Human cell lines regulate the redox state (Eh) of the cysteine/cystine (Cys/CySS) couple in culture medium to approximately −80 mV, a value similar to the average Eh for Cys/CySS in human plasma. The mechanisms involved in regulation of extracellular Eh of Cys/CySS are not known, but GSH is released from tissues at rates proportional to tissue GSH concentration, and this released GSH could react with CySS to contribute to maintenance of this balance. The present study was undertaken to determine whether depletion of cellular GSH alters regulation of extracellular Cys/CySS Eh. Decrease of GSH in HT-29 cells by inhibiting synthesis with l-buthionine-[S,R]-sulfoximine showed no effect on the rate of reduction of extracellular CySS to achieve a stable Eh for Cys/CySS in the culture medium. Limiting Cys and CySS in the culture medium also substantially decreased cellular GSH but resulted in no significant effect on extracellular Cys/CySS Eh. Addition of CySS to these cells showed that extracellular Cys/CySS Eh approached −80 mV at 4 h while cellular GSH and extracellular GSH/GSSG Eh recovered more slowly. Together, these results show that HT-29 cells have the capacity to regulate the extracellular Cys/CySS Eh by mechanisms that are independent of cellular GSH. The results suggest that transport systems for Cys and CySS and/or membranal oxidoreductases could be more important than cellular GSH in regulation of extracellular Cys/CySS Eh.
- oxidative stress
- thiol/disulfide redox control
- amino acid deficiency
- amino acid transport
cysteine (Cys) is usually a nonessential amino acid in the diet of mammals because dietary methionine (Met) is a metabolic precursor for cysteine and, if present in an adequate amount, can completely support the Cys requirement (10). However, the conversion of Met to Cys does not occur in most mammalian cells so that Cys, or its disulfide form, cystine (CySS), is needed to support individual cellular requirements (32). CySS is the predominant form in the plasma, and extensive research has characterized Cys and CySS transport systems (2) and utilization of Cys and CySS for cell growth (3, 27) and protection against oxidative stress (4). Inhibition of Cys incorporation into GSH by l-buthionine-[S,R]-sulfoximine (BSO) sensitizes many cell types to anticancer therapy (6), and supply of N-acetyl-l-cysteine and other Cys precursors is a common approach to enhance antioxidant defenses (21).
In addition to the understanding of the utilization of the specific chemical forms, Cys and CySS, recent data show that the oxidation-reduction potential (hereafter termed “redox state” 1) of the Cys/CySS couple, representing the electromotive force relative to a standard hydrogen electrode (EhCySS), can have important effects on cell phenotype. For instance, human colon epithelial (Caco-2) cells proliferate more rapidly at a more reduced EhCySS in cell culture media under conditions in which the cellular GSH/GSSG redox state (EhGSSG) is unaffected (13). Cell growth signaling through a MAP kinase pathway was found to be sensitive to a thiol reagent that was impermeant to cells, indicating control by an extracellular redox-sensitive protein (25). The redox-sensitive pathway was inhibited by a metalloproteinase inhibitor and by a blocking antibody to TGF-α, showing that cell proliferation can be signaled by extracellular Cys/CySS redox state through effects on extracellular protein thiols. In contrast to stimulation of proliferation by a more reduced Eh, binding of human THP-1 monocytes to vascular endothelial cells, an early step of atherogenesis, was increased at more oxidized extracellular EhCySS (8). The mechanism involved transcriptional activation of cell adhesion molecules (ICAM, platelet endothelial cell adhesion molecule, P-selectin) signaled by NF-κB. NF-κB was activated by an oxidized extracellular EhCySS through increased reactive oxygen species (ROS) under conditions in which cellular GSH was unaffected. EhCySS-dependent ROS generation was inhibited by nonpermeant thiol-reactive agents, indicating that the redox dependence was due to an extracellular redox-sensitive protein. Together, these data show that variation in extracellular EhCySS can be an important determinant of cell function.
