Abstract
TR6/DcR3 is a secreted molecule belonging to the TNFR family. Its ligands are LIGHT, Fas ligand, and TL1A, all TNF family members. TR6 is expressed in some tumors and is hypothesized to endow tumor cells with survival advantages by blocking Fas-mediated apoptosis. It can also inhibit T cell activation by interfering with two-way T cell costimulation between LIGHT and HveA. In this study, we discovered a novel function of TR6: inhibition of T cell chemotaxis. Human T cells pretreated with soluble or solid-phase TR6-Fc showed compromised migration toward CXCL12/stromal cell-derived factor 1α in vitro in a Transwell assay. Such an effect could also be observed in T cells pretreated with soluble or solid-phase HveA-Fc or anti-LIGHT mAb, suggesting that LIGHT reverse signaling was likely responsible for chemotaxis inhibition. TR6 pretreatment also led to T cell chemotaxis suppression in vivo in the mice, confirming in vivo relevance of the in vitro observation. Mechanistically, a small GTPase Cdc42 failed to be activated after TR6 pretreatment of human T cells, and further downstream, p38 mitogen-activated protein kinase activation, actin polymerization, and pseudopodium formation were all down-regulated in the treated T cells. This study revealed a previously unknown function of TR6 in immune regulation, and such an effect could conceivably be explored for therapeutic use in controlling undesirable immune responses.
TR6/DcR3, a TNFR family member, is a secreted protein due to the lack of the transmembrane domain in its sequence (1, 2). TR6 has three known ligands, Fas ligand (L)3 (3), LIGHT (1, 4), and TL1A (5), all of which are TNF family members. FasL binding by TR6 interferes with the interaction between Fas and FasL. Consequently, FasL-induced apoptosis of lymphocytes or tumor cells can be repressed by TR6 (2). The consequence of interaction between TR6 and LIGHT is more complex. LIGHT is a ligand of HveA and lymphotoxin β receptor (LTβR) (1, 4, 6) in addition to TR6. It is expressed on the surface of activated T cells (6) and immature dendritic cells (7). Our recent study has shown that LIGHT is expressed in resting T cells as well according to confocal microscopy (8). LIGHT costimulates T cell responses via HveA (7, 9, 10). We have demonstrated that LIGHT, although being a ligand, also reversely transduces costimulatory signals into T cells (8, 11). Moreover, LIGHT induces apoptosis in cells expressing both HveA and LTβR (12) or LTβR alone (13). Soluble TR6 blocks the two-way costimulation between HveA and LIGHT among T cells (1), while TR6 on solid phase costimulates T cells via LIGHT reverse signaling (8, 11). TR6 can modulate dendritic cell maturation (14), whether this effect occurs via LIGHT reverse signaling remains a matter of debate. TR6 also blocks LIGHT-induced apoptosis of LTβR-bearing cells (4). Since LTβR is not expressed on lymphocytes, such an effect does not have significant relevance in the immune system. The third TR6 ligand, TL1A, is predominantly expressed on endothelial cells (5). TL1A is also a ligand of DR3 and can enhance T cell alloresponses via DR3 in vivo (5). TR6 strongly inhibits TL1A-augmented T cell proliferation and lymphokine secretion and also inhibits TL1A-triggered tumor cell apoptosis (5).
TR6 mRNA and protein are detectable in some malignant tumors (15). The inhibition of apoptosis and repression of T cell costimulation obviously represent plausible mechanisms for TR6-secreting tumors to gain survival advantage by avoiding apoptosis and evading immune surveillance. In this study, we discovered yet another function of TR6 in the immune system. We demonstrated that TR6 inhibited T cell chemotaxis both in vitro and in vivo, and such an effect was probably mediated by reverse signaling through LIGHT. A cascade of signaling and effector events such as Cdc42 activation, p38 mitogen-activated protein kinase (MAPK) activation, actin polymerization, and pseudopodium formation in CXCL12/stromal cell-derived factor 1α (SDF-1α)-stimulated T cells was inhibited by TR6 pretreatment. The physiological significance of these findings is discussed.
Materials and Methods
Recombinant proteins and mAb
The recombinant proteins, i.e., TR6 fused with Fc (TR6-Fc), HveA fused with Fc (HveA-Fc), and TR11/GITR fused with Fc (TR11-Fc), were described in our previous publication (1). Because TR11-Fc had no effect on T cell migration, compared with PBS or normal human IgG, it was used as a control protein for TR6-Fc. The Fc sequence in the recombinant protein was mutated, so that it did not bind to FcγR on the cell surface. The endotoxin levels in all of the recombinant proteins were below 10 endotoxin units/mg (1). Human mAb against LIGHT was described in our earlier publication (8).
