Metallothionein mediates leukocyte chemotaxis

Initiation of an immune response is accompanied by physiological changes that can produce a stressful environment for both the cells involved in the immune response, and for bystander cells that are part of adjacent but uninvolved tissues. These stresses can be further increased by the presence of infectious microorganisms. The changes to the environment include increases in reactive oxygen and reactive nitrogen species, products of cellular metabolism, and agents that initiate apoptotic or necrotic cell death.

Cells react to stressful environments with a broad range of different homeostatic responses. These responses can include the synthesis of a host of stress response proteins, including the heat shock proteins, acute phase cytokines, and metallothionein. Metallothionein is a novel member of this type of response with a unique biochemistry and an intriguing array of physiological roles. Metallothionein is small (about 7 kDa) and extremely thiol-rich [1]. The thiols participate in complexing with divalent metal cations [2]. When metallothionein binds to essential divalent metals (e.g. zinc and copper) it may serve as a metal reservoir for apoenzymes and zinc-finger transcription regulators [3, 4]. Metallothionein that is induced by other divalent metal cations (e.g. mercury, cadmium,) protects essential cellular functions [5] and enhances the survival of both cells and whole organisms that are exposed to toxic heavy metals. The thiol-rich nature of metallothionein also enables it to regulate the redox potential of cells, and thus serves as a way of indirectly regulating redox-sensitive transcription via NF-kB [6]. There are also reports that link metallothionein to a much more direct interaction with NF-kB [7, 8]. Metallothionein has also been found to be released to the extracellular environment in a number of different compartments, including cell culture media, serum, urine, bronchoalveolar spaces, liver sinusoids, and inflammatory lesions [9‐12]. Extracellular metallothionein has been shown to have significant immunomodulatory effects both in vivo and in vitro [13‐16] however the molecular mechanism(s) of this effect have yet to be elucidated.

Leukocyte movement is an essential component of the normal response to inflammatory signals. A variety of chemotactic agents can be produced by local immune cells, damaged bystander cells, and by invading microorganisms. In aggregate, these soluble signals determine the infiltration and departure of cells that participate in the inflammation, and serve as essential regulatory components of the immune response. Stress responses alter these patterns of leukocyte trafficking in various ways. For example, psychological stress in humans has been shown to increase both the magnitude of the cellular influx at an inflammatory site and the chemotactic index of peripheral blood mononuclear cells [17]. Restraint stress in hamsters has similarly increased leukocyte trafficking and delayed type hypersensitivity responses [18]. Xenobiotics may alter leukocyte trafficking in similar ways to diminish immune competence.

Cysteines present in the primary amino acid sequence of metallothionein are arranged in cys, cys-cys, cys-X-cys, and cys-X₃-cys motifs. These motifs are also found in chemokine molecules, and serve to differentiate chemokine families from one another. A comparison of the sequences of metallothionein and the chemokine Ccl17 is shown in Figure 1A. In addition to having similar cysteine motifs, Ccl17 is also located near the MT genes in both mice and humans (Figure 1B). Combined with our previously published observations of the impact extracellular metallothionein has on developing T-dependent humoral immunity [15] information suggested that metallothionein might display chemotactic activity.

Naïve splenocytes and thioglycollate-elicited leukocytes respond chemotactically to metallothionein in the Boyden chamber assay at a level similar to that observed when the cells are exposed to guinea pig serum used as a source of activated complement (Figure 2A). To avoid the heterogeneity of primary cell populations, the chemotactic dose response of cells to metallothionein was characterized using Jurkat T cells. Previous studies have shown that these T cells express CXCR4 (the stromal derived factor-1α (SDF-1α) receptor), and exposure to an SDF-1α gradient has previously been shown to induce a chemotactic response [19]. Using the Boyden chamber assay, we found that these cells also respond to metallothionein (Figure 2B) in a dose-dependent manner. The dose response curve shows a peak response at 14.3 μM, and the shape of the curve is consistent with that of other chemokines [20].

