Portal de Periódicos da CAPES

Erythrocytes Reduce Extracellular Ascorbate Free Radicals Using Intracellular Ascorbate as an Electron Donor

2000; Elsevier BV; Volume: 275; Issue: 36 Linguagem: Inglês

10.1074/jbc.m910281199

1083-351X

Martijn M. VanDuijn, Karmi Tijssen, John VanSteveninck, Peter J.A. Van den Broek, Jolanda Van der Zee,

Burn Injury Management and Outcomes

Ascorbate is readily oxidized in aqueous solution by ascorbate oxidase. Ascorbate radicals are formed, which disproportionate to ascorbate and dehydroascorbic acid. Addition of erythrocytes with increasing intracellular ascorbate concentrations decreased the oxidation of ascorbate in a concentration-dependent manner. Concurrently, it was found, utilizing electron spin resonance spectroscopy, that extracellular ascorbate radical levels were decreased. Control experiments showed that these results could not be explained by leakage of ascorbate from the cells, inactivation of ascorbate oxidase, or oxygen depletion. Thus, this means that intracellular ascorbate is directly responsible for the decreased oxidation of extracellular ascorbate. Exposure of ascorbate-loaded erythrocytes to higher levels of extracellular ascorbate radicals resulted in the detection of intracellular ascorbate radicals. Moreover, efflux of dehydroascorbic acid was observed under these conditions. These data confirm the view that intracellular ascorbate donates electrons to extracellular ascorbate free radical via a plasma membrane redox system. Such a redox system enables the cells to effectively counteract oxidative processes and thereby prevent depletion of extracellular ascorbate. Ascorbate is readily oxidized in aqueous solution by ascorbate oxidase. Ascorbate radicals are formed, which disproportionate to ascorbate and dehydroascorbic acid. Addition of erythrocytes with increasing intracellular ascorbate concentrations decreased the oxidation of ascorbate in a concentration-dependent manner. Concurrently, it was found, utilizing electron spin resonance spectroscopy, that extracellular ascorbate radical levels were decreased. Control experiments showed that these results could not be explained by leakage of ascorbate from the cells, inactivation of ascorbate oxidase, or oxygen depletion. Thus, this means that intracellular ascorbate is directly responsible for the decreased oxidation of extracellular ascorbate. Exposure of ascorbate-loaded erythrocytes to higher levels of extracellular ascorbate radicals resulted in the detection of intracellular ascorbate radicals. Moreover, efflux of dehydroascorbic acid was observed under these conditions. These data confirm the view that intracellular ascorbate donates electrons to extracellular ascorbate free radical via a plasma membrane redox system. Such a redox system enables the cells to effectively counteract oxidative processes and thereby prevent depletion of extracellular ascorbate. ascorbate free radical dehydroascorbic acid electron spin resonance phosphate-buffered saline tris(ethylenediamine)nickel(II) chloride 2-hydrate 4-hydroxy-2,2,6,6-tetramethylpiperidin-1-oxyl high pressure liquid chromatography Ascorbate, a well known antioxidant in biological systems, is capable of reducing a variety of oxidative compounds, especially free radicals (1Buettner G.R. Arch. Biochem. Biophys. 1993; 300: 535-543Crossref PubMed Scopus (2061) Google Scholar, 2Rose R.C. Bode A.M. FASEB J. 1993; 7: 1135-1142Crossref PubMed Scopus (421) Google Scholar). The importance of sufficient levels of ascorbate has been well established in the past (3Padh H. Biochem. Cell Biol. 1990; 68: 1166-1173Crossref PubMed Scopus (293) Google Scholar, 4Gershoff S.N. Nutr. Rev. 1993; 51: 313-326Crossref PubMed Scopus (126) Google Scholar). Primates and guinea pigs lack the ability to synthesize ascorbate from glucose and are dependent on dietary intake. Ascorbate deficiency may lead to scurvy and to oxidative injury resulting in necrosis or apoptosis. It may also cause malignant proliferation of cells as a consequence of oxidative DNA damage (5Block G. Am. J. Clin. Nutr. 1991; 54: 1310s-1314sCrossref PubMed Scopus (123) Google Scholar). The oxidation of ascorbate is a two-step reaction in which single electrons are transferred. The first step yields a relatively stable radical, the ascorbate free radical (AFR).1 In the second step, AFR donates a second electron, yielding dehydroascorbic acid (DHA). These steps are reversible, but DHA can irreversibly be hydrolyzed to diketo-gulonic acid, which degrades further to potentially toxic compounds (6Bianchi J. Rose R.C. Toxicology. 1986; 40: 75-82Crossref PubMed Scopus (47) Google Scholar, 7Pence L.A. Mennear J.H. Toxicol. Appl. Pharmacol. 