Redox Remodeling as an Immunoregulatory Strategy^†

Redox Potentials in the Intra- and Extra-Cellular Compartments

The cytoplasmic and extracellular redox potentials are vastly different and influence the structure, stability and function of the macromolecules that reside in each compartment. Within the intracellular compartments, several redox buffers exist e.g. thioredoxin (Trx), glutathione (GSH) and cysteine and the relative concentrations of their oxidized versus reduced species sets the ambient redox poise for the system. Interestingly, the individual redox systems appear to be under kinetic control, are not in equilibrium with each other and independently regulate the redox status of their client redox partners (17, 18). Quantitatively, GSH is the major intracellular redox buffer and is found at concentrations ranging from 0.5-10 mM in mammalian cells (19). The intracellular GSH/GSSG (glutathione disulfide) redox potential in dividing cells is estimated to range from -260 mV to -230 mV and is progressively more oxidized in cells undergoing differentiation/growth arrest (-220 mV to -190 mV) or apoptosis (-170 mV to -150 mV) (17) (Fig. 2). The redox potentials for Trx1 in cytoplasm and nuclei are ∼-280 mV and -300 mV, respectively while the redox potential of mitochondrial Trx2 is estimated to range from -360 mV to -340 mV (18).

Extracellularly, the cysteine/cystine couple represents the major thiol/disulfide redox buffer. Plasma cystine and cysteine concentrations are reported to be 100-200 μM and 10-25 μM respectively and a redox potential of ∼-80 mV for this couple has been estimated for plasma in healthy humans (20). Paralleling the changes in the intracellular GSH/GSSG redox potential, increasing extracellular cysteine/cystine potentials are associated with cells undergoing proliferation (<-80 mV), differentiation/growth arrest (∼-80 mV) or apoptosis (0 mV to -80 mV) (Fig. 2). An age-dependent increase in the extracellular redox potential has been reported, which is also influenced by lifestyle choices such as smoking and by diseases such as AIDS (17). In contrast to the intracellular compartment, the GSH concentration in the extracellular space is very low (2-4 μM in human plasma). A major fate of secreted GSH is cleavage to its component amino acids, glutamate, cysteine and glycine. The cysteine thus released is a major source of extracellular cysteine and cystine, which is formed rapidly in the oxidizing milieu of this compartment (21).

Although dynamic regulation of the extracellular redox potential, which is linked to intracellular metabolism, has an important bearing on cell function, it is less well-studied and appreciated than intracellular redox control and its perturbations in pathological states. Reactive cysteines on proteins can be reversibly oxidized to sulfenic acids or form disulfide bonds, which can induce changes in their structure and function and elicit downstream effects in redox signaling pathways (22, 23). Disulfide bonds on ectodomains of membrane proteins and in secreted soluble and matrix proteins form a dynamic scaffold that can be reorganized by their shuffling or by their reduction (24). It has been proposed that a general loosening of the extracellular disulfide crosslink scaffold might precede cell division (25, 26). Cancer cells typically have higher membrane thiol levels in comparison to nontransformed cells, and it is speculated that this might facilitate higher proliferative rates (27). The redox status of specific membrane proteins influences their transport or receptor activity (17). For instance, CD4 (cluster of differentiation 4), a glycoprotein found on the surface of helper T cells that is used as a receptor by HIV-1 for gaining entry, has a redox sensitive disulfide bond in one of its four immunoglobulin-like domains (D2). T cell activation shifts the equilibrium from the disulfide to the dithiol state (28). Locking the dithiols in the D2 domain by chemical modification blocks HIV-1 entry indicating that a redox-linked conformational change in CD4 is critical for viral penetration into T cells (28).

T Cell Induced Extracellular Redox Remodeling by Dendritic Cells

The physiological relevance of extracellular reductive modeling during an adaptive immune response is supported by the dramatic increase in free thiols in lymphoid tissue following immunization (29). Under these conditions, enhanced nonprotein thiol staining is observed both inside cells and in the extracellular space. In contrast, Peyer's patches from the gut show virtually no staining for nonprotein thiols under these conditions, consistent with the antigenic hyporesponsiveness of this intestinal microenvironment (11, 12).

