Enhancement of T Cell Receptor Signaling by a Mild Oxidative Shift in the Intracellular Thiol Pool¹

In vitro, T cell activation and IL-2 production are readily induced by adequate concentrations of ligands or Abs that bind to the TCR and the costimulatory receptor CD28 (1). Induction of IL-2 gene expression through CD28 was shown to require activation of 5-lipoxygenase and is associated with a decrease in the intracellular glutathione (GSH)⁶ level (2). However, in typical infections in vivo, the ligand concentration for the two types of receptors is expected to be at least initially suboptimal, and it may be critical for the survival of the infected host that the specific immune response is induced before optimal Ag doses have accumulated. Therefore, for the immunologist it would be important to know whether and how the threshold for the triggering of the signal cascades may be lowered under physiological conditions by auxiliary mechanisms.

A large body of evidence indicates that the activation of certain signal cascades and transcription factors can be induced or enhanced by reactive oxygen species or other oxidants (for examples, see Refs. 3, 4, 5, 6, 7, 8, 9, 10, 11). This phenomenon was mostly interpreted as an oxidative stress response that protects the cells against potentially lethal stress. Various types of oxidative stress were found to induce in certain bacteria and mammalian cells the expression of proteins with cytoprotective activity against oxidative stress (3, 4, 10, 12, 13). When the oxidative stress was induced with unphysiologically high millimolar concentrations of hydrogen peroxide (HP), it was found that the activity of several protein tyrosine kinases (PTKs) such as Lck (14, 15, 16), ZAP 70 (17), and Syk (18) was strongly enhanced. This enhancement may result either from direct oxidative activation or indirectly from the oxidative inhibition of a tyrosine phosphatase that normally down-regulates these PTKs (19, 20). Millimolar concentrations of HP were found to cause also the activation of the mitogen-activated protein kinases (MAPKs) extracellular signal-related kinase (ERK) 1 and ERK2 in Jurkat T cells and in cardiac myocytes (21, 22). In view of the unphysiologically high concentrations of HP that have been used in these earlier studies, it was not clear, however, whether and how these redox effects may contribute to the regulation of the immune system under physiological conditions.

The physiological relevance of redox effects in the immune response against environmental pathogens in vivo was suggested by the markedly increased susceptibility to Listeria infection of mice lacking the gp91 protein of the NADPH oxidase (23, 24). In the physiological microenvironment of T cells, HP is produced by activated macrophages at an estimated rate of 2–6 × 10⁻¹⁴ μmol/h per cell and may reach 10–100 μM in the vicinity of these cells (25, 26, 27). Physiologically relevant concentrations of HP were previously shown to augment IL-2 production by mitogenically stimulated T cells in different experimental systems (28, 29). In one of these studies, 200 μM HP was shown to cause an increase in IL-2 gene expression, in the AP-1 transcription factor activity and in the expression of c-Jun but not c-Fos mRNA (29). Treatment of Jurkat cells with 200 μM HP in the absence of a mitogenic stimulus was found to stimulate the expression of c-Jun and the transient expression of c-Fos but not the production of IL-2 (30). Concentrations of 30–100 μM HP were also shown to induce NF-κB transcription factor activity in one subline of Jurkat cells but not in others (7).

IL-2 production by ex vivo derived mitogenically stimulated murine lymphocytes was shown to be enhanced also by high but physiologically relevant concentrations of l-lactate (LAC; 10–30 mM), i.e., another metabolite from activated macrophages (31, 32). Because these concentrations of LAC were found to cause also a decrease in the intracellular GSH level, and the enhancement of IL-2 production was reversed by the addition of exogenous GSH, it was suggested that the IL-2 production is modulated by the intracellular GSH level or by the GSH/GSH disulfide (GSSG) ratio (32). Taken together, these studies suggested the possibility that even moderate changes in the GSH status of T cells may play an important immunopotentiating role in the immune system. Therefore, the studies in this report have been focused mainly on the regulatory role of the intracellular thiol status. To investigate the effect of a mild oxidation of the intracellular GSH pool on the signal cascades in T cells in more detail, we used mainly the GSH reductase inhibitor 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU; Refs. 33, 34) as an alternative to HP and LAC because it provided a more selective means to modulate the intracellular thiol status. Previous studies have shown that treatment of T cells with BCNU causes a dose-dependent decrease in the intracellular GSH level and a corresponding increase in GSSG, a profound activation of the transcription factor AP-1, and a moderate activation of the transcription factor NF-κB (35).

