The transport systems of animal and human tissues for vitamin C are reviewed with respect to their properties. It emerges that pure diffusion plays only a very minor role while a variety of more or less specific transporters is found on cellular membranes. Although most tissues prefer the reduced ascorbate over the oxidized dehydroascorbic acid and have high-affinity transporters for it, there are several examples for the reversed situation. Special attention is given to similarity or identity with glucose transporters, especially the GLUT-1 and the sodium-dependent intestinal and renal transporters, and to the very widespread dependence of ascorbate transport on sodium ions. The significance of ascorbate transport for vitamin C-requiring and nonrequiring species as well as alterations in states of disease can be seen from ample experimental evidence.

A transmembrane monodehydroascorbate reductase activity with a high affinity in the subpicomolar concentration range of the free radical can be measured at the surface of erythroleukemic cells using a ferricyanide-driven redox cycle. The activity is dependent on the membrane potential and can therefore only be found in intact cells. It is independent of the glutathione content of the cells. Thenoyltrifluoroacetone is an efficient inhibitor of the activity, whereas ouabain, monensin and tetraethylammonium show no effect.Cells are able to generate ascorbate from dehydroascorbic acid. This explains why both forms of vitamin C show practically the same affinity for the redox cycle but why it does not drive the redox cycle by itself because it is much slower and is not inhibited by thenoyltrifluoroacetone.The reductase activity is independent of the degree of differentiation of the leukemic cells.NADH-monodehydrascorbate reductase is present in the outer mitochondria1 membranes of several animal tissues [ 1, 21, but its activity has also been identified in plasma membranes [3, 41 and coated vesicles [5] as well as in microsomal vesicles of not completely defined origin [6].Evidence for transmembrane orientation of this activity in cultured cells has been obtained in two ways; HL-60 cells stabilize ascorbate in the surrounding medium [7, 81 and ascorbate has been shown to stimulate reduction of extracellular ferricyanide by K562 and U937 cells [9]. The latter process is independent of an uptake and release cycle of oxidized and reduced ascorbate and has high affinity for ascorbate. Dehydroascorbate (DHA) has a similar effect with lower maximal rates. This paper describes an attempt to characterize this stimulatory effect and discriminate it from other transmembrane electron-transport activities in K562 cells. It also provides a possible explanation for the difficulties arising in isolating this enzyme from mammalian cell plasma membranes.The monodehydroascorbate-reducing activity and the DHA-reducing property of the cells also described here might contribute to the ability of cellular systems to keep the antioxidant ascorbate [lo] MATERIALS AND METHODS Cell cultureK562 cells [ l l ] were grown in RPMI 1640 media supplemented with penicillinktreptomycin (100 U/ml), 2 mM glutamine and 10% fetal calf serum in a humidified atmosphere with 5 % CO, at 37°C. They were kept in the logarithmic growth phase by appropriate dilution twice a week.Differentiation was induced by addition of 50 pM hemin [ 121. The amount of hemoglobin-producing cells was estimated by benzidine staining [13] and counting the blue cells in a counting chamber.Glutathione depletion was achieved by overnight incubation with buthionine sulfoximine (BSO) [14, 151 and verified by the determination of the total glutathione by the enzymecycling assay [16]. AssaysReduction of ferricyanide was carried out in Tris/NaCI/ glucose medium (150 mM NaCl, 20 mM Tris, 7 mM glucose, pH 7.4) as described [9]. Reagents and inhibitors were dissolved in appropriate sol...

Transmembrane reduction of extracellular oxidants by K562 and U937 leukemic cells was stimulated by catalytic amounts of ascorbate or dehydroascorbate. This stimulation was not due to transport of ascorbate in different redox states in and out of the cells. The membrane redox cycle was strictly dependent on the presence of the cells at every stage, and showed high affinity for ascorbate with simple linear kinetics. Metabolic inhibitors and sulfhydryl reagents inhibited this stimulation. Ascorbate uptake was also dependent on oxidation, but in a very different manner and with much lower affinity for ascorbate. The uptake was non-saturable in the concentration range used. There was some release of ascorbate from the cells, which cannot account for an appreciable part of the reduction of extracellular electron acceptors.Stimulation by external electron acceptors has been implicated as a possible control element of cell proliferation [l]. The oxidants used to establish this hypothesis have been mainly ferricyanide ([Fe(CN>, [2] and a mixture of ascorbate and dehydroascorbic acid, yielding an ascorbate free radical (monodehydroascorbate; AFR) 131. These oxidants are believed to act via a trans-plasma-membrane oxidoreductase, depleting the intracellular NAD + /NADH pool of reducing equivalents and alkalinizing the cytoplasm by secreting protons via the N a f / H + antiport [4-81.The AFR-mediated stimulation of cell proliferation, in particular, was attributed to the action of NADH: monodehydroascorbate oxidoreductase residing in the plasma membrane [9], however the reaction has not been defined for this system. It is known that some cell types, notably erythrocytes, can also reduce external oxidants via export of ascorbic acid and uptake of dehydroascorbic acid [lo, 111 which is reduced by intracellular glutathione [12].This study demonstrates that catalytic amounts of ascorbate or dehydroascorbic acid stimulate the reduction of external ferricyanide by two lines of suspension-cultured human tumor cells, in a manner precluding export of reducing equivalents as a feasible mechanism. The ascorbate/ferricyanide mixture may provide a tool for studying redox stimulation of cell proliferation over short times. The instability of ferricyanide, especially in serum-containing media, prohibits experiments of this kind over larger time periods. MATERIALS AND METHODS CellsK562 human erythroleukemia [13] and U937 histiocyticlymphoma monoblastic cells [14] were cultivated to logarithmic growth phase in RPMI (Roswell Park Memorial Institute) 1640 bicarbonate-buffered medium supplemented with 2 mM L-glutamine, 100 U/ml penicillin/streptomycin and 10% (by vol.) fetal calf serum (FCS).For experiments, the cells were harvested by centrifugation, washed twice in unsupplemented RPMI and the buffer medium used for the particular experiment, in which the cells were finally resuspended.Fresh complete medium was added to log-phase cells, and the supernatant harvested 24 h later for use as cell-conditioned medium.

The influence of conjugation of doxorubicin to holotransferrin on the receptor-mediated endocytosis of and on the iron uptake from transferrin was studied using K562 cells. '251-labelled transferrin and doxorubicin-transferrin conjugates were used in the binding, dissociation, and ligandexchange experiments at 0 "C, and 5gFe,'251-labelled (double-labelled) ligands were used in the endocytosis, iron uptake, and recycling experiments at 37 "C. The binding affinity of conjugates was about half of that of transferrin. Binding of 1251-labelled ligands was blocked by both unlabelled ligands to the same degree, however, it was not blocked at all by an 8000-fold excess of doxorubicin.After saturation bindings, slightly more '251-labelled conjugates dissociated from the surface of cells than transferrin. Exchange of '251-labelled ligands for unlabelled ligands resulted in different EC,, values (defined as the concentration of unlabelled ligand at which half as much radioligand is exchanged for unlabelled ligand as would be exchanged at infinitely high concentration of unlabelled ligand under similar assay conditions). While transferrin exchanged transferrin with an EC,, value close to the binding affinity, conjugates exchanged conjugates with much lower efficiency. The heterolog exchange experiments yielded EC,, values inbetween the two extrema. For studying iron uptake, K562 cells were loaded with the double-labelled ligands either at 37 "C (endosome-loading only) or at 0°C (surface-loading only). Results obtained for the endocytosis of, the iron uptake from, and the recycling of double-labelled ligands indicate that (a) the rate of iron uptake is smaller from conjugates than from transferrin, (b) there are at least two parallel recycling processes for both ligand . receptor complexes, and (c) each time constant characterizing the different steps of iron uptake via receptor-mediated endocytosis is smaller for conjugates than for transfenin (or, the half times characterizing the different steps are higher for conjugates than for transferrin). Endocytosis and iron uptake were unaffected by free doxorubicin (12.5 pM) or colchicin (1 mh4). Benzyl alcohol (30 mM) slowed down the rate of both endocytosis and iron uptake, while dithiothreitol (5 mM) decreased the rate of iron uptake and increased the rate of endocytosis. N-Ethylmaleimide (1 mMj completely stopped both endocytosis and iron uptake. The results suggest that the binding of conjugates to the surface of cells is governed by the binding of the transfenin part of conjugates to the transferrin receptor. However, conjugation of doxorubicin to transferrin seems to influence all properties of transfenin. A minor alteration in the structure may modify the binding affinity and stability of conjugates as well as decrease the accessibility of competing ligand to displace the bound ligand. Either of them could be responsible for the longer endocytosis and recycling time for conjugates. Our results also point to the importance of both the sulfhydryl groups and the fluidity of p...

K562 erythroleukaemic cells produced ascorbate when incubated with dehydroascorbic acid. The reduction depended on the number of cells and on the concentration of dehydroascorbic acid. The observed rate consists of a high affinity (apparent Km 7 mu M, Vmax 3 center dot 25 pmol min-1 (10(6) cells)-1 and a low affinity component, which was non-saturable up to 1 mM of DHA (rate increase of 0 center dot 1 pmol min-1 (10(6) cells)-1 (1 mu M of DHA-1). The rate was dependent on temperature and was stimulated by glucose and inhibited by phloretin, N-ethylmaleimide, parachloro-mercuribenzoate and the noyltrifluoroacetone. Although uptake of DHA proceeded at a higher rate than its extracellular reduction, the generation of extracellular ascorbate from DHA cannot be accounted for by intracellular reduction and the release of ascorbate, since the latter was not linear with time and had an initial rate of approximately 3 pmol min-1 (10(6) cells-1). At a concentration of DHA of 100 mu M this is 25 per cent of the observed reduction.