R ecently our laboratory has cloned both the rat canalicular and sinusoidal GSH transporters (RcGshT and RsGshT, respectively; Yi, J., S. Lu, J. Fernandez-Checa, and N. Kaplowitz. 1994. J. Clin. Invest. 93:1841-1845; and 1995. Proc. Natl. Acad. Sci. USA. 92:1495-1499). The current work characterized GSH transport and the expression of these two GSH transporters in various mammalian cell lines. The average cell GSH levels (nmol/10(6) cells) were 25, 22, 32, 13, and 13 in HepG2, HeLa, CaCo-2, MDCK, and Cos-1 cells, respectively. GSH efflux was temperature dependent and averaged 0.018, 0.018, 0.012, 0.007, and 0.019 nmol/10(6) cells/min from HepG2, HeLa, CaCo-2, MDCK, and Cos-1 cells, respectively. Dithiothreitol (DTT), which stimulates rat sinusoidal GSH efflux, stimulated GSH efflux only in HepG2 and HeLa cells which was partially reversed by subsequent cystine treatment. GSH uptake (1 mM plus 35S-GSH) was temperature dependent, linear up to 45 min, and Na+-independent with average rates of 1.12, 0.91, 0.45, and 0.45 nmol/10(6) cells/30 min for HepG2, HeLa, CaCo-2, MDCK, and Cos-1 cells, respectively. BSP-GSH (2mM), which cis-inhibits sinusoidal GSH uptake in rat liver and HepG2 cells, inhibited GSH uptake only in HeLa cells. mRNA and polypeptide of RcGshT are expressed in all cells whereas those of RsGshT are expressed only in HepG2 and HeLa cells. In conclusion, bidirectional GSH transport, mediated by the "canalicular" GSH transporter, is ubiquitous in mammalian cells. Sinusoidal GSH transporter expression is more restricted, being present in HepG2 and HeLa cells. DTT and BSP-GSH affect GSH transport only in cells expressing the sinusoidal transporter confirming their selective action on this transporter.