The magnitude of extracellular cyteine accumulation during activation of T cells increases with time and with the DC to T cell ratio and requires sustained contact between DCs and T cells. Increased cell surface thiols on T cells is correlated with increased production of the cytokine, IL-2, in vitro and enhanced proliferation in vivo (30). Naïve T cells require cysteine for GSH synthesis. However, cysteine is the least abundant of all amino acids in circulation (31) and naïve T cells are unable to import cystine efficiently due to the absence of the cystine transporter, x_C^- (32), thus creating a metabolic dependence on antigen presenting cells to meet their cysteine needs. Antigen presenting cells possess the x_c^- antiporter that uses the glutamate gradient to drive import of cystine, which is subsequently converted to cysteine in the reducing intracellular milieu and is ultimately secreted into the extracellular space. In addition to stimulating cysteine secretion, the interaction between antigen presenting cells and T cells results in the appearance of extracellular Trx1 (9). Trx1 is secreted by several cell types via a nonclassical leaderless secretory pathway under conditions of oxidative stress and inflammation (33). Secreted Trx1 does not appear to play a direct role in reduction of extracellular cystine leading to cysteine accumulation during T cell activation (10). Extracellular Trx1 interacts in a redox-sensitive manner with the TNF receptor superfamily member 8 (34) and exhibits proinflammatory effects by stimulating cytokine release and proliferation of lymphocytes (33, 35).

The pathway for extracellular cysteine accumulation during co-culture of DCs and naïve T cells has been mapped recently (10). In principle, two metabolic routes could be considered to lead to enhanced cysteine accumulation outside the cell (Fig. 3): (i) the transsulfuration pathway (blue), which provides an avenue for conversion of methionine to cysteine, and (ii) import of cystine into the cell where it is rapidly reduced to cysteine and converted to GSH, which is subsequently secreted and degraded by the ectoenzymes γ-glutamyltranspeptidase and a dipeptidase. Expression of the catalytic subunit of the system x_c- transporter is induced when DCs during co-cultivation with naïve T cells and is correlated with increased extracellular cystine clearance (10). Metabolic labeling and pharmacological inhibition studies have established the involvement of the convoluted metabolic pathway originating in cystine and culminating in GSH-derived cysteine as the source of extracellular cysteine provided by DCs (10). This pathway demonstrates the dynamic interplay between the intra- and extra-cellular compartments for redox homeostasis via interconnected but independent redox nodes, i.e., GSH and cysteine.

The extracellular cysteine/cystine redox potential for DCs in culture is ∼-80 mV, a value that is consistent for cells experiencing growth arrest (10). Naive T cells in culture that have not received activation signals are fated to undergo apoptosis and exhibit an extracellular cysteine/cystine redox potential of ∼-45 mV. In contrast, when naïve T cells receive activation signals during co-culture with DCs, a more reducing extracellular environment reflected in a redox potential of -110 mV (at 36 h), results. This redox potential change is consistent with conditions that are conducive for T cell proliferation (10).

In addition to triggering intracellular signaling pathways, engagement of DCs and T cells during activation leads to dynamic changes in the redox status of exofacial proteins in both cell types. A 30 mV potential shift is expected to lead to a 10-fold change in the ratio of reduced:oxidized cysteines in proteins. Indeed, enhanced cell surface labeling of protein thiols with the fluorescent dye, Alexa-maleimide, is seen during co-culture of DCs and T cells (by ∼4 and 8-fold respectively) as visualized by confocal microscopy and quantified by FACS analysis (10).

Redox Signaling During T cell Activation

Paralleling reductive remodeling of the extracellular redox poise with consequent effects on the exofacial protein thiol status and intracellular redox metabolism, is the initiation of a flurry of redox-active signaling across the immune synapse. The timing and balance between oxidative and reductive responses to engagement of antigen presenting cells and T cells are important for modulating activation, proliferation and apoptosis of T cells. At low levels, ROS (reactive oxygen species) e.g. H₂O₂ and O₂^•-, are considered to be mitogenic and their downstream effects are commonly mediated via changes in protein phosphorylation and/or activation/inhibition of transcription factors (36). Crosslinking of the TCR and the co-stimulatory molecule, CD28, results in enhanced intracellular H₂O₂ production that is needed for NF-κB activation and IL-2 and IL-2 receptor α chain gene transcription (37) and is consistent with an important role for ROS in the immediate early events during activation. Significant sources of ROS include membrane-bound NADPH-dependent oxidase, lipoxygenase and the mitochondrial respiratory chain (Fig. 4A, red arrows). However, sustained pro-oxidant conditions inhibit T cell proliferation and promote apoptosis (38).