The two signaling pathways derived from the TCR and CD28 receptor were shown to involve a complex scenario of intracellular biochemical events that finally result in multiple cellular responses. The two signaling pathways eventually merge and synergistically stimulate the activity of different members of the MAPK family, i.e., Jun-N-terminal kinase (JNK) and p38 MAPK, and induce the transcription factor NF-κB and the expression of IL-2 by a combination of transcriptional activation and mRNA stabilization (36, 37). The activation of Src family PTKs such as Lck and Fyn is one of the earliest events in TCR signal transduction. Lck and Fyn knockout mice display severe immunological defects (38). Once activated, Lck and/or Fyn induce signal cascades that include among other components PLCγ (39), PKC family members (40, 41) such as PKCΘ (42), Vav1 (39), Rac (43, 44), and finally JNK and c-Jun (45), i.e., a common component of the transcription factors AP-1 and NF-AT (46). Because stimulation at the TCR in the presence of the costimulatory CD28 signal leads to T cell activation and IL-2 production, whereas repeated TCR stimulation in the absence of the costimulatory signal leads to anergy and peripheral tolerance (47, 48), we addressed in a first set of experiments the question whether exposure of T cells to physiologically relevant concentrations of HP or LAC may modulate selectively TCR/CD3 signaling, CD28 signaling, or both.

BCNU was obtained from Bristol Arzneimittel (Munich, Germany). Wortmannin, Trolox, cyclosporin A, MKK886, N-acetylcysteine (NAC), and pyrrolidine dithiocarbamate (PDTC) (Sigma, St. Louis, MO), and glycerol trinitrate (GTN) (Merck, Darmstadt, Germany) were purchased from the indicated suppliers. All other reagents were obtained either from Sigma or from Roche Molecular Biochemicals (Mannheim, Germany). The Abs were obtained from the indicated sources: anti-phospho-p38 and anti-phospho-p42/44, New England Biolabs (Beverly, MA); anti-p38, anti-Fyn (FYN3), and anti-Lck (2102), Santa Cruz Biotechnology (Santa Cruz, CA); anti-IκB kinases (IKK)α and anti-JNK, PharMingen (San Diego, CA); anti-Flag (M2), Sigma; anti-hemagglutinin (HA) (12CA5), Roche Molecular Biochemicals; and anti-Myc (9E10), Upstate Biotechnology (Lake Placid, NY). Anti-TCR(CD3) (OKT3) and anti-CD28 Abs were derived from hybridomas and purified. Rac cDNAs provided by Dr. S. Gutkind (National Institutes of Health, Bethesda, MD) (49) were Myc tagged and inserted into pEF-BOS-derived vectors. The expression vectors encoding VavΔ319–356 (50), MEKK1Δ DN (51), MLK3 KR (52), PKC-Θ KR (53), HA-JNK (54), and MKK7 DN (55) were previously described. The reporter plasmids (κB)₃-luc (56), 4XRE/AP-luc (57), and the IL2 promoter luciferase construct (58) have been described previously. The IL-2 luciferase reporter construct contains the human IL-2 promoter (residues 577 to +53) inserted into MluI and HindIII cloning sites of pGL2-Basic (Promega, Madison, WI).

