The actions of adenosine, adenosine deaminase, the adenosine uptake blocker, S-(p-nitrobenzyl)-6-thioinosine, and of the adenosine deaminase inhibitor, erythro-9(2-hydroxy-3-nonyl)adenine, on electrically evoked [3H]acetylcholine release were investigated in rat phrenic nerve-hemidiaphragm preparations. Adenosine deaminase (0.25-2.5 U/ml) increased [3H]acetylcholine release. S-(p-Nitrobenzyl)-6-thioinosine (3-30 microM) and erythro-9(2-hydroxy-3-nonyl)adenine (25 nM-50 microM) caused biphasic effects on [3H]acetylcholine release: at low concentrations S-(p-nitrobenzyl)-6-thiomosine (5 microM) and erythro-9(2-hydroxy-3-nonyl)adeNine (50 nM) decreased [3H]acetylcholine release, and at concentrations higher than 10 microM S-(p-nitrobenzyl)-6-thioinosine and 0.5 microM for erythro-9(2-hydroxy-3-nonyl)adenine facilitated [3H]acetylcholine release. Both S-(p-nitrobenzyl)-6-thioinosine-induced inhibition and facilitation of [3H]acetylcholine release resulted from extracellular endogenous adenosine accumulation, because they were blocked after inactivation of endogenous adenosine with adenosine deaminase (0.5 U/ml). The inhibitory actions of both S-(p-nitrobenzyl)-6-thioinosine (5 microM) and erythro-9(2-hydroxy-3-nonyl)adenine (50 nM) were antagonized by the A1 adenosine receptor antagonist, 1,3-dipropyl-8-cyclopentylxanthine (2.5 nM), whereas the blockade of A2a adenosine receptors with PD 115,199 (25 nM) prevented the facilitatory effects of S-(p-nitrobenzyl)-6-thioinosine (30 microM) and erythro-9(2-hydroxy-3-nonyl)adenine (50 microM). The adenosine deaminase inhibitor, erythro-9(2-hydroxy-3-nonyl)adenine (25 nM), potentiated the effect of S-(p-nitrobenzyl)-6-thioinosine (3-30 microM), and this adenosine uptake blocker, when applied at a concentration (3 microM) that by itself was devoid of effect, potentiated both the inhibitory (25 nM) and excitatory (0.5 microM) effects of erythro-9(2-hydroxy-3-nonyl)adenine, on evoked [3H]acetylcholine release. Exogenously applied adenosine (10-500 microM) had biphasic effects similar to those of S-(p-nitrobenzyl)-6-thioinosine and erythro-9(2-hydroxy-3-nonyl)adenine. Adenosine (30 microM) reduction of evoked [3H]acetylcholine release was prevented after pretreatment with 1,3-dipropyl-8-cyclopentylxanthine (2.5 nM); when applied at high concentrations (100-500 microM), adenosine consistently increased evoked [3H]acetylcholine release in a PD 115,199 (25 nM)-sensitive manner. It is concluded that both uptake and deamination are effective in removing extracellular endogenous adenosine that tonically activates both inhibitory (A1) and excitatory (A2a) adenosine receptors, regulating the A1/A2a adenosine receptors' activation balance.