Preparation and culture of human T cells and mouse spleen cells
Adult PBMC were prepared by Lymphoprep gradient (Nycomed, Oslo, Norway), and T cells were prepared from PBMC by negative selection (deletion of cells positive for CD11b, CD16, CD19, CD36, and CD56) with magnetic beads (Miltenyi Biotec, Auburn, CA) according to the manufacturer’s instructions. Blood samples from different donors were used to prepare T cells for repeated experiments. Mouse total spleen cells were prepared by lysing RBC with 0.84% NH4Cl among the cells flushed out from the spleen. The cells were cultured in serum-free medium.
In vitro T cell migration assay
6 T cells pretreated (16-h preincubation) with TR6-Fc, a control fusion protein, TR11-Fc (both at 20 μg/ml), or a p38 MAPK inhibitor, SB203580, or its nonfunctional structural analog, SB202474 (both at 5 μM). In some experiments, cross-linked human anti-LIGHT mAb was also used: the cells were preincubated with anti-LIGHT mAb (clone 1.2.2, 20 μg/ml) along with goat anti-human IgG (10 μg/ml). The Transwell ensemble was then incubated at 37°C for 2 h, and T cells migrating into the lower chamber were counted by flow cytometry. All the samples were in duplicate.
Flow cytometry
For F-actin staining, 1 × 106
In vivo T cell migration assay
BALB/c spleen cells were incubated overnight with serum-free medium containing TR6-Fc or TR11-Fc (both at 20 μg/ml). The cells were then labeled with CFSE (5 μM) for 10 min at room temperature. After washing, 6 × 107 labeled cells in 0.2 ml of HBSS were injected i.v. into the tail vein of a BALB/c mouse and 1 μg of CXCL12/SDF-1α in 2 ml of PBS was injected i.p. 2 h later. The mice were sacrificed after another 20 h. Their peritoneal cavities were rinsed with 8 ml of PBS, and PEC in PBS were analyzed by flow cytometry for total cell number, total CD3+ cell number, and CD3+CSFE+ cell number. For each experiment, three mice were used in each group and three independent assays were performed.
Confocal microscopy
Human T cells were cultured overnight in the presence of TR6-Fc or a control protein-Fc (TR11-Fc, both at 20 μg/ml). These cells were then stimulated with CXCL12/SDF-1α (80 ng/ml) for 1 min at 37°C and fixed with 3.7% Formalin for 30 min at room temperature. They were stained with Alexa Fluor-488-conjugated phalloidin (Molecular Probes) for 30 min on ice and mounted on slides with Prolong Anti-fade Mounting Medium (Molecular Probes) for examination by confocal microscopy. Digital images were processed with Photoshop (Adobe, Seattle, WA).
Immunoblotting
Human T cells were cultured overnight in serum-free medium in 12-well plates at 5 × 106 cells/well in the presence of TR6-Fc or TR11-Fc (both at 20 μg/ml). After washing, the cells were stimulated with SDF-1α (80 ng/ml) in serum-free medium at 37°C for 10 min. The remainder of the immunoblotting procedure was detailed in our previous publication (16). Briefly, the cells were washed and lysed in lysis buffer for 10 min; the cleared lysates were resolved by 10% SDS-PAGE. The proteins were then transferred to Immobilon-P (Millipore, Bedford, MA) membranes, which were sequentially hybridized with rabbit anti-phospho-p38 MAPK and anti-p38 MAPK Ab (New England Biolabs, Mississauga, Ontario, Canada). Immunoreactive protein bands were visualized by HRP-conjugated goat anti-rabbit IgG followed by ECL.
Rac-1 and Cdc42 activity assays
Twenty million T cells were preincubated overnight with TR6-Fc or TR11-Fc (both at 20 μg/ml). After washing, the cells were stimulated with CXCL12/SDF-1α (80 ng/ml) for 5 min at 37°C. The cells were lysed and Rac-1 and Cdc42 activity in the lysates was measured by PAK-1 pull-down assays (Upstate Biotechnology, Lake Placid, NY) according to the manufacturer’s instructions. Briefly, GTP-loaded active Rac-1 and Cdc42 were precipitated with PAK-1 PBD-conjugated agarose beads and resolved in 12% SDS-PAGE. They were then detected with mouse mAb anti-Rac-1 and Cdc42, respectively, in immunoblotting.
Results
TR6 inhibits T cell chemotaxis in vitro
In the course of our study into TR6’s effect on T cells, we noticed that TR6-treated T cells presented an altered morphology upon mitogen stimulation. This prompted us to examine whether T cell mobility was affected by TR6. T cells from PBMC were pretreated with TR6-Fc or a control fusion protein TR11-Fc overnight in serum-free medium and then loaded in the upper Transwell chamber. A chemokine, CXCL12/SDF-1α, which is known to induce T cell chemotaxis, was added to the bottom chamber at different concentrations. T cells were allowed to migrate for 2 h at 37°C. As illustrated in Fig. 1⇓A, CXCL12/SDF-1α dose-dependently induced T cell migration into the lower chamber. At 80 ng/ml CXCL12/SDF-1α, T cells precultured in plain medium or in the presence of TR11-Fc showed similar migration rates, but those precultured in the presence of TR6-Fc manifested a significantly reduced rate. Decreased migration of TR6-precultured T cells was also observed with CXCL12/SDF-1α at 20 and 5 ng/ml (p < 0.01 for all three concentrations, one-way ANOVA followed by Tukey’s multiple comparisons test).