In addition to the Boyden chamber assay, chemotaxis was assessed using a recently developed technology called ECIS/taxis. This assay employs a miniature under-agarose chemotaxis chamber in which cell arrival on a surface microelectrode is measured by changes to electrical current flow through the electrode [21]. Unlike the Boyden chamber assay, the under-agarose configuration allows the establishment of stable and shallow chemotactic gradients that more accurately reflect the subtle gradients found in vivo than the Boyden chamber assay. In ECIS/taxis measurements, the total normalized resistance to current flow is proportional to the number of cells that occupy the surface of the target electrode. If no chemoattractant is added, few cells move out of the well and those that do never reach the target electrode and no change in resistance is recorded (Figure 3). In the presence of a gradient of metallothionein, the initial arrival of a small number of cells at the target electrode is indicated by the appearance of small, rapid resistance fluctuations. Over time as more cells accumulate on the electrode, a gradual increase in resistance is observed. The concentration of metallothionein added to the chemoattractant well that elicited the highest resistance increase (Figure 3) and the fastest movement of responding Jurkat T cells (data not shown) was 14.3 μM. This dose optimum is similar to that measured with the Boyden chamber assay. The average speed of the fastest Jurkat T cells, 1.56 ± 0.12 μm/min, was calculated from the time of arrival of the first cells at the target electrode. This speed is comparable to the speed of the Jurkat T cell response to a gradient of SDF-1α (1.43 ± 0.1 μm/min, Figure 3). We also tested metallothionein's effect on WBC 264-9C cells. These cells have been shown to exhibit chemotactic movement toward a source of activated complement [22]. In the ECIS/taxis assay, WBC 264-9C cells are chemotactic toward both activated complement and metallothionein (Figure 4). The metallothionein response was dose-dependent, and the optimal dose was similar to that found with Jurkat T cells (data not shown). The average speed of WBC 264-9C cells responding to a gradient of guinea pig serum used as a source of activated complement was 1.32 ± 0.2 μm/min, compared to 0.75 ± 0.03 μm/min for cells responding to metallothionein. However, the population response was nevertheless robust since total resistance (indicating the absolute number of responding cells) ultimately reached a level that approximates that produced by cells responding to activated complement.

In light of the potential for contaminants (e.g. small liver peptides that co-purify with the MT) in the commercially available metallothionein preparations, we affinity-purified metallothionein from commercial preparations using an anti-metallothionein monoclonal antibody (UC1MT) coupled to CNBr-activated Sepharose. The purified metallothionein also stimulates Jurkat T cell chemotaxis (Figure 5A). In addition, the chemotactic response to the original metallothionein (data not shown) or the affinity-purified metallothionein (Figure 5A) could be blocked by pre-incubation of the purified metallothionein with either UC1MT or E9 monoclonal anti-metallothionein antibodies. Neither of the two anti-metallothionein antibodies nor the isotype-matched IgG₁ (MOPC 21) stimulated cell movement to the target electrode on their own (data not shown).

G protein activation has been shown to play a role in the chemotactic response and this pathway can be inhibited by cholera toxin (CTX) [23] or pertussis toxin (PTX) [24, 25]. Jurkat T cells (10⁶ cells/ml) were pre-incubated with 0.133 μg/ml CTX or 200 ng/ml pertussis toxin and then placed in a metallothionein gradient. The chemotactic effect of metallothionein on Jurkat T cells could be blocked by CTX (figure 5B) and by PTX (figure 5C). PTX was also capable of blocking the chemotaxis of Jurkat cells to SDF-1α.

