Microbial stimuli such as bacterial lipopolysaccharide (LPS) or glycosylphosphatidylinositol-mucins derived from Trypanosoma cruzi trypomastigotes (tGPI-mucins) are effective stimulators of the synthesis of cytokines by macrophages. Here, we evaluated the ability of cyclic AMP mimetic or elevating agents to modulate TNF-alpha and IL-12 synthesis by murine inflammatory macrophages. Cholera Toxin (ChTx) inhibited tGPI-mucins (2.5 nM) or LPS (100 ng ml(-1)) induced TNF-alpha and IL-12(p40) synthesis in a concentration-dependent manner. Similarly, the cyclic AMP mimetics, 8-bromo cyclic AMP or dibutyryl cyclic AMP, or prostaglandin (PG) E2 inhibited the synthesis of both cytokines by macrophages exposed to microbial stimuli. The protein kinase A inhibitor H-89 partially reversed the inhibitory effects of dibutyryl cyclic AMP and PGE2 on both IL-12(p40) and TNF-alpha synthesis. Pretreatment of macrophages with dibutyryl cyclic AMP or ChTx augmented the synthesis of IL-10 triggered by microbial products. Elevation of cyclic AMP inhibited the synthesis of TNF-alpha, but not IL-12(p40), by inflammatory macrophages from IL-10 knockout mice. Kinetic studies showed that synthesis of both TNF-alpha and IL-10 peaked at 8 h and IL-12 at 24 h after stimulation with microbial stimuli. Together, our findings favour the hypothesis that the cyclic AMP inhibitory activity on IL-12(p40) but not on TNF-alpha synthesis is dependent on de novo protein synthesis, most likely involving IL-10, by macrophages stimulated with microbial products. Accordingly, dibutyryl cyclic AMP inhibited IL-12(p40) synthesis only when added before or at the same time of the stimuli. In contrast, the effect of this cyclic AMP analogue on TNF-alpha synthesis was protracted and observed even 2 h after the addition of the stimuli.