TR6-Fc inhibits human T cell chemotaxis in vitro. A, T cells pretreated with TR6-Fc showed compromised chemotaxis toward SDF-1α. Human T cells were preincubated overnight in the absence (Med) or presence of soluble TR6-Fc or a control fusion protein TR11-Fc (both at 20 μg/ml) in serum-free medium. The cells were then loaded in the upper Transwell chambers (0.3 × 106/well), with the lower chamber containing different concentrations of CXCL12/SDF-1α. After a 2-h incubation at 37°C, cells in the lower chamber were collected and counted by flow cytometry. Samples were in duplicate in each experiment. Means ± SD of four measurements from two independent experiments are shown. The asterisks indicate highly significant differences (p < 0.01, one-way ANOVA followed by Tukey’s multiple comparisons test) between TR6-Fc- and control TR11-Fc-treated samples. B, T cells pretreated with TR6-Fc can be normally activated by mitogens. T cells were pretreated with medium (Med), TR11-Fc, or TR6-Fc for 24 h and then stimulated overnight further (24–40 h) with or without soluble anti-CD3 (0.2 μg/ml), as indicated. At 40 h after the initiation of culture, the expression of a T cell activation marker, CD69, was analyzed with flow cytometry. The experiments were performed three times, and histograms of a representative experiment with percentages of CD69-positive T cells are shown. The percentage of CD69-positive cells in the second, third, and fourth panels has no statistical difference according to pooled data from three independent experiments (p > 0.05, one-way ANOVA analysis followed by Tukey’s test for 2 × 2 comparison on all groups). C, TR6 pretreatment does not affect the expression of the CXCL12/SDF-1α receptor CXCR4. Fresh T cells or T cells cultured in the absence (Med) or presence of TR11-Fc or TR6-Fc overnight were stained with FITC-conjugated anti-CXCR4 mAb. Percentages of CXCR4-positive cells are indicated. The experiments were performed three times and a representative set of data is shown. The percentage of CXCR4-positive cells in second, third, and fourth panelshas no statistical difference according to pooled data from three independent experiments (p > 0.05, one-way ANOVA followed by Tukey’s test for 2 × 2 comparison on all groups).
The reduced migration rate was not due to the toxic effects of TR6 pretreatment. The viability of TR6-Fc- or TR11-Fc-precultured T cells, according to trypan blue dye exclusion, was similarly between 90 and 95% (data not shown). To further prove this point, T cells precultured in the presence of TR6-Fc or TR11-Fc or precultured in plain medium for 24 h were washed and restimulated with soluble anti-CD3 (0.2 μg/ml). After additional culture for 16 h, all of the cells were gated and analyzed for their expression of a T cell activation marker, CD69. As seen in Fig. 1⇑B, CD69 expression in all anti-CD3-stimulated groups was increased drastically compared with cells without stimulation (medium (Med) 24–40 h), but there was no statistical difference among cells pretreated with plain medium, TR11-Fc, or TR6-Fc (p > 0.05, one-way ANOVA followed by Tukey’s test for 2 × 2 comparison of all groups). This indicates that preculturing with TR6-Fc did not affect T cell viability. Was the inhibited migration due to repression of CXCL12/SDF-1α receptor CXCR4 expression on TR6-Fc-treated T cells? Approximately 40% of fresh T cells expressed low intensity CXCR4 (Fig. 1⇑C). After overnight culture in plain medium, CXCR4+ cells increased to ∼90% with a concomitant rise in CXCR4 intensity. T cells treated with either TR6-Fc or TR11-Fc showed similar up-regulation of CXCR4 as T cells cultured in plain medium Fc (p > 0.05, one-way ANOVA followed by Tukey’s test for 2 × 2 comparison of all groups), indicating that TR6-Fc treatment did not affect CXCR4 expression. This result also confirms that T cell viability was not impaired by TR6-Fc pretreatment.
The effect of TR6 on T cell migration was likely mediated by LIGHT
TR6 can bind to FasL and LIGHT, both of which are expressed on T cells. Our recent study revealed that TR6 is capable of triggering reverse signaling through LIGHT and such an effect results in T cell costimulation (8, 11). We investigated here whether the inhibition of T cell migration by TR6 involved LIGHT. For this purpose, human anti-LIGHT mAb was cross-linked by goat anti-human IgG in solution (human anti-LIGHT mAb at 20 μg/ml and goat anti-human IgG at 10 μg/ml were added to the culture at the same time) and was used in place of TR6-Fc during preincubation. As shown in Fig. 2⇓A, cross-linked anti-LIGHT mAb pretreatment also suppressed T cell migration in response to CXCL12/SDF-1α in the Transwell assay compared with cross-linked normal human IgG as a control (p < 0.01, one-way ANOVA followed by Tukey’s multiple comparisons test). This suggests that such an effect involves LIGHT.