During activation, increased ROS levels launch an early pro-oxidant response that is relayed via signaling pathways in antigen presenting cells and in T cells and result in activation of protein tyrosine kinases (e.g. Fyn, Src and Lck in T cells (Fig. 4B)), oxidative inhibition of protein tyrosine phosphatases e.g. SHP1 and activation of transcription factors e.g. NF-κB (39). The NF-κB pathway regulates the expression of various inflammatory genes including cytokines, chemokines and costimulatory molecules. We speculate that as a consequence of an initial increase in ROS levels by mechanisms that are not clear, NF-κB is activated in DCs and stimulates GSH biogenesis (via activation of γ-glutamylcysteine ligase (40)) (Fig. 4A, red arrows). Increased GSH synthesis is both an autocorrective reaction to oxidizing conditions and initiates the next response phase, i.e. an antioxidant wave (Fig. 4A, blue arrows). We hypothesize that the NF-κB signaling pathway is important for stimulating extracellular cysteine accumulation (10). The combined effect of these cellular responses would be the initiation of an antioxidant response leading to a reductive milieu both in the intra- and extra-cellular space that is conducive to T cell proliferation. The importance of plasticity in redox remodeling during T cell activation is supported by the observation that deficiency of Ncf1 encoding neutrophil cytosolic factor 1 (or P47phox), the activating protein in the NADPH oxidase complex, results in a reduced capacity for ROS genesis, increased cell surface thiols and enhanced T cell autoreactivity in an arthritis model (30).

GSH serves as an important proliferative signal in T lymphocytes (41) and is required for cell cycle progression from the G1 to S phase (42). It is needed for the activity of ribonucleotide reductase and therefore, for DNA synthesis (43). Furthermore, the activities of telomerase (44) and of key transcriptional factors, e.g. NF-κB and AP1 (45), and cell cycle proteins e.g. Id2 and E2F4 (44) are redox regulated. GSH is concentrated in the nucleus during the early phase of cell proliferation and becomes more evenly distributed in confluent cells (46). GSH regulates nuclear protein function via glutathionylation and protects DNA from oxidative damage during the key stage of replication (46). GSH affects ROS levels in cells, which can either activate or inactivate specific redox sensitive targets at cell cycle checkpoints, thereby influencing cell fate (47). Interestingly, increased synthesis and nuclear sequestration of GSH and decreased sensitivity to apoptosis were observed in response to overexpression of the B cell leukemia/lymphoma 2 (Bcl-2) protein in HeLa cells (48).

Redox-sensitive signaling cascades are also elicited in T cells upon activation. For instance, a 10-30% decrease in intracellular GSH in peripheral T lymphocytes completely abrogates T cell receptor-stimulated calcium signaling (49). The adaptor protein linker for activation of T cells (LAT) (Fig. 4B), a membrane protein that plays a central role in signal transduction during T cell activation, is also influenced by the intracellular redox status (50). Marked diminution in intracellular GSH level as seen under chronic oxidative stress conditions, causes a conformational change in LAT, apparently via formation of an intramolecular disulfide bond, and results in its displacement from the membrane (50). This cytoplasmic relocalization results in failure to phosphorylate in response to T cell activation and derails the signal transduction cascade that leads eventually to expression of IL-2 and other genes. This redox-sensitive conformational displacement is associated with the hyporesponsive phenotype of synovial T cells in rheumatoid arthritis because of their depleted antioxidant capacity resulting from the chronic inflammation associated with this disease of the joints (50).

In summary, redox responsive signaling networks during T cell priming involves dynamic and spatially regulated changes in the intra- and extra-cellular compartments and comprises both small molecules (e.g. ROS and redox-active metabolites) and proteins. Redox signaling has several important implications for T cell biology (36). Hypoxic conditions as encountered in poorly oxygenated tumors might limit the efficiency of T cell priming and contribute to their anergic phenotype in this environment. Alternatively, a pro-oxidant environment resulting from ROS production by active neutrophils might facilitate priming of T cells but, if overwhelming, impair signaling via inhibitory signals such as tyrosine phosphatases or the NF-κB inhibitor, IkB. Additionally, redox signaling appears to influence T cell commitment to the Th1, Th2 and regulatory T cell phenotypes (51, 52).

Regulatory T Cells Interfere with Redox Remodeling by Dendritic Cells

The immune system balances the host's needs for microbial and tumor immunity with keeping autoimmunity in check (2). To achieve self-tolerance, T cells are “educated” in the thymus and autoreactive T cells are destroyed. However, a small fraction of self-reactive T cells escape from the thymus into the periphery, and if left unchecked, can cause autoimmune diseases (53). Naturally occurring CD4⁺CD25⁺Foxp3⁺ regulatory T cells, which comprise about 5%-10% of total CD4⁺ T cell population, suppress autoreactive T cells to maintain immune tolerance (54).