The Cys/CySS redox couple is the most abundant low-molecular-weight thiol/disulfide system in human plasma, with a total concentration in Cys equivalents ranging from ∼100 μM in young healthy adults to >200 μM in older individuals (18). Although the EhGSSG is about −230 mV in tissues (23) and the Eh for reduction of O2 to H2O is >600 mV, the EhCySS in plasma of young healthy individuals is −80 ± 9 mV (16). The relatively small variation for EhCySS despite the large difference between the GSH/GSSG and H2O/O2 couples indicates that mechanisms exist to control extracellular EhCySS (7).
GSH is exported from liver and other tissues by multiple transport systems (28), and in liver this transport rate is dependent on cellular concentration (1) and largely accounts for GSH turnover. Reed and Beatty (29) showed that GSH efflux from hepatocytes reduced extracellular CySS to generate Cys, thereby providing a mechanistic basis for regulation of extracellular EhCySS by the cellular GSH/GSSG couple. Moreover, the EhGSSG in plasma of young healthy adults (−138 ± 9 mV) was found to be intermediate between tissue EhGSSG and plasma EhCySS, further suggesting that release of tissue GSH could serve as a mechanism to reduce the extracellular EhCySS (16). Thus cellular release of GSH could provide a mechanism to maintain extracellular EhCySS. If so, this could provide an important connection between tissue oxidative stress and phenotypic effects resulting from variations in plasma EhCySS.
The purpose of the present study was to determine whether regulation of extracellular EhCySS is dependent on cellular GSH. Experiments were performed with HT-29 cells, a moderately differentiated human colon carcinoma cell line in which cellular EhGSSG and EhCySS have been characterized under different conditions of growth, inhibition of GSH synthesis, and culture with Cys- and CySS-limiting media (19, 22). In the present study, cellular GSH was varied by inhibition of the first enzyme in the GSH biosynthetic pathway, glutamate:cysteine ligase, with BSO and by culture of cells in media with limiting concentrations of Cys and CySS (22). Each treatment caused extensive loss of cellular GSH but had no effect on the rate of reduction of the extracellular Cys/CySS couple. The data show that in this cell model the extracellular EhCySS is not regulated by the cellular GSH pool. These results imply that extracellular EhCySS in vivo, such as that measured in plasma, is not likely to reflect tissue GSH status but rather to reflect a balance of other processes, possibly including transport and oxidation/reduction.
MATERIALS AND METHODS
Except as indicated, all chemicals were purchased from Sigma (St. Louis, MO). Distilled, deionized water was used throughout. HPLC-quality solvents were used for HPLC.
HT-29 cells were obtained from American Type Culture Collection (Gaithersburg, MD) and maintained at 37°C and 95% air-5% CO2 in McCoy's 5A complete medium, 10% (vol/vol) fetal bovine serum, and 5% (vol/vol) penicillin-streptomycin (Atlanta Biologicals; Atlanta, GA). To examine the effect of inhibition of GSH synthesis by BSO on the rate of reduction of extracellular cystine, HT-29 cells were pretreated with 0, 10, or 100 μM BSO for 24 h. At this time, cells were washed once with phosphate-buffered saline (PBS), and medium was changed to Cys-free McCoy's medium supplemented with CySS and including BSO at the same concentrations as used during the preincubation. Aliquots of the medium were removed and analyzed at 0, 1, 4, or 12 h after medium change.
For experiments with Cys deficiency, cells were seeded in McCoy's complete medium on 60-mm plates at a density of 6 × 105 cells/plate and maintained for 24 h. At this time, cells were washed once with PBS and treated with Cys-free McCoy's medium for 48 h; time-matched controls received McCoy's medium. To examine recovery after culture in Cys-free medium, medium was changed to Cys-free McCoy's medium with addition of 100 μM CySS; controls similarly received Cys-free McCoy's medium with addition of 100 μM CySS. Samples were harvested and analyzed immediately and at 4, 12, and 24 h after the last medium change. Processing time for Cys deficiency experiments was 10 min, so initial measurements do not closely reflect added concentrations. Under all conditions, i.e., BSO treatments and Cys-deficient medium, cell viability was >99%.