A more direct assessment of chemotaxis was done using time-lapse video microscopy of cells moving in the presence of a metallothionein or SDF-1α gradient. When Jurkat T cells were exposed to a gradient similar to that present in the ECIS/taxis assay, they could be observed to move out of the cell well and continue up the gradient toward the chemoattractant well. Tracks of the outlines of these cells show persistent directional movement (Figure 6D, E). In the absence of a gradient few cells exit the cell well (data not shown), and those that do show little directional movement (Figure 6A–C). The speed, persistence and chemotactic indices of individual cell movements are consistent with the speeds calculated using the ECIS/taxis measurements of population movement (Table 1). In order to assess the role of chemokinesis in this process, cells were overlaid with a pre-formed agarose sheet containing a uniform concentration of metallothionein, SDF-1α or medium alone. The MT-exposed cells moved more rapidly than control cells, but the movement lacked directional persistence and was much slower than movement in a spatial gradient (Table 1 and Figure 6A–C). Similar results were obtained using a checkerboard analysis of cell movement in the Boyden chamber format (Table 2). Metallothionein added to the same side of the filter as the cells, or to both sides of the filter in equal concentration resulted in fewer cells reaching the lower surface of the filter than in wells where the metallothionein was added to the opposite side of the filter from the cells. This data supports the conclusion that metallothionein induces both chemotaxis and chemokinesis in Jurkat T cells.

Another hallmark of cellular responses to chemokines is a change in the amount and distribution of polymerized actin. Signal transduction through G protein coupled receptors causes reorganization of the actin cytoskeleton, leading to the formation of new F-actin rich lamellipods that extend in the direction of movement. This reorganization can be assessed by in vitro measurements of polymerized actin from cell extracts of stimulated cells with phalloidin [26]. Metallothionein stimulated a 19% increase in total F-actin within 30 seconds and a 79% increase by 2 minutes (Figure 7). The extent and timing of this response is consistent with receptor activation in other cell types [27, 28].

Cells of the immune system operate in a complex microenvironment where they are presented with a host of different and often conflicting signals [29, 30]. The ways in which cells integrate and respond to these signals can ultimately govern the way in which the immune system will respond to antigen exposure. In some cases, the outcome is an activated immune response that is designed to eliminate the source of antigen. In other instances, the cells become anergic or undergo apoptosis and thus fail to initiate or participate in an immune response to that antigen. An early aspect of many immune responses is the directional movement of cells toward a site of infection or other injury. This directional movement is a response to chemotactic factors produced by some infectious organisms, to chemokines produced by cells already at the site of inflammation, or to other agents. Cells that express receptors for these signals can detect the gradient(s) of diffusing chemoattractants, and move toward the source of the agent. This chemotactic response is an essential aspect of lymphocyte trafficking.

In this report, we show that metallothionein can direct the chemotaxis of primary and transformed leukocytes. While metallothionein has been historically thought of as an intracellular protein, there are numerous reports that describe its presence in serum, urine [31], broncho-alveolar spaces [10], liver sinusoids [32], and other extracellular locations. While the mechanism(s) by which metallothionein is released from cells has yet to be determined, heat shock protein 70 [33], Interleukin 1β [34] and fibroblast growth factor [35] are among a set of proteins that lack signal sequences and nevertheless are released from cells by a non-traditional secretory mechanism. These results indicate that stress response proteins may gain access to the extracellular environment via mechanisms other than cell lysis, and suggest that a thorough understanding of the immunomodulatory roles played by metallothionein must include the extracellular compartment.

Metallothionein is synthesized in response to acute phase cytokines (e.g. IL-1, IL-6, and TNF-a) that are secreted at sites of inflammation [36‐38] in a variety of contexts in which immune activities are changing. Metallothionein is also synthesized in cells exposed to glucocorticoids, a signal that is often associated with stressful environments [39]. Furthermore, metallothionein can be induced by reactive oxygen species, by endotoxin [40], and in cells exposed to divalent metal cations [1]. With all of these different initiators, it is not surprising that elevated metallothionein levels are detected in the context of neoplastic disease [41, 42], autoimmune disease [43], chronic inflammation [44], and infection [45]. Previous work from our laboratory and others has shown that metallothionein can have significant immunomodulatory activities. For example, metallothionein can diminish T dependent humoral responses and it can alter the proliferative capacity of lymphocytes [16], diminish cytotoxic T cell function [46], and it can alter the effector function of macrophages [47]. Inadequate expression of metallothionein in the context of inflammatory disease can dramatically shorten life span [43], and exogenous metallothionein can diminish the severity of a collagen-induced arthritis [48].