Chemotaxis inhibition by TR6 is mediated by LIGHT and is not specific to CXCL12/SDF-1α. Samples were in duplicate in each experiment and the experiment was performed twice. Means ± SD of four measurements from two independent experiments are shown. The asterisks indicate highly significant differences (p < 0.01, one-way ANOVA followed by Tukey’s multiple comparisons test) between the test group (treated with anti-LIGHT mAb, HveA-Fc, or TR6-Fc) and control normal human IgG (NhIgG)-treated samples. A, Cross-linked anti-LIGHT mAb in solution inhibits T cell chemotaxis. T cells were cultured overnight in the absence (Med) or presence of human anti-LIGHT mAb (20 μg/ml) plus goat anti-human IgG (10 μg/ml), and chemotaxis of these cells toward CXCL12/SDF-1α (80 ng/ml) was tested by the Transwell assay as described in Fig. 1⇑A. Normal human IgG was similarly cross-linked by goat anti-human IgG and was used as control. B, Soluble HveA-Fc inhibits T cell chemotaxis similar to TR6-Fc, and these proteins lost their function after heat inactivation. T cells were cultured overnight in the absence or presence of soluble HveA-Fc or TR6-Fc (both at 20 μg/ml). In some samples, HveA-Fc and TR6-Fc were heat inactivated for 5 min at 100°C before use. Chemotaxis of the treated T cells toward CXCL12/SDF-1α (80 ng/ml) was assayed by the Transwell assay. C, Solid-phase TR6-Fc, HveA-Fc, and anti-LIGHT mAb inhibit T cell chemotaxis. TR6-Fc, HveA-Fc, anti-LIGHT mAb or normal human IgG were coated on wells at 1 μg/100 μl/well overnight at 4°C and the wells were then washed. T cells were cultured in these precoated wells overnight and then used in the Transwell assay using CXCL12/SDF-1α (80 ng/ml) in the lower chamber. D, Media conditioned in TR6-Fc-, HveA-Fc, or anti-LIGHT mAb-coated wells had no effect in T cell migration toward CXCL12/SDF-1α. Complete medium was incubated in TR6-Fc-, HveA-Fc-, or anti-LIGHT mAb-coated wells for 24 h, The media were harvested and used to culture T cells in uncoated wells overnight. The cells were then used in Transwell assay using CXCL12/SDF-1α (80 ng/ml) in the lower chamber. E, TR6-Fc preculturing inhibits T cell chemotaxis toward CCL21. T cells were cultured overnight in the absence (Med) or presence of soluble TR6-Fc or TR11-Fc (both at 20 μg/ml) as indicated. Chemotaxis of these cells toward CCL21 (400 ng/ml) in the lower chamber was examined by the Transwell assay as described in Fig. 1⇑A.
To further prove the involvement of LIGHT, HveA, which binds to LIGHT but not to FasL (6), was used to pretreat the T cells. As shown in Fig. 2⇑B, when HveA-Fc was used in solution to pretreat the T cells, it could inhibit T cell migration provoked by CXCL12/SDF-1α (p < 0.001, one-way ANOVA followed by Tukey’s multiple comparisons test), similar to TR6-Fc. This suggests that LIGHT is implicated in the process.
To understand whether such inhibition was due to TR6-triggered reverse signaling through LIGHT or due to disruption of existing interaction between LIGHT and HveA, we used solid-phase anti-LIGHT mAb, TR6-Fc, or HveA-Fc to condition the T cells. As shown in Fig. 2⇑C, when T cells were cultured in well coated with these reagents, their subsequent migration toward CXCL12/SDF-1α in the Transwell assay were significantly reduced (p < 0.01 for all three reagents, one-way ANOVA followed by Tukey’s multiple comparisons test)
The amount of reagents coated on the wells was minute, and there was unlikely sufficient molecules to leak from solid phase into the solution to block the LIGHT and HveA interaction. This was proved as follows. Media were put in wells coated with TR6-Fc, HveA-Fc, αLIGHT mAb, or control normal human IgG for 24 h, and these conditioned media were then used to culture T cells overnight. However, as shown in Fig. 2⇑D, T cells cultured in these conditioned media had no difference in their migration capability toward CXCL12/SDF-1α (p > 0.001, one-way ANOVA followed by Tukey’s multiple comparisons test). This confirms that the inhibitory effect seen with solid-phase TR6-Fc, HveA-Fc, and anti-LIGHT mAb is not due to leaching of these molecules from the solid phase and suggests that TR6 triggers reverse signaling through LIGHT on the T cell surface and results in compromised T cell migration.
