cysteine/cystine couple is a newly recognized node in the ...for buthionine sulfoximine (bso)...

The FASEB Journal express article10.1096/fj.03-0971fje. Published online June 4, 2004.

Cysteine/cystine couple is a newly recognized node in the circuitry for biologic redox signaling and control Dean P. Jones,*,§ Young-Mi Go,* Corinna L. Anderson,|| Thomas R. Ziegler,*,§ Joseph M. Kinkade, Jr.,†,‡ and Ward G. Kirlin¶

Departments of *Medicine, †Biochemistry, and ‡Epidemiology, §Center for Clinical and Molecular Nutrition, and the Graduate Programs of ||Nutrition Health Sciences, Emory University, Atlanta, Georgia 30322, and ¶Department of Pharmacology and Toxicology, Morehouse School of Medicine, Atlanta, GA 30310

Corresponding author: Dean P. Jones, Department of Medicine, 4131 Rollins Research Center, Emory University, Atlanta, GA 30322. E-mail: [email protected]

ABSTRACT

Redox mechanisms function in control of gene expression, cell proliferation, and apoptosis, but the circuitry for redox signaling remains unclear. Cysteine and methionine are the only amino acids in proteins that undergo reversible oxidation/reduction under biologic conditions and, as such, provide a means for control of protein activity, protein-protein interaction, protein trafficking, and protein-DNA interaction. Hydrogen peroxide and other reactive oxygen species (ROS) provide a mechanism to oxidize signaling proteins. However, oxidation of sulfur-containing side chains of cysteine and methionine by ROS can result in oxidation states of sulfur (e.g., sulfinate, sulfonate, sulfone) that are not reducible under biologic conditions. Thus, mechanisms for oxidation that protect against over-oxidation of these susceptible residues and prevent irreversible loss of activity would be advantageous. The present study shows that the steady-state redox potential of the cysteine/cystine couple (Eh = –145 mV) in cells is sufficiently oxidized (>90 mV) relative to the GSH/GSSG (–250 mV) and thioredoxin (Trx1, –280 mV) redox couples for the cysteine/cystine couple to function as an oxidant in redox switching. Consequently, the cysteine/cystine couple provides a means to oxidize proteins without direct involvement of more potent oxidants. A circuitry model incorporating cysteine as a redox node, along with Trx1 and GSH, reveals how selective interactions between the different thiol/disulfide couples and reactive protein thiols could differentially regulate metabolic functions. Moreover, inclusion of cysteine/cystine as a signaling node distinct from GSH and Trx1 significantly expands the redox range over which protein thiol/disulfide couples may operate to control physiologically relevant processes.

Key words: proliferation • differentiation • apoptosis • phase 2 enzymes • oxidative stress

ysteine is a critical component of many structural, catalytic, and regulatory domains of proteins, including DNA binding regions of transcription factors, cysteine-rich domains of receptor tyrosine kinases and docking proteins, and active sites of cysteine proteases,

phosphatases, and ATPases. At least three related systems function catalytically in redox control, thioredoxin-1/thioredoxin reductase-1 (Trx1/TR1), glutaredoxin-1/GSH/glutathione disulfide

C

(page number not for citation purposes)

reductase (Grx1/GSH/GR), and oxidase-linked protein disulfide isomerases (PDI) (1–3). The Trx1- and GSH-dependent systems rely upon NADPH for function and generally serve to reduce protein disulfides as well as cysteine sulfenic acid and methionine sulfoxide (4), but they have different activities with different proteins. In contrast, PDIs in the cisternae of the secretory pathway function to introduce disulfides (5), driven by a flavin-containing, oxygen-dependent system (6). Additional enzymes (thiol oxidase, sulfhydryl oxidases) oxidize low molecular weight thiols and sulfhydryl groups of proteins (7, 8). In addition, activation of numerous cytokine and growth factor signaling pathways results in reactive oxygen species (ROS) generation (9–12), and a family of NADPH oxidases that produce ROS in signaling of cell growth and angiogenesis has been identified (13). Although details of how these systems interact remain unclear, accumulating evidence shows that they are part of a signaling network in which reversible oxidation-reduction of protein thiols serve as switches for redox control circuits to regulate antioxidant defenses and specific cell growth and apoptotic pathways.