Analysis of GSH, GSSG, Cys, CySS, and CySSG.
Three hundred-microliter aliquots of medium were added to 300 μl of ice-cold 10% (wt/vol) perchloric acid solution containing 0.2 M boric acid and 10 μM γ-glutamylglutamate (γ-Glu-Glu) as internal standard (15). The remaining medium was aspirated, and cells were washed once with 1 ml of PBS and immediately treated with 500 μl of ice-cold 5% (wt/vol) perchloric acid solution containing 0.2 M boric acid and 10 μM γ-Glu-Glu and placed on ice. After 5 min, cells were scraped and transferred into microcentrifuge tubes. Samples were stored at −20°C until derivatization with iodoacetic acid and dansyl chloride. For HPLC analysis, derivatized samples were centrifuged, and 20 μl of the aqueous layer was applied to the Supelcosil LC-NH2 column (25 cm × 4.6 mm; Supelco, Bellefonte, PA). Derivatives were separated with a sodium acetate gradient in methanol-water and detected by fluorescence. Concentrations of thiols and disulfides were determined by integration relative to the internal standard. Extracellular redox states (i.e., half-cell reduction potentials, Eh) of the GSH/GSSG and Cys/CySS pools were calculated from concentrations of GSH, GSSG and Cys, CySS in molar units with the following forms of the Nernst equation for pH 7.4 (7, 15): GSH/GSSG, Eh = −264 + 30 log([GSSG]/[GSH]2), Cys/CySS, Eh = −250 + 30 log([CySS]/[Cys]2).The standard potential at pH 7.0 (Eo) for GSH (30) was adjusted to pH 7.4 assuming a pH dependence of 59 mV/pH unit (5). The Eo value for Cys/CySS was estimated from the Eo value for GSH/GSSG and Keq = 3 for 2 GSH + CySS ↔ GSSG + 2 Cys.
Results are expressed as means ± SE. Statistical analyses were performed by ANOVA, with time and treatment as the main effects. The Tukey test was used to test for differences between treatments. Significance was set at a P value <0.05.
Effect of BSO on cellular GSH and GSSG.
During 24-h incubation of HT-29 cells under control conditions with McCoy's medium, no significant changes in cellular GSH or GSSG concentration occurred. Addition of BSO resulted in both concentration-dependent and time-dependent decreases in GSH (Fig. 1A), with a sevenfold decrease in GSH at 24 h with 100 μM BSO (P < 0.0001). GSSG decreased with time under all conditions, but no significant effect of BSO was detected (Fig. 1B). Under these conditions, there was no loss of cell viability as measured by Trypan blue exclusion. Calculation of cellular EhGSSG showed that the changes in GSH and GSSG concentrations were associated with a significant 30 mV (P < 0.01) oxidation (from −240 to −210 mV) with 100 μM BSO at 24 h (Fig. 1C). Thus inhibition of GSH synthesis was sufficient to result in a substantial decrease in cellular GSH concentration and an oxidation of the cellular GSH/GSSG pool.
Effect of BSO on extracellular GSH, GSSG, and CySSG.
Extracellular GSH concentrations were very low (submicromolar) under all conditions and were unaffected by BSO pretreatment (Fig. 2A). Extracellular GSSG concentrations averaged 0.08 μM and also did not vary significantly over time or by BSO treatment (Fig. 2B). Calculations of the redox state of the extracellular GSH/GSSG pool showed that it was considerably oxidized (range −35 to −65 mV) relative to the cellular values and was unaffected by BSO treatment (Fig. 2C).
In contrast to GSH and GSSG, where no changes were detected with time, the disulfide of GSH and cysteine, CySSG, increased as a function of time in control cells (Fig. 3). This increase is consistent with the results of Reed and Beatty (29), who showed that GSH released from cells reacts with CySS to form CySSG. Analysis of cell extracts showed that CySSG was not detectable, further indicating that the CySSG was formed in the extracellular medium. BSO treatment significantly decreased CySSG in the culture medium (Fig. 3), with CySSG concentration in media from BSO-treated cells (1.8–2.1 μM) at 12 h being <50% of that in matched controls (4.2 μM; P < 0.05). These results support the interpretation that cellular GSH release and reduction of CySS occur. However, in combination with the observation that there is no change in extracellular EhGSSG, the results indicate that the rate of GSH release was insufficient to change the extracellular EhGSSG in the presence of the extracellular CySS concentration.