The observed migration inhibition by TR6-Fc and HveA-Fc was not due to contaminating endotoxin in the protein preparation, because their inhibitory effect could be destroyed by 5-min boiling (Fig. 2⇑B; p > 0.05 for boiled TR6-Fc vs medium and for boiled HveA-Fc vs medium, one-way ANOVA followed by Tukey’s multiple comparisons test). To assess whether TR6 affected the general mobility of T cells and the observed effect was only restricted to CXCL12/SDF-1α stimulation, we tested the T cell response to another chemokine, CCL21. After incubation with TR6-Fc but not with a control fusion protein (TR11-Fc), T cells showed a significantly reduced migration rate toward CCL21 in the Transwell assay (Fig. 2⇑E; p < 0.01, one-way ANOVA followed by Tukey’s multiple comparisons test). Therefore, it seems that TR6 pretreatment inhibited the general mobility of T cells.
TR6 inhibits T cell chemotaxis in vivo
To confirm that the observed inhibitory effect of TR6 on T cell migration in vitro had physiological relevance, we established an in vivo T cell chemotaxis model. BALB/c spleen cells were first precultured overnight in the presence of TR6-Fc or TR11-Fc and then labeled with CFSE. The labeled cells were injected i.v. into syngeneic BALB/c mice. Two hours later, CXCL12/SDF-1α was injected i.p. After another 20 h, PEC were analyzed for their total number, total T cell number, total CFSE-labeled T cell number, and percentage of CFSE-labeled T cells among total T cells. As shown in Fig. 3⇓, CXCL12/SDF-1α injection into the peritoneal cavities resulted in a drastic increase of total PEC and T cells in PEC (Fig. 3⇓, A and B). There was no significant difference in these parameters in mice that received TR6-Fc-, TR11-Fc-, or plain medium-precultured cells (p > 0.05, one-way ANOVA followed by Tukey’s multiple comparisons test), reflecting the fact that the majority of the exodus cells were of host origin (as CFSE-labeled donor cells showed difference, to be elaborated below). When CFSE-labeled exodus T cells were analyzed, there was a significant decrease in their number in mice receiving cells pretreated with TR6-Fc compared with mice receiving cells pretreated with TR11-Fc or plain medium (Fig. 3⇓C; p < 0.001, one-way ANOVA followed by Tukey’s multiple comparisons test). Such a reduction in T cell number was also reflected in the percentage abatement of CFSE-labeled T cells among total exodus T cells (Fig. 3⇓D; p < 0.001, one-way ANOVA followed by Tukey’s multiple comparisons test), because the number of the latter remained similar in mice receiving spleen cells with the different treatments (Fig. 3⇓B). These data showed the in vivo relevance of compromised T cell chemotaxis after exposure to TR6.
TR6-Fc inhibits mouse T cell chemotaxis in vivo. BALB/c spleen cells were cultured overnight in medium (Med) or in the presence of TR6-Fc or TR11-Fc (both at 20 μg/ml). The cells were then labeled with CFSE for 10 min at room temperature and injected i.v. into BALB/c mice. After 2 h, 1 μg of CXCL12/SDF-1α was administered i.p. The mice were sacrificed after another 20 h, and their total PEC (A), the percentages of total T cells among total PEC (B), CFSE-labeled peritoneal exudate T cells (C), and percentages of CFSE-labeled T cells among total peritoneal exudate T cells (D) were analyzed. Each group consisted of three mice; similar results were obtained from three independent experiments; and a set of representative data are presented. PBS: mice receiving CSFE-labeled spleen cells cultured in medium alone without subsequent CXCL12/SDF-1α but with PBS administration; Med→CXCL12: mice receiving CFSE-labeled spleen cells precultured in medium, followed by CXCL12 injection (1 μg/mouse) i.p; TR11-Fc→CXCL12: mice receiving CFSE-labeled spleen cells precultured in the presence of soluble TR11-Fc (20 μg/ml), followed by CXCL21 injection i.p.; TR6-Fc→CXCL21: mice receiving CFSE-labeled spleen cells precultured in the presence of TR6-Fc (20 μg/ml), followed by CXCL12 injection i.p.
TR6 pretreatment of T cells represses their actin polymerization upon CXCL12/SDF-1a stimulation
T cell chemotaxis depends on cytoskeleton reorganization, for which actin polymerization plays an essential role. We, therefore, examined actin polymerization upon SDF-1α stimulation in T cells with or without TR6 pretreatment. The F-actin staining of T cells precultured in plain medium without subsequent CXCL12/SDF-1α stimulation was considered as background (0% positive, Fig. 4⇓A). When such cells were stimulated by CXCL12/SDF-1α for 1 min, 65.9% of them became F-actin positive. Preculturing the cells in the presence of TR11-Fc overnight had no apparent effect on actin polymerization upon subsequent CXCL12/SDF-1α stimulation, and 65% of the cells were positive for F-actin staining. However, overnight TR6-Fc pretreatment significantly reduced F-actin-positive cells to 29.4% (p < 0.01, one-way ANOVA followed by Tukey’s test for 2 × 2 comparison of all groups). This result shows that TR6 conditioning of T cells leads to compromised actin polymerization.