1979; 50: 57-65Crossref PubMed Scopus (14) Google Scholar, 8Rose R.C. Choi J.L. Bode A.M. Life Sci. 1992; 50: 1543-1549Crossref PubMed Scopus (32) Google Scholar). The degradation of DHA is very fast, with a half-life of approximately 8 min at 37 °C (9Van Duijn M.M. Van der Zee J. VanSteveninck J. Van den Broek P.J.A. J. Biol. Chem. 1998; 273: 13415-13420Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar, 10Bode A.M. Cunningham L. Rose R.C. Clin. Chem. 1990; 36: 1807-1809Crossref PubMed Scopus (169) Google Scholar). Under oxidative stress, the consumption of ascorbate can be high, and without regeneration, ascorbate would soon be depleted. However, in order to restore ascorbate levels, systems exist that reduce AFR and/or DHA. Most of these systems are located in the cell. Reactions proceed spontaneously, such as glutathione-mediated ascorbate regeneration (11May J.M. Qu Z.C. Whitesell R.R. Cobb C.E. Free Radic. Biol. Med. 1996; 20: 543-551Crossref PubMed Scopus (171) Google Scholar), or are mediated by enzymes, such as thioredoxin reductase (12May J.M. Mendiratta S. Hill K.E. Burk R.F. J. Biol. Chem. 1997; 272: 22607-22610Abstract Full Text Full Text PDF PubMed Scopus (280) Google Scholar, 13May J.M. Cobb C.E. Mendiratta S. Hill K.E. Burk R.F. J. Biol. Chem. 1998; 273: 23039-23045Abstract Full Text Full Text PDF PubMed Scopus (163) Google Scholar, 14Mendiratta S. Qu Z.C. May J.M. Free Radic. Biol. Med. 1998; 25: 221-228Crossref PubMed Scopus (74) Google Scholar), glutaredoxin (15Park J.B. Levine M. Biochem. J. 1996; 315: 931-938Crossref PubMed Scopus (110) Google Scholar), and protein-disulfide isomerase (16Gilbert H.F. Methods Enzymol. 1998; 290: 26-50Crossref PubMed Scopus (120) Google Scholar). Moreover, an NADH:AFR reductase has been described that reduces intracellular AFR (17Diliberto Jr., E.J. Dean G. Carter C. Allen P.L. J. Neurochem. 1982; 39: 563-568Crossref PubMed Scopus (83) Google Scholar). These pathways only convert oxidized ascorbate species that are present in the cell. Regeneration of extracellular ascorbate is more complicated. Reduction of extracellular DHA can involve transport of DHA into the cell, e.g. by the GLUT-1 glucose transporter (9Van Duijn M.M. Van der Zee J. VanSteveninck J. Van den Broek P.J.A. J. Biol. Chem. 1998; 273: 13415-13420Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar, 18Vera J.C. Rivas C.I. Velasquez F.V. Zhang R.H. Concha I.I. Golde D.W. J. Biol. Chem. 1995; 270: 23706-23712Abstract Full Text Full Text PDF PubMed Scopus (157) Google Scholar, 19Rumsey S.C. Kwon O. Xu G.W. Burant C.F. Simpson I. Levine M. J. Biol. Chem. 1997; 272: 18982-18989Abstract Full Text Full Text PDF PubMed Scopus (399) Google Scholar, 20Goldenberg H. Schweinzer E. J. Bioenerg. Biomembr. 1994; 26: 359-367Crossref PubMed Scopus (78) Google Scholar). This regeneration pathway, however, cannot prevent depletion of extracellular ascorbate. Ascorbate is trapped inside the cell after reduction and will only slowly leak back to the extracellular space. It is, therefore, essential that other regenerating pathways exist. Recently, Himmelreich et al. (21Himmelreich U. Drew K.N. Serianni A.S. Kuchel P.W. Biochemistry. 1998; 37: 7578-7588Crossref PubMed Scopus (65) Google Scholar) reported that in erythrocytes, extracellular DHA can be reduced without entering the cell. Also, transmembrane AFR reductases have been described in the plasma membrane of liver cells (22Villalba J.M. Canalejo A. Rodriguez-Aguilera J.C. Buron M.I. Morre D.J. Navas P. J. Bioenerg. Biomembr. 1993; 25: 411-417Crossref PubMed Scopus (33) Google Scholar) and red blood cells (23May J.M. Qu Z.C. Cobb C.E. Biochem. Biophys. Res. Commun. 2000; 267: 118-123Crossref PubMed Scopus (52) Google Scholar, 24Goldenberg H. Grebing C. Low H. Biochem. Int. 1983; 6: 1-9PubMed Google Scholar). It was suggested that the latter reduce extracellular ascorbate radicals using intracellular NADH, thereby generating ascorbate and NAD+. Previous observations indicated that alternative systems might exist in the plasma membranes of cells. It was found that intracellular ascorbate acted as electron donor for transmembrane electron transport, with ferricyanide as extracellular electron acceptor (9Van Duijn M.M. Van der Zee J. VanSteveninck J. Van den Broek P.J.A. J. Biol. Chem. 1998; 273: 13415-13420Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar, 25May J.M. Qu Z.C. Whitesell R.R. Biochim. Biophys. Acta. 1995; 1238: 127-136Crossref PubMed Scopus (50) Google Scholar, 26Orringer E.P. Roer M.E.S. J. Clin. Invest. 1979; 63: 53-58Crossref PubMed Scopus (115) Google Scholar, 27May J.M. FASEB J. 1999; 13: 995-1006Crossref PubMed Scopus (185) Google Scholar). However, the actual substrate is still unknown, as ferricyanide is not a physiological compound. Because ascorbate is a natural compound in, for example, blood plasma, we hypothesized that under physiological conditions, AFR and/or DHA might act as the extracellular electron acceptor. Therefore, we investigated this hypothesis using erythrocytes as a model system. The data in this paper provide the first experimental evidence for the presence of a system in the erythrocyte membrane that reduces ascorbate free radicals in the extracellular space using intracellular ascorbate as an electron donor. Chemicals were obtained from Sigma or Baker, unless indicated otherwise. l-Ascorbate oxidase (EC 1.10.3.3) was purchased as sticks containing 17 units of enzyme (Roche Diagnostics, Almere, The Netherlands) and was dissolved in phosphate-buffered saline (PBS). Tris(ethylenediamine)-nickel(II) chloride 2-hydrate (Ni(en)32+) was synthesized according to State (28State H.M. Inorg. Synth. 