Metallothionein regulates intracellular zinc signaling during CD4⁺ T cell activation

Metallothioneins (MT) are low molecular weight, high cysteine content proteins that are expressed in most mammalian cells. Their upregulation in response to increases in intracellular zinc ion concentration ([Zn²⁺]_i), reactive oxygen species (ROS), pro-inflammatory cytokines, and as part of proliferation and differentiation [1] suggests that MT may play an important role during the development of immune responses. Specifically, we hypothesize that upregulated MT expression can affect both intracellular redox and zinc signaling events that occur during CD4⁺ T helper cell activation. The release of zinc ions by MT in response to redox signaling may play a role in immune responses by directly influencing zinc signaling pathways during CD4⁺ T helper cell activation.

The intracellular labile zinc concentration has recently been recognized as an important component of T cell activation [2]. Modest increases in [Zn²⁺]_i (in the range of 200pM–1nM) can have profound effects on signaling pathways originating from the T cell receptor (TcR) [3], cytokine receptors [4, 5] and toll-like receptors [6]. Increases in [Zn²⁺]_i inhibit the activity of selected intracellular kinases [7–9] and phosphatases [10, 11] which, in turn, affect STAT and MAPK signaling networks and their associated transcription factor activity. Regulating [Zn²⁺]_i, therefore, has the potential to dramatically influence the development of CD4⁺ T cell effector function [12–14] and diseases that arise from abnormal CD4⁺ T cell activity.

MT binds and releases zinc under physiologically relevant conditions and, in concert with the families of cell-specific zinc importers (Zip) and exporters (ZnT), is part of a network that tightly controls intracellular [Zn²⁺]_I [15]. The intracellular pool of MT is a mixture of oxidized, reduced and zinc bound forms of the protein [16, 17], and the dynamic equilibrium of these forms regulates zinc availability in response to intracellular ROS. Interestingly, zinc release from MT is possible even in the reducing environment of the cytosol through reactions with selenium redox catalysts [18, 19]. MT contains 20 cysteine residues which impart a very low redox potential (−366 mV), making it a preferential target for oxidation when compared with other free thiols including glutathione (GSH) [20, 21]. Of the seven zinc binding sites in MT, three have a comparatively weak metal binding affinity (logK = 10–7.7) [22] which facilitates zinc release upon thiol oxidation and results in a downstream effect on enzyme activity [23]. This property uniquely positions MT to transduce ROS signals into zinc signals within the narrow ranges of redox and [Zn²⁺]_i fluctuations that occur during CD4+ T cell activation.

Naïve CD4⁺ T cells express low levels of MT [24, 25]. Upon activation, [Zn²⁺]_i increases over time as a result of the release of zinc stored in vesicles into the cytoplasm [12] and from increased zinc import from the extracellular environment [3, 26]. This sustained zinc signal [27] induces the expression of MT which supports T cell proliferation [26]. During primary CD4⁺ T cell activation and differentiation, extracellular cytokines, including TNF-alpha, IL-1, IL-6, and IL-27 [28–30] and glucocorticoids [31], also influence the expression of MT and simultaneously drive T cell differentiation. Cells with an increased concentration of cytoplasmic MT are then positioned to more efficiently convert ROS signals into zinc signals [32] during subsequent activation events, which has the potential to affect CD4⁺ T helper cell function.

In this report, we confirm that MT expression following the primary activation of naïve CD4⁺ T cells is influenced by the cytokine IL-27 [30]. Furthermore, we show that expression of MT provides CD4⁺ T cells with a redox-sensitive pool of intracellular zinc that can be mobilized under conditions of oxidative stress or in response to intracellular ROS generation following signaling through the TcR. In the absence of functional Mt1 and Mt2 genes (MT ^-/- ), the increase in [Zn²⁺]_i following redox signaling is reduced, and this results in decreased p38 activation in MT ^-/- cells which can be rescued by pharmacologically increasing [Zn²⁺]_i. These results demonstrate that MT plays a role in CD4⁺ T cell activation by transducing ROS signals into an increased [Zn²⁺]_i that subsequently affects downstream effector function.

