The factors regulating relaxin production by the primate CL during spontaneous menstrual cycles and in early pregnancy are poorly understood. Since the CL produces steroids, notably progesterone, and expresses progesterone receptors, luteal progesterone may act locally to regulate relaxin production. For the current study, either trilostane (600 mg daily)--a 3 beta-hydroxysteroid dehydrogenase inhibitor--or vehicle was administered to rhesus monkeys during the midluteal phase (Days 6 and 7) of spontaneous menstrual cycles to examine the effects of reduced luteal progesterone synthesis on relaxin secretion. Trilostane treatment reduced serum concentrations of progesterone within 3 h of initial administration and maintained low levels typical of the follicular phase (< 1 ng/ml), causing premature menses without significant alteration in serum bioactive LH levels. Nevertheless, the patterns and levels of circulating relaxin, as measured by homologous macaque ELISA, were not different between trilostane- and vehicle-treated monkeys, with relaxin levels peaking in both groups by Day 13 of the luteal phase. To determine if chorionic gonadotropin (CG) injections simulating early pregnancy could stimulate relaxin production in a progesterone-depleted environment, trilostane or vehicle was administered as described above, followed by injections in increasing dosages of human (h) CG beginning 3 days after initial trilostane administration. Serum progesterone levels in trilostane-treated animals were significantly reduced prior to and during hCG treatment when compared with vehicle-treated animals. However, serum relaxin levels were comparable between these groups; relaxin levels peaked approximately 10-fold above pre-hCG levels in both trilostane- and vehicle-treated animals.(ABSTRACT TRUNCATED AT 250 WORDS)