Alzheimer's disease (AD) is associated with gliosis, neuroinflammation and higher levels of prostaglandins. Conflicting roles for cyclooxygenases and prostaglandins in the etiopathology of AD have been reported. We hypothesized that PGE2 signaling through EP2 and EP4 G-protein-coupled receptors could protect against amyloid beta-peptide (Abeta) neurotoxicity by increasing the cAMP signaling cascade. Using primary neuronal cultures, we investigated the presence of EP receptors (EP1-4) and the action of PGE2 and EP receptor agonists on neuronal susceptibility to Abeta1-42 toxicity. Low concentrations (1 microm) of PGE2, butaprost (EP2 agonist), and 1-hydroxy-PGE1 (EP4/EP3 agonist) were neuroprotective against Abeta1-42 toxicity, while sulprostone (EP3/EP1 agonist) at similar doses had no detectable effects. EP2 and EP4 receptor-mediated neuroprotection would involve changes in cAMP levels, as both EP2 and EP4 agonists increased intracellular cAMP concentration by approximately doubling basal levels, and both exhibited neuroprotective actions against Abeta-induced toxicity. The protein kinase A (PKA) inhibitor RpcAMPS significantly attenuated the neuroprotection by butaprost, but not that by 1-hydroxy-PGE1, implying differences between EP2 and EP4 receptor protective mechanisms. Additionally, the increase in reactive oxygen species generated following exposure to Abeta was reduced by stimulation of both EP2 and EP4 receptors. Together, these results indicate that PGE2 can protect neurons against Abeta toxicity by acting on given receptors and stimulating a cascade of intracellular events, including the cAMP-PKA pathway. We propose that development and testing of specific PGE2 receptor agonists downstream of cyclooxygenase could lead to therapeutic applications.