Activation and proliferation of CD4⁺ T cells is associated with an increase in the concentration of intracellular labile zinc ions ([Zn²⁺]_i) [33] and the expression of metallothioneins (MT) [25]. Manipulating [Zn^2+] _i [13, 34] or MT expression [30, 35] during activation affects cell signaling networks and cytokine secretion patterns. In elderly populations, a decreased ability to regulate increases in [Zn²⁺]_i following CD4⁺ T cell activation results in increased MT expression and altered T cell function [26, 36]. This suggests that zinc and MT are coordinately regulated during activation and this allows CD4⁺ T cells to respond appropriately in different environments.

To determine the degree to which CD4⁺ T cells regulate [Zn²⁺]_i and MT expression during activation and effector cell development, naïve CD4⁺ T cells were stimulated using anti-CD3 and anti-CD28 antibodies in the presence or absence of IL-27 to promote the development of Tr1 or Th0 phenotypes, respectively [37, 38]. In both culture conditions, expression of CD25 served as an activation marker and was increased by 24 h post-stimulation (Fig. 1a). After 6 days of culture, CD25 was expressed by >95 % of CD4⁺ T cells in both conditions, (Fig. 1b) indicating cell activation was not reduced in the absence of IL-27 signaling.

An increase in [Zn²⁺]_i in CD4⁺ T cells activated under Th0 inducing conditions was observed after 6 days (Fig. 1c). This increase followed the expression of CD25, indicating naïve CD4⁺ T cells maintain zinc homeostasis in the time period following initial activation events. In the presence of IL-27, [Zn²⁺]_i was increased by 4 days post activation (Fig. 1c) and after 6 days was significantly higher than in cells induced under Th0 conditions (Fig. 1d), suggesting a link between IL-27 signaling and the regulation of [Zn²⁺]_i during primary activation. Next, we measured intracellular MT expression during the same activation window. A modest increase in MT expression was observed 6 days after activation under Th0 inducing conditions while MT expression was increased significantly when cells were activated in the presence of IL-27 (Fig. 1e). This increased MT expression persisted for at least two days after cells were transferred and cultured in media alone (Fig. 1f), identifying a 48 h temporal window where altered MT expression could potentially play a role during CD4⁺ T cells re-activation. Interestingly, the increase in intracellular MT in naïve CD4⁺ T cells 3 days after activation preceded a measurable increase in [Zn²⁺]_i, indicating MT induction did not require an increase in [Zn²⁺]_i under these culture conditions. However, the highest level of MT expression and [Zn²⁺]_i both occurred on day 6 (Fig. 1c, e), supporting the well-established link between increased [Zn²⁺]_i and MT induction.

To determine if MT gene dose affects intracellular zinc homeostasis during the development of Tr1 cells, the [Zn²⁺]_i in Mt1Mt2 knockout cells (MT^-/-) was compared with wildtype congenic cells (MT^+/+) during naïve CD4⁺ T cell activation. The same pattern of [Zn²⁺]_i increase following activation was observed in both MT^+/+ and MT^-/- CD4⁺ T cells at each of the stages of Tr1 cell differentiation (Fig. 1g), indicating that MT did not significantly affect intracellular labile zinc homeostasis under these activation conditions. [Zn²⁺]_i was reduced when proliferating lymphoblasts were resuspended in fresh media with no additional stimulation for 2 days, demonstrating that the increase in the [Zn²⁺]_i above 500 pM is transient and associated with activation and the lymphoblast phenotype.

