Specificity of Ascorbate Analogs for Ascorbate Transport

Cellular ascorbic acid accumulation occursin vitro by two distinct mechanisms: transport of ascorbate itself or transport and subsequent intracellular reduction of its oxidized product, dehydroascorbic acid. It is unclear which mechanism predominates in vivo. An easily detectable compound resembling ascorbate but not dehydroascorbic acid could be a powerful tool to distinguish the two transport activities. To identify compounds, 21 ascorbate analogs were tested for inhibition of ascorbate or dehydroascorbic acid transport in human fibroblasts. The most effective analogs, competitive inhibitors of ascorbate transport withK i values of 3 μm, were 6-deoxy-6-bromo-, 6-deoxy-6-chloro-, and 6-deoxy-6-iodo-l-ascorbate. No analog inhibited dehydroascorbic acid transport. Using substitution chemistry, [125I]6-deoxy-6-iodo-l-ascorbate (1.4 × 104 mCi/mmol) was synthesized. HPLC detection methods were developed for radiolabeled and nonradiolabeled compounds, and transport kinetics of both compounds were characterized. Transport was sodium-dependent, inhibited by excess ascorbate, and similar to that of ascorbate. Transport of oxidized ascorbate and oxidized 6-deoxy-6-iodo-l-ascorbate was investigated usingXenopus laevis oocytes expressing glucose transporter isoform GLUT1 or GLUT3. Oxidation of ascorbate or its analog in media increased uptake of ascorbate in oocytes by 6–13-fold compared with control but not that of 6-deoxy-6-iodo-l-ascorbate. Therefore, 6-deoxy-6-iodo-l-ascorbate, although an effective inhibitor of ascorbate transport, either in its reduced or oxidized form was not a substrate for dehydroascorbic acid transport. Thus, radiolabeled and nonradiolabeled 6-deoxy-6-iodo-l-ascorbate provide a new means for discriminating dehydroascorbic acid and ascorbate transport in ascorbate recycling.

Ascorbic acid (vitamin C, ascorbate) is a required nutrient for humans. Ascorbate is a cofactor in eight intracellular mammalian enzymatic reactions and serves a primary role in the defense against oxidant radicals in vivo (1)(2)(3). Loss of two electrons leads to formation of dehydroascorbic acid, which can be reduced back to ascorbate enzymatically or nonenzymatically (4 -8).
Ascorbate accumulation in vivo and in cells occurs by two mechanisms, collectively termed ascorbate recycling (3,9,10). Ascorbate is transported as such or is oxidized extracellularly to dehydroascorbic acid, transported by glucose transporters, and reduced internally (3,11,12). Under physiologic conditions in the absence of oxidation, ascorbate is the primary, if not the only, substrate available (13,14). Nevertheless, it remains unknown whether normal intracellular ascorbate accumulation occurs solely due to ascorbate transport itself or whether some fraction of ascorbate is oxidized at the cell membrane to dehydroascorbic acid, transported, and reduced intracellularly. A serious disadvantage in using ascorbate itself to discriminate these mechanisms of transport is that ascorbate can be accumulated by both pathways. To characterize the mechanisms of ascorbate accumulation physiologically, a compound is needed that is similar to ascorbate but is transported by only one mechanism. Such a compound has not been available.
An ideal compound should meet several criteria. It should be similar to ascorbate because this is the substrate found physiologically. The compound should be measurable in small amounts. Its utility would be greatly enhanced if the compound could be detected easily, for example by HPLC or, especially, by radiochemical detection. Current radioactive detection of ascorbate is problematic because of low specific activity of available radiolabeled ascorbate, emphasizing the need for a much higher specific activity radiolabeled analog. Finally, an ideal compound to study transport and accumulation would be specific for mechanisms of ascorbate transport only and would show no activity toward dehydroascorbic acid transport mechanisms.