Jurkat T leukemia cells were grown at 37°C in RPMI 1640 medium containing 10% (v/v) heat-inactivated FCS, 10 mM HEPES, 1% (v/v) penicillin/streptomycin, and 2 mM glutamine (all obtained from Life Technologies, Gaithersburg, MD). Jurkat cells were grown in an incubator at 37°C and 5% CO₂ and transfected by electroporation using a gene pulser (Bio-Rad, Richmond, CA) at 250V/950 μF. Stimulation of Jurkat cells was performed in a final volume of 500 μl by adding anti-CD3 (final concentration 10 μg/ml, clone OKT3) and/or anti-CD28 (final concentration 10 μg/ml, clone 9.3) Abs.

PBLs (95% pure) were prepared from heparinized blood of healthy donors by density centrifugation on Ficoll gradients (Lymphoprep Nycomed Pharma, Oslo, Norway) as described (59). After lysis of B cells with sheep erythrocytes, the cells were stimulated with anti-CD3 plus anti-CD28 Abs and/or BCNU (20 μM) for the indicated periods of time. Secreted IL-2 was determined in the supernatant by a kit from BioSource (Nivelles, Belgium) according to the instructions of the manufacturer.

Proteins were extracted from cells in Nonidet P-40 lysis buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM PMSF, 10 mM NaF, 0.5 mM sodium vanadate, leupeptin (10 μg/ml), 1% (v/v) Nonidet P-40, and 10% (v/v) glycerol). Equal amounts of protein were separated by SDS-PAGE before blotting to a polyvinylidene difluoride membrane (Millipore, Bedford, MA). The membrane was then incubated in a small volume of TBST containing various dilutions of the primary Abs, followed by the detection of the respective proteins with an appropriate secondary Ab coupled to HRP. Secondary Abs were visualized by enhanced chemiluminescence according to the instructions of the manufacturer (Amersham Lifescience, Freiburg, Germany).

Intracellular GSH and GSSG levels were determined as described (35).

Cells were lysed in Nonidet P-40 lysis buffer and the IKK, JNK, p56Lck, and p59Fyn proteins contained in the cell lysate were immunoprecipitated. The precipitate was washed three times in lysis buffer. IKK and JNK precipitates were washed two times in kinase buffer (20 mM HEPES/KOH pH 7.4, 25 mM β-glycerophosphate, 2 mM DTT, and 20 mM MgCl₂). The kinase assay was performed in a final volume of 20 μl kinase buffer containing 2 μg of bacterially expressed GST-c-Jun 5–89(5–89) or GST-IκB-α 1–54(1–54) protein, 5 μCi [γ-³²P]ATP, and 20 μM ATP for 20 min at 30°C. Immunoprecipitated p56Lck and p59Fyn were washed two times with kinase buffer (20 mM PIPES, 10 mM MnCl₂) and incubated in the presence of different concentrations of GSH or GSSG. After treatment, immune complexes were washed once with kinase buffer and suspended in a total reaction volume of 20 μl kinase buffer containing 20 mM PIPES pH 7.2, 10 mM Cl₂, 5 μCi [γ-³²P]ATP, and 20 μM ATP for 20 min at 30°C. The reaction was stopped by the addition of 5 × SDS sample buffer. The reaction products were separated by SDS-PAGE, autoradiographed, and quantified using a phosphorimager.

Reporter assays were performed essentially as described (60). Briefly, cells were washed with isotonic buffer and lysed in 100 μl of lysis buffer (Promega, Mannheim, Germany). The luciferase assays were performed according to the manufacturer’s instructions (Promega) and quantified in a Duo Lumat LB 9507 (Berthold, Wildbad, Germany). The results were normalized to the activity of β-galactosidase expressed by a cotransfected lacZ gene under the control of a constitutive Rous sarcoma virus promoter.

The recombinant Fyn protein was isolated from baculovirus-infected Sf9 (Spodoptera frugiperda) insect cells, and purified by affinity chromatography on a phosphotyrosine column plus anion exchange chromatography (61).