Effect of BSO on rate of reduction of extracellular CySS.
Cys concentration increased with time in the culture medium of cells incubated with 100 μM CySS (Fig. 4A). However, instead of an expected inhibition of the rate of Cys appearance due to decreased cellular GSH concentration, there was no significant effect in response to BSO (Fig. 4A). The rate of CySS loss from the culture medium was also unaffected by BSO (Fig. 4B), indicating that the uptake and use of CySS by the cells was unaffected by the rate of utilization of CySS to supply Cys and synthesize GSH within the cells. The EhCySS was regulated to a value between −90 and −80 mV by 12 h, and the rate of approach to this steady state was unaffected by BSO treatment (Fig. 4C). Consequently, the results show that despite a substantial effect of BSO on cellular GSH and EhGSSG, these changes were not associated with extracellular EhCySS, yet extracellular EhCySS was regulated to a value similar to that observed in vivo in human plasma.
Effect of Cys deficiency and readdition on extracellular GSH and EhGSSG.
Previous research showed that culture of HT-29 cells for 2 days under Cys- and CySS-limiting conditions caused a marked decrease in cellular GSH concentration and an oxidation of the cellular EhGSSG (22). Measurement of the extracellular GSH and GSSG concentrations under these conditions showed no significant effect of Cys-limiting medium on the extracellular GSH and GSSG concentrations and only a modest oxidation of extracellular EhGSSG (Table 1). To determine whether the extracellular GSH pool changed in association with the recovery of the cellular GSH pool, we measured GSH and GSSG in the culture medium after addition of fresh medium containing 100 μM CySS. Both extracellular GSH and GSSG increased with time (Fig. 5, A and B; P < 0.0001), and extracellular EhGSSG was adjusted to −90 mV by 12 h (Fig. 5C; P < 0.0001).
Extracellular Cys/CySS redox state during Cys deficiency.
Medium from cells cultured in Cys-limiting medium had detectable Cys and CySS, apparently due to the content in bovine serum and release from cells. However, somewhat surprisingly, there was no significant difference in EhCySS values when cells cultured in Cys-limiting medium were compared with cells in control medium (Fig. 6). This result indicates that the HT-29 cells do not completely remove Cys from the extracellular compartment despite the limitation of Cys for maintenance of protein and GSH synthesis. When CySS was added to the culture medium in Cys-deficient cells, Cys in the extracellular medium increased significantly (P < 0.01; Fig. 6A). The pattern of increase was similar to that seen in controls that had not been deprived of Cys for the preceding 2 days, but Cys levels in Cys-deficient cells receiving CySS were significantly lower at all time points (P < 0.01; Fig. 6A), suggesting an increased use of CySS in the cells that had previously been deprived of Cys. In agreement with this, CySS levels were significantly lower in cells that had previously been cultured in Cys-limiting conditions (Fig. 6B; P < 0.01). In both control and Cys-limiting media, the rates of CySS loss from the medium, expressed in Cys equivalents, were more than fourfold the rate of Cys appearance (Fig. 6, A and B), indicating that the rate of conversion of CySS to Cys, as measured in the extracellular medium, was only a fraction of the total amount of CySS cleared by the cells. This rapid loss of CySS was not investigated but could indicate a rapid S-cysteinylation of proteins. Extracellular EhCySS achieved a steady-state value of about −90 mV by 4 h, and no significant difference due to prior Cys deficiency was detected (Fig. 6C). This was more rapid than EhGSSG, which achieved a steady state only after 12 h (Fig. 5C). Because the Eh of the Cys/CySS pool recovered more readily than the redox of the extracellular GSH/GSSG pool, and depletion of cellular GSH due to Cys deficiency had no effect on the recovery of extracellular EhCySS, the data confirm the BSO experiments in showing that cellular GSH does not have a central role in maintaining extracellular EhCySS.
