Previous studies from this laboratory have shown that progesterone (PROG) treatment in ovariectomized rats produces an anti-anxiety response similar to that observed after the administration of prototypical anxiolytic benzodiazepine (BDZ) compounds, The FROG-induced anxiolytic response was highly correlated with an increased level of 34-hydroxy-5 alpha-pregnan-20-one (allopregnanolone) in the blood and brain, and was also associated with a facilitation of GABA-stimulated chloride ion (Cl-) influx in cortical synaptoneurosomes. This correlative evidence suggested that the anxiolytic effect of FROG was a result of its in vivo reduction to the neuroactive steroid, allopregnanolone. In this report, a series of studies was conducted to determine the mechanism(s) by which FROG alters behavior in animal models of anxiety. In the first experiment, ovariectomized rats were injected with FROG (1 mg/0.2 ml, SC) 4 h prior to a test in the elevated plus-maze, Some animals also received an injection of picrotoxin (0.75 mg/kg, IP), a GABA(A) receptor-gated Cl- channel antagonist, whereas other animals were pretreated with RU 38486 (5 mg/0.2 ml, SC), a progestin receptor antagonist. FROG elicited anxiolytic behavior in the plus-maze, an effect that was blocked by picrotoxin administration. Pretreatment with RU 38486 was not effective in altering FROG-induced anxiolytic behavior in the plus-maze, in a second experiment, the effect of FROG on behavior in the plus-maze was determined in the presence of N,N-diethyl-4-methyl-3-oxo-4-aza-5 alpha-androstane-17 beta-carboxamide (4-MA; 10 mg/0.2 ml, SC), a 5 alpha-reductase inhibitor. The enzyme inhibitor was potent in preventing the anxiolytic effect observed in the plus-maze after FROG administration. In the defensive burying paradigm, FROG treatment also produced anxiolysis by reducing the duration of burying behavior, and this effect was prevented by 4-MA pretreatment, but not by RU 38486 administration, After the completion of the behavioral assays, analysis of blood allopregnanolone levels revealed a marked increase in FROG-treated females. The FROG-induced elevation in circulating allopregnanolone was blocked by pretreatment with 4-MA. In cortical synaptoneurosomes, the sensitivity (inverse of the EC(50)) and the maximal response (Emax) in GABA-stimulated Cl-36(-) uptake were increased in FROG-treated females, The potentiation of FROG on both of these neurochemical measures was not observed in animals pretreated with 4-MA. Together, these studies provide evidence that the anxiolytic effect of FROG is not associated with an intracellular steroid receptor that initiates genomic-mediated responses. The evidence is consistent with a nongenomic mechanism whereby FROG is metabolized to allopregnanolone, a neuroactive steroid that potentiates GABA(A) receptor-mediated responses.