Sakaguchi and coworkers made the groundbreaking discovery of this distinct T cell subpopulation in 1995 and demonstrated that depletion of the CD25⁺ population from the CD4⁺ T cells induced autoimmunity when T cells were transferred to the immunodeficient nude mice (1). In contrast, transfer of the CD4⁺CD25⁺ T cells together with the CD4⁺CD25^- T cells prevented autoimmune diseases. Besides the role of regulatory T cells in controlling autoimmunity, they also play important roles in controlling anti-microbial, anti-tumor responses and transplantation immunity (54). Regulatory T cells mature in the thymus, migrate to lymph nodes and are activated by self or nonself antigen-presenting cells. The homing receptors on regulatory T cells enable them to traffic to sites of infection to control immune responses (2). Regulatory T cells also suppress the activation and proliferation of B cells, DCs and natural killer cells by mechanisms that remain to be fully elucidated (55).

Some of the strategies used by regulatory T cells for mediating their suppressive effects (2, 3) are shown in Fig. 5 and include: (i) secretion of inhibitory cytokines viz. TGFβ, IL-35 and IL-10 (56-58), (ii) cytolytic suppression by secretion of the proteases granzyme-A or granzyme-B (59, 60), (iii) metabolic disruption e.g. by direct transfer of cAMP to effector T cells (61) or by secretion of pericellular adenosine (62), which inhibits effector T cell functions and enhances induced regulatory T cell generation, (iv) suppression of DC maturation and/or function (63) by induction of indoleamine 2,3-dioxygenase, which catalyzes the rate-limiting step in tryptophan catabolism and creates in turn, a shortage of this essential amino acid for effector T cells (64) and (v) by interfering with extracellular reductive redox remodeling by DCs during T cell activation (10). The panoply of suppressive strategies identified to date for regulatory T cell raises questions about their relative importance and how they are integrated in vivo. In the proposed “hierarchical” model, one or few master mechanisms govern regulatory T cell suppressive functions in various physiological settings (3). Alternatively, in the “contextual” model, the microenvironment and tissue compartment govern the suppressive strategy that is deployed resulting in the differential contribution of a given mechanism in different disease models (3).

In contrast to naïve T cells, co-culture of regulatory T cells with DCs does not affect extracellular cysteine concentration. However, regulatory T cells suppress cysteine accumulation in the extracellular compartment when added to co-cultures of DCs and naive T cells. As a consequence, both intracellular (diminished GSH levels in T cells) and extracellular (diminished cell surface thiol labeling on T cells and on DCs) perturbations in the redox status result (10). Remarkably, although regulatory T cells are known to mediate their suppressive functions by multiple strategies, provision of a single reagent, i.e. exogenous cysteine at concentrations seen under DC-T cell co-culture conditions, alleviates inhibition of T cell proliferation (10). This observation begs the question as to whether redox regulation serves as a master switch in the multipronged suppressive action of regulatory T cells.

We posit that the redox changes in the intra- and extra-cellular compartments influence one or more of the well known regulatory T cell suppressive mechanisms. For instance, the anti-inflammatory cytokine IL-10 has antioxidant properties (65) and TGFβ, a multifunctional cytokine is redox regulated (66). Activation of latent TGFβ requires reductive cleavage of a disulfide bond that links it to the latency-associated peptide, but over-reduction, leads to formation of inactive TGFβ monomers. Thus activation and inactivation of TGFβ are subject to redox control and dynamic changes in the extracellular redox milieu might be important for regulating TGFβ activity (67). Furthermore, granzyme A, the cytolytic T cell protease, cleaves redox factor 1 (Ref-1), which in turn enhances cell death (68). Thus, redox control may be integral to regulatory T cell mediated suppression mechanisms and could be more pervasive than previously recognized.

In Vivo Studies and Therapeutic Implications

The redox status of secondary lymphoid organs such as lymph nodes and spleens are more reduced than non-lymphoid organs (29). The nonprotein thiol content in lymphoid tissues is reported to increase in response to immunization with DCs, B cells and macrophages, contributing to the reductive remodeling (29, 69). It is speculated that the reducing microenvironment might protect lymphoid organs from oxidative stress during T cell activation and antibody production (70, 71). However, low levels of ROS are essential for the onset of the immune response. In vivo treatment of mouse models with catalytic antioxidants (manganese porphyrin derivatives) causes inefficient CD4⁺ T cell activation and proliferation by inhibiting ROS generation in antigen presenting cells (72). The catalytic antioxidants inhibit DNA binding by NF-κB and subsequent production of proinflammatory cytokines (73). Redox modulation by catalytic antioxidants also suppresses CD8⁺ T cell functions such as proliferation and lysis of target cells (74).