The increased intracellular pool of zinc-MT that is present after the development of the CD4⁺ Tr1 cell effector phenotype is a potential reservoir of zinc that can be mobilized during reactivation. We assessed whether this MT-bound zinc pool could be released in response to reactive oxygen species (ROS) and affect [Zn²⁺]_i. Exposing CD4⁺ Tr1 cells to hydrogen peroxide (H₂O₂), a biologically relevant oxidant associated with T cell activation [39], for 7 min resulted in an increase in [Zn²⁺]_i in a dose dependent manner (Fig. 2a). This size and timing of this increase was unaffected by the addition of 20 μM ZnSO₄ to the media during the time tested (Additional file 1: Figure S1), confirming that the increase in [Zn²⁺]_i was not due to an influx of extracellular zinc. Oxidant-induced release of zinc was significantly greater in MT^+/+ CD4⁺ T cells than in MT^-/- CD4⁺ T cells during the initial 7 min of exposure when compared by linear regression analysis (p < .05) (Fig. 2b and c).

To confirm that the oxidant-releasable zinc was sequestered via a thiolate bond, the thiol-specific oxidant DTDP was added to cells and the release of zinc was measured over time. As with H₂O₂, the release of Zn²⁺ following DTDP exposure was greater in MT^+/+ Tr1 cells than in MT^-/- Tr1 cells (p < .05) (Fig. 2d). The robust release of Zn²⁺ following either oxidant exposure, irrespective of MT gene dose, confirms that MT is not the only redox-sensitive mobilizable pool of intracellular zinc (Fig. 2d). When a pool of zinc-MT is present, however, exposure to ROS results in a faster and larger increase in intracellular [Zn²⁺]_I, indicating that MT contributes to a more efficient transduction of an oxidant signal into a zinc signal compared with other zinc-bound proteins. This additional labile zinc has the potential to inhibit the activity of several important T cell protein tyrosine phosphatases [4, 10] and affect downstream cell signaling.

Global changes in intracellular redox buffering capacity also have the potential to affect oxidant-induced zinc release from protein thiols [40]. To determine if MT gene dose influences the total intracellular redox buffering capacity, total reduced cellular thiols were quantified in MT^-/- and wildtype control MT^+/+ CD4⁺ naïve and effector T cells by CPM assay [41]. The concentration of reduced thiols was higher in CD4⁺ Tr1 effector cells compared with naïve lymphocytes (p < .05), but was not different between MT^-/- and MT^+/+ Tr1 cells (Fig. 3a), indicating that the expression of MT did not significantly alter the total intracellular free thiol concentration. This is not unexpected, given that reduced glutathione exists in >1000 fold molar excess compared to MT in CD4⁺ T cells [25, 42]. Additionally, MT expression did not contribute to a significant difference in the total redox buffering capacity, as shown by the similar level of oxidation of the fluorescent probe CM-H₂DCF in MT^-/- and MT^+/+ Tr1 cells after exposure to H₂O₂ (Fig. 3b). These results suggest that MT is not a significant source of total intracellular free thiols under these cell culture conditions and that CD4⁺ T cells maintain their intracellular redox potential in the absence of MT. However, when MT is present as part of the intracellular pool of free thiols, cells maintain a greater zinc release potential in response to ROS. In this way, MT serves to amplify zinc signals that are transduced by an initiating oxidant signal without affecting total intracellular redox buffering capacity or the oxidant signal itself.

Intracellular ROS generation is an important component of effector CD4⁺ T cell signaling following TcR activation [39, 43, 44] and is a requirement for effector function and expression of important regulatory molecules including Fas ligand (FasL). Hydroxyl radicals, superoxides, hydrogen peroxides, and nitric oxides [45], have all been shown to release Zn²⁺ from MT either directly [20, 21, 46], or indirectly via a mechanism where reduced glutathione (GSH) is oxidized to glutathione disulfide (GSSG), which then oxidizes MT [47]. To determine if ROS production is affected by the presence of MT, CD4⁺ Tr1 cells were stimulated with anti-CD3 cross-linked with anti-IgG, which is sufficient to stimulate ROS production in CD4⁺ T cells [43]. CD4⁺ Tr1 cells produced a robust intracellular ROS signal within 30 min of TcR stimulation (Fig. 4a) and ROS persisted for at least 2 h indicated by the oxidation of the CM-H₂DCF probe. Oxidation of the probe after activation was not significantly higher in MT^-/- cells compared with MT^+/+ control cells, further supporting the conclusion that MT is part of a larger intracellular redox-buffering network and is not essential for maintaining cellular redox buffering capacity.