Change in the steady-state redox of the GSH/GSSG system, a two-electron couple, between rapidly proliferating cells (–250 mV) and differentiated cells (–200 mV) is sufficient to exert control in such pathways through either protein dithiol/disulfide motifs or cysteine residues subject to S-thiylation (14–16). The GSH/GSSG redox couple is further oxidized by 30 mV or more during apoptosis (16, 17), sulfur amino acid starvation (18), inhibition of GSH synthesis (19), and exposure to phase II enzyme inducers in differentiated cells (15). Thus, in agreement with the abundant data implicating GSH in redox signaling, the GSH/GSSG redox couple varies sufficiently to control biologic functions through protein switches.

The redox state of Trx1 is maintained at –270 to –280 mV in proliferating cells (16, 20) and is oxidized little in response to differentiation (16), apoptosis (21), or sulfur amino acid deficiency (22). Thus, despite the common and overlapping biologic functions of the GSH and Trx1 systems, the redox responses of GSH and Trx1 are distinct. Consequently, GSH and Trx1 can serve as independent nodes within the redox circuitry to organize and control redox-switching patterns of different proteins. This character can be viewed in analogy to the electron carriers NADH/NAD+ and NADPH/NADP+, which are maintained at different steady-state redox potentials in cells (23), and used by specific oxidoreductase systems to provide different pathways of electron flow for catabolism and anabolism within the same aqueous compartment.

Studies of human plasma show that the two predominant low molecular weight thiol-disulfide pools, GSH/GSSG and cysteine/cystine (Cys/CySS), are not in redox equilibrium and that the Cys/CySS redox is ~60 mV more oxidized than the redox of GSH/GSSG (24, 25). This raises the possibility that if Cys/CySS is similarly oxidized in cells, Cys/CySS could provide a third redox node for differential control of the thiol/disulfide redox state of specific proteins.

The purpose of this study was to determine the redox state of cellular Cys/CySS and to determine whether the redox state of this couple is equilibrated with or independent of the redox of the cellular GSH/GSSG couple. Results show that the steady-state Cys/CySS redox in HT29 cells was –145 mV, at least 90 mV more oxidized than the GSH/GSSG pool. GSH/GSSG and Cys/CySS couples were both oxidized during cell differentiation, but to different extents. In contrast, Cys deficiency or inhibition of GSH synthesis specifically oxidized the GSH/GSSG couple without an effect on the Cys/CySS couple, and induction of GSH synthesis in


proliferating cells resulted in oxidation of Cys/CySS with little effect on GSH/GSSG redox. The results demonstrate that the cellular Cys/CySS redox couple is independent of GSH/GSSG and sufficiently oxidized to provide an oxidative signal similar to ROS in redox signaling and control. Together with previous data, these findings establish that Cys, GSH, and Trx1 constitute three distinct nodes in the circuitry for thiol/disulfide redox signaling.

MATERIALS AND METHODS

Materials

N-β-Maleimidopropionic acid (BMPA) was from Pierce Chemicals (Rockford, IL). Except as indicated, all other chemicals were purchased from Sigma (St. Louis, MO). Distilled, deionized water was used throughout. HPLC quality solvents were used for HPLC.

Cell culture

HT29 cells, a human colonic epithelial cell line, were obtained from American Type Culture Collection (ATCC; Gaithersburg, MD), and were maintained at 37°C and 95% air, 5% CO2 in McCoy’s 5A complete medium, 10% (v/v) fetal bovine serum, and 5% (v/v) penicillin/streptomycin (Atlanta Biologicals, Atlanta, GA). This medium contains 0.2 mM cysteine, 1.5 mM glutamine, 0.1 mM glycine, and 0.0016 mM GSH. McCoy’s 5A Cys-free medium did not contain Cys but was otherwise identical to McCoy’s 5A complete medium. For experiments with Cys-free medium, cells were seeded in McCoy’s medium on 60 mm plates at 6 × 105 cells/plate and maintained for 24 h before changing to Cys-free medium. A small amount of CySS was present under these conditions, estimated to be ~5 µM, due to the inclusion of fetal bovine serum. For buthionine sulfoximine (BSO) experiments, cells were seeded on six-well plates at a density of 5 × 105 cells/well and maintained for 24 h before addition of BSO. Under all conditions used in the present study, including incubations in Cys-free medium, cell viability was >99%.

Analysis of GSH and GSSG

Cells were washed once with PBS and immediately treated with 500 µl of ice-cold 5% (w/v) 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 and analysis by HPLC with fluorescence detection (14). Thiols and disulfides were quantified by integration relative to the internal standard (14). Cellular concentrations were calculated using the measured cell volume (7 µl/106 cells; 15). The redox states (Eh) of GSH/GSSG and Cys/CySS were calculated using the Nernst equation with Eo adjusted for the measured cytosolic pH (14).

Statistical analysis

Results are expressed as mean ± SE. Differences were compared across groups and time using one-way or two-way ANOVA, as appropriate. Specific differences between treatments or time and their interactions were compared using Tukey’s pairwise comparisons. Differences were considered to be significant when P < 0.05.


RESULTS

Redox state of Cys/CySS in proliferating HT29 cells

The redox state of Cys/CySS was calculated from the concentrations of Cys and CySS in cells using iodoacetic acid and methodology previously validated for GSH and GSSG determinations (14). Cellular cysteine was determined to be 126 ± 13 µM and cystine was 31 ± 5 µM. The Cys/CySS redox was –145 ± 3 mV, a value considerably more oxidized than the value of –249 ± 2 mV measured for the GSH/GSSG couple under the same conditions (Table 1). Because of the large difference between redox states of these two couples, three possible artifacts in determining the Cys/CySS redox state were considered: 1) artifactual oxidation during sampling, 2) carryover of cysteine or cystine from the culture medium during extraction, or 3) sequestration of cysteine or cystine into subcellular compartments so that the total cellular value provided an erroneous measure of the cytosolic pool.

To address the first of these issues, an approach was developed to avoid perchloric acid extraction and to use a thiol-modifying reagent, N-β-maleimidopropionic acid (BMPA), which reacted with thiols at least 20-fold faster than iodoacetic acid (data not shown). The results with the more reactive BMPA were 138 ± 20 µM and 24 ± 5 µM, respectively. The calculated redox state was –151 ± 7 mV, not significantly different from the value obtained with the standard method with iodoacetic acid (–145 ± 3 mV). Values for GSH and GSSG were determined simultaneously and were also similar for the two methods (Table 1). Thus, the results show that there is not a substantial artifactual oxidation during analysis and that the redox states of Cys/CySS and GSH/GSSG associated with the cell extracts differed by ~100 mV.

To determine the extent of contamination due to carryover of culture medium, cells were cultured for 24 h, medium was aspirated, and cells were washed with phosphate-buffered saline either 0, 1, or 2 times before extraction. Results showed no significant difference in either Cys content or CySS content for these extraction conditions (Table 1). Calculated redox states for Cys/CySS after one or two washes were ~110 mV more oxidized than corresponding values for GSH/GSSG. Measurements of Cys and CySS concentrations in the culture medium during the first 24 h after medium change showed that these concentrations decrease rapidly so that washes were needed to prevent significant carryover during the first 16 h after medium change (data not shown). Thus, carryover from the culture medium was readily controlled and did not contribute to significant error.

To examine the possibility that the relatively oxidized state of Cys/CySS could be due to CySS sequestration in the secretory pathway (26) or other subcellular compartments, we did experiments with digitonin to selectively permeabilize the plasma membrane (27). This treatment allows release of cytosolic contents without release of contents from the cisternae of the secretory pathway. Aliquots of digitonin were added, and cells were visually inspected by phase contrast microscopy to obtain conditions providing lysis of 97–99% of the cells. Analysis of thiols and disulfides showed that the fraction released upon treatment with digitonin (cytosolic fraction) was similar to cellular values (Table 1). In contrast, the remaining sequestered pools had less cysteine. Calculated cytosolic redox state for Cys/CySS was 10 mV more reduced than the cellular value, but this difference was not statistically significant (Table 1). In contrast, the Eh


for sequestered Cys/CySS was –134 mV, significantly more oxidized than the cytosolic value (Table 1). No differences were detected for GSH/GSSG pools. Thus, the results show that the sequestered pools do not create a large error in estimation of redox state and therefore establish that Cys/CySS redox is 90–100 mV more oxidized than the GSH/GSSG couple.

Oxidation of Cys/CySS redox during differentiation

Differentiation of human colonic epithelial cells results in a substantial oxidation of the cytosolic GSH/GSSG redox state (15, 16). To determine whether differentiation induced a similar oxidation of Cys/CySS, we treated HT29 cells with the differentiating agent sodium butyrate. At 48 h after addition of 5 mM butyrate, Cys and CySS concentrations were substantially decreased compared with controls and the Eh for Cys/CySS was oxidized by ~30 mV (Fig. 1). The time course for oxidation of Cys/CySS was delayed relative to GSH/GSSG, and the magnitude of oxidation of Cys/CySS was not as great, but the general patterns of oxidation were similar. Thus, even though the pools are not in redox equilibration, they respond similarly to differentiation.

Cys/CySS redox during Cys deficiency

Culture of HT29 cells in Cys-free medium resulted in a marked decrease in cellular GSH concentration and an associated oxidation of GSH/GSSG couple (18). To determine the effect of Cys deficiency on Cys/CySS redox, cells were cultured for 48 h in Cys-free McCoy’s medium. Results showed a decrease in cellular Cys and CySS concentrations, with little change in Cys/CySS redox (Fig. 2). Addition of either 100 µM Cys or 50 µM CySS resulted in rapid recovery of redox of the GSH/GSSG couple. However, under these conditions, there was a modest oxidation of the Cys/CySS redox. Thus, under conditions of Cys depletion and recovery, the redox states of Cys/CySS do not change in parallel with that of GSH/GSSG.

Inhibition of GSH synthesis with buthionine sulfoximine (BSO)

To determine whether maintenance of steady-state redox of Cys/CySS was dependent upon cellular GSH, we treated cells with buthionine sulfoximine, an inhibitor of glutamate:cysteine ligase, the first enzyme in the pathway for GSH synthesis. Inhibition of GSH synthesis with BSO caused a marked decline in GSH concentration and an associated 40 mV oxidation of GSH/GSSG redox. Under these conditions, there was no effect on the Cys/CySS redox (Fig. 3). Thus, in addition to the above data showing that GSH/GSSG and Cys/CySS are not equilibrated, these results show that the redox state of the Cys/CySS couple is not determined by the redox state of the GSH/GSSG couple in HT29 cells. This implies that the redox state of Cys/CySS in cells may be independently regulated and provide a distinct redox couple for signaling and control.

Effect of the GSH synthesis inducer benzyl isothiocyanate (BIT)

BIT is a compound that enhances GSH synthesis and increases GSH concentration in proliferating HT29 cells with only 8–10 mV oxidation of the GSH/GSSG redox couple (15). Because enhanced Cys utilization for GSH synthesis and/or reaction of Cys with BIT could result in changes in Cys/CySS independently of GSH/GSSG redox, we examined whether BIT treatment altered Cys/CySS redox. Results showed that 25 µM BIT decreased Cys concentration


within 15 min of addition, and this decrease was associated with a 25–30 mV oxidation of Cys/CySS that was sustained for 24 h (Fig. 4). These results show that Cys/CySS redox is considerably more oxidized than the GSH/GSSG redox and that GSH/GSSG redox can be oxidized without effect on the Cys/CySS redox. The finding that Cys/CySS can be oxidized with little effect on cellular GSH/GSSG confirms that the redox state of the Cys/CySS couple is controlled independently of the redox of the GSH/GSSG couple. Given that Cys metabolism and CySS uptake are highly regulated processes, these results show that the Cys/CySS couple should be viewed as a distinct node in the circuitry for redox signaling and control.

DISCUSSION

Previous studies have shown that metabolism and transport of Cys and CySS are regulated. For instance, Cys oxidation (28) and incorporation into GSH (29) are under metabolic regulation, and expression of one of the key transport systems for CySS uptake is under transcriptional control (30). Moreover, the Cys/CySS redox in plasma is tightly controlled (24) and maintained in the steady state at a value ~60 mV more oxidized than the GSH/GSSG couple (24, 25). However, no previous estimates of cellular Cys/CySS redox state are available. The current values, ranging from about –160 to –125 mV, show that Cys/CySS redox in cells is considerably oxidized relative to the GSH/GSSG couple. The data further show that a large error due to contamination by extracellular medium or oxidation during extraction does not occur and that differential subcellular compartmentation of CySS can contribute an error of only ~10 mV. Use of previous literature data on Cys and CySS concentrations (31) to estimate Cys/CySS redox state provides a value of –122 mV, confirming the current interpretation that the steady-state redox of Cys/CySS in cells is considerably more oxidized than that for the GSH/GSSG couple. Combined with the recent data showing that GSH/GSSG redox is controlled independently of the Trx1 redox state (16, 20), the results show that these central thiol/disulfide redox couples can provide three distinct nodes for interaction of redox-sensitive components in cells.

This interpretation is unexpected given the well-known exchange reactions of thiols and disulfides under physiologic conditions. Mammalian cells in culture can grow with CySS even though no enzymes are known that specifically function to reduce CySS. Data of Reed and Beatty (32) provide a possible explanation in terms of thiol-disulfide exchange. Their results showed that hepatocytes could utilize CySS in the culture medium because GSH, which was released from cells, reacted with CySS to generate Cys. While this reaction clearly occurs, a consideration of the rate constants for such chemical reactions reveals that the reaction is slow compared with the rates of Cys metabolism in cells. The apparent thiol-disulfide exchange rate in cells with 3 mM GSH and 30 µM cystine can be estimated to be ~2 µM·min–1, assuming a second-order rate constant of 20 M–1·min–1 (33). The reaction may be catalytically enhanced by enzymes such as glutaredoxin or Trx (34). Mannervik et al. (35) estimated from in vitro studies that in rat liver, 18% of the total cystine reduction was due to Trx. In their study, the maximal rate of reduction with 3 mM CySS and otherwise physiologic concentrations of reagents was 700 µM·min–1. If one assumes that the rate is first order in CySS concentration, then the rate of reduction would only be ~7 µM·min–1 with 30 µM CySS in cells (i.e., comparable to the above-calculated apparent rate of 2 µM·min–1). In comparison, the rate of utilization of Cys for net protein synthesis during proliferation in HT29 cells can be estimated from data of Kirlin et al. (15) to be 60 µM·min–1 while the maximal rate of utilization by HT29 cells for GSH synthesis


can be estimated from Miller et al. (18) to be ~50 µM·min–1. This means that maximal Cys utilization is in the range of 100 µM·min–1, that is, at a rate considerably faster than the expected CySS reduction rate. Thus, the rate of CySS reduction is likely to be relatively slow and, therefore, consistent with the observed displacement of Cys/CySS redox state from that for GSH/GSSG.

Previous research has largely focused on ROS, GSH, and Trx1 as key components of the network for redox signaling and control (Fig. 5). ROS are diffusible oxidants that can be generated by enzymes such as NADPH oxidases (13) and exert effects through oxidation of specific proteins; such signaling is opposed by Trx1- and GSH-dependent reduction (Fig. 5). In the model in Figure 5 incorporating these components, activity of a protein can be switched by either 1) a dithiol/disulfide switch or 2) a thiol/S-thiyl switch, with directionality controlled by coupling to a reduced (Trx or GSH) or oxidized (ROS, O2, or CySS) redox node:

1) Pr-SS + reductant → Pr-(SH)2 (“on” or “off,” depending upon protein)

Pr-(SH)2 + oxidant → Pr-SS (activity state reversed from above)

2) Pr-SS-R + R'SH → Pr-SH + RSSR’ (“on” or “off,” depending upon protein)

Pr-SH + RSSR → Pr-SS-R (activity state reversed from above)

In this model, six different classes of signaling/control proteins can exist in terms of the reductant, oxidant combination: 1, Trx, ROS; 2, Trx, O2; 3, Trx, CySS; 4, GSH, ROS; 5, GSH, O2; and 6, GSH, CySS (Fig. 2). Peroxiredoxins are reduced by Trx and oxidized by H2O2 and have the character of a class 1 redox signaling protein. Class 4 proteins include some glutathione S-transferases, which have a peroxidase activity that allows for sulfhydryl switching, which could control processes such as binding to JunK. Sulfhydryl oxidases could function in signaling through sulfhydryl switching proteins coupled with either Trx (class 2) or GSH (class 5).

The present results show that oxidation of dithiol moieties by cystine, as well as S-cysteinylation of thiols as shown below, is likely to provide new classes of signaling proteins with Trx (class 3) or GSH (class 6) as a reductant:

PrSH + CySS → PrSS-Cys + Cys (activity “on” or “off”)

Pr-SS-Cys + GSH → PrSH + CySSG (activity opposite of above).

For S-cysteinylation, CySSG, which is maintained at a very low concentration in cells, would be rapidly removed by reaction with GSH, Trx, or GSSG reductase.

In such a network, Trx1 and Cys/CySS provide boundary conditions for thiol/disulfide redox activity. The redox states of both Trx1 and Cys/CySS are relatively stable to changes in cell growth conditions, with Trx1 being maintained in the range of –280 to –270 mV (16, 20–22) whereas that of Cys/CySS is maintained in the range of –160 to –125 mV (current research). The redox state of GSH/GSSG lies intermediate to these values, being similar to the Trx1 redox state in rapidly proliferating cells, similar to the Cys/CySS redox state in apoptotic cells, and


intermediate between the two in differentiated and growth-arrested cells (14–19). Consideration of the redox states of these nodes under different growth and apoptotic conditions means that the same nodes can provide different circuits for redox signaling, effectively providing a context-dependent regulation that can limit proliferative functions to proliferative cells, differentiated functions to differentiated cells, and apoptotic functions to apoptotic cells (14).

Many studies have used Cys precursors to investigate functions of GSH in redox signaling. The observation that Cys redox is distinct from that for GSH suggests that additional investigation is warranted to separate possible effects due to supply of Cys for GSH synthesis and effects on Cys redox state. For instance, N-acetylcysteine (NAC), a precursor of Cys and GSH, protects against apoptosis. This has been interpreted to mean that enhanced cellular GSH protects against apoptosis (36, 37). However, high concentrations of NAC can shift both the GSH/GSSG and Cys/CySS redox states. Indeed, studies with N-acetyl-D-cysteine, a thiol compound with the same chemical redox properties of NAC but that is not a precursor for GSH synthesis, showed that GSH synthesis was not needed for protection because the nonphysiologic D isomer showed comparable protection (38). Thus, although many of the effects of NAC and other Cys precursors are likely to be mediated by GSH, some may be due to effects on Cys or Cys/CySS redox.

In summary, the present results show that the redox state of the cellular Cys/CySS couple is considerably more oxidized than that for the GSH/GSSG couple and varies independently of the GSH/GSSG couple. Together, with recent findings that the redox state of Trx1 varies independently of that for the GSH/GSSG couple, and that the GSH/GSSG redox varies according to cellular growth conditions, the results indicate that a Cys/CySS node must now be added to the Trx1 and GSH/GSSG nodes in the thiol-disulfide redox circuitry of cells. The third node provides additional connections to a limited number of oxidases or peroxidases for the expansion of a dynamic redox network. However, it is the qualitative implications of this nodal expansion that should provide new ways of thinking about how redox signaling events regulate physiologically relevant processes.

ACKNOWLEDGMENTS

We thank Prof. M. W. Anders for his thoughtful discussions and critical comments concerning this research. Support was provided by National Institutes of Health grants ES09047, ES011195, ES09313, and DK55850.

REFERENCES

1. Powis, G., and Montfort, W. R. (2001) Annu. Rev. Pharmacol. Toxicol. 41, 261–295

2. Holmgren, A., and Bjornstedt, M. (1995) Meth. Enzymol. 252, 199–208

3. Ferrari, D. M., and Soling, H. D. (1999) Biochem. J. 339, 1–10

4. Stadtman, E. R., Moskovitz, J., Berlett, B. S., and Levine, R. L. (2002). Mol. Cell. Biochem. 234-235, 3–9


5. Frand, A. R., Cuozzo, J. W., and Kaiser, C. A. (2000) Trends Cell Biol. 10, 203–210

6. Sevier, C. S., Cuozzo, J. W., Vala, A., Aslund, F., and Kaiser, C. A. (2001) Nat. Cell Biol. 3, 874–882

7. Lash, L. H., and Jones, D. P. (1986) Arch. Biochem. Biophys. 247, 120–130

8. Thorpe, C., Hoober, K. L., Raje, S., Glynn, N. M., Burnside, J., Turi, G. K., and Coppock, D. L. (2002) Arch. Biochem. Biophys. 405, 1–12

9. Kreuzer, J., Viedt, C., Brandes, R. P., Seeger, F., Rosenkranz, A. S., Sauer, H., Babich, A., Nurnberg, B., Kather, H., and Krieger-Brauer, H. I. (2003) FASEB J. 17, 38–40

10. Bae, Y. S., Kang, S. W., Seo, M. S., Baines, I. C., Tekle, E., Chock, P. B., and Rhee, S. G. (1997) J. Biol. Chem. 272, 217–221

11. Garg, A. K., and Aggarwal, B. B. (2002) Mol. Immunol. 39, 509–517

12. Irani, K., Xia, Y., Zweier, J. L., Sollott, S. J., Der, C. J., Fearon, E. R., Sundaresan, M., Finkel, T., and Goldschmidt-Clermont, P. J. (1997) Science 275, 1649–1652

13. Lambeth, J. D. (2002) Curr. Opin. Hematol. 9, 11–17

14. Jones, D. P. (2002) Meth. Enzymol 348, 93–112

15. Kirlin, W. G., Cai, J., Thompson, A. S., Diaz, D., Kavanaugh, T. J., and Jones, D. P. (1999) Free Rad. Biol. Med 27, 1208–1218

16. Nkabyo, Y. S., Ziegler, T. R., Gu, L. H., Watson, W. H., and Jones, D. P. (2002) Am. J. Physiol. 283, G1352–G1359

17. Cai, J., and Jones, D. P. (1998) J. Biol. Chem. 273, 11401–11404

18. Miller, L. T., Watson, W. H., Kirlin, W. G., Ziegler, T. R., and Jones, D. P. (2002) J. Nutr. 132, 2303–2306

19. Lawrence, C. L., Watson, W. H., Herzog, B., Miller, L. T., Ziegler, T. R., and Jones, D. P. (2002) FASEB J. 16, A996

20. Watson, W. H., Pohl, J., Montfort, W. R., Stuchlik, O., Powis, G., and Jones, D. P. (2003) J. Biol. Chem., 278, 33408–33415

21. Watson, W. H., and Jones, D. P. (2003) FEBS Lett. 543, 144–147

22. Miller, L. T. (2001) Effect of cysteine starvation and repletion on thiol-disulfide redox in HT29 cells. M.S. Thesis, Emory University, Atlanta


23. Sies, H. (1982) Nicotinamide nucleotide compartmentation. In Metabolic Compartmentation (Sies, H., ed) pp. 205–231, Academic Press, London

24. Jones, D. P., Carlson, J. L., Mody, V. C., Jr., Cai, J., Lynn, M. J., and Sternberg, P., Jr. (2000) Free Rad. Biol. Med. 28, 625–635

25. Jones, D. P., Mody, V. C., Jr., Carlson, J. L., Lynn, M. J., and Sternberg, P., Jr. (2002) Free Rad. Biol. Med 33, 1290–1300

26. Hwang, C., Sinskey, A. J., and Lodish, H. F. (1992) Science 257, 1496–1502

27. Fayos, B. E., and Wattenberg, B. W. (1997) Endothelium 5, 339–350

28. Kwon, Y. H., and Stipanuk, M. H. (2001) Am. J. Physiol. 280, E804–E815

29. Dickinson, D. A., and Forman, H. J. (2002) Ann. N. Y. Acad. Sci. 973, 488–504

30. Sato, H., Kuriyama-Matsumura, K., Hashimoto, T., Sasaki, H., Wang, H., Ishii, T., Mann, G. E., and Bannai, S. (2001) J. Biol. Chem. 276, 10407–10412

31. Noda, T., Iwakiri, R., Fujimoto, K., Rhoads, C. A., and Aw, T. Y. (2002) Cell Prolif. 35, 117–129

32. Reed, D. J., and Beatty, P. W. (1978) The role of the cystathionine pathway in glutathione regulation by isolated hepatocytes. In Functions of Glutathione in Liver and Kidney (Sies, H., and Wendel, A., eds.) pp. 13–21, Springer-Verlag, Berlin

33. Gilbert, H. F. (1990) Adv. Enzymol. Relat. Areas Mol. Biol. 63, 69–172

34. Ziegler, D. M. (1985) Annu. Rev. Biochem. 54, 305–329

35. Mannervik, B., Axelsson, K., Sundewall, A. C., and Holmgren, A. (1983) Biochem. J. 213, 519–523

36. Malorni, W., Rivabene, R., Santini, M. T., and Donelli, G. (1993) FEBS Lett. 327, 75–78

37. Sandstrom, P. A., Mannie, M. D., and Buttke, T. M. (1994) J. Leuk. Biol. 55, 221–226

38. Jones, D. P., Maellaro, E., Jiang, S., Slater, A. F. G., and Orrenius, S. (1995) Immunol. Lett. 45, 205–209

Received February 5, 2004; accepted April 23, 2004.


Table 1 Evaluation of methodology to measure redox state of cysteine/cystine in HT29 cells

Cys (µM)

CySS (µM)

Eh (Cys) (mV)

GSH (mM)

GSSG (µM)

Eh (GSH) (mV)

Experiment 1, efficiency of trappinga IAA (n=4) 126±13 31±4.8 –145±3 3.2±0.11 21±3 –249±2

BMPA (n=4) 138±20 24±5 –151±7 3.2±0.2 19±6 –252±4

Experiment 2, carry-over from mediumb 0 washes (n=4)

110±7 29±0.81 –142±1 3.3±0.16 16±2 –253±2

1 wash (n=4)

107±7 28±1.2 –142±2 3.2±0.11 13±2 –256±1.8

2 washes (n=4)

92±9 24±2.4 –140±2 3.0±0.14 12±2 –254±1.17

Experiment 3, selective releasec Cytosolic (n=3)

137±8 17.7±1.7 –155±1 3.4±0.06 14±1.3 –256±1.1

Sequestered (n=3)

62±8d 20±6 –134±3d 3±1 14.7±3 –250±4

aExperiment 1 compared two analytical approaches to measure thiols and disulfides. The first (IAA) used perchloric acid extraction followed by derivatization with iodoacetic acid and dansyl chloride and separation by HPLC. The second (BMPA) used a methanol extraction, derivatization with N-β-maleimidopropionic acid (BMPA) and dansyl chloride, and separation by HPLC. No significant differences were observed. bExperiment 2 compared repeated washes with phosphate-buffered saline. No significant differences were observed. cExperiment 3 compared the thiol and disulfide content released following digitonin permeabilization (cytosolic fraction) to the content remaining with the cell cytoskeleton (sequestered pools). dResults showed a significant difference only for Cys and Eh for Cys/CySS.


Fig. 1

Figure 1. Cys, CySS, and Cys/CySS redox state in HT29 cells following treatment with sodium butyrate. Cells were cultured in McCoy’s medium with or without 5 mM sodium butyrate for the times indicated and then analyzed by HPLC for Cys (top left panel) and CySS (top right panel) contents. Differentiation was assessed by expression of alkaline phosphatase and γ-glutamyltranspeptidase; cells approached maximal differentiation by 72 h (15). Cellular redox state (Eh) values (bottom panel) were calculated using the Nernst equation and are plotted along with similarly determined values for redox state of the GSH/GSSG couple.


Fig. 2

Figure 2. Cys, CySS, and Cys/CySS redox changes following culture in Cys-free medium and readdition of Cys. Cells were cultured in McCoy’s Cys-free medium for 2 days, treated with 100 µM Cys or 50 µM CySS for the times indicated, and then analyzed by HPLC for Cys (top panel) and CySS (middle panel) contents. Cellular redox state (Eh) values (bottom panel) were calculated using the Nernst equation and are plotted along with similarly determined values for redox state of the GSH/GSSG couple. No significant differences for Eh for Cys/CySS were detected, but Eh for GSH/GSSG at 0 h was significantly different from controls maintained in the Cys-containing culture medium and from cells at 1, 2, or 5 h after readdition of Cys or CySS.


Fig. 3

Figure 3. Effect of the GSH synthesis inhibitor on Cys, CySS, and Cys/CySS redox state. Cys/CySS redox state in HT29 cells following treatment with different concentrations of buthionine sulfoximine (BSO). Cells were cultured in McCoy’s medium with or without BSO at concentrations of 1, 10, or 100 µM for the times indicated and then analyzed by HPLC for Cys (top right panel) and CySS (top left panel) contents. Cellular redox state (Eh) values (bottom panel) were calculated using the Nernst equation and are plotted along with similarly determined values for redox state of the GSH/GSSG couple.


Fig. 4

Figure 4. Changes in Cys, CySS, and Cys/CySS redox in HT29 cells following treatment with benzyl isothiocyanate (BIT). Proliferating cells were cultured in McCoy’s medium with or without 25 µM BIT for the times indicated and then analyzed by HPLC for Cys (top left panel) and CySS (top right panel) contents. Cellular redox state (Eh) values (bottompanel) were calculated using the Nernst equation and are plotted along with similarly determined values for redox state of the GSH/GSSG couple.


Fig. 5

Figure 5. Circuitry model for cellular redox signaling and control by reversible oxidation of thiols. Reversible oxidation-reduction of possible classes of signaling proteins (labeled 1–6) can occur via combinations of opposing reduction and oxidation reactions. In this scheme, electron flow is from the most reduced component (NADPH) on the top to electron acceptors on the bottom. Reduction is driven by well-characterized NADPH-dependent reductases, thioredoxin reductase 1 (TR1), and glutathione disulfide reductase (GR) through redox control nodes represented by Trx1 and GSH. The Trx1 and GSH are located schematically according to respective measured steady-state redox values, given in the scale on the left. Oxidation of signaling proteins can occur by reactive oxygen species (ROS) or by O2 catalyzed by sulfhydryl oxidases (SO). The present study shows that the Cys/CySS couple can provide a third thiol/disulfide node for cellular redox signaling in which oxidation can occur without direct coupling to ROS or O2. Oxidation of Cys/CySS occurs extracellularly catalyzed by a membranal thiol oxidase (TO) (7). Redox circuits for proteins of classes 1 and 2 include Trx1 as the reducing node and ROS or O2 as the oxidant, respectively. Redox circuits for proteins of classes 4 and 5 depend on GSH as the reductant and ROS or O2 as the oxidant. Circuits for proteins of classes 3 and 6 depend on Trx1 and GSH for reduction and are coupled to Cys redox state for oxidation. Because GSH redox state changes as a function of cell growth, proteins with intermediate Eo values (6a and 6b) could be differentially regulated by the change in GSH/GSSG redox state during cell growth and differentiation.


cysteine/cystine couple is a newly recognized node in the ...for buthionine sulfoximine (bso)...

Documents