To address these issues, we synthesized or obtained 21 structural analogs of ascorbate having substitutions or conformational changes at positions 2-6. We studied the ability of each analog to inhibit ascorbate transport in cultured human fibroblasts, cells in which ascorbate transport kinetics have been previously characterized (15). Based on these observations, we synthesized 6-deoxy-6-iodo-L-ascorbate and [ 125 I]6-deoxy-6iodo-L-ascorbate, a high specific activity ascorbate analog, and we characterized their transport characteristics in fibroblasts and in Xenopus oocytes expressing glucose transporter isoforms GLUT1 or GLUT3. The results suggested that 6-deoxy-6-iodo-L-ascorbate and [ 125 I]6-deoxy-6-iodo-L-ascorbate may be ideal compounds for quantitation of ascorbate-specific transport. 14 C]Ascorbic acid (8.0 mCi/mmol) and [ 125 I]NaI (25 mCi in 71 l of 10 Ϫ5 M NaOH) were purchased from NEN Life Science Products (Boston, MA). Ascorbic acid, 2-deoxyglucose, 3-O-methyl glucose, serine, leucine, threonine, bovine serum albumin, dithiothreitol * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Reagents-[
¶ To whom correspondence should be addressed: Bldg. 10 (DTT), 1 and EDTA were obtained from Sigma, and methanol was purchased from Baker (Philipsburg, NJ). Other reagents were of the highest available commercial grade.
[ 125 I]6-Deoxy-6-iodo-L-ascorbic acid was synthesized by an exchange reaction between methyl 2,3-isopropylidene-6-deoxy-6-iodo-L-gulonate and [ 125 I]NaI (NEN Life Science Products). [ 125 I]NaI (71 l of the aqueous original solution, 25 mCi) was added to methyl 2,3-isopropylidene-6-deoxy-6-iodo-L-gulonate (1 mg in 200 l of acetone), and the resulting solution was evaporated to dryness with nitrogen to remove water. Acetone was added (200 l), and the vial was sealed and heated at 60°C for 3 days. After evaporation of the solvent with nitrogen, the residue was dissolved in ethyl acetate (400 l) and extracted with H 2 O (2 ϫ 200 l) to remove unbound free [ 125 I]iodide. The organic layer was evaporated with nitrogen, and a solution of concentrated HCl (50 l) in methanol (400 l) was added. This solution was heated in the same sealed vial at 60°C for 2 h and then evaporated to dryness under a slow stream of nitrogen to yield [ 125 I]6-deoxy-6-iodo-L-ascorbic acid (0.67 mg, specific activity 1.4 ϫ 10 4 mCi/mM). The dry radiolabeled analog was stored at Ϫ70°C and used within 1 week. Before use the analog was diluted with ultrapure water containing 1 mM DTT to form a stock concentration of 3.5 mM.
Fibroblast Culture-Human skin fibroblasts, strains CRL 1501 and CRL 1497, were obtained from the American Type Culture Collection (Manassas, VA). Fibroblasts were cultured in Eagle's minimum essential medium supplemented with 1% nonessential amino acids, 2 mM glutamine, and 10% heat-inactivated fetal bovine serum (Life Technologies, Inc.). Cells were cultured at 37°C in a 5% CO 2 humidified atmosphere. Cultures were subdivided using 0.05% trypsin with 0.53 mM EDTA (Life Technologies). Experiments were performed using cultures with passage numbers 5-15.