1960; 6: 200-201Google Scholar). Erythrocytes were obtained from 1-day-old citrate-anticoagulated blood collected from healthy human volunteers by the Bloodbank Leiden/Haaglanden. The cells were washed three times with PBS. The buffy coat of white cells was removed carefully with each wash. Erythrocytes were loaded with ascorbate by resuspension of the cells to a hematocrit of 20% in PBS containing 2.5 mm adenosine as energy source (29May J.M. Qu Z.C. Whitesell R.R. Biochemistry. 1995; 34: 12721-12728Crossref PubMed Scopus (121) Google Scholar) and DHA at concentrations as indicated under "Results." Control erythrocytes were treated similarly but without DHA present. After 30 min of incubation at room temperature, erythrocytes were washed three times with PBS and used within 1 h for subsequent experiments. All experiments were performed in PBS at room temperature. Ascorbate concentrations in erythrocytes were determined by HPLC. Loaded, packed erythrocytes were mixed with 3 volumes of 7 mm potassium phosphate, 1 mm EDTA, pH 4.0, and frozen in liquid nitrogen. After thawing, hemoglobin was removed using an Amicon micropartition system with 30-kDa cutoff ultrafiltration membranes (Millipore, Etten-Leur, The Netherlands). 100 μl of 0.5 mm EDTA, 450 μl of methanol, and 5 μl of concentrated HCl were added to 200 μl of ultrafiltrate, and 100 μl of this mixture was injected on an Adsorbosphere SAX column (250 × 4.6 mm, Alltech, Breda, The Netherlands). Ascorbate was eluted with 7 mm potassium phosphate, 7 mm KCl, pH 4.0, at a rate of 1.5 ml/min. The LKB 2140 detector (LKB, Bromma, Sweden) was set at 265 nm, and chromatograms were integrated on a personal computer, using Nelson 2600 revised 4.1 software (Nelson Analytical, Cupertina, CA). After the elution of ascorbate, the column was regenerated with 0.25 m potassium phosphate and 0.5 m KCl, pH 5.0. Ascorbate concentrations are expressed relative to the water content of packed erythrocytes, which is 70% of the packed cell volume (26Orringer E.P. Roer M.E.S. J. Clin. Invest. 1979; 63: 53-58Crossref PubMed Scopus (115) Google Scholar). Ascorbate concentrations were determined by measuring the absorption at 265 nm in a Beckman DU-65 spectrophotometer (extinction coefficient, 14.500 m−1cm−1) or by HPLC analysis. For HPLC analysis, 200 μl of the solution was diluted with 100 μl of 0.5 mm EDTA, 450 μl of methanol, and 5 μl of concentrated HCl and subsequently injected and analyzed as described above. When present, erythrocytes were removed by centrifugation prior to the determination of ascorbate. Ascorbate depletion from erythrocytes was carried out by treating cells with 1 mm 4-hydroxy-2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPOL) as described by May et al. (30May J.M. Qu Z.C. Mendiratta S. Arch. Biochem. Biophys. 1998; 349: 281-289Crossref PubMed Scopus (283) Google Scholar) and Mehlhorn (31Mehlhorn R.J. J. Biol. Chem. 1991; 266: 2724-2731Abstract Full Text PDF PubMed Google Scholar). Briefly, 20% erythrocytes were incubated in PBS with 1 mm TEMPOL for 5 min at 37 °C. After subsequent centrifugation, this treatment was repeated twice, followed by three washes with PBS alone. NADH levels in erythrocytes were modulated by incubation of 10% erythrocytes with 10 mm pyruvate, xylitol, or glucose at 37 °C for 45 min. After centrifugation, samples were taken for the NAD/NADH assay. The remaining erythrocytes were resuspended in their respective buffers, including pyruvate, glucose, or xylitol, and immediately used for an ascorbate oxidation assay. NAD and NADH levels were determined using an alcohol dehydrogenase cycling assay modified from Wagner and Scott (32Wagner T.C. Scott M.D. Anal. Biochem. 1994; 222: 417-426Crossref PubMed Scopus (77) Google Scholar). Briefly, 200 μl of packed erythrocytes were diluted in 1.3 ml of extraction buffer, frozen in liquid nitrogen, thawed, and filtered using an Amicon micropartition system with 30-kDa cutoff ultrafiltration membranes. The filtrate was further analyzed as described by Wagner and Scott (32Wagner T.C. Scott M.D. Anal. Biochem. 1994; 222: 417-426Crossref PubMed Scopus (77) Google Scholar). ESR spectra were recorded on a JEOL-RE2X spectrometer operating at 9.36 GHz with a 100 kHz modulation frequency and equipped with a TM110 cavity. Samples were transferred to a quartz flat cell using a rapid sampling device, which allowed the recording of spectra to be started within seconds after mixing. ESR spectrometer settings were as follows: microwave power, 40 mW; modulation amplitude, 1 G; time constant, 0.3 s; scan time, 5 min; scan width, 15 G. Loading of erythrocytes withl-[carboxyl-14C]ascorbate (Amersham Pharmacia Biotech) was carried out by incubating a 20% suspension with 500 μm14C-labeled ascorbate and 1.7 units/ml ascorbate oxidase. After 30 min, the erythrocytes were washed three times with PBS and resuspended at 10% hematocrit. After the experiment, the suspensions were centrifuged, and the supernatants were transferred to vials for liquid scintillation counting and for immediate HPLC analysis, as described above. Fractions were collected and analyzed by liquid scintillation counting. The scintillation mixtures used were Emulsifier Scintillator 299 and Flo-Scint IV (Packard). 14C-Labeled DHA was prepared by incubation of [14C]ascorbate with bromine, according to Washkoet al. (33Washko P.W. Wang Y. Levine M. J. Biol. Chem. 1993; 268: 15531-15535Abstract Full Text PDF PubMed Google Scholar), with every step on ice. Immediately after preparation, 1 mm14C-labeled DHA was added to either PBS or to a 10% suspension of ascorbate-loaded or control erythrocytes. After incubation, supernatants were collected, separated by HPLC, and analyzed by liquid scintillation counting. Oxygen consumption was measured using a YSI 5300 biological oxygen monitor (YSI Inc., Yellow Springs, OH). All experiments were performed at least three times and were evaluated using Student's t test where applicable. Ascorbate can easily be oxidized in aqueous solution by spontaneous, metal-ion, or enzyme-catalyzed reactions. In this study, the enzyme ascorbate oxidase was used to oxidize ascorbate. Incubation of ascorbate with ascorbate oxidase resulted in a decrease of the ascorbate concentration that was linear for 15 min (Fig. 1). After 60 min, approximately 60% of the ascorbate was oxidized. Ascorbate oxidation decreased upon addition of erythrocytes, and decreased even further by addition of erythrocytes, loaded with ascorbate. The addition of cells with high levels of intracellular ascorbate resulted in complete protection (Fig. 1). This suggested that intracellular ascorbate played an important role in the effect of erythrocytes on the oxidation of ascorbate. To test this, ascorbate-loaded cells were treated with TEMPOL, which depleted intracellular ascorbate by more than 90% (data not shown). As expected, the effect of ascorbate-loaded cells on ascorbate oxidation could be reversed by treatment with TEMPOL (Fig.1). In subsequent experiments, erythrocytes were preincubated with various concentrations of DHA to determine the correlation between the intracellular ascorbate concentration and extracellular ascorbate oxidation. Fig. 2 shows that the rate of ascorbate oxidation in the external medium was strongly dependent on the intracellular ascorbate concentration: with increasing intracellular ascorbate concentrations, the oxidation of extracellular ascorbate decreased.Figure 2Correlation between the intracellular ascorbate concentration and the oxidation rate of extracellular ascorbate. Erythrocytes were loaded with ascorbate by treating cells with various concentrations of DHA, as described under "Experimental Procedures." Subsequently, a 10% suspension was incubated with 100 μm ascorbate and 4 milliunits/ml ascorbate oxidase. After 0 and 15 min, the amount of ascorbate in the supernatant was determined by measuring A 265 and by HPLC, and the rate of ascorbate oxidation was calculated. Results ofA 265 measurements are presented as the average of three experiments ± S.E. Measurements using the HPLC method are not shown, but they produced similar data. The intracellular ascorbate concentration was determined as described under "Experimental Procedures." Inset, intracellular ascorbate concentration resulting from preincubation with various concentrations of DHA.View Large Image Figure ViewerDownload (PPT) Fig. 1 shows that addition of control erythrocytes (i.e. not loaded with ascorbate) decreased ascorbate oxidation by 30–40%. To test whether NADH contributed to this effect, erythrocytes were incubated with various compounds that are known to perturb the NADH/NAD+ ratio (34Alvarez J. Camaleno J.M. Garcia-Sancho J. Herreros B. Biochim. Biophys. Acta. 1986; 856: 408-411Crossref PubMed Scopus (8) Google Scholar, 35Momsen G. Arch. Biochem. Biophys. 1981; 210: 160-166Crossref PubMed Scopus (20) Google Scholar). A decrease in the intracellular NADH concentration resulted in an increase in the ascorbate oxidation rate, whereas an increase in NADH resulted in a concomitant decrease in the ascorbate oxidation rate (Table I). This shows that NADH, as well as ascorbate, plays a role in the regeneration of extracellular ascorbate. To measure the relative contribution of endogenous ascorbate to the effect of control erythrocytes on ascorbate oxidation, cells were treated with TEMPOL. However, it was found that TEMPOL also reduced intracellular NADH levels (Table I). It was, therefore, not possible to separate the effects of endogenous ascorbate and NADH. It should be noted that TEMPOL depleted NADH only in the case of control erythrocytes, and not in erythrocytes loaded with ascorbate.Table IEffect of energy sources and TEMPOL on NADH levels and the ascorbate oxidation rateAdditionsAscorbate oxidation rateNADHμm/min% reduced nucleotideNone1.30 ± 0.0526 ± 7Adenosine1.06 ± 0.0978 ± 5Glucose1.08 ± 0.1052 ± 9Xylitol1.14 ± 0.0758 ± 9Pyruvate1.39 ± 0.0419 ± 5TEMPOL1.39 ± 0.0614 ± 5Erythrocytes that had not been loaded with ascorbate were preincubated with the additions listed. Subsequently, NADH levels and oxidation rates of extracellular ascorbate were determined. All treatments and assays were performed as described under "Experimental Procedures." NADH levels are expressed as a percentage of total nucleotides,i.e. NAD+ + NADH (48 nmol/μl packed cells). All results are shown as an average ± S.E. Open table in a new tab Erythrocytes that had not been loaded with ascorbate were preincubated with the additions listed. Subsequently, NADH levels and oxidation rates of extracellular ascorbate were determined. All treatments and assays were performed as described under "Experimental Procedures." NADH levels are expressed as a percentage of total nucleotides,i.e. NAD+ + NADH (48 nmol/μl packed cells). All results are shown as an average ± S.E. The effect of erythrocytes on ascorbate oxidation could be explained by inhibition of ascorbate oxidase by the erythrocytes. To study this hypothesis, control or ascorbate-loaded erythrocytes were incubated with ascorbate oxidase for 0 or 15 min. After removal of the erythrocytes, 100 μm ascorbate was added to the supernatants, and the rate of ascorbate oxidation was determined. It was found that incubation of ascorbate oxidase with control or ascorbate-loaded erythrocytes did not affect the activity of the enzyme. Oxygen is also required for ascorbate oxidase activity. Under our experimental conditions, oxygen levels did not decrease more than 20%, which showed that oxygen levels were not limiting during the experiments (data not shown). A possible explanation for the protection by ascorbate-loaded erythrocytes could be leakage of ascorbate from the erythrocytes. More than 1 mm ascorbate was accumulated intracellularly after incubation with DHA (Fig. 2, inset). Leakage of the accumulated ascorbate could result in an extracellular ascorbate concentration of up to 100 μm. To investigate leakage of intracellular ascorbate, erythrocytes were loaded with14C-labeled ascorbate. TableII shows that incubation of these erythrocytes caused some leakage of 14C-labeled material, both in control cells and after addition of ascorbate, DHA, or ascorbate oxidase. The addition of 100 μm ascorbate and 4 milliunits of ascorbate oxidase/ml resulted in a small increase in leakage of radioactivity. This amounted to an extracellular concentration of 3 μm14C-labeled material. From the data in Fig. 2, it can be deduced that complete protection against oxidation of ascorbate would need leakage of 20 μm ascorbate. This means that efflux of ascorbate could maximally account for 15% of the observed effect. Only in the presence of a higher ascorbate oxidase concentration did leakage become substantial. Furthermore, HPLC analysis revealed that most of the radioactivity was released from the cells as [14C]DHA, and not as [14C]ascorbate (Fig.3). Leakage of DHA could be inhibited by the addition of cytochalasin B, an inhibitor of the GLUT-1 transporter (Table II). Thus, it can be concluded that leakage of ascorbate is not responsible for the decrease in ascorbate oxidation.Table IILeakage of radioactivity from erythrocytes containing [ 14 C]ascorbateExtracellular additionsLeakage% of total contentNone1.1 ± 0.3100 μm ascorbate1.3 ± 0.1100 μmDHA1.5 ± 0.14 milliunits/ml ascorbate oxidase1.2 ± 0.214 milliunits/ml ascorbate oxidase0.8 ± 0.1100 μm ascorbate, 4 milliunits/ml ascorbate oxidase3.0 ± 0.52-ap < 0.01 compared to control using Student's t test.100 μmascorbate, 14 milliunits/ml ascorbate oxidase12.4 ± 1.22-ap < 0.01 compared to control using Student's t test.100 μmascorbate, 4 milliunits/ml ascorbate oxidase, 20 μmcytochalasin B1.8 ± 0.3100 μm ascorbate, 14 milliunits/ml ascorbate oxidase, 20 μm cytochalasin B4.2 ± 0.42-ap < 0.01 compared to control using Student's t test.Erythrocytes were loaded with 14C-labeled ascorbate, as described under "Experimental Procedures," washed, and incubated as a 10% suspension with the additions indicated in the table. Leakage after 15 min was determined by centrifugation of the suspensions. Radioactivity in the supernatant was determined by liquid scintillation counting. Values are shown as mean ± S.E. (n = 8), relative to the amount of 14C that would be released by lysis of all cells.2-a p < 0.01 compared to control using Student's t test. Open table in a new tab Erythrocytes were loaded with 14C-labeled ascorbate, as described under "Experimental Procedures," washed, and incubated