Experiments have been performed to characterize guinea-pig peritoneal eosinophil cyclic nucleotide phosphodiesterase (PDE) activity and establish whether it is involved in regulating superoxide (.O2-) generation. Eosinophils were found to contain a predominantly membrane-bound cAMP PDE(s) (92.5 +/- 2.4% of total activity) which was resistant to solubilization with Triton X-100 (1%). This particulate PDE exhibited complex kinetics (K(m) = 1.3 and 31.4-mu-M) and was unaffected by cGMP (IC50 > 100-mu-M) or CaCl2 (2 mM) + calmodulin (10 units/mL). Little cGMP PDE activity was detected in either the soluble or particulate fractions. Inhibitions of the Ro-20-1724-inhibited (Type IV) cAMP PDE, namely Ro-20-1724 (IC50 = 0.92 +/- 0.43-mu-M), rolipram (IC50 = 0.20 +/- 0.04-mu-M) and denbufylline (IC50 = 0.20 +/- 0.01-mu-M), potently inhibited the particulate cAMP PDE, as did the non-selective inhibitors trequinsin (IC50 = 0.11 +/- 0.02-mu-M) and AH-21-132 (IC50 = 2.57 +/- 0.02-mu-M). Eosinophil cAMP PDE was resistant to SK&F 94120 (IC50 > 1000-mu-M), the cGMP-inhibited (Type III) cAMP PDE inhibitor, and the cGMP PDE (Type I) inhibitor, zaprinast, was only weakly active (IC50 = 35.33 +/- 10.74-mu-M). .O2- release from resting cells was potently inhibited by rolipram (IC50 = 0.05 +/- 0.03-mu-M) and denbufylline (IC50 = 0.06 +/- 0.04-mu-M) but surprisingly, in view of its potent cAMP PDE inhibitory activity, was only weakly decreased by trequinsin (IC50 = 8.0 +/- 2.7-mu-M). AH-21-132 (IC50 > 10-mu-M), SK&F 94120 (IC50 > 10-mu-M) and zaprinast (IC50 > 10-mu-M) were without effect. Rolipram and denbufylline alone exerted little effect on cAMP in intact cells but, in the presence of 10-mu-M isoprenaline, potently increased intracellular accumulation (EC50 = 0.45 +/- 0.16 and 0.28 +/- 0.08-mu-M, respectively). Trequinsin and AH-21-132 only weakly enhanced isoprenaline-stimulated cAMP accumulation. Although it induced a marked rise in cAMP only in the presence of isoprenaline, rolipram (50-mu-M) alone was able to increase the activity ratio of cAMP-dependent protein kinase from 0.24 to 0.84. The results suggest that Ro-20-1724-inhibited cAMP PDE plays a role in regulating eosinophil .O2- generation. The poor correlation between the PDE inhibitory actions of certain compounds and their effectiveness in elevating cAMP and inhibiting .O2- suggests the existence of a barrier impeding access to the enzyme.