These studies identify an organic solute transporter (OST) that is generated when two novel gene products are co-expressed, namely human OST␣ and OST␤ or mouse OST␣ and OST␤. The results also demonstrate that the mammalian proteins are functionally complemented by evolutionarily divergent Ost␣-Ost␤ proteins recently identified in the little skate, Raja erinacea, even though the latter exhibit only 25-41% predicted amino acid identity with the mammalian proteins. Human, mouse, and skate OST␣ proteins are predicted to contain seven transmembrane helices, whereas the OST␤ sequences are predicted to have a single transmembrane helix. Human OST␣-OST␤ and mouse Ost␣-Ost␤ cDNAs were cloned from liver mRNA, sequenced, expressed in Xenopus laevis oocytes, and tested for their ability to functionally complement the corresponding skate proteins by measuring transport of [ 3 H]estrone 3-sulfate. None of the proteins elicited a transport signal when expressed individually in oocytes; however, all nine OST␣-OST␤ combinations (i.e. OST␣-OST␤ pairs from human, mouse, or skate) generated robust estrone 3-sulfate transport activity. Transport was sodium-independent, saturable, and inhibited by other steroids and anionic drugs. Human and mouse OST␣-OST␤ also were able to mediate transport of taurocholate, digoxin, and prostaglandin E 2 but not of estradiol 17␤-D-glucuronide or p-aminohippurate. OST␣ and OST␤ were able to reach the oocyte plasma membrane when expressed either individually or in pairs, indicating that co-expression is not required for proper membrane targeting. Interestingly, OST␣ and OST␤ mRNAs were highly expressed and widely distributed in human tissues, with the highest levels occurring in the testis, colon, liver, small intestine, kidney, ovary, and adrenal gland.

Methylmercury (MeHg) readily crosses cell membrane barriers to reach its target tissue, the brain. Although it is generally assumed that this rapid transport is due to simple diffusion, recent studies have demonstrated that MeHg is transported as a hydrophilic complex, and possibly as an L-cysteine complex on the ubiquitous L-type large neutral amino acid transporters (LATs). To test this hypothesis, studies were carried out in Xenopus laevis oocytes expressing two of the major L-type carriers in humans, LAT1-4F2 heavy chain (4F2hc) and LAT2-4F2hc. Oocytes expressing LAT1-4F2hc or LAT2-4F2hc demonstrated enhanced uptake of [(14)C]MeHg when administered as the L-cysteine or D,L-homocysteine complexes, but not when administered as the D-cysteine, N -acetyl-L-cysteine, penicillamine or GSH complexes. Kinetic analysis of transport indicated that the apparent affinities ( K (m)) of MeHg-L-cysteine uptake by LAT1 and LAT2 (98+/-8 and 64+/-8 microM respectively) were comparable with those for methionine (99+/-9 and 161+/-11 microM), whereas the V (max) values were higher for MeHg-L-cysteine, indicating that it may be a better substrate than the endogenous amino acid. Uptake and efflux of [(3)H]methionine and [(14)C]MeHg-L-cysteine were trans -stimulated by leucine and phenylalanine, but not by glutamate, indicating that MeHg-L-cysteine is both a cis - and trans -substrate. In addition, [(3)H]methionine efflux was trans -stimulated by leucine and phenylalanine even in the presence of an inwardly directed methionine gradient, demonstrating concentrative transport by both LAT1 and LAT2. The present results describe a major molecular mechanism by which MeHg is transported across cell membranes and indicate that metal complexes may form a novel class of substrates for amino acid carriers. These transport proteins may therefore participate in metal ion homoeostasis and toxicity.

N-Acetylcysteine (NAC) and dimercaptopropanesulfonate (DMPS) are sulfhydryl-containing compounds that produce a dramatic acceleration of urinary methylmercury (MeHg) excretion in poisoned animals, but the molecular mechanism for this effect is unknown. NAC and DMPS are themselves excreted in urine in high concentrations. The present study tested the hypothesis that the complexes formed between MeHg and these anionic chelating agents are transported from blood into proximal tubule cells by the basolateral membrane organic anion transporters (Oat) 1 and Oat3. Xenopus laevis oocytes expressing rat Oat1 showed increased uptake of [ 14 C]MeHg when complexed with either NAC or DMPS but not when complexed with L-cysteine, glutathione, dimercaptosuccinate, penicillamine, or ␥-glutamylcysteine. In contrast, none of these MeHg complexes were transported by Oat3-expressing oocytes. The apparent K m values for Oat1-mediated transport were 31 Ϯ 2 M for MeHg-NAC and 9 Ϯ 2 M for MeHg-DMPS, indicating that these are relatively high-affinity substrates. Oat1-mediated uptake of [ 14 C]MeHg-NAC and [ 14 C]MeHg-DMPS was inhibited by prototypical substrates for Oat1, including p-aminohippurate (PAH), and was trans-stimulated when oocytes were preloaded with 2 mM glutarate but not glutamate. Conversely, efflux of [ 3 H]PAH from Oat1-expressing oocytes was trans-stimulated by glutarate, PAH, NAC, DMPS, MeHg-NAC, MeHg-DMPS, and a mercapturic acid, indicating that these are transported solutes. [3 H]PAH uptake was competitively inhibited by NAC (K i of 2.0 Ϯ 0.3 mM) and DMPS (K i of 0.10 Ϯ 0.02 mM), providing further evidence that these chelating agents are substrates for Oat1. These results indicate that the MeHg antidotes NAC and DMPS and their mercaptide complexes are transported by Oat1 but are comparatively poor substrates for Oat3. This is the first molecular identification of a transport mechanism by which these antidotes may enhance urinary excretion of toxic metals.

Rat Oatp1 (Slc21a1) is an organic anion-transporting polypeptide believed to be an anion exchanger. To characterize its mechanism of transport, Oatp1 was expressed in Saccharomyces cerevisiae under control of the GAL1 promoter. Protein was present at high levels in isolated S. cerevisiae secretory vesicles but had minimal posttranslational modifications and failed to exhibit taurocholate transport activity. Apparent molecular mass (M) of Oatp1 in yeast was similar to that of unmodified protein, approximately 62 kDa, whereas in liver plasma membranes Oatp1 has an M of approximately 85 kDa. To assess whether underglycosylation of Oatp1 in yeast suppressed functional activity, Oatp1 was expressed in Xenopus laevis oocytes with and without tunicamycin, a glycosylation inhibitor. With tunicamycin, M of Oatp1 decreased from approximately 72 to approximately 62 kDa and transport activity was nearly abolished. Mutations to four predicted N-glycosylation sites on Oatp1 (Asn to Asp at positions 62, 124, 135, and 492) revealed a cumulative effect on function of Oatp1, leading to total loss of taurocholate transport activity when all glycosylation sites were removed. M of the quadruple mutant was approximately 62 kDa, confirming that these asparagine residues are sites of glycosylation in Oatp1. Relatively little of the quadruple mutant was able to reach the plasma membrane, and most remained in unidentified intracellular compartments. In contrast, two of the triple mutants tested (N62/124/135D and N124/135/492D) were present in the plasma membrane fraction yet exhibited minimal transport activity. These results demonstrate that both membrane targeting and functional activity of Oatp1 are controlled by the extent of N-glycosylation.