as a 10% suspension with the additions indicated in the table. Leakage after 15 min was determined by centrifugation of the suspensions. Radioactivity in the supernatant was determined by liquid scintillation counting. Values are shown as mean ± S.E. (n = 8), relative to the amount of 14C that would be released by lysis of all cells. The oxidation of ascorbate by ascorbate oxidase results in the formation of ascorbate free radical, which can easily be detected by ESR spectroscopy. The ESR spectrum consists of a doublet with hyperfine splitting, a H4 = 1.8 G (Fig. 4 A). Control experiments showed that AFR peak intensities were constant for at least 15 min. The data in Fig. 4 were obtained with a scan time of 5 min and therefore represent steady state levels of AFR. No signal was observed when ascorbate was omitted from the incubation mixture (data not shown). Removal of ascorbate oxidase resulted in a small AFR signal, which amounted to 15% of the signal in Fig. 4 A (data not shown). This signal was due to autoxidation and transition metal-mediated oxidation of ascorbate (36Van Duijn M.M. Van der Zee J. Van den Broek P.J.A. Protoplasma. 1998; 205: 122-128Crossref Scopus (17) Google Scholar). Addition of erythrocytes decreased AFR signal intensity by 10%, whereas the addition of ascorbate-loaded erythrocytes decreased the AFR concentration considerably (compare Fig. 4, A, C, and E). In the presence of ascorbate-loaded erythrocytes, AFR could be generated both inside and outside the cell. To distinguish between intra- and extracellular AFR, the line broadening compound Ni(en)32+ was used. This compound induces line broadening of the AFR signal, as is illustrated in Fig.4 B, without influencing its redox properties (37Wakefield L.M. Cass A.E. Radda G.K. J. Biol. Chem. 1986; 261: 9746-9752Abstract Full Text PDF PubMed Google Scholar). Because Ni(en)32+ cannot cross the plasma membrane, only the intensity of the extracellular AFR signal will be affected. Addition of 5 mmNi(en)32+ to an incubation of control or ascorbate-loaded erythrocytes in the presence of ascorbate and ascorbate oxidase resulted in the line broadening of the AFR signal (Fig. 4, D and F). This shows that the AFR signal detected under these experimental conditions came from ascorbate located outside the cell. To determine the correlation between the intracellular ascorbate concentration and extracellular AFR, cells were preincubated with various DHA concentrations. Fig. 5 shows the effect of increasing intracellular ascorbate concentrations on external AFR signal intensities. Ascorbate-loaded erythrocytes reduced the AFR signal in the extracellular medium in a concentration-dependent manner. The effect was similar to the effect observed in Fig. 2. So far, our data seem to indicate that extracellular AFR is reduced by intracellular ascorbate. If the hypothesis is correct that intracellular ascorbate donates an electron to extracellular AFR, it can be expected that intracellular AFR is generated as a consequence of this reaction. The data in Fig. 4suggest that under the present conditions, only extracellular AFR can be observed. The absence of an intracellular AFR signal is most likely due to the efficient regeneration of ascorbate in the erythrocyte. To overwhelm the reductive capacity of the cell, extracellular AFR levels were increased using high concentrations of ascorbate and ascorbate oxidase. Incubation of 1 mmascorbate with 20 milliunits/ml ascorbate oxidase in the presence of 5 mm Ni(en)32+ resulted in a broadened AFR signal (Fig.6 A). Subsequently, erythrocytes were added (20% suspension), and the spectrum in Fig.6 B was obtained, which is identical to the spectrum obtained without erythrocytes. When ascorbate-loaded erythrocytes were used, on the other hand, the ESR spectrum showed additional, slightly sharper peaks (Fig. 6 C). This indicates that an additional AFR signal was formed that could not be broadened by Ni(en)32+. Subtraction of curve A (Fig.6) from curve C generated the spectrum given in Fig. 6 D,which is typical for AFR in the absence of Ni(en)32+ (simulated in Fig.6 E). This shows that at high levels of extracellular AFR, intracellular AFR can indeed be detected. Although the ESR experiments indicated that extracellular AFR was reduced by the erythrocytes, we also investigated whether reduction of extracellular DHA played a role. [14C]DHA was added either to PBS or to a 10% suspension of ascorbate-loaded or control erythrocytes. After 10 min, the supernatants were collected and injected on a HPLC system. A small peak at the retention time of ascorbate was found to develop during the incubation of DHA alone (Fig. 7). The presence of erythrocytes resulted in a slightly larger peak, but no differences were found between ascorbate-loaded and control erythrocytes. Addition of erythrocytes with increasing intracellular ascorbate concentrations decreased the oxidation rate of extracellular ascorbate by ascorbate oxidase (Fig. 1). At intracellular concentrations of 1 mm, ascorbate degradation was even completely prevented (Fig. 2). Removal of intracellular ascorbate by TEMPOL reversed the process (Fig. 1). In control experiments, inactivation of ascorbate oxidase and leakage of intracellular ascorbate could be eliminated as explanations for these findings. Thus, intracellular ascorbate directly influences the oxidation rate of extracellular ascorbate, even in the physiological range of 20–60 μm ascorbate (Fig. 2) (38Evans R.M. Currie L. Campbell A. Br. J. Nutr. 1982; 47: 473-482Crossref PubMed Scopus (178) Google Scholar). Addition of control erythrocytes, i.e. erythrocytes with only endogenous ascorbate, decreased the ascorbate oxidation by 30–40%. NADH also contributed to this effect, as the ascorbate oxidation rate was readily affected by changes in the concentration of NADH in the erythrocytes (Table I). Unfortunately, the relative contribution of endogenous ascorbate and NADH could not be established experimentally, as depletion of endogenous ascorbate by TEMPOL also caused depletion of NADH (Table I). Nevertheless, it can be concluded that the degradation of extracellular ascorbate is prevented by reactions that are driven by intracellular NADH and ascorbate. In erythrocytes loaded with ascorbate, the ascorbate-driven reaction prevails. Ascorbate oxidase produces AFR from ascorbate in a direct reaction, followed by disproportionation of two AFR molecules to one molecule of ascorbate and one molecule of DHA (39Yamazaki I. Piette L.H. Biochim. Biophys. Acta. 1961; 50: 62-69Crossref PubMed Scopus (161) Google Scholar). In order to restore ascorbate levels, either AFR or DHA has to be reduced to ascorbate. There have been reports on the reduction of extracellular DHA by K-562 cells (40Schweinzer E. Mao Y. Krajnik P. Getoff N. Goldenberg H. Cell Biochem. Funct. 1996; 14: 27-31Crossref PubMed Scopus (13) Google Scholar) and by erythrocytes (21Himmelreich U. Drew K.N. Serianni A.S. Kuchel P.W. Biochemistry. 1998; 37: 7578-7588Crossref PubMed Scopus (65) Google Scholar), but our data did not indicate an important role for this process. A small amount of ascorbate or an ascorbate-like compound was formed on incubation of DHA in PBS (Fig. 7). Jung and Wells (41Jung C.H. Wells W.W. Arch. Biochem. Biophys. 1998; 355: 9-14Crossref PubMed Scopus (45) Google Scholar) described a degradation product of DHA that can reduce DHA to ascorbate. In addition, degradation of DHA results in erythroascorbic acid, an analogue of ascorbate that could elute very similar in our HPLC system. These processes could contribute to the formation of the ascorbate-like peak at 5.5 min. The peak increased slightly in the presence of erythrocytes, irrespective of the presence of intracellular ascorbate. This small increase may be caused by a NADH-dependent redox system. Thus, although our data indicate that erythrocytes reduce some extracellular DHA, it is evident that this does not depend on the intracellular ascorbate concentration. Reduction of ascorbate radical levels could be established using ESR spectroscopy (Figs. 4 and 5). A correlation was found between increasing intracellular ascorbate concentrations and the decrease in extracellular ascorbate oxidation, reflected either by the absorbance of ascorbate in the extracellular medium or by steady state levels of AFR (Figs. 2 and 5). This suggests that there is a close relationship between the intracellular ascorbate concentration and the reduction of extracellular oxidation products. A plausible explanation is that intracellular ascorbate supplies an electron that reduces extracellular AFR to extracellular ascorbate. In this process, intracellular ascorbate will be oxidized to an ascorbate radical. Either this radical can be reduced to ascorbate via intracellular NADH-AFR reductases or two radicals can disproportionate to ascorbate and DHA. At first, we were unable to detect intracellular AFR. This can be explained by the fact that with low levels of extracellular AFR, the cell is able to regenerate intracellular ascorbate efficiently, and intracellular levels of AFR will be too low to be detected. Higher levels of AFR will cause a more severe oxidative stress over the plasma membrane, and intracellular regeneration of ascorbate might be overwhelmed. This was indeed the case. Using Ni(en)32+ to line broaden extracellular AFR signals, a small intracellular AFR signal could be detected (Fig.6). This strongly suggests that intracellular ascorbate donates an electron to extracellular AFR, forming extracellular ascorbate and intracellular AFR. The efflux of 14C-labeled material from the erythrocytes provided additional evidence (Fig. 3 and Table II). HPLC analysis of the extracellular medium revealed that 14C-labeled material emerged as [14C]DHA, rather than [14C]ascorbate (Fig. 3). The efflux of14C-labeled material was increased in the presence of ascorbate and ascorbate oxidase. This confirmed that external