Despite similar levels of ROS generation following activation, the increase in [Zn²⁺]_i after 1 h of stimulation was greater in MT^+/+ Tr1 cells (Fig. 4b) compared with MT^-/- Tr1 cells. Similar to exogenous hydrogen peroxide exposure (Fig. 2), this increase was not affected by the addition of 20 μM ZnSO₄ during the initial 3 h time period, indicating the difference in zinc levels was not due to a zinc influx from the extracellular environment (data not shown). The increase in [Zn²⁺]_i following reactivation of Tr1 cells with anti-CD3 continued for at least 48 h (Fig. 4c) and the peak [Zn²⁺]_i was higher during restimulation than during primary activation (Figs. 1d and 4d). After 8 h of restimulation, the increased [Zn²⁺]_I was no longer greater in MT^+/+ Tr1 cells compared with MT ^-/- Tr1 cells. This suggests that while immediate increases in [Zn²⁺]_i following reactivation are affected by the release of Zn²⁺ from the intracellular stores, [Zn²⁺]_i 12 h after restimulation Zn²⁺ is predominantly regulated by ion transporters associated with T cell activation [12, 33]. Interestingly, CD4⁺ effector cells differentiated under Th1 conditions achieved an even higher [Zn²⁺]_i after 48 h restimulation compared with either Tr1 or Th0 cells, although there was no difference associated with MT genotype (Additional file 2: Figure S2). This suggests that different CD4⁺ T helper subsets regulate [Zn²⁺]_i differently to mediate effector function.

Intracellular labile zinc can influence T cell signaling by directly affecting both the STAT [9, 48] and MAPK pathways [2, 4]. The co-stimulatory effect of Zn²⁺ on p38 MAPK signaling in CD4⁺ T cells has recently been observed following signaling through both the T cell receptor [7] and the IL-1 receptor [49]. During activation, relatively small increases in [Zn²⁺]_i can inhibit protein tyrosine phosphatase activity [3, 10], which results in increased phosphorylation of p38 MAPK [50]. Other MAPK proteins are also affected by [Zn²⁺]_i following stimulation through the TcR including ERK1/2, but we chose to investigate p38 because the inhibitory phosphatase that regulates p38 phosphorylation is sensitive to the picomolar increases in [Zn²⁺]_i that were observed when MT was present following restimulation of Tr1 cells. To determine if a co-stimulatory effect of Zn²⁺ on p38 activation is associated with a ROS-mediated release of Zn²⁺ from MT, phosphorylated p38 was measured in MT^+/+ or MT^-/- Tr1 cells during re-stimulation through the TcR in vitro. Phospho-p38 levels peaked at 10 min post re-stimulation with anti-CD3 antibody, irrespective of MT gene dose (Fig. 5a). This increase was significantly higher in MT^+/+ Tr1 cells compared with MT^-/- Tr1 cells and was not a result of differences in total p38 levels between the two strains (Additional file 3: Figure S3).

An association between MT gene dose and increased p38 activation was also observed when cells were stimulated with the protein kinase C activator PMA. Peak levels of phospho-p38 were observed after 20 min of PMA stimulation in both MT^+/+ and MT^-/- Tr1 CD4⁺ cells and were again significantly higher in cells that express MT (Fig. 5b). PMA treatment did not appear to induce a zinc influx from the extracellular environment because p38 activation was not affected by increasing the extracellular [Zn²⁺] during stimulation with PMA (Fig. 5c). However, PMA does promote ROS generation [44] and a subsequent increase in [Zn²⁺]_i [51] in CD4⁺ T effector cells. This suggests zinc released from MT by ROS increases p38 activation and represents a novel mechanism for regulating p38 activation through expression of MT.

Differences in p38 activation mediated by MT were overcome with the addition of ZnSO₄ together with the zinc ionophore pyrithione, indicating MT^-/- Tr1 cells were capable of achieving levels of p38 activation similar to MT^+/+ cells when sufficient intracellular free zinc was present. Furthermore, differences in MT expression did not affect the activation of p38 when the available pool of intracellular labile zinc was removed by adding the zinc chelator TPEN (Fig. 5c). The observation that p38 activation was not affected by MT gene dose under conditions of similar intracellular zinc availability, via addition of a zinc specific ionophore or chelator, suggests that MT expression affects p38 activation through the regulation of available intracellular labile zinc.

The downstream effects of increased [Zn²⁺]_i can include altered effector phenotype and cytokine secretion [9, 13, 48]. There are reports that MT gene dose can influence IL-10 expression, but these effects have not been consistently observed [30, 35]. To determine if MT gene dose affects the expression of cytokines, we measured secretion of IL-10 and IFN-gamma from MT^+/+ and MT^-/- Tr1 cells. Secretion of IL-10 was detected in the media by 6 h post stimulation and was greater in MT^-/- Tr1 cell cultures compared with MT^+/+ controls at 24 and 48 h post stimulation (Fig. 6a). Secretion of IFN-gamma during the same time period was not significantly affected by MT gene dose (Fig. 6b), suggesting that the influence of MT on IL-10 secretion is cytokine specific and not due to a more general effect on the secretion of all cytokines.

To determine if the elevated IL-10 secretion levels observed in MT^-/- cells were sensitive to [Zn²⁺]_i, the resting zinc level was increased in Tr1 CD4⁺ cells by pre-incubating cells in media containing ZnSO₄ [20 μM] for 24 h before restimulation (Additional file 4: Figure S4). This enabled us to test the effect of an increased [Zn²⁺]_i in Tr1 cells during the early activation window (0-6 h) where MT associated increases in the labile zinc pool had been observed (Fig. 4b and c). This concentration of ZnSO₄ did not affect cell viability and in the presence of 10 % FBS (which buffers labile Zn²⁺ ions), is estimated to increase the extracellular labile [Zn²⁺] in media to 10 μM [34]. Increasing the [Zn²⁺]_i during restimulation significantly decreased IL-10 secretion from MT^-/- but not MT^+/+ Tr1 cells (Fig. 6c). During the same time period, increasing [Zn²⁺]_i did not significantly affect IFN-gamma secretion, further supporting the conclusion that regulation of effector function by Zn²⁺ released from MT is cytokine specific. This is in line with previous studies which found that pre-incubation in 25 μM zinc aspartate reduced IL-2, IL-10, and IL-17 secretion following anti-CD3 stimulation in naïve splenocytes [13]. Our observations support this finding and suggest that MT mediated differences in IL-10 secretion are mediated through the regulation of [Zn²⁺]_I in CD4⁺ Tr1 cells.

The relationship between MT expression and immune function is complex. Both extracellular and intracellular pools of MT play important roles in modulating immune responses and influence a range of cellular behaviors including chemotaxis [52, 53], apoptosis [54], proliferation [55], differentiation [30, 56], and resistance to cellular stress [57] and infection [58]. Furthermore, MT expression is increased in many pathological conditions [59] including several types of cancer [60] and inflammatory auto-immune diseases [61] and significantly effects the development of both humoral [62] and cellular adaptive immune responses [35].

In this study, we have specifically investigated CD4⁺ Tr1 cell responses in vitro to assess the effect of intracellular MT expression on zinc signaling. The results of this study demonstrate that intracellular MT plays an important role as a signal transducer by converting ROS signals into a transient increase in labile [Zn²⁺]_i following Tr1 cell re-activation. The amplitude of this zinc signal is sufficient to regulate zinc-sensitive signaling pathways and influence effector function. The ability of MT to release free zinc ions in response to different types of cellular stress is well established. However, this is the first time, to our knowledge, that MT expression has been shown to strengthen the zinc signal resulting from T cell activation and have an effect on p38 MAPK signaling.