The application of chemical concepts of redox potential to biological systems has been outlined for the GSH/GSSG couple by Schafer and Buettner (31). Systems interact with the couple in two ways, corresponding to one-electron and two-electron transfer reactions. The former is represented by S-glutathionylation of proteins and is termed a type I switch, while the latter involves internal disulfide formation from vicinal thiols and is termed a type II switch. Evidence is available for both types of sulfur switches, with the former having a characteristic 3.2-fold change in —SS—/—SH per 30 mV and the latter having a 10-fold change per 30 mV (14). For Cys/CySS, the same principles apply, but little information is available concerning cysteinylation of proteins. On the other hand, EhCySS measured in human plasma ranges from about −120 to −20 mV, indicating that substantial redox-dependent changes could occur for proteins in the plasma or in the cell membrane that interact with the Cys/CySS couple.
In an important study of optimal cell culture conditions, Hwang and Sinskey (11) found that pH, partial pressure of O2, and redox potential (Eh) of the culture medium were each important to achieve maximal cell density. In 22 cell lines, they found that maximal cell density was obtained with Eh maintained at −60 mV. Redox potential was measured with a potentiometric electrode, which does not provide a measure of the Eh for any individual redox couple but rather provides a collective measurement of all redox couples interacting with the electrode. However, in their studies, they found that Eh could be readily adjusted by addition of Cys, and our calculations of EhCySS in culture medium based on measured concentrations of Cys, CySS, and pH show that the measured and calculated values are in reasonable agreement.
A redox effect on maximum cell density could occur through stimulated proliferation or inhibited cell death. To examine whether cell proliferation was regulated by redox and whether this effect could occur specifically by EhCySS of the culture medium, Jonas et al. (13) studied Caco-2 cell proliferation as a function of EhCySS. The results showed that a more reduced EhCySS increased proliferation rate and also showed that the Caco-2 cells altered the extracellular EhCySS to −80 mV. The present study shows that another intestinal cell line, HT-29, controls EhCySS to a very similar value (−90 mV). Other studies have shown that normal human retinal pigment epithelial cells (12) and bovine aortic endothelial cells (8) also adjust extracellular EhCySS to values in the range of −80 to −90 mV. Together with the data of Hwang and Sinskey (11), these data indicate that regulation of extracellular Eh, and specifically EhCySS, is a common feature of mammalian cells in culture.
The range of steady-state EhCySS values maintained in cultured cells is identical to the EhCySS in human plasma (−80 ± 9 mV) (7, 16, 18). Of potential importance, plasma EhCySS becomes oxidized in association with age (18) and oxidative stress (24), raising the possibility that plasma EhCySS could be important in determining the balance of cell populations in vivo. Hwang and Sinskey (11) found that an oxidation of culture medium Eh was associated with a decline in cell density. Similarly, Nkabyo et al. (26) found that medium EhCySS became oxidized as Caco-2 cells became confluent and proliferation decreased. Jiang et al. (12) found that cells exposed to more oxidized extracellular EhCySS were more sensitive to oxidant-induced apoptosis. If these in vitro findings are relevant to in vivo control of cell populations, then mechanisms controlling extracellular EhCySS could be important determinants of tissue function during aging and disease.
Interorgan Cys supply is intimately linked to GSH metabolism (28). After ingestion of meals, Cys is converted to GSH in the liver, and probably other tissues, to provide a short-term store for Cys (9). GSH is exported from liver and other tissues via specific transport systems. After release, GSH and disulfides formed from GSH by thiol/disulfide exchange (e.g., CySSG) are hydrolyzed to constituent amino acids by γ-glutamyltransferase and dipeptidases, principally in the kidneys and small intestine (28). Thus there is a well-established precursor-product relationship for interconversion of GSH and Cys to maintain Cys homeostasis.