AFR induced an oxidative stress over the membrane, which resulted in the oxidation of intracellular ascorbate. At high levels of AFR, the production of DHA exceeds the capacity of the cell for regeneration to ascorbate, resulting in efflux of DHA via the GLUT-1 transporter (TableII). Although there is good evidence for ascorbate-dependent electron transport across the membrane, it is still not clear what the functional components of this system are. Possible mechanisms are transport chains of a physicochemical nature, e.g. an α-tocopherol electron shuttle, and protein-mediated systems such as an integral plasma membrane protein, complexes of multiple proteins, or proteins using a co-factor to traverse the membrane. However, the standard reduction potentials (Eo′) of α-tocopherol and ascorbate with their respective radicals differ by more than 200 mV, which makes the transfer of an electron from α-tocopherol to AFR unlikely (1Buettner G.R. Arch. Biochem. Biophys. 1993; 300: 535-543Crossref PubMed Scopus (2061) Google Scholar). Another putative co-factor is coenzyme Q, but its capability to move freely across the lipid bilayer has been questioned (27May J.M. FASEB J. 1999; 13: 995-1006Crossref PubMed Scopus (185) Google Scholar, 42Ulrich E.L. Girvin M.E. Cramer W.A. Markley J.L. Biochemistry. 1985; 24: 2501-2508Crossref PubMed Scopus (77) Google Scholar). Moreover, capsaicin and dicumarol, inhibitors of coenzyme Q-mediated electron transport, did not inhibit the decrease in the level of AFR in the presence of ascorbate-loaded erythrocytes (data not shown). The most likely form appears to be a single protein or a protein complex, accepting electrons from intracellular ascorbate and transporting them to extracellular AFR. Interestingly, a cytochrome with a similar function has been described in chromaffin cells from the adrenal medulla. In chromaffin vesicles inside these cells, norepinephrine is synthesized at the expense of large amounts of ascorbate. A cytochrome b 561 is present in the vesicle membrane, transporting electrons from cytoplasmic ascorbate to intravesicular AFR (43Srivastava M. Duong L.T. Fleming P.J. J. Biol. Chem. 1984; 259: 8072-8075Abstract Full Text PDF PubMed Google Scholar, 44Wakefield L.M. Cass A.E. Radda G.K. J. Biol. Chem. 1986; 261: 9739-9745Abstract Full Text PDF PubMed Google Scholar). A similar or identical cytochrome could explain the ascorbate:AFR oxidoreductase activity in the erythrocyte. Indeed, evidence was found for the presence of a b cytochrome in the erythrocyte membrane with spectral characteristics similar to those of cytochromeb 561. 2M. M. VanDuijn, J. T. Buijs, J. Van der Zee, and P. J. A. Van den Broek, submitted for publication. However, it has been reported that cytochromeb 561 is not present in the erythrocyte (45Pruss R.M. Shepard E.A. Neuroscience. 1987; 22: 149-157Crossref PubMed Scopus (31) Google Scholar). This was confirmed in recent work by our group.2 It is conceivable that a cytochrome similar to cytochromeb 561 is responsible for the reduction of AFR by the erythrocyte, but so far, no proteins with structural homology to cytochrome b 561 have been described. The various processes in the erythrocyte are summarized in Fig.8. Extracellular AFR is reduced by intracellular ascorbate via an ascorbate:AFR reductase (1). Concurrently, intracellular ascorbate is oxidized to AFR, which can be reduced to ascorbate via intracellular NADH-AFR reductases or disproportionate to ascorbate and DHA. Efflux of DHA proceeds through the GLUT-1 transporter (3), which facilitates bidirectional transport of DHA over the plasma membrane. In addition, extracellular AFR can be reduced by a NADH-dependent AFR reductase (2). Presently, it is not known whether the ascorbate-dependent AFR reductase discussed in this paper is related in any way to other AFR reductases known from the literature. Generation of intracellular AFR, as well as efflux of DHA, was found by May et al. (11May J.M. Qu Z.C. Whitesell R.R. Cobb C.E. Free Radic. Biol. Med. 1996; 20: 543-551Crossref PubMed Scopus (171) Google Scholar, 29May J.M. Qu Z.C. Whitesell R.R. Biochemistry. 1995; 34: 12721-12728Crossref PubMed Scopus (121) Google Scholar) upon incubation of erythrocytes with ferricyanide. However, recently May and Qu (46May J.M. Qu Z.C. Biochim. Biophys. Acta. 1999; 1421: 19-31Crossref PubMed Scopus (40) Google Scholar) observed that pCMBS partly inhibited ascorbate-dependent ferricyanide reduction in erythrocytes, whereas it did not influence ferric iron reduction. We were unable to find any effect of pCMBS on the AFR reductive capacity (data not shown). This suggests that AFR and ferricyanide are not reduced by the same reductase, but it could imply that the erythrocyte AFR reductase also catalyzes ferric iron reduction. Further studies are, however, required to identify the transmembrane reductase or reductases that are responsible for the ascorbate-dependent reduction of extracellular oxidants.