Evidence that neurosteroids are potent modulators of the action of GABA at GABAA receptors has prompted the investigation of the mechanism that controls brain neurosteroid synthesis by glial cell mitochondria in vivo. In vitro studies suggest that the interaction of the diazepam binding inhibitor (DBI)--a polypeptide that is abundant in steroidogenic cells--with glial mitochondrial DBI receptors (MDRs) is a crucial step in the physiological regulation of neurosteroid biosynthesis. MDRs bind 4'-chlorodiazepam (4'-CD), N,N-di-n-hexyl-2-(4-fluorophenyl)-indol-3-acetamide (FGIN-1-27), and the isoquinoline carboxamide PK 11195 with high affinity, and these ligands have been used to investigate whether the stimulation of glial MDRs increases brain pregnenolone production in vivo. Adrenalectomized and castrated (A-C) male rats (to eliminate peripheral sources of pregnenolone) were pretreated with trilostane (to prevent pregnenolone metabolism to progesterone), and the pregnenolone content in brain regions dissected after fixation with a 0.8-s exposure to microwave irradiation focused to the head was determined by HPLC followed by specific radioimmunoassay. The forebrain and cerebellum of A-C rats contained 4-7 ng of pregnenolone/g of tissue, and the olfactory bulb contained 10-14 ng/g. These concentrations of brain pregnenolone are only 30-40% lower than those of sham-operated rats. In contrast, the plasma pregnenolone content of sham-operated rats was 2-3 ng/ml, but it was only 0.15-0.20 ng/ml in the plasma of A-C rats. In A-C rats, treatment with the MDR ligands 4'-CD and FGIN-1-27 increased the pregnenolone content in the brain but failed to change the plasma or peripheral tissue content of this steroid. The effect of 4'-CD on brain pregnenolone content was maximal (70-100% increase) at the dose of 18 mumol/kg, 5-10 min after intravenous injection. The effect of oral administration of FGIN-1-27 on brain pregnenolone content was maximal (80-150% increase) at doses of 400-800 mumol/kg and peaked at approximately 1 h. That this effect of FGIN-1-27 was mediated by the MDR was documented by pretreatment with the MDR partial agonist PK 11195 (100 mumol/kg, i.p.). PK 11195 did not affect basal brain pregnenolone content but prevented the accumulation of brain pregnenolone induced by FGIN-1-27. FGIN-1-27 and 4'-CD failed to increase the brain concentration of dehydroepiandrosterone in A-C rats. These data suggest that glial cell MDRs play a role in neurosteroid biosynthesis in vivo.