A comparison of N-methyl-D-aspartate-evoked release of adenosine and [3H]norepinephrine from rat cortical slices.

Tetrodotoxin reduced N-methyl-D-aspartate (NMDA)-evoked release of adenosine by 35% but virtually abolished [3H]norepinephrine release. Although [3H]norepinephrine release from rat cortical slices evoked by 500 microM NMDA was abolished by 1.2 mM Mg++, which produces a voltage-sensitive, uncompetitive block of NMDA-channels, adenosine release was increased in the presence of Mg++. Partial depolarization with 12 mM K+ relieved the Mg++ block of 500 microM NMDA-evoked [3H]norepinephrine release but did not affect adenosine release, indicating that a Mg++ requirement for the adenosine release process per se cannot account for this discrepancy. NMDA was 33 times more potent in releasing adenosine than [3H]norepinephrine. At submaximal concentrations of NMDA (10 and 20 microM), adenosine release was augmented in Mg+(+)-free medium. Although a high concentration of the uncompetitive NMDA antagonist MK-801 [(+)-5-methyl-10,11,dihydro-5H-dibenzo[a,d]cyclohepten-5-10-imine maleate] (3 microM) blocked NMDA-evoked release of [3H]norepinephrine and adenosine, a lower concentration (300 nM) decreased NMDA-evoked [3H]norepinephrine release by 66% without affecting adenosine release. These findings suggest that maximal adenosine release occurs when relatively few NMDA receptors are activated, raising the possibility that spare receptors exist for NMDA-evoked adenosine release. Rather than acting as a protectant against excessive NMDA excitation, released adenosine might provide an inhibitory threshold which must be overcome for NMDA-mediated neurotransmission to proceed.