The class of nuclear hormone receptors known as peroxisome proliferator-activated receptors (PPARs) have three isoforms
PPARα, PPARβ and PPARγ and are expressed in a wide variety of tissues. PPARs act as transcription factors, regulating expression of target genes upon binding of a suitable ligand and heterodimerization with the retinoid X receptor. A diverse range of compounds activate PPARs. PPARα ligands include endogenous fatty acids and the fibrate class of drugs. The insulin sensitizing agents known as thiazolidinediones activate PPARγ as do some prostaglandins including the purported endogenous PPARγ ligand, 15-deoxy
Δ12,14-prostaglandin J2. An endogenous ligand for PPARβ has not been discovered, however this isoform is activated non-specifically by some fatty acids and prostaglandin I2 analogues. Physiological roles of PPARs include control of lipid homeostasis, cellular differentiation, inflammation and cell cycle control. Apoptosis in neurons could be due to inappropriate activation of cellular machinery normally used by proliferative cell types to progress through the cell cycle. Therefore, given that PPARs have been implicated with cell cycle control mechanisms and these same mechanisms are utilized in neuronal cell apoptosis; it was hypothesized that activation of PPARs in neurons would affect cellular viability. Rat cerebellar granule
neurons (CGNs) cultured in neurobasal medium containing a depolarizing (25 mM) concentration of KCl were used throughout our toxicity studies because this model allows the rapid generation of homogenous cultures containing almost exclusively neuronal cells. Furthermore, such cultured neurons undergo apoptosis when switched to media containing a physiological KCl concentration, allowing assessment of potential neuroprotection. The presence of mRNA from the three PPAR isoforms within our cultured neurons was confirmed using RT-PCR and in situ hybridization. Cultured CGNs were exposed to ligands specific for each PPAR isoform. Wy-14,643 was used to activate PPARα, GW0742 activated PPARβ while either ciglitazone or 15-deoxy-Δ12,14- prostaglandin J2 (15-deoxy PGJ2) were used as PPARγ agonists. Lactate dehydrogenase release (LDH assay) or the ability to bioreduce a tetrazolium salt (MTS assay) were used as measures of toxicity. Wy-14,643 was not inherently toxic while GW0742 displayed inherent toxicity after 48 hours but not after a 24 hour exposure. TUNEL staining showed that GW0742-mediated neuronal death was apoptotic. Both ciglitazone and 15-deoxy PGJ2 were inherently toxic with 15-deoxy PGJ2 exerting toxic effects after shorter exposure times and at lower concentrations than ciglitazone. TUNEL staining showed that the cell death caused by either PPARγ ligand was apoptotic. To see
if PPAR activation provides neuroprotection, CGN cultures were exposed to PPAR agonists at the same time as apoptosis was initiated by switching to media containing a reduced KCl concentration. Wy-14,643, despite exhibiting no inherent toxicity, significantly augmented lowered-KCl induced apoptosis. Treatment with GW0742 provided some transient neuroprotection. The presence of ciglitazone also provided some timedependent neuroprotection, however 15-deoxy PGJ2 significantly enhanced cell death initiated by KCl withdrawal. To determine the effect of PPARα or PPARβ activation on apoptosis initiated by a different mechanism, Wy-14,643 and GW0742 were added to cultures in which apoptosis was initiated by treatment with either cytosine arabinoside (Ara-c) or sodium nitroprusside (SNP). Both Wy-14,643 and GW0742
augmented Ara-c or SNP induced cell death. Immunoblotting experiments were performed to determine the role of the immediate early gene, c-Jun following exposure to GW0742, ciglitazone or 15-deoxy PGJ2. Increased c-Jun protein expression was shown 6 hours after treatment with GW0742 whereas no increases were seen after 12 or 48 hours. Ciglitazone had little effect on c-Jun expression either in the presence of 25 mM KCl or following KCl withdrawal. Expression of c-Jun was markedly increased 12 and 24 hours after addition of 15-deoxy PGJ2 in the presence of 25 mM KCl. Significant inhibition of c-Jun expression was provided by 15-deoxy PGJ2 3 and 6 hours after KCI withdrawal compared to control cultures subjected to KCI withdrawal alone. This initial inhibition of c-Jun expression was followed by a promotion of c-Jun expression after 24 hours. The results obtained show that the effect of PPAR
activation in neurons is isoform specific and are also influenced by ligand concentration and exposure time. Given that PPAR agonists have been shown to have both neurotoxic and neuroprotective properties in our in vitro study, further investigations into the potential role of PPARs in mammalian neurodegeneration, neuronal function and development are warranted.