Jurkat cells (5 × 10⁶) were treated as indicated, and nuclear extracts were prepared essentially as described (62). Briefly, cells were washed twice with TBS buffer (25 mM Tris-HCl pH 7.4, 137 mM NaCl, 5 mM KCl, 0.7 mM CaCl₂, 0.1 mM MgCl₂). Cells were centrifuged and the pellet was resuspended in 200 μl cold buffer A (10 mM HEPES/KOH pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, and 0.5 mM PMSF) by gentle pipetting. After incubation for 10 min on ice, 5 μl of 10% Nonidet P-40 was added, and cells were lysed by vortexing. The homogenate was centrifuged for 30 s in a microfuge, and the pellet containing the cell nuclei was dissolved in 30 μl buffer C (20 mM HEPES/KOH pH 7.9, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 0.5 mM PMSF, and 1% (v/v) aprotinin). The extract was centrifuged for 5 min in a microfuge at 4°C, and 5 μg of proteins contained in the supernatant was used for band shift assays.

Binding of NF-κB to its cognate DNA (60) and of CD28RE/AP-binding proteins (57) were measured as described. The sense sequences of the oligonucleotides were 1) NF-κB: 5′-AGTTGAGGGGACTTTCCCAGGC-3′; and 2) CD28RE/AP: 5′-TCTGGTTTAAAGAAATTCCAAAGAGTCATCAG-3′.

To determine the effects of the macrophage products LAC and HP on TCR-mediated signals and costimulatory signals, we transiently transfected Jurkat T cells with a luciferase reporter gene construct fused to the human IL-2 promoter. The transfected cells were incubated for 18 h in medium and then stimulated for 8 h with anti-CD3, anti-CD28 Abs, or the combination of both. If indicated, the cultures were also treated with HP at a final concentration of 50 μM for 1 h or LAC at a final concentration of 10–30 mM for 8 h before Ab stimulation. The results showed that treatment with HP or LAC had relatively little effect on the transcriptional activity after stimulation with anti-CD3 Abs in the absence of anti-CD28 Abs but caused a substantial enhancement in the presence of anti-CD28 Abs alone or in combination with anti-CD3 Abs (Fig. 1). Considering that anti-CD3 Abs caused in the presence of anti-CD28 an ∼3-fold increase in transcription, the data indicated that exposure to HP or LAC can substitute at least partly for the TCR-mediated signal but not for the costimulatory signal. Therefore, we stimulated the Jurkat cells in all subsequent experiments with anti-CD28 Abs in combination with anti-CD3 Abs or PMA.

To determine the influence of the thiol redox state on the transcriptional activity under the control of the IL-2 promoter, we measured the transcription from the reporter gene construct after treatment of the transfected Jurkat cells with various anti-CD28 Abs in combination with anti-CD3 Abs or PMA in the presence or absence of BCNU. In a first set of experiments we showed that BCNU causes a dose- and time-dependent decrease in the intracellular GSH/GSSG ratio of the Jurkat cells (Fig. 2,A) and aggravates the decrease in the GSH/GSSG level that is seen after treatment with anti-CD3/CD28 Abs alone (Fig. 2,B). These changes were similar to the changes seen with the physiologically relevant concentration of 50 μM HP and with 100 μM of the NO donor glycerol trinitrate (GTN) (Fig. 2,B). BCNU caused only a moderate enhancement of transcription from the IL-2 promoter in the absence of the stimulatory agents, but a substantial enhancement in the presence of these agents (Fig. 3,A). The 2- to 3-fold enhancement after BCNU treatment was again in the same order of magnitude as the relative enhancement that was achieved in anti-CD28-stimulated cells by anti-CD3 Abs (Fig. 1).