Because GSH released from tissues can reduce CySS via thiol/disulfide exchange (28), tissue release of GSH could also function to counter extracellular oxidation of Cys in maintenance of extracellular EhCySS. The present results, however, show that two independent mechanisms to deplete cellular GSH by >85% had no significant effect on steady-state extracellular EhCySS in HT-29 cell cultures. While there is a possibility that at extremely low levels of GSH regulation could be impaired, the data are sufficient to conclude that extracellular EhCySS is largely independent of cellular GSH status. Thus other mechanisms for control of EhCySS must be considered.
In isolated, vascularly perfused small intestine, a Cys-CySS shuttle mechanism was proposed to function in regulation of luminal redox state when GSSG was added (7). When millimolar amounts of GSSG were placed in the lumen, redox state recovered within 20 min, with EhCySS always being more reduced than EhGSSG. This could be explained by CySS uptake, CySS reduction to Cys, and Cys release to function as a reductant for the extracellular GSSG. If this mechanism occurs in HT-29 cells, then the mechanism for intracellular CySS reduction would appear to require a system other than GSH. Previous studies showed that cellular EhCySS was not equilibrated with EhGSSG or with the thioredoxin-1 system in HT-29 cells (17, 22). However, each of these systems, along with others, could partially contribute to the rate of CySS reduction in cells. In liver, the thioredoxin system has been estimated to catalyze 18% of the reduction of CySS (20). Additional studies are needed to address this important question.
Estimates of rates for individual steps are summarized in Fig. 7. The rate of Cys incorporation into protein (step 1) in HT-29 was estimated to be ∼0.42 nmol·106 cells−1·min−1 from the data of Kirlin et al. (19), and the maximal rate of incorporation into GSH (step 2) was estimated to be ∼0.12 nmol·106 cells−1·min−1 from the data of Miller et al. (22). From Figs. 3 and 5, the rate of GSH release (total forms detected; step 3) was considerably less, i.e., ∼0.004 nmol·106 cells−1·min−1. The rate of CySS clearance from the medium (step 4) was ∼0.72 nmol·106 cells−1·min−1 (expressed in Cys equivalents) for control cells in Figs. 4 and 6. This value is about half of the estimated protein synthesis rate, suggesting that another process, such as protein S-cysteinylation, may also occur. For both Figs. 4 and 6, the rate of Cys appearance in the culture medium (step 5) was ∼0.10 nmol·106 cells−1·min−1 under control conditions, i.e., considerably slower than the rate of CySS loss and estimated rate of protein synthesis. As previously reported (17), the nonenzymatic thiol/disulfide exchange rate for the reaction of CySS with GSH is too slow (0.05 nmol·106 cells−1·min−1) to accommodate the needed rate of reduction of CySS to Cys (step 6) in cells. At the prevailing conditions in the extracellular medium, the rate of reaction of GSH at the highest concentration detected (200 nM) with the maximal CySS concentration (100 μM) used (step 7), with the conservative assumption that the second-order rate constant is 20 M−1min−1, is also too slow to account for Cys appearance (0.0004 nmol·ml−1·min−1). Thus the experimental data in combination with the rate calculations show that nonenzymatic reaction of CySS with GSH does not have a major role in controlling extracellular EhCySS.
In conclusion, the present study shows that HT-29 cells regulate oxidized extracellular EhCySS to a value comparable to that found in vivo in human plasma. This mechanism controlling extracellular EhCySS is not sensitive to decreased cellular GSH due to inhibition of synthesis or limitation of Cys and CySS in the culture medium. Consequently, processes that are not dependent on GSH, e.g., transport and oxidation-reduction reactions, function in maintenance of extracellular EhCySS.
This research was supported by National Institutes of Health Grants ES-011195 and ES-09047 (D. P. Jones) and DK-5580 (T. R. Ziegler).
We thank L. T. Miller for her contributions to experiments with cysteine-deficient culture media.
↵1 “Redox state” is used in preference to “redox potential” to avoid implication of thermodynamic equilibrium. Values are calculated from relevant concentrations of the reduced and oxidized components of the redox couple with the Nernst equation.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society