TR6-Fc inhibits CXCL12/SDF-1α-induced T cell actin polymerization and pseudopodium formation. Human T cells were cultured overnight in the absence (Med overnight (ON)) or presence of TR11-Fc (TR11-Fc ON) or TR6-Fc (TR6-Fc ON) (both at 20 μg/ml). After washing, the cells were incubated with CXCL21/SDF-1α (80 ng/ml) for 1 min, stained with Alexa Fluor-488-conjugated phalloidin and analyzed by flow cytometry as well as confocal microscopy. A, Flow cytometry analysis of F-actin. F-actin staining of T cells cultured overnight in medium without subsequent CXCL12/SDF-1α simulation was used as a negative control, with its F-actin intensity set at 0%. Percentages of F-actin-positive cells among cells with different treatments, as indicated, are shown. The experiments were performed three times and a set of representative histograms are presented. The percentage of F-actin-positive cells in panel 4 is significantly lower than that of panels 2 and 3 according to pooled data from the three independent experiments (p < 0.01, one-way ANOVA followed by Tukey’s test for 2 × 2 comparison on all groups). B, Confocal microscopy of T cell actin polymerization and morphology. The same set of T cells as described in A was examined with confocal microscopy. The pseudopodium formation in cells precultured in medium or TR11-Fc was similar upon SDF-1α stimulation (upper right panel vs lower left panel, respectively), but preculturing cell in the presence of TR6-Fc reduced pseudopodium formation compared with cell pretreated with TR11-Fc (lower left panel vs lower right panel). The experiments were performed three times and a set of representative data are presented. C, Quantitative assessment of T cell pseudopodium formation. The cells in B were quantified for pseudopodium-positive cells. Three view fields (containing ∼100 cells/field) per sample were examined, and the means ± SD of the percentages of pseudopodium-positive cells among total cells are indicated. The asterisk indicates a highly significant difference between the test group and the TR11-Fc control (p < 0.01, one-way ANOVA followed by Tukey’s test for 2 × 2 comparison on all groups). The experiment was performed twice and pooled data from the two independent experiments are shown.
Consistent with the status of actin polymerization, CXCL12/SDF-1α rapidly (within 1 min) induced pseudopodium formation in T cells (Fig. 4⇑B, comparing the upper left panel with upper right panel; quantitative assessment of Fig. 4⇑B is shown as a bar graph in Fig. 4⇑C). T cells cultured overnight in medium or TR11-Fc had a similar degree of pseudopodium formation upon CXCL12/SDF-1α stimulation (Fig. 4⇑B, upper right panel vs lower left panel), but TR6-Fc pretreatment of the cells significantly inhibited this process (Fig. 4⇑B, lower left panel vs lower right panel; p < 0.001, one-way ANOVA followed by Tukey’s multiple comparisons test). Thus, it appears that T cells receiving TR6 preconditioning have modified signaling pathways that lead to compromised actin polymerization, pseudopodium formation, and, eventually, chemotaxis.
TR6 pretreatment of T cells compromises p38 MAPK activation
p38 MAPK is an upstream signaling molecule essential for actin polymerization (17). Its activity in TR6-treated T cells was investigated. T cells were precultured overnight in the absence (Med) or presence of TR6-Fc or TR11-Fc. The cells were washed and stimulated with 80 ng/ml CXCL12/SDF-1α for 10 min. p38 MAPK tyrosine phosphorylation, which represents p38 MAPK activation and total p38 MAPK in these cells, were assessed by immunoblotting. As shown in Fig. 5⇓A, CXCL12/SDF-1α stimulation resulted in similar up-regulation of p38 MAPK tyrosine phosphorylation in T cells precultured in plain medium or in the presence of TR11-Fc. However, preculturing in the presence of TR6-Fc prevented such up-regulation.
Compromise of p38 MAPK activation by TR6 treatment resulted in chemotaxis inhibition. A, Immunoblotting of activated and total p38 MAPK. Human T cells were cultured overnight in the absence (Med) or presence of TR11-Fc or TR6-Fc (both at 20 μg/ml). After washing, the cells were stimulated for 10 min with CXCL12/SDF-1α (80 ng/ml). The cell lysates were resolved in 10% SDS-PAGE and blotted membranes were sequentially hybridized with anti-phospho-p38 MAPK and anti-total p38 MAPK. Arrows indicate 38-kDa phospho-p38 MAPK and total p38 MAPK. The experiments were performed three times and a set of representative data is presented. B, p38 MAPK inhibitor represses CXCL12/SDF-1α-induced T cell chemotaxis in a Transwell assay. Human T cells were cultured in serum-free medium for 2 h in the presence of 5 μM p38 MAPK inhibitor SB203580 or its nonfunctional structural analog SB202474. After washing, the Transwell assay was performed as described in Fig. 1⇑A.
To ascertain that p38 MAPK activation was a relevant event that controlled T cell chemotaxis upon CXCL12/SDF-1α stimulation, T cells were treated with a p38 MAPK-specific inhibitor, SB20358 (5 μM), for 2 h and then evaluated for their chemotaxis by the Transwell assay. The p38 MAPK inhibitor but not its nonfunctional structural analog, SB202474, suppressed the T cell chemotaxis upon CXCL12/SDF-1α stimulation (Fig. 5⇑B), as expected, indicating that the compromised p38 MAPK activation in TR6-treated T cells is relevant and responsible for T cell chemotaxis inhibition.