To identify the redox-responsive regions of the IL-2 promoter, we determined also the effect of BCNU on the expression of the luciferase reporter gene under the control of multimeric binding sites of different immunologically important transcription factors. Jurkat T cells were transfected with a luciferase reporter construct containing four repeats of the CD28RE/AP binding site (4×CD28RE/AP-Luc) and were then stimulated again in the presence or absence of BCNU with anti-CD28 Abs in combination with PMA or anti-CD3 Abs. These experiments revealed that the CD28RE/AP-dependent transcription was again strongly augmented by BCNU (Fig. 3,B). Analogous experiments with a reporter gene under the control of NF-κB binding sites gave similar results (data not shown). The transcription of a reporter gene under control of NF-AT binding sites, in contrast, was not markedly affected by BCNU treatment irrespective of whether transcription was induced by costimulation (data not shown). To investigate the effect of BCNU on the binding of nuclear proteins to the CD28RE/AP element in unstimulated and stimulated Jurkat cells, we analyzed also the DNA/protein complexes by EMSA. The results showed that the stimulatory agents induced the binding of proteins in complex I and enhanced the binding of complexes II and III proteins (Fig. 3 C). In the absence of the stimulating agents, BCNU had no effect on complex I, but enhanced the binding of proteins in complex II and III. If applied in combination with anti-CD3/CD28 Abs, BCNU did not further enhance the binding of complex III proteins, but strongly augmented the binding of proteins in complexes I and II. Supershift experiments using subunit-specific Abs revealed the occurrence of the NF-κB p65 protein in complexes I and II and the predominant localization of c-Rel within complex I (data not shown).

Because the CD28RE/AP element is contacted also by proteins of the NF-κB/Rel family (63), and because NF-κB activity was previously shown to be enhanced by BCNU (35), we examined the effect of BCNU on the activation of NF-κB in more detail. In a first set of experiments, we determined whether BCNU may induce or enhance the NF-κB DNA-binding activity whether administered alone or in combination with PMA, anti-CD3, and anti-CD28 Abs. The DNA-binding activity of NF-κB was determined by EMSA. BCNU was found to enhance the induction of NF-κB DNA-binding by the stimulating agents but had little effect by itself (Fig. 4,A). To test whether the enhanced induction of the NF-κB DNA-binding activity may result from an increased activation of IKKs or from IKK-independent mechanisms as reported for the UV-induced NF-κB activation (64), the IKK complex was isolated from extracts of BCNU-treated or untreated Jurkat cells by immunoprecipitation with monoclonal anti-IKKα Abs. The subsequent analysis by immunocomplex kinase assays with recombinant GST-IκBα as substrate revealed that the CD3/CD28-induced IKK activity was strongly enhanced by BCNU (Fig. 4 B). This was associated with an increased phosphorylation of IKKα itself, indicating that the BCNU treatment acted upon a target further upstream of the IKKs.

The three known families of MAPKs, i.e., ERK, p38 MAPK, and JNK (65), play a key role in immune responses by inducing the transactivation potential of several immunologically important transcription factors (66). Their activity is controlled by another set of protein kinases (MAPK kinases), which are activated in turn by various signals including cytokines (66). To analyze the effect of BCNU on the activation of the three MAPK families, we treated Jurkat T cells with graded concentrations of BCNU and determined in different aliquots of the cell extracts the JNK activity, the phosphorylation of p38 MAPK, and the phosphorylation of ERK1/2 using phospho-specific Abs. The results (Fig. 5,A) showed that the kinase activity of endogenous JNK, as measured by the immunocomplex kinase assay, is strongly augmented by 10–100 μM BCNU but inhibited at higher concentrations. The analysis of p38 MAPK and ERK1/2 phosphorylation by immunoblotting with phospho-specific Abs revealed that the phosphorylation of p38 MAPK was similarly enhanced by 10–100 μM BCNU and inhibited at higher concentrations, whereas p42/44 (i.e., ERK1/2) was not detectably phosphorylated in the presence of BCNU (Fig. 5,A). Concentrations of 10–100 μM BCNU decreased the intracellular GSH/GSSG ratio of costimulated Jurkat cells typically to values <10 (see Fig. 1).

