In rats the sinusoidal glutathione (GSH) carrier transports GSH bidirectionally, and its activity is influenced by the thiol-disulfide status; the Vmax of sinusoidal GSH efflux was increased by dithiothreitol (DTT) and decreased by cystine. In the present work we examined the specificity and directionality of the thiol effect. Using in situ perfused livers, we found that 1 mM DTT and other dithiols, including 1,2-ethanedithiol, 1,3-propanedithiol, and 1,4-butanedithiol, stimulated sinusoidal GSH efflux by 200-500% but dihydrothioctic acid, which is negatively charged, had no effect. Uncharged or positively charged monothiols (2 mM), such as dimercaprol, monothioglycerol, 2-mercaptoethanol, 3-mercapto-2-butanol, 1-mercapto-2-propanol, and cysteamine, also exerted a stimulatory effect on sinusoidal GSH efflux. In contrast, monothiols containing a negatively charged substituent, such as penicillamine, captopril, N-acetylcysteine, mercaptopropionylglycine, mercaptoethanesulfonic acid, mercaptoacetic acid, and mercaptopropionic acid, had no effect. The thiol moiety was essential for activity, inasmuch as ethanol, propanol, propanediol, and glycerol had no effect on sinusoidal GSH efflux. The effect of DTT or cystine pretreatment (2 mM or 0.5 mM, respectively, for 30 min) on GSH uptake was then examined using cultured rat hepatocytes. The linear rate of [35S]GSH uptake and the concentration dependence were measured after cells were pretreated with acivicin (0.5 mM, for 15 min) and buthionine sulfoximine (10 mM, 15 min), to prevent breakdown and resynthesis of GSH from precursors, respectively. Uptake buffer also contained 20 mM alpha-(methylamino)isobutyric acid and 20 mM threonine (inhibitors of amino acid transport systems A and ASC, respectively), to prevent uptake of cysteine. Pretreatment with DTT decreased the Vmax of GSH uptake by approximately 50% (control Vmax value, 24 nmol/10(6) cells/30 min), whereas the Km remained unaffected (approximately 8 mM). Cystine pretreatment had no influence on GSH uptake but inhibited efflux. In conclusion, the presence of at least one thiol group and the absence of negative charge are required to stimulate sinusoidal GSH efflux. The direction of GSH transport is modulated by the thiol-disulfide status, so that thiol reduction changes the GSH transporter from a bidirectional GSH transporter into a preferentially unidirectional (outward) transporter by inhibiting uptake while stimulating efflux and thiol oxidation favors inward transport by inhibiting only efflux.