Differential desensitization of ionotropic non-NMDA receptors having distinct neuronal location and function

The release of tritium from rat hippocampal synaptosomes prelabeled with [3H]noradrenaline ([3H]NA) or [3H]5-hydroxytryptamine ([3H]5-HT) and from rat neocortex synaptosomes prelabeled with [3H]choline and the release of endogenous GABA and glutamate from rat neocortex synaptosomes were monitored during superfusion with media containing varying concentrations of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) or kainic acid. Concentration-dependent release potentiations were elicited by both excitatory amino acids (EAAs) in all the transmitter systems investigated. The releases evoked by 100 microM AMPA were, in all cases, almost totally dependent on external Ca2+ and sensitive to 6.7-dinitroquinoxaline-2,3-dione (DNQX), indicating involvement of non-NMDA receptors. When cyclothiazide, a drug able to prevent desensitization of AMPA-preferring receptors, was added to the superfusion medium (at 1 or 10 microM) concomitantly with 100 microM AMPA or kainate, the EAA-evoked release of [3H]NA was significantly enhanced. Concanavalin A, a lectin thought to prevent desensitization of kainate-preferring receptors, had no effect (up to 10 microM) on the release of [3H]NA evoked by AMPA or kainate. The effect of cyclothiazide was lost if, after an 8-min pretreatment, the drug was removed just before the AMPA stimulus. When added concomitantly with the EAAs, cyclothiazide potentiated the release of [3H]5-HT elicited by AMPA and, less so, that evoked by kainate. Concanavalin A was ineffective. Neither cyclothiazide (1 or 10 microM) nor concanavalin A (3 or 10 microM) could affect the release of [3H]ACh or endogenous GABA provoked by 100 microM AMPA or kainate, suggesting that the receptors involved do not desensitize. Exposure of neocortex synaptosomes to AMPA or kainate concomitantly with cyclothiazide caused endogenous glutamate release that did not differ from that evoked by the EAAs alone. In contrast, the effects of AMPA and kainate were potentiated by concanavalin A. The activity of the lectin (3 microM) persisted when it was applied for 8 min and then removed before the AMPA or kainate (100 microM) pulse. When hippocampal synaptosomes prelabeled with [3H]NA were subjected to three subsequent AMPA (100 microM) stimuli (S1, S2 and S3), the release of [3H]NA decreased dramatically from S1 to S3 (S3/S1 = 0.14 +/- 0.04); a significant 'protection' of the AMPA effect was offered by 1 microM cyclothiazide (S3/S1 = 0.36 +/- 0.06). This value did not differ from the S3/S1 ratio (0.38 +/- 0.04) obtained in parallel experiments with 12 mM K+. The release evoked by high-K+ was insensitive to cyclothiazide. Finally, the effect of AMPA on the release of [3H]ACh did not respond to cyclothiazide also during three subsequent stimuli with 100 microM AMPA. To conclude: a) ionotropic non-NMDA receptors mediating enhancement of NA, 5-HT, ACh, GABA and glutamate release exist on the corresponding nerve terminals; b) the receptors present on noradrenergic and serotonergic neurons are AMPA-preferring receptors, whereas the glutamate autoreceptors resemble most the kainate-preferring subtype; the receptors mediating ACh and GABA release can not be subclassified at present; c) desensitization may not be a property of all non-NMDA ionotropic receptors. The receptors here characterized represent five models of native non-NMDA receptors suitable for pharmacological and molecular studies.