To determine whether a decrease of the GSH/GSSG ratio by other experimental strategies would lead also to a synergistic enhancement of the CD3/CD28-induced activation of JNK and p38 MAPK phosphorylation, we treated Jurkat cells with moderate concentrations of HP or the NO donor GTN, which were previously shown to decrease also the GSH/GSSG ratio if applied in combination with anti-CD3 and anti-CD28 Abs (see Fig. 2). The results showed that both agents enhanced the activation of JNK and the phosphorylation of p38 MAPK in anti-CD3/anti-CD28-stimulated Jurkat cells to a degree that was similar to that seen with BCNU (Fig. 5 B). In contrast, higher concentrations of GTN (≥0.5 mM) or HP (≥0.5 mM) inhibited the activity of JNK and the phosphorylation of p38 MAPK (data not shown), indicating that these signal pathways are enhanced by moderately oxidative conditions but inhibited by strongly oxidative conditions.

To ensure that the regulatory redox effect operates also in primary T cells, we stimulated human peripheral T lymphocytes with anti-CD3 plus anti-CD28 Abs together with or without 20 μM BCNU. The results (Fig. 6) showed that the kinase activity of the endogenous JNK as measured by c-Jun phosphorylation as well as the phosphorylation of p38 MAPK were markedly enhanced by BCNU treatment (Fig. 6,A). In addition, BCNU treatment was found to cause a substantial increase in IL-2 production (Fig. 6 B).

To characterize the regulatory redox effects in more detail, we treated Jurkat T cells with BCNU in the absence or presence of the antioxidants NAC or PDTC, the 5-lipoxygenase inhibitor MK886, the phosphatidylinositol 3-kinase antagonist wortmannin, the calcineurin inhibitor cyclosporin A, the PTK inhibitor herbimycin A, and the vitamin E-analogous antioxidant trolox that prevents peroxidation of membrane lipids. These experiments revealed that the BCNU-induced JNK activation and p38 MAPK phosphorylation were not affected by trolox, MK886, wortmannin, or cyclosporin A, but were profoundly inhibited by NAC, PDTC, or herbimycin A (Fig. 7 A), indicating that one or more tyrosine kinase species may be involved in the activation cascade. The inhibitory effect of the structurally unrelated antioxidants NAC and PDTC provided additional support for the interpretation that the GSH system modulates the signaling cascade by redox regulation.

To determine the contribution of several immunologically important signaling proteins to the redox-sensitive signal process, we transfected Jurkat T cells with an expression vector encoding HA-tagged JNK together with dominant negative forms of Rac, Vav, MEKK1, MLK3, PKCΘ, and MKK7. These experiments showed that the activation of JNK by BCNU was markedly inhibited by all dominant negative proteins under test (Fig. 7 B). Because Vav, Rac, and PKCΘ are activated early during T cell activation (67), it is suggested that the redox-sensitive components are located upstream of these components at a very early point in the signal cascade. This conclusion was further supported by the fact that the activation of JNK by ectopic expression of Rac was not blocked by NAC (data not shown).

Because the previous experiments suggested collectively that the redox-sensitive target may be a relatively early component of the signal cascade, we studied the effect of BCNU on PTKs and on the tyrosine phosphorylation of signal proteins. Because the two PTKs Lck and Fyn were previously shown to be activated in T cells with a sulfhydryl-reactive reagent or high (i.e., physiologically not relevant) concentrations of HP (14, 15, 16, 68), we treated Jurkat cells again for various time periods with BCNU and determined at first the kinase activity of the endogenous Lck by immunoprecipitation and in vitro kinase assay. These experiments showed that BCNU stimulated (within 5 min) the autophosphorylation of Lck (Fig. 8,A) and the phosphorylation of its substrate protein enolase (data not shown). Similarly, BCNU stimulated also the autophosphorylation (Fig. 8,B) and kinase activity (data not shown) of Fyn. When cell extracts from BCNU-treated Jurkat T cells were analyzed by Western blotting with anti-phosphotyrosine Abs, another protein of 74 kDa was also found to be phosphorylated within 5 min after BCNU treatment followed by two proteins of 95 and 56 kDa (Fig. 8,C). The protein of 95 kDa was identified by immunoblotting and immunoprecipitation as the Vav protein (data not shown), which is known to be inducibly phosphorylated by Lck (43, 44) and Fyn (69). To determine whether GSH may interact directly with the Fyn kinase protein, we studied the effect of GSSG on rFyn in a cell-free system. The results showed that GSSG enhanced strongly the autophosphorylation of rFyn in a dose-dependent fashion (Fig. 8 D), whereas HP (0.5–2 mM), thioredoxin (10–150 μg/ml), or butylated hydroxyanisol (100–500 μM) had no conspicuous effect (data not shown).

