The enzyme 4-ene-3-ketosteroid-5a-oxidoreductase [5a-reductase; 3-oxo-5a-steroid A4-dehydrogenase, 3-oxo-5a-steroid: (acceptor) EC 1.3.99.5] plays a key role in androgen-dependent target tissues, where it catalyzes the conversion of testosterone to the biologically active dihydrotestosterone. The regulation of 5a-reductase expression has not been studied at the molecular level as the enzyme is a membrane protein that is labile in cell-free homogenates. We developed a sensitive bioassay of the enzyme activity expressed in Xenopus oocytes microinjected with rat liver and prostate mRNA. After microinjection, incubation of intact oocytes in the presence of [3H]testosterone revealed the in ovo appearance of active 5a-reductase. Polyadenylylated RNA was fractionated by sucrose gradient centrifugation, and the enzymatic activity was shown to be encoded by a 1600-to 2000-base-pair fraction of hepatic poly(A) + RNA. 5a-Reductase mRNA was most efficiently translated when up to 80 ng of RNA was injected per oocyte. In the injected oocytes, 5a-reductase mRNA was found to be a short-lived molecule (til2 = 2 hr), whereas its in ovo translatable 5a-reductase protein exhibited stable enzymatic activity for over 40 hr. Moreover, the levels oftranslatable tissue-specific 5a-reductase mRNAs as monitored in the Xenopus oocytes correlated with the variable 5a-reductase activities in female rat liver, male rat liver, and prostate homogenates; the ratio of their specific activities was of 2500:630:1, respectively. Altogether, these results provide supporting evidence in favor of the transcriptional control of 5a-reductase expression in rat tissues.ical inactivation of 4-ene-3-ketosteroids, mainly adrenal corticosteroids (19).Previous biochemical studies of 5a-reductase have revealed kinetic parameters, such as affinity constants to steroid hormones, as well as the enzyme subcellular distribution in the rough endoplasmic reticulum and/or nuclei of various tissues (1,(20)(21)(22)(23)(24). Because the activity of Sareductase is extremely unstable in cell-free preparations (1, 20), it has not been purified, and therefore the regulation of its expression could not be studied at the molecular level. To approach this issue, we pursued an experimental model in which 5a-reductase biosynthesis can be studied in intact living cells. Toward this aim, we designed a bioassay for the detection of newly synthesized 5a-reductase produced in mRNA-microinjected Xenopus oocytes. The oocytes efficiently translate microinjected heterologous mRNA, perform correct posttranslational processing and, most importantly for 5a-reductase, direct newly synthesized proteins into the correct subcellular compartments (for review, see ref. 25).Here, we report that Xenopus oocytes are able to synthesize catalytically active 5a-reductase, directed by microinjected poly(A) + RNA prepared from female and male rat liver. For yet unknown reasons, there are remarkable differences between the specific activities of 5a-reductase in female liver, male liver (16, 17)...