In PC12 cells, forskolin as well as the adenosine receptor agonist 5'-N-ethylcarboxamidoadenosine (NECA) increased intracellular adenosine-3',5'-cyclic monophosphate (cyclic AMP) levels, which peaked at 45-60 minutes and declined thereafter. Maximum levels were 3000 and 1700 pmol/10(6) cells during treatment with 10 mu M forskolin or 0.1 mu M NECA, respectively. Extracellular cyclic AMP rose with time, at mean rates of 24.7 (forskolin) and 11.3 (NECA) pmol/min/10(6) cells. With either drug, a linear correlation was obtained between the calculated time integral of intracellular cyclic AMP and the measured extracellular cyclic AMP levels, indicating that the outflow of cyclic AMP was sustained by a nonsaturated transport system. The ability of forskolin to increase intracellular and extracellular cyclic AMP levels was hindered in a concentration-dependent manner by 8-(p-sulfophenyl)theophylline (8-SPT). A similar inhibition was exerted by other two adenosine receptor antagonists, 8-cyclopentyl-1,3-dipropylxanthine and 3,7-dimethyl-1-propargylxanthine. The concentration-response curve to adenosine was shifted to the right by 25 mu M 8-SPT, whereas that of forskolin was shifted downwards. Adenosine deaminase (ADA, EC 3.5.44, 1 U/mL) reduced the intracellular cyclic AMP response to forskolin by 68%, whereas the adenosine transport inhibitor, dipyridamole (10 mu M), significantly increased 1 and 10 mu M forskolin-dependent cyclic AMP accumulation. Erythro-9-(2-hydroxy-3-nonyl)adenine (10 mu M), an inhibitor of ADA, and alpha,beta-methyleneadenosine 5'-diphosphate (100 mu M), an inhibitor of ecto-5'-nucleotidase, did not alter forskolin activity. These results demonstrate that a cyclic AMP extrusion system operates in PC12 cells during adenylyl cyclase stimulation by forskolin and that this stimulation involves a synergistic interaction with endogenous adenosine. However, extruded cyclic AMP does not appear to significantly contribute to the formation of the endogenous adenosine pool. (C) 1999 Elsevier Science Inc.