Associative stimulation of N-methyl-D-aspartate (NMDA) receptors and quisqualate ionotropic receptors (Qi) induces long-term potentiation at particular glutamatergic synapses. Release of arachidonic acid as a result of stimulation of NMDA receptors has been proposed to play a part in the establishment of long-term potentiation. But long-term plasticity events at some other glutamatergic synapses do not involve activation of NMDA receptors. Here we report that in mature striatal neurons in primary cultures, quisqualate can release arachidonic acid by associatively activating both quisqualate metabotropic receptors coupled to phospholipase C (Qp) and Qi receptors. Independent activation of these two receptor types with specific agonists did not stimulate arachidonic acid release. These results support a role for the associative activation of Qp and Qi receptors in synaptic plasticity events, including long-term potentiation at particular synapses.

The release of arachidonic acid (ArA) metabolites from mouse neurons and astrocytes in primary culture has been studied in response to ionomycin or glutamate stimulation. Cells were preincubated with [3H]ArA for 24 h and the radioactivity released was examined by HPLC. In striatal, cortical and hippocampal neurons, glutamate and ionomycin strongly stimulated the release of ArA, but neither prostaglandins (PGs) nor hydroxyeicosatetraenoic acids (HETEs) could be detected. If they were released, these latter compounds represented < 0.02% of the amount of ArA. In contrast, in astrocyte cultures, ionomycin (but not glutamate) strongly stimulated the release of PGs and HETEs as well as ArA. Reversed- and straight-phase HPLC analysis revealed the presence of PGD2, PGE2, PGF2alpha, 12-hydroxyheptadeca-5,8,10-trienoic acid (HHT) and HETEs (15-HETE, 11-HETE and 5-HETE). Indomethacin inhibited the release of PGs and HHT, but also that of 11- and 15-HETE, indicating that these two HETEs may be produced through the cyclooxygenase pathway. Metabolism of [3H]ArA was also examined in cellular homogenates. Although > 50% of the [3H]ArA was metabolized to PGF2alpha, PGE2, PGD2, HHT, 15- and 11-HETE in cultured astrocyte homogenates, no [3H]ArA metabolism could be detected in cultured striatal neuron homogenates. Moreover, neuronal homogenates did not inhibit the metabolism of [3H]ArA observed in either astrocyte or platelet homogenates. These results indicate that central neurons in primary culture possess very low lipoxygenase and cyclooxygenase activities. They emphasize the need to identify the cellular source of ArA metabolites in the brain, particularly when considering the multiple new messenger roles proposed for these molecules, such as that of retrograde messengers involved in synaptic plasticity phenomena.