Compromised Cdc42 activation in T cells pretreated with TR6-Fc
Small G proteins Rac-1 and Cdc42, which are signaling molecules upstream of p38 MAPK, are important in cytoskeleton reorganization, which in turn controls pseudopodium formation and cell mobility (18). The activity of Rac-1 and Cdc42 of TR6-treated T cells was examined using a PAK-1 pull-down assay. T cells were preincubated overnight in the presence of TR6-Fc or TR-11-Fc and then stimulated with CXCL12/SDF-1α for 5 min. The activated GTP-associated Rac-1 or Cdc42 was pulled down using solid-phase (agarose beads) PAK-1, which is a substrate of and binds to GTP-associated Rac-1 and Cdc42. As shown in Fig. 6⇓, Rac-1 activation, which was reflected by the levels of GTP-associated Rac-1, was not modulated by SDF-1 stimulation; TR11-Fc or TR6-Fc pretreatment did not affect its activity. In contrast, Cdc42 was rapidly activated by CXCL12/SDF-1α; TR6-Fc but not control TR11-Fc pretreatment prevented its activation. The selective inhibition of Cdc42 activation, which is important in filopodium formation (18), by TR6 thus represents a plausible mechanism for the repressive effect of TR6 on T cell chemotaxis.
TR6 inhibits Cdc42 activation in T cells. T cells were precultured overnight in serum-free medium or in the presence of TR6-Fc or a control fusion protein TR11-Fc (both at 20 μg/ml) and were then stimulated with CXCL12/SDF-1α (80 ng/ml) for 5 min. The activated GTP-binding Rac-1 and Cdc42 was absorbed by PAK-1 PBD-conjugated agarose beads and resolved in 12% SDS-PAGE. Rac-1 and Cdc42 were then detected with anti-Rac-1 and Cdc42 mAbs, respectively, in immunoblotting. Five percent of the input lysates were examined to confirm similar levels of total Cdc42 in each sample. The experiments were performed three times and a set of representative data is presented.
Discussion
In this study, we showed for the first time that TR6 could inhibit T cell chemotaxis both in vitro and in vivo. Mechanistically, such an effect was likely mediated by reverse signaling from TR6 to LIGHT on the T cell surface. T cells treated with TR6 failed to mobilize the signaling pathway leading from Cdc42 to p38 MAPK to actin polymerization upon encountering CXCL12/SDF-1a.
There are two possible mechanisms to explain the T cell migration inhibition caused by TR6. In the first possible mechanism, constitutive interaction between LIGHT and HveA on T cells promotes T cell mobility; soluble TR6-Fc, anti-LIGHT mAb, or HveA-Fc interferes with such interaction and leads to inhibited T cell mobility. To examine the plausibility of this mechanism, we coated TR6-Fc, anti-LIGHT mAb, or HveA-Fc on solid phase. In this case, only a minute amount of those molecules were actually adsorbed to the plate surface, and there were unlikely sufficient molecules leaking into the solution to block the LIGHT and HveA interaction (indeed, media conditioned in TR6-Fc-, anti-LIGHT mAb-, or HveA-Fc-coated wells had no effect on T cell migration). Moreover, these molecules fixed on a flat surface would probably not be able to interfere with the three-dimensional interaction between LIGHT and HveA on the cell surface. We showed that those solid-phase molecules could still inhibit T cell chemotaxis, suggesting that this first mechanism does not seem plausible.
The second possible mechanism dictates that TR6 triggers a negative signal reversely transmitted through a TNF family member(s) on the cell surface, most likely through LIGHT, and such signaling results in reduced T cell mobility. There are a couple of lines of evidence supporting this mechanism. First, we have recently proved that TR6 can transduce costimulation signals reversely via LIGHT into T cells (8, 11); such reverse signaling is not entirely out of the norm and has been reported for several other TNF family members, e.g., CD40L (19), CD30L (20), TNF-α (21), TNF-related activation-induced cytokine (22), FasL (23), and TRAIL (24). Second, when TR6-Fc, anti-LIGHT mAb, or HveA-Fc was put on solid phase, they could also inhibit T cell chemotaxis, as described above, likely by cross-linking their binding partners on the T cell surface. These data strongly suggest that the second mechanism is in operation. If so, how do we explain that TR6-Fc, HveA-Fc, and anti-LIGHT mAb in solution could inhibit T cell chemotaxis? It is possible that a low degree of cross-linking by the bivalent TR6-Fc or HveA-Fc, or their aggregates, is enough to trigger reverse signaling. For the anti-LIGHT mAb in solution, it was only effective after being cross-linked by antihuman IgG and without cross-linking, the mAb was not effective in solution (data not shown), suggesting the necessity of LIGHT cross-linking for the T cell migration inhibition.
