The effects of guanine nucleotides and divalent cations on the activation of rabbit luteal adenylyl cyclase by the diterpene forskolin were investigated. Saturating concentrations of forskolin elicited 10- to 15-fold stimulation of adenylyl cyclase activity in the absence of added guanine nucleotide. No lag was observed in the time course of forskolin-induced activation. Addition of 10 microM guanosine triphosphate (GTP) and guanyl-5'-yl imidodiphosphate [GMP-P(NH)P] inhibited forskolin activation by 10-15% and 30-40%, respectively, in the presence of 3.0 mM MgCl2. GMP-P(NH)P was more potent than GTP in inhibiting forskolin activation of adenylyl cyclase having an IC50 of 36 nM compared to 610 nM for GTP. In contrast, the Kact for stimulating adenylyl cyclase activity by both GMP-P(NH)P and GTP were similar, 1.00 and 0.86 microM, respectively. GMP-P(NH)P-induced inhibition of the forskolin-activated enzyme was not due to the hysteretic nature of GMP-P(NH)P activation of luteal adenylyl cyclase, as addition of GMP-P(NH)P to an enzyme that had been treated 5 min earlier with forskolin resulted in the immediate inhibition of enzymatic activity. Addition of GMP-P(NH)P to a concentration-effect curve for forskolin increased the Kact value for forskolin form 7.18 to 26.8 microM. There was a different MgCl2 concentration requirement for maximal stimulation of luteal cyclase by GMP-P(NH)P (8 mM MgCl2) and maximal inhibition of forskolin-stimulated activity by GMP-P(NH)P (0.5-0.6 mM MgCl2). Further, MnCl2 concentrations above 1.0 mM completely abolished the inhibitory action of GMP-P(NH)P on forskolin activation of luteal cyclase. In fact, at 2.0 mM MnCl2, adenylyl cyclase activity in the presence of GMP-P(NH)P plus forskolin was greater than that of forskolin alone. Thus taken together, these findings suggest that the rabbit corpus luteum contains an inhibitory guanine nucleotide-binding regulatory component in addition to a stimulatory regulatory component. Further, these components demonstrate differing requirements for both guanine nucleotides and divalent cations in order to interact with the catalytic moiety of adenylyl cyclase. The existence of such an inhibitory component suggests the presence of an inhibitory receptor in the corpus luteum which could negatively regulate adenylyl cyclase resulting in the inhibition of cAMP production and reduced progesterone output from the corpus luteum under normal physiological conditions.