(1) To investigate the interaction between upstream and downstream enzymes involved in brain prostaglandin (PG) synthesis, we examined the expression and activity of cyclooxygenase (COX)-1, of different phospholipase A2 (PLA2) enzymes, and of prostaglandin E2 synthase (PGES) enzymes in COX-2 deficient mice. The PGE2 level was decreased by 52% in the COX-2 knockout mice brain, indicating a significant role of COX-2 in formation of PGE2. However, when we added exogenous arachidonic acid (AA) to brain homogenates, COX activity was increased in the COX-2 deficient mice, suggesting a compensatory increase in COX-1 expression and an intracellular compartmentalization of the COX isozymes. Activity and expression of cytosolic cPLA2 and secretory sPLA2 enzymes, supplying AA to COX, were significantly increased. Our results indicate that compensatory mechanisms exist in COX-2 deficient mice and that microsomal PGES-2 is functionally coupled with COX-2. Thus, this pathway might represent a novel target for anti-inflammatory and neuroprotective drugs. Then, we further examined how COX-2 deficiency was affecting the NF-kappaB pathway, which controls COX-2 expression. We found a decrease in NF-kappaB DNA-protein binding activity, which was accompanied by a reduction of the phosphorylation state of both I-kappaBalpha and p65 proteins in the COX-2 deficient mice. The mRNA and protein levels of p65 were also reduced in COX-2 deficient mice, whereas total cytoplasmic I-kappaB protein level was not significantly changed. Taken together, these changes may be responsible for the observed decrease in NF-kappaB DNA binding activity. NF-kappaB DNA binding activity was selectively affected in the COX-2 deficient mice compared to the wild type as there was no significant change in NFATc DNA binding activity. Overall, our data indicate that constitutive brain NF-kappaB activity is decreased in COX-2 deficient mice and suggest a reciprocal coupling between NF-kappaB and COX-2. (2) To determine the specific role of COX-1 in brain AA cascade, we examined the expression and activity of COX-2 as well as the expression and activity of the PLA2 enzymes (group IV cPLA2 and group V sPLA2), which generate AA, and of known downstream PGES isoforms, which generate biologically active PGE2, in brain from COX-1 deficient mice. The expression and activity of brain cPLA2 and sPLA2 were significantly increased in COX-1-deficient mice compared to the wild-type. We also find a compensatory increase in COX-2 expression, accompanied by an up-regulation of the NF-kappaB pathway. Downstream enzymes also were affected, as the protein levels of mPGES-1 and -2, but not cPGES, were decreased in COX-1-deficient mice. Brain PGE2 level was increased and thromboxane B2 level was decreased in COX-1-deficient mice, suggesting that these end-products are specifically derived from COX-1 and COX-2 metabolism of AA, respectively. Taken together, these data are consistent with our hypothesis that the COX-1 deficiency results in the altered expression of the remaining enzyme that regulate mobilization and conversion of AA to prostaglandins. (3) COX-2 is expressed under basal conditions in areas of the brain susceptible to excitotoxicity, a form of toxicity that occurs from over activation of excitatory neurotransmitter systems such as glutamate. While many studies have attempted to determine the role of COX-2 in excitotoxicity through pharmacological inhibition of the enzyme, the results are controversial. We attempted to further study the role of COX in excitotoxicity by testing the susceptibility of mice deficient in either COX-1 or COX-2 to kainic acid (KA) excitotoxicity. Mice deficient in either COX-1 or COX-2 and their wild-types were injected intraperitoneally with saline or 10 mg/kg KA, a sublethal dose which induced seizure activity, and video recorded for 2 hours after the injection. Median Racine seizure score (RSS) was significantly elevated in KA-exposed COX-2 deficient mice compared to wild type and heterozygous mice. COX-1 deficient mice did not differ from wild type mice in the magnitude of KA-induced seizures. Only COX-2 deficient mice exhibited neurons positive for Fluoro-Jade B (FJB), a histochemical stain that detects neuronal degeneration, 24 hours after KA injection. FJB positive neurons were detected in areas known to be affected by KA such as the CA1 and CA3 regions of the hippocampus, amygdala and thalamus. In summary, COX-2 deficient, but not COX-1 deficient, mice exhibit an increased sensitivity to KA-induced seizure activity and are more susceptible than wild type mice to excitotoxic neuronal damage, suggesting that COX-2 may be protective against excitotoxicity.