Our first set of experiments has shown that moderately oxidative conditions as they are induced by 50 μM HP or 30 mM LAC enhance the transcription from the IL-2 promoter if (and only if) the cells are simultaneously exposed to a CD28 costimulatory signal together with or without an additional signal from the TCR/CD3 receptor. This synergism between costimulatory signal and redox effect may be important for the activation of the immune system by small amounts of Ag under natural conditions. Therefore, our more detailed analysis of the signal cascade was performed on cells that have been stimulated mostly with anti-CD28 Ab together with anti-CD3 Ab or PMA.

With these experimental conditions, we have then shown 1) that the transcription from the IL-2 promoter can be enhanced by moderate oxidative changes in the intracellular thiol pool of the T cells, and 2) that such moderately oxidative conditions are associated with the stimulation of a signal pathway that leads to the selective activation of the MAPKs JNK and p38 MAPK but not ERK/1 and ERK/2. The redox-sensitive components of this signal cascade were shown to be located at a very early point in the cascade and tentatively identified as p56 Lck and p59 Fyn. Most of these experiments were performed with a Jurkat T cell line, but similar results were also obtained with primary human T cells (Fig. 6). The TCR expression was not markedly altered in Jurkat or primary human T cells by BCNU at the relevant concentrations (25–100 μM) and after various time intervals (6–36 h) as tested cytofluorometrically with anti-CD3 Abs (data not shown).

The changes that were seen under these moderately oxidative conditions resembled only partly the effects of the strongly oxidative conditions that have been reported previously. In line with our results in this report, millimolar concentrations of HP were previously shown to enhance the activity of Lck (16). However, at the relatively high concentration used by these authors and especially if applied in combination with vanadate, HP may enhance indirectly the activated and autophosphorylated state of Lck and other PTKs of the Src family by inhibiting corresponding protein tyrosine phosphatases (19). Beiqing et al. (30) reported that HP caused a substantial increase in the expression of c-Jun and AP-1 DNA-binding activity in Jurkat cells but failed to activate the transcription factor NF-AT. HP even inhibited the transcription under control of NF-AT or the IL-2 promoter in cells that had been stimulated with PMA plus PHA. Because NF-AT activity requires the phosphorylation of c-Jun, this inhibitory phenomenon may be related to our observation that the activation of JNK and the phosphorylation of p38 MAPK were inhibited by higher (i.e., 0.5 mM) concentrations of GTN or HP (data not shown). Others have found that millimolar concentrations of HP cause also the activation of ERK1 and ERK2 in Jurkat T cells and cardiac myocytes (21, 22). These MAPKs were not detectably activated under our conditions. Activation of JNK and p38 MAPK in the absence of ERK1/2 activation was previously observed also in human 293 cells after treatment with the S-alkylating agent methyl methanesulfonate, and there was suggestive evidence for a determining role of the intracellular GSH pool also in those experiments (70). Taken together, these earlier studies and our experiments with 50 μM BCNU lead to the conclusion that the moderate oxidation or depletion of intracellular thiols has relatively selective effects on cellular signal cascades. BCNU inhibits the GSH reductase enzyme by interacting with a functionally important sulfhydryl group, but is not a nonspecific sulfhydryl reactive reagent (34). Because an experimentally induced decrease in the GSH:GSSG ratio by three different agents, i.e., by BCNU, HP, or the NO donor GTN, had essentially the same effect on the activation of c-Jun and p38 MAPK, it is suggested that the redox modulation of T cell activation involves a shift of the GSH redox state and/or the oxidative modification of certain thiol groups in signal proteins. Activation of JNK without concomitant activation of ERK2 has also been observed in T cells stimulated with natural CD28 ligands (71); a similar activation pattern was found in human skin fibroblasts after irradiation with UV-A (72), suggesting that UV-A may cause a similar type of oxidative stress with similar effects on redox-sensitive signal proteins.