There are a multitude of studies which show that different forms of stress originating from external sources can alter normal immune function [49]. Psychological, physical and chemical agents which induce stress each affect the immune system. In some instances, these stressors suppress effective immune functioning, which renders the individual susceptible to infectious pathogens. In other instances, the immune modifications result in undesirable increases in immune recognition of self antigens, ultimately resulting in autoimmune disease. These stressors are known to induce metallothionein synthesis and may alter immune functions in part via their effect on metallothionein.

We have demonstrated metallothionein-induced chemotactic cell movement in the traditional Boyden chamber assay, by computerized analysis of time-lapse images of cell movement, and using the ECIS/taxis assay. We have shown that the response of Jurkat T cells to a metallothionein gradient corresponds well with chemotactic responses of leukocytes to other agents. Jurkat T cells and WBC 264-9C migrate in response to a metallothionein gradient at speeds which are similar to those found in other systems [50, 51]. In addition, the pattern of the dose response to metallothionein is a bell shaped curve similar to other classical chemokines [20]. One important consideration is whether the chemotactic response to metallothionein occurs at physiologically relevant concentrations. Chemoattractants can act over an extremely wide concentration range (e.g. 4 logs) [52, 53] because the cells sense the local spatial differential in chemoattractant concentration. Our work shows that 1 to 10 μM metallothionein can stimulate chemotaxis of cells in both Boyden and under-agarose assays. Higher concentrations of metallothionein used in the ECIS/taxis assay refer to the concentrations added to the micro-volume chemoattractant wells, which are then diluted in the process of diffusion away from the source. The metallothionein amounts used in these experiments represent biologically reasonable concentrations, given that metallothionein has been measured at concentrations of 1 μM in serum (which would be substantially diluted from the source tissue concentration) in normal patients, and in individuals undergoing some form of stress (inflammation, cancer, toxicant exposure, etc.) [54].

The chemotactic response to metallothionein can be blocked by monoclonal antibodies to metallothionein while isotype-matched antibody has no effect. This blockade of the response is an important control, since commercial metallothionein preparations contain contaminating peptides from the liver tissue source (D. Lawrence, personal communication). Since both cholera toxin and pertussis toxin block the metallothionein-initiated chemotaxis, it is likely that G protein mediated signaling is involved in the response. Another common aspect of chemotactic signaling is an activation of the actin polymerization machinery in response to a sharp increase in chemoattractant concentration [26, 55]. Metallothionein causes an increase both in total F-actin content and in peripheral F-actin. It will be of great interest to determine the receptor for metallothionein and the signal transduction pathway that leads to actin polymerization.

It is intriguing to speculate that once outside the cell, metallothionein serves as one of the many signals designed to draw immunocompetent cells to sites of cellular stress. Our observation(s) that metallothionein and anti-metallothionein injections modify immune activity in vivo suggest that there is an appropriate range of extracellular metallothionein in which leukocytes ordinarily function [14, 15]. A pair of recent reports suggest that cytosolic constituents of apoptotic cells are released to the extracellular compartment and support the progression of the inflammatory process [56, 57]. These reports further suggested that release of the cytosolic components of these dying cells might represent one of the signals central to the "Danger Hypothesis" proposed by Matzinger et al. [58, 59]. This hypothesis holds that an active immune response cannot be mounted without a signal indicating that cellular damage has occurred. While other reports have suggested that heat shock proteins can fill this role [60], metallothionein is another potential candidate for the danger signal.