The x_c^- cystine transporter, which transports cystine using the glutamate gradient, plays an important role in redox-based immunoregulation. Under normal conditions, lamina propria macrophages are unable to transport cystine and secrete cysteine because they lack the x_c^- transporter (12). In inflammatory bowel disease, local recruitment of peripheral blood monocytes which exhibits high expression of the x_c^- transporter leads to extracellular cysteine accumulation and hyper-reactivity of lamina propria T cells (12). Furthermore, lymphoma cells, which cannot import cystine like naïve T cells, depend on tumor-associated somatic cells such as activated macrophages and DCs for their cysteine supply. Inhibition of the x_c^- transporter by sulfasalazine inhibits growth of lymphoma cells and tumor progression (75). Overexpression of the x_c^- transporter in lymphoma cells greatly increases intracellular and extracellular cysteine levels, protecting cells from oxidative stress induced cell death (76).

Redox modulation as a strategy for immunoregulation has been used in several diseases. HIV infects and kills CD4⁺ T cells, leading to a significant decrease in functional CD4⁺ T cells in AIDS. HIV infected individuals have lower cellular and plasma GSH levels compared with healthy controls, which correlates with low T cell numbers and deficient function. Administration of N-acetyl cysteine, a cysteine precursor, restores intracellular GSH levels and has shown benefits for HIV-infected individuals (77). Sulfasalazine is used in the treatment of T cell mediated autoimmune diseases such as inflammatory bowel disease and rheumatoid arthritis. It decreases proliferation of autoreactive T cells by inhibiting the x_c^- cystine transporter on antigen presenting cells thereby perturbing the redox environment (78).

Since regulatory T cells play a central role in suppression of various immune responses, manipulation of their function is an important strategy for immune intervention. Enhancing regulatory T cell function in autoimmunity, allergy, transplantation and pregnancy disorders can diminish unwanted immune responses. On the other hand, attenuating regulatory T cell function in cancer and microbial infection may be desirable (79). The recent identification of a novel immunosuppressive strategy deployed by regulatory T cells, which impacts the intra- and extracellular redox environments during T cell activation (10), illuminates a new therapeutic target.

Concluding Remarks

Redox modulation has emerged as a key regulatory strategy in the adaptive immune system. It has long been known that T cell activation and proliferation require a reducing milieu. This reducing microenvironment is shaped primarily by the metabolic activity of antigen presenting cells, especially DCs. Interaction of DCs with naïve T cells stimulate cystine consumption and cysteine accumulation in the extracellular space, which produces an extracellular redox potential suitable for T cell proliferation (9, 10). A more reducing extracellular redox potential is reflected in the increased T cell surface thiol status (10). The specific membrane targets of redox remodeling and their effects on T cell biology, i.e. activation and proliferation, remain to be elucidated. The greater availability of extracellular cysteine also influences the intracellular antioxidant capacity within T cells since cysteine limits GSH biosynthesis. Consequently, intracellular GSH levels rise and in turn, influences T cell signal transduction pathways and gene expression. The choreography of GSH localization and the GSH/GSSG redox potential changes during T cell activation, and their correlation with the onset and operation of signaling pathways and cell cycle progression await elucidation.

Modulation by regulatory T cells of the extracellular redox microenvironment during T cell activation could be mediated by one or more mechanisms. For instance, by limiting cysteine availability, regulatory T cells deprive effector T cells of a building block needed for protein and GSH synthesis. Alternatively, by perturbing the redox environment, regulatory T cells can have both indirect effects by enhancing other suppressive mechanisms used by them (as discussed above) and direct effects on T cell activation and proliferation targets, which are sensitive to the redox potential and the redox status of key signaling proteins. Many questions remain to be addressed regarding how regulatory T cells inhibit reductive remodeling by DCs. For instance, do they interfere with cystine uptake, inhibit the cysteine secretion pathway or simply compete with effector T cells for the extracellular cysteine pool? Is there a connection between the mechanism for perturbing redox remodeling and Foxp3 expression and what is the extent of cross-talk between the other suppressive mechanisms and redox remodeling? And finally, what is the physiological relevance of the redox remodeling mechanism in normal and disease states? The answers to these questions will help illuminate the biology of regulatory T cell suppressive mechanisms and identify potential therapeutic targets.