Furthermore, our analysis of the signaling events in BCNU-treated cells indicated that the oxidative modification of certain thiol groups in one or a few early components of the signal cascade leads to the activation of Vav, NF-κB, JNK, and p38 MAPK. This has the interesting implication that many Vav/Rac-regulated processes, including the organization of cytoskeletal proteins (73), may be subject to regulation by redox changes in the cellular thiol pool. Our experiments with pathway-specific inhibitors excluded the possibility that these redox-mediated effects involve lipid peroxidation, 5-lipoxygenase, phosphatidylinositol 3-kinase, or Ca²⁺-dependent signaling processes. The redox-mediated activation of JNK and p38 MAPK appears to involve PTKs because the activation was completely inhibited by herbimycin A and by dominant negative versions of Vav, PKCΘ, and Rac, i.e., signaling components that are located downstream on PTKs.

Our experiments on the oxidative activation of rFyn by GSSG in a cell-free system (Fig. 8) suggest that Fyn is (one of) the redox-sensitive targets responsible for the redox modulation of T cell signaling. P59^fyn and p56^lck are PTKs of the Src family that are predominantly expressed in lymphoid cells and involved in Ag receptor signaling (74, 75, 76). It has been shown in various cells that oxidative stress causes formation of mixed disulfides between low m.w. thiols and sulfhydryl groups of certain cytosolic proteins (S-thiolation). Mixed disulfide formation with GSH (i.e., S-glutathiolation) was found to stimulate certain enzymes and to inhibit others (77, 78, 79, 80). The details of the chemical modification p59^fyn and GSSG remain to be analyzed but may involve the motif CPxxxxxxMxxCW (see Ref. 81) that is shared by the PTKs, p59^fyn, p56^lck, and p60^c-src. Similar motifs occur also in p69^ltk (82) and p72^syk (83).

The mild oxidative changes in the intracellular thiol pool occur not only in pathological conditions of oxidative stress but are likely to play an important role in the normal immune response under physiological conditions. In the immediate vicinity of activated macrophages, T cells are exposed to ∼10⁻⁴ M HP (25, 26, 27), i.e., a concentration that was found to cause a similar change in the GSH/GSSG ratio (Fig. 2,B) and a similar increase in IL-2 promoter activity (Fig. 1,A) as the treatment with BCNU (Figs. 2 and 3). Moreover, several types of leukocytes including activated macrophages and granulocytes perform glycolysis and release LAC into the extracellular space even under aerobic conditions (84, 85). The interstitial fluid in the immediate vicinity of activated macrophages is likely to contain 20–40 mM LAC (31). LAC concentrations may even systemically reach concentrations of 20–30 mM in conditions of lactic acidosis (86). Exposure of T cells to 10–30 mM lac was previously shown to decrease the intracellular GSH level, enhance the production of IL-2 (31, 32), and enhance also the activation of CTL (87). By amplifying the TCR-mediated signal transduction, the exposure of T cells to HP and LAC from activated macrophages may enable the infected host to start the immune response against an invading pathogen long before optimal doses of Ag accumulate. The decrease in the signal threshold may be critically important for the immune system to win the race with rapidly multiplying pathogens. Decreasing the threshold may also increase the risk of autoimmune processes. T cells isolated from the synovial fluid of patients with rheumatoid arthritis were recently found to have predominantly a decreased intracellular GSH level and the “primed” CD45RO phenotype (88).

Taken together, our experiments indicate that moderate changes in the GSH status as they are typically induced by several different macrophage products of small m.w. may amplify the TCR-mediated signal pathway at a very early point and, thereby, play an important auxiliary role in the immune system of higher organisms. The contribution of these oxidative changes to the functional changes of T cells in autoimmune processes and other pathological states remains to be investigated.

We thank E. Strohmeier, N. Erbe, and I. Fryson for their technical assistance.