TR6 could in theory reversely signal through LIGHT or FasL to achieve its effect in chemotaxis inhibition. Which one mediates such an effect of TR6? Our following observation is worth mentioning in this regard: 1) HveA-Fc, which binds to LIGHT but not FasL, had a similar effect as TR6-Fc (Fig. 2⇑, B and C), suggesting that the effect could be achieved through LIGHT without FasL. 2) Anti-LIGHT mAb could inhibit T cell migration (Fig. 2⇑, A and C), further suggesting that LIGHT mediated the effect. 3) LIGHT is expressed on resting T cells as we demonstrated before (8), and this provides a basis for TR6 to act on resting T cells. 4) The binding of TR6-Fc to T cells was predominately through LIGHT, because TR6-Fc could associate with >80% wild-type T cells and only with <20% LIGHT−/− T cells (data not shown). Taking together, these data suggest that LIGHT but not FasL is the main mediator for the observed effect of TR6 in chemotaxis inhibition. With that said, we cannot totally exclude the possibility that reverse signaling through FasL, or maybe some so far unidentified TR6-binding partner on the T cell surface, contributes to a lesser extent to the observed T cell chemotaxis inhibition by TR6. In this regard, it is worth mentioning that soluble TR6 have been shown to suppress soluble FasL-induced chemotaxis of microglial cells in a Transwell assay (25). In that system, the proposed mechanism is the blockage of forward signaling from FasL to Fas on the microglial cell surface by soluble TR6. However, it will be interesting to know whether such inhibition is the result of the general decrease of cell mobility by TR6 treatment.
We have identified signaling alternation in T cells treated with TR6. Selective change was revealed in one of the small GTPases, Cdc42, known to be essential in filopodium formation (18). This was followed by inability of p38 MAPK activation, which is essential for actin polymerization. Largely remaining unknown is how signals are initiated from LIGHT in the first place. LIGHT has a short featureless cytoplasmic tail (6) and certainly has no signaling capability by itself. It is reasonable to hypothesize that signaling from LIGHT depends on the adaptor molecules it associates with and such a hypothesis is under investigation.
TR6 is secreted by several kinds of tumors (15). In addition to its effect of blocking Fas-mediated apoptosis of tumor cells (15) and interfering with the two-way T cell costimulation mediated by HveA and LIGHT, TR6 now has a newly found function: inhibition of T cell chemotaxis. Consistent with this finding, in the rat glioma with forced expression of exogenous TR6, tumor-infiltrating T cells are significantly reduced compared with tumors without TR6 expression (25); our additional study discovered that serum TR6 levels of gastric cancer patients were inversely correlated to degree of lymphocyte infiltration in their tumor mass (data not shown). Therefore, T cell chemotaxis inhibition seems to be an additional strategy TR6-secreting tumors employ to evade immune surveillance by keeping potential tumor-infiltrating T cells at bay.
Although tumors might deploy TR6 to promote their survival, TR6 could also be used for a good cause. We previously demonstrated, in a mouse heart transplantation model, that TR6 administration results in prolonged heart allograft survival (1). Such a beneficial effect was initially explained by inhibition of LIGHT to HveA forward costimulation (1) and later by inhibition of the two-way costimulation between HveA and LIGHT on T cells by TR6 (1, 8). Obviously, our current finding provides an additional mechanism to explain the prolonged graft survival. Under the influence of TR6, alloresponsive T cells have reduced chemotaxis toward the allograft or its draining lymphoid organs and this will repress T cell activation, which eventually translates into reduced graft rejection force. Such a property of TR6 could be exploited in the future to repress local or systemic immune responses.
Acknowledgments
We acknowledge the editorial assistance and statistical analysis provided by Ovid Da Silva and Robert Boileau, respectively, of the Research Support Office, Research Centre, Centre Hospitalier de l’Université de Montréal, Montreal, Canada.
Footnotes
-
↵1 This work was supported by grants from the Canadian Institutes of Health Research (Grants MT-15673, MOP57697, and PPP57321), the Canadian Institutes of Health Research/Canadian Blood Service Partnership Program, the Kidney Foundation of Canada, the Heart and Stroke Foundation of Quebec, the Roche Organ Transplantation Research Foundation, Switzerland (Grant ROTRF 474950960), the Juvenile Diabetes Research Foundation (Grant 5-2001-540), and the J-Louis Levesque Foundation (to J.W.) J.W. is a National Scholar of the Fonds de la Recherche en Santé du Québec.
-
↵2 Address correspondence and reprint requests to Dr. Jiangping Wu, Laboratory of Immunology, Research Centre, Notre Dame Hospital, Centre Hospitalier de l’Université de Montréal, Pavilion DeSève, Room Y-5616, 1560 Sherbrooke Street East, Montreal, Quebec H2L 4M1, Canada. E-mail address: jianping.wu{at}umontreal.ca
-
↵3 Abbreviations used in this paper: L, ligand; LT, lymphotoxin; MAPK, mitogen-activated protein kