl-D-aspartate (NMDA)/post-synaptic density protein (PSD)-95 signalling is neuroprotective. Disruption to NMDA/PSD95 also alters nNOS signalling, this highlights the importance of further investigations into the exact role that NOS and other inflammatory pathways are playing in cell death.