The relationship between p38 MAPK signaling and expression of IL-10 is poorly understood. While it has been reported that pharmacological inhibition of p38 activation down-regulates IL-10 expression [63, 64], much less is known about the effect of changes in p38 activation that are confined to the 30 min following stimulation, which were observed in CD4⁺ T cells that differed in MT expression. Our results indicate that reduced p38 activation in cells that do not express MT does not reduce IL-10 or IFN-gamma expression. It is therefore likely that other zinc sensitive signaling pathways including ERK1/2 [7], which have been shown to regulate IL-10 expression [65], may play a role during Tr1 cell restimulation and secretion of IL-10.

The control of intracellular labile zinc is a complex process where the active transport, sequestration, and release of zinc ions from intracellular thiols in response to different environmental stimuli are continuously regulated. The maintenance of zinc homeostasis and the development of normal T cell responses in CD4⁺ T cells from MT knockout mice suggest that MT is only part of a larger network of zinc regulation. However, recent evidence has revealed that even small changes in [Zn²⁺]_i can significantly affect signaling networks, especially those involving intracellular protein tyrosine phosphatases. This suggests that the contribution of MT in amplifying oxidant-induced zinc signals has downstream consequences, even in the presence of other regulators of zinc homeostasis.

The appropriate upregulation of MT expression in the presence of cytokines such as IL-27 and IL-6 provides a mechanism by which MT primes CD4⁺ T helper cells to differentially respond to ROS signals following re-activation and/or exposure to oxidative stress by transiently increasing the [Zn²⁺]_i. The timing and magnitude of this amplified zinc signal are ideally positioned to regulate signaling cascades that immediately follow activation through the T cell receptor. In this way, oxidant-induced zinc release from MT, in combination with other pathways of zinc mobilization across plasma and vesicular membranes, create dynamic intracellular zinc signals that shape appropriate immune responses. In contrast to transient MT expression, chronic inflammation leads to sustained MT overexpression which is associated with low intracellular zinc ion bioavailability and impaired T cell pathway activation in older populations [66, 67]. These observations underscore a balance where appropriate increases in MT lead to increased zinc signaling, proliferation, and immune function while aberrant overexpression during chronic inflammation leads to decreased zinc signaling and immune depression.

Previous studies have observed that human CD4⁺ T cells express higher levels of MT than other peripheral blood cell populations and that individual levels of MT induction in lymphocytes varies considerably within the human population [68, 69]. Differential transduction and amplification of zinc signals in CD4⁺ T cells may help to explain the association of MT expression with autoimmune disease, infection and cancer via regulation of CD4⁺ T cell signaling. This is especially true when the consequences of increased MT expression in inflammatory autoimmune disease are divergent. For example, MT is associated with increased inflammation and pathogenesis in inflammatory bowel disease models [70] but suppression of inflammation in models of autoimmune arthritis [35, 71]. The unique contribution of MT to ROS-mediated zinc signal transduction provides further insight into the complex relationship between MT and inflammation and could represent a novel therapeutic target for the manipulation of immune responses.

[Zn²⁺]_i, intracellular labile zinc concentration; CM-H₂DCF, chloromethyl dichlorodihydrofluorescein; DTDP, dithiodipyridine; FasL, Fas Ligand; FBS, fetal bovine serum; GSH, reduced glutathione; GSSG, glutathione disulfide; H₂O₂, hydrogen peroxide; IgG, immunoglobulin G; MAPK, mitogen-activated protein kinase; MT, metallothioneins; PMA, phorbol 12-myristate 13-acetate; ROS, reactive oxygen species; STAT, signal transducer and activator of transcription; TcR, T cell receptor; Th0, T-helper cell type 0; Th1, T-helper cell type 1; TPEN, N,N,N',N'-tetrakis(2-pyridylmethyl)ethane-1,2-diamine; Tr1, type-1 regulatory T cell; Zip, Zrt- and Irt-like proteins; ZnT, zinc transporters.