Fibroblast Transport Assays-Fibroblasts were plated in 12-well tissue culture dishes (Costar, Cambridge, MA) and grown until confluent, approximately 7 days. Fifteen minutes before experiments, culture medium was removed and replaced with two changes of balanced salt solution (BSS) containing 150 mM NaCl, 5 mM KCl, 1.9 mM KH 2 PO 4 , 1.1 mM Na 2 HPO 4 , 0.3 mM MgSO 4 ⅐7H 2 O, 1 mM MgCl 2 ⅐6H 2 O, 1.5 mM CaCl 2 ⅐2H 2 O, 10 mM HEPES, adjusted to pH 7.4 with NaOH. Inhibition of ascorbic acid transport was determined by adding the indicated concentration of inhibitor to plated cells along with [ 14 C]ascorbic acid, final concentration 10 M. Cells were incubated at 37°C for the times specified in the legends. After incubation, cells were placed on ice and washed three times with 2 ml of ice-cold phosphate-buffered saline. Intracellular ascorbic acid was measured either by solubilizing cells with 0.1 N NaOH, 1% CHAPS or by extracting intracellular ascorbic acid with 0.5 ml of ice cold 60% methanol, 1 mM EDTA. Both methods provided similar results. The methanol/EDTA was gently swirled in the culture wells for 2 min, removed, and immediately frozen on dry ice for later HPLC analysis. An aliquot of either NaOH cell lysate or methanol/ EDTA extraction solution was removed, and radioactivity was quantified using scintillation spectrometry. Two analogs, 3-O-benzyl-L-ascorbate and 3-O-octadecyl-L-ascorbate, were initially solubilized in ethanol because of poor water solubility. Final concentrations of ethanol in the assay buffer were Յ 1% (v/v), and control values were obtained from cells incubated with ascorbate and ethanol without inhibitor. All other inhibitors were suspended in HPLC grade H 2 O at final assay concen- Transport and accumulation of 6-deoxy-6-iodo-L-ascorbic acid and [ 125 I]6-deoxy-6-iodo-L-ascorbic acid were measured in a similar manner as ascorbate above except that 0.1 mM DTT was added to the incubation medium, and 1 mM DTT was added to the 60% methanol, 1 mM EDTA extraction solution. The DTT addition was necessary to prevent analog oxidation during extraction.
Cell protein was quantified from NaOH/CHAPS cell lysates using the bicinchoninic acid protein assay (Pierce) (17), using bovine serum albumin (Sigma) as the standard.
Results were calculated as pmol of ascorbate/g of cellular protein and converted to micromolar amounts based on an intracellular volume for human fibroblasts of 11.0 Ϯ 2.0 l/mg of cell protein (15). Experimental points represent the mean Ϯ S.D. of at least three samples. Error bars were omitted when the S.D. was less than the size of the symbol.
HPLC Analysis-Ascorbate mass was measured by reverse-phase HPLC using a 5, 250 ϫ 4.6-mm C-18 column (Columbus, Phenomenex, Torrance CA) with coulometric electrochemical detection as described previously (18). 6-Deoxy-6-iodo-L-ascorbate was measured by HPLC with electrochemical detection using the same detection system and settings, but the mobile phase used for ascorbate was modified to contain 55% methanol. [ 125 I]6-Deoxy-6-iodo-L-ascorbic acid was measured using the HPLC separation conditions for unlabeled compound, with the addition of an online scintillation spectrometry detection system after the coulometric detector (12).
Xenopus Oocyte Transport Assay-Xenopus oocytes were isolated and injected with mRNA coding for the glucose transporter isoforms GLUT1 or GLUT3 as described (12). Three days postinjection, individual oocytes were incubated for 15 min at room temperature with 50 M of either [ 14 C]ascorbic acid or [ 125 I]6-deoxy-6-iodo-L-ascorbate in the presence or absence of either hydrogen peroxide (H 2 O 2 ) (8 mM) or xanthine (50 M) and xanthine oxidase (0.05 units/ml) (Sigma) and 36 M DTT. The presence of DTT was due to dilution of DTT present in original stock material. After 15-min incubations, oocytes were washed four times with ice-cold phosphate-buffered saline containing 1 M phloretin. Individual oocytes were solubilized with 10% SDS, and internalized radioactivity was quantified by scintillation spectrometry (12).

RESULTS
Inhibition of Ascorbate Transport by Analogs-Ascorbic acid analogs with stereochemical inversion or substitutions on carbons 2, 3, 4, 5, and 6 were examined individually for their ability to inhibit ascorbic acid uptake in cultured fibroblasts (Table I). Without analog, ascorbate was accumulated at least 8-fold against a concentration gradient (not shown). Little inhibition of ascorbic acid transport was observed with compounds containing substitutions on carbon 2 or 3. Analogs examined included 2-amino-L-ascorbic acid, 2-sulfato-L-ascorbic acid, 3-O-benzyl-L-ascorbic acid, and 3-O-octadecyl-L-ascorbic acid (Table I). Two other analogs, 2-O-octadecyl-L-ascorbic acid and 2-O-myristyl-L-ascorbic acid, were also tested, but these compounds decreased adherence of fibroblasts to tissue culture plates, making interpretation of results difficult. The addition of these same compounds to human neutrophils did not inhibit ascorbate transport without toxicity as measured by trypan blue dye exclusion (not shown).
D-Ascorbic acid, the enantiomer of L-ascorbic acid, was a poor inhibitor of ascorbate transport. L-Isoascorbic acid and D-isoascorbic acid, with stereochemical inversions at carbons 4 and 5, respectively, were only slightly more effective. 5,6-Isopropylidene-L-ascorbic acid was a moderate inhibitor, but inhibition was lost with the addition of a methoxymethyl group onto carbon 3. A series of 5-deoxy-and 6-deoxy-L-ascorbic acid analogs all effectively inhibited ascorbate transport (Table I). This series included 5-and 6-deoxy ascorbic acid, 5-deoxy and 6-deoxy-6-halo-L-ascorbic acids, and several 6-O-phenyl ethers. The 6-deoxy-6-halo compounds were by far the best inhibitors tested. The chloro, bromo, and iodo analogs were better inhibitors than L-ascorbate itself (Table I).
Based on these data, we examined inhibition kinetics of the 6-halo series in greater detail. All of the 6-deoxy-6-halo-L-ascorbic acid analogs inhibited ascorbic acid transport competitively (Fig. 1, A-E). For each analog, an inhibition constant (K i ) was determined by replotting the slopes of the lines against inhibitor concentration (Table II). Without analog, ascorbate was accumulated at least 4.7-fold against a concentration gradient (not shown).
Effects of the 6-halo series on dehydroascorbic acid transport were also examined. Incubations of fibroblasts with 40 M radiolabeled [ 14 C]dehydroascorbic acid in the presence or absence of 1 mM analog showed no difference in dehydroascorbic acid transport or ascorbate accumulation (not shown).
Lack of inhibition of ascorbate transport by analogs with alterations at position 2 or 3 suggests that these positions are needed for functional interaction with the ascorbate transporter. The specificity of the stereochemical configuration at carbons 4 and 5 is indicated by the minimal inhibition observed with the ascorbate stereochemical isomers D-ascorbate, D-isoascorbate, and L-isoascorbate. In contrast, substitutions of L-ascorbate at carbon 5 or 6 did not appreciably affect the ability of compounds to inhibit ascorbate transport. As noted above, the 6-deoxy-6-halo analogs had lower K i values for ascorbic acid transport than ascorbic acid itself. Other structurally related compounds, including 2-deoxyglucose, 3-Omethyl glucose, serine, leucine, and threonine, did not inhibit ascorbate transport (not shown).
Synthesis and Detection of [ 125 I]6-Deoxy-6-iodo-L-ascorbic Acid-As shown above, 6-deoxy-6-iodo-L-ascorbate was a potent and specific inhibitor of ascorbate transport. We chose this analog for further study because of two potential advantages in biochemical and cell studies. Because of the double bond at C-2-C-3, 6-deoxy-6-iodo-L-ascorbate had the possibility of being measured by HPLC with electrochemical detection. Of particular interest, it was conceivable that this compound could be made in radioiodinated form.
Both nonradiolabeled and 125 I-labeled 6-deoxy-6-iodo-Lascorbate were measured using HPLC with electrochemical detection. Also, the radiolabeled compound was separated by HPLC and quantitated using on-line scintillation spectrometry. Chromatograms of each compound are shown in Fig. 2. The purity of [ 125 I]6-deoxy-6-iodo-L-ascorbate was approximately 90% after synthesis as measured by HPLC. Two small peaks, representing potential degradation products, eluted at approximately 4.3 and 5 min. No detectable free iodide was measured during HPLC analysis. Specific activity of the radiolabeled analog was over 1000-fold higher than that of commercially   available 14 C-labeled ascorbate, resulting in a marked increase in measurement sensitivity. This is illustrated in Fig. 2C, where the HPLC profile of 35 pmol of [ 125 I]6-deoxy-6-iodo-Lascorbate is compared with that of 1000 pmol of [ 14 C]ascorbate. Using HPLC, we examined the stability of both radiolabeled and nonradiolabeled 6-deoxy-6-iodo-L-ascorbate in water; 60% methanol, 1 mM EDTA; and BSS. Stability was maintained for 6 h at 4°C, and Ͻ5% loss occurred after 2 h at 37°C (not shown). 0.1 mM DTT prevented loss at 37°C and was added in experiments examining kinetics of cellular uptake. Although degradation was minimal in 60% methanol, 1 mM EDTA at 4°C alone, some degradation occurred during cell extraction (not shown). Therefore, 1 mM DTT was added during cell extraction with 60% methanol/EDTA to prevent analog loss.
We determined whether transport and accumulation of 6-deoxy-6-iodo-L-ascorbate are sensitive to the presence of sodium, as is ascorbate transport. Fibroblasts were incubated with analog 5-130 M in balanced salt solution with or without sodium, which was replaced by equimolar amounts of choline and potassium. Without sodium, transport and accumulation of the ascorbate analog were inhibited as much as 95% (Fig. 3).
Transport kinetics of 6-deoxy-6-iodo-L-ascorbate were determined by incubating fibroblasts with analog 3.75-900 M for 1 h at 37°C (Fig. 4A). Data analysis demonstrated two component transport, an initial high affinity component and a sec-ondary linear component, within the range of analog concentrations used. The high affinity component was calculated by subtracting a line extrapolated from the linear portion of the curve (15). As calculated by Eadie-Hofstee transformation of these data, the apparent K m was 6.7 M and the apparent V max was 51.5 M/h (Fig. 4B). Calculation of transport kinetics by nonlinear curve fitting gave comparable results (not shown).
We predicted that transport properties of the radiolabeled analog would be similar to those of the unlabeled compound.

TABLE II
Inhibition constants (K i ) of 6-halo analogs for ascorbate transport Inhibition constants were determined by plotting the slopes of the lines in the double reciprocal plots (Fig. 1, A-E) versus inhibitor concentration. Compound K i M 6-Deoxy-L-ascorbic acid 140 6-Deoxy-6-fluoro-L-ascorbic acid 32 6-Deoxy-6-chloro-L-ascorbic acid 2.9 6-Deoxy-6-bromo-L-ascorbic acid 2.  3. Sodium dependence of 6-deoxy-6-iodo-L-ascorbate uptake in human fibroblasts. Fibroblasts were incubated with increasing concentrations of analog at 37°C for 1 h in the presence (q) or absence (f) of sodium ions. After incubations, cells were washed, and intracellular 6-deoxy-6-iodo-L-ascorbate was extracted using 60% methanol 1 mM EDTA as described under "Materials and Methods." Data are mean Ϯ S.D. expressed as M accumulation based on previously determined intracellular fibroblast volume (15).

FIG. 4. Apparent transport kinetics of 6-deoxy-6-iodo-L-ascorbate in human fibroblasts.
A, fibroblasts were incubated with analog for 1 h at 37°C in BSS medium, and intracellular accumulation was measured as described under "Materials and Methods." B, data plotted were derived from A by extrapolation of a line from the linear portion of the dose-response curve in A and subtraction from each data point. Eadie-Hofstee transformation (inset) of these data is also shown. Apparent transport kinetics derived in this manner were K m of 6.7 M and V max of 51.5 M/h. Transport and accumulation of [ 125 I]6-deoxy-6-iodo-L-ascorbate (10 or 20 M) in cultured fibroblasts were inhibited by L-ascorbate in a concentration-dependent fashion (Fig. 5A). Maximal inhibition of 87 and 81% for 10 and 20 M, respectively, occurred with an external concentration of 1 mM ascorbate. Transport was examined with and without sodium (Fig. 5B). Without sodium, [ 125 I]6-deoxy-6-iodo-L-ascorbate transport was inhibited as much as 93%.
Apparent transport kinetics of [ 125 I]6-deoxy-6-iodo-L-ascorbate were studied by incubating cultured fibroblasts with analog 3.75-800 M for 1 h at 37°C (Fig. 6). Similar to results with nonradiolabeled compound, data analysis demonstrated a two component transport process: an initial high affinity component and a later nonsaturating linear component. The slope of the linear portion was identical to that seen with transport without sodium (not shown). Derivation of the data, by subtracting a line extrapolated from the linear portion of the curve followed by Eadie-Hofstee transformation, resulted in an apparent K m of 5.2 M and an apparent V max of 17.2 M/h. Nonlinear curve fitting yielded similar results (not shown). High affinity sodium-dependent and low affinity sodium-independent [ 125 I]6-deoxy-6-iodo-L-ascorbate transport are consistent with ascorbate transport data in fibroblasts (15).
Dehydroascorbic Acid Transporters and [ 125 I]6-Deoxy-6-iodo-L-ascorbic Acid-Interpretations of experiments examining transport of ascorbic acid and dehydroascorbic acid are complicated by the reversible nature of substrate oxidation and reduction. The ability to distinguish ascorbate transport from dehydroascorbic acid transport with a specific analog would be a useful tool. As described above, the 6-haloascorbate analogs did not inhibit dehydroascorbic acid transport in fibroblasts. Unlike ascorbic acid, the 6-deoxy-6-haloascorbate analogs when oxidized cannot form a stable cyclic hemiketal structure in solution because of the absence of a free 6-OH group. If, as seems possible, it is this stabilized cyclic form of dehydroascorbic acid that is transported, oxidized 6-halo analogs, including [ 125 I]6-deoxy-6-iodo-L-ascorbate, should not be transported by mechanisms that transport dehydroascorbic acid.
Dehydroascorbic acid is transported by glucose transport isoforms GLUT1 or GLUT3 expressed in Xenopus laevis oocytes (12). To examine transport of [ 125 I]6-deoxy-6-iodo-L-ascorbate, oocytes expressing either GLUT1 or GLUT3 were incubated with [ 14 C]ascorbate or [ 125 I]6-deoxy-6-iodo-L-ascorbate in the presence or absence of the oxidizing agents hydrogen peroxide or xanthine/xanthine oxidase (Fig. 7, A and B). Consistent with previously reported data demonstrating dehydroascorbic acid transport by GLUT1 and GLUT3 (12,19), oxidation of [ 14 C]ascorbate by hydrogen peroxide or xanthine/xanthine oxidase resulted in a 6-or 13-fold increase in oocyte radiolabel uptake (Fig. 7A). In the absence of oxidation, transport of ascorbate was not different from that in control oocytes not  Fig. 3. Subsequent to incubations, cells were washed and solubilized with NaOH/CHAPS, and radioactivity was quantified by scintillation spectrometry as described under "Materials and Methods." Data represent mean Ϯ S.D.

FIG. 6. Apparent transport kinetics of [ 125 I]6-deoxy-6-iodo-Lascorbate in fibroblasts.
A, human fibroblasts were incubated with radiolabeled analog in BSS medium for 1 h at 37°C, and intracellular accumulated analog was measured as described in Fig. 5 and under "Materials and Methods." B, data plotted were derived from A by subtraction of an extrapolated line, as in Fig. 4. Eadie-Hofstee transformation of the derived data is also shown (inset). Calculated apparent transport kinetics were K m of 5.2 M with a V max of 17.2 M/h. expressing glucose transport proteins. Transport of [ 125 I]6-deoxy-6-iodo-L-ascorbate was also not different from control uninjected oocytes. In contrast to the results with [ 14 C]ascorbate, oxidation of [ 125 I]6-deoxy-6-iodo-L-ascorbate resulted in no increased uptake (Fig. 7B).
To verify our interpretation of these results, oxidation of [ 14 C]ascorbic acid and [ 125 I]6-deoxy-6-iodo-L-ascorbate during incubations was measured by HPLC. Because DTT was present in the [ 125 I]6-deoxy-6-iodo-L-ascorbate stock solution, all incubations were adjusted to contain similar amounts of DTT (37 M), and oxidants were adjusted accordingly to assure adequate substrate oxidation. Initial concentrations of [ 14 C]ascorbic acid and [ 125 I]6-deoxy-6-iodo-L-ascorbate were 50 M. Fifteen minutes after the addition of xanthine/xanthine oxidase, [ 14 C]ascorbic acid concentration decreased 60%, and [ 125 I]6deoxy-6-iodo-L-ascorbate concentration decreased 55% (not shown). After the addition of H 2 0 2, [ 14 C]ascorbic acid decreased 85%, and [ 125 I]6-deoxy-6-iodo-L-ascorbate decreased 90% (not shown). No oxidation occurred in control samples (not shown). Therefore, [ 125 I]6-deoxy-6-iodo-L-ascorbate and its oxidized product are not transported by either GLUT1 or GLUT3-mediated mechanisms of dehydroascorbic acid transport, irrespective of the oxidation state of the analog. DISCUSSION Goals of these experiments were to identify an ascorbate analog that is transported exclusively by ascorbate transporters, with no activity toward dehydroascorbic acid transport mechanisms. As a first step, the ability of ascorbate analogs to inhibit ascorbate transport was examined. These results provided insights into structure/activity relationships of transport inhibition. Inhibition of ascorbate transport required a C-4S absolute configuration in a five-member reduced ring with no substitutions on carbon 2 or 3. Alterations at positions 2-4 effectively eliminated or greatly reduced the ability of these compounds to interact with the ascorbate transport processes. In contrast, several compounds with substitutions on carbons 5 and 6 were inhibitors of ascorbate transport. Even analogs with comparatively large substituents, including 6-O-aryl ethers, were effective inhibitors. Of all analogs tested, 6-deoxy-6-halo-L-ascorbic acids were the best inhibitors. None of the 6-halo analogs inhibited dehydroascorbic acid transport.
These initial data suggested that 6-deoxy-6-iodo-L-ascorbic acid would be a promising candidate for further development.
To that end it was necessary to develop detection systems for this analog, characterize its transport characteristics as an ascorbate-like compound, and examine its selectivity toward ascorbate and dehydroascorbic acid transporters. We established a new method to quantitate 6-deoxy-6-iodo-L-ascorbic acid by HPLC with electrochemical detection. Sensitivity and specificity of quantitative detection of the analog by this method were comparable with the results obtained for ascorbic acid. This method was used to show that 6-deoxy-6-iodo-Lascorbate was accumulated in fibroblasts with transport kinetics comparable with those previously described for ascorbate (15). Ascorbate transport in fibroblasts has a Na ϩ -dependent high affinity component (K m of 6 M) (15), and 6-deoxy-6-iodo-L-ascorbic acid showed a Na ϩ -dependent high affinity component with similar kinetics (K m of 5.2 M).
The possibility of incorporating radioiodine into 6-deoxy-6iodo-L-ascorbate sparked our interest in developing this analog as a new tool to study ascorbate transport and function. Synthesis was performed by incorporating [ 125 I]iodide into methyl 2,3-isopropylidene-6-deoxy-6-iodo-L-gulonate, followed by acidcatalyzed rearrangement to the radiolabeled derivative. The specific activity of the final compound was approximately 1000fold greater than that of commercially available [ 14 C]ascorbate. Radiolabeled and nonradiolabeled analogs had identical HPLC elution times. Cellular transport of the radiolabeled compound was sodium-dependent, was inhibited by ascorbate, and had an apparent transport affinity similar to that of ascorbate.
Although stability of both radiolabeled and nonradiolabeled compounds was less than that of ascorbic acid, this was addressed by utilizing either low temperatures (Յ4°C during HPLC analysis) or by the addition of 0.1-1 mM DTT. Greater instability of these compounds may be related to their inability to form stable hemiketal structures after oxidation as discussed below.
Data from several laboratories (9, 10, 20 -22) strongly support the existence in cells of separate transport processes for ascorbate and dehydroascorbic acid. Because of the dual nature of ascorbate accumulation and because ascorbate can be oxidized to dehydroascorbic acid, experiments examining transport and accumulation of these compounds can be difficult to perform. Specific inhibitors of either transport activity would be very useful in the analysis of the relative roles of ascorbate and dehydroascorbic acid in cellular accumulation. Although glucose and similar molecules inhibit dehydroascorbic acid transport (23), in some cells glucose also inhibits ascorbate transport (24). In the current studies, the 6-halo analogs were very effective inhibitors of ascorbate transport but did not affect dehydroascorbic acid transport.
Dehydroascorbic acid exists in solution predominantly as a hydrated hemiketal (4), which has a half-life of approximately 6 min at pH 7.4 at 37°C (25). The 6-halo analogs do not have the necessary OH group required for cyclization to form this structure. Based on this consideration, we reasoned that oxidation of 6-dexy-6-iodo-L-ascorbic acid would lead to compounds that could not utilize transport mechanisms known to transport dehydroascorbic acid. To test this hypothesis, we expressed in Xenopus oocytes GLUT1 and GLUT3, glucose transport proteins that efficiently transport dehydroascorbic acid (12). Oxidation of 6-deoxy-6-iodo-L-ascorbic acid by either hydrogen peroxide or xanthine/xanthine oxidase failed to induce transport in this system. However, when ascorbate was oxidized similarly, transport of newly formed dehydroascorbic acid resulted in a 6 -13-fold increase in ascorbic acid accumulation. Thus, although 6-dexy-6-iodo-L-ascorbic acid was a potent competitive inhibitor of ascorbate transport, its oxidation did not lead to its transport and accumulation via pathways similar to those of dehydroascorbic acid. This observation demonstrates that 6-deoxy-6-iodo-L-ascorbic acid can be used to characterize the specific contribution of ascorbate transport pathways to total intracellular ascorbate accumulation.
In addition to sodium-dependent transport, ascorbate is transported in fibroblasts by a low affinity sodium-independent temperature-sensitive component (15). The low affinity transport activity could not be accounted for by diffusion and appeared to be carrier-mediated (15). In this paper, we found that 6-deoxy-6-iodo-L-ascorbate also displayed sodium-independent low affinity transport, which appeared linear up to approximately 1 mM external substrate concentration. These new findings with 6-iodoascorbate provide further evidence that the low affinity transport activity exists. One interpretation of this activity is that it is due to external ascorbate oxidation, dehydroascorbic acid transport, and internal reduction (15). The results presented here make this interpretation unlikely, because oxidation of 6-deoxy-6-iodo-L-ascorbate did not result in its transport in oocytes expressing GLUT1 or GLUT3. Additional mechanisms to explain the low affinity sodium-independent transport activity include secondary transport processes mediated by the sodium-dependent transport protein or transport mediated by a distinct protein. Experiments to test these possibilities may be feasible using 6-deoxy-6-iodo-L-ascorbate together with a functionally expressed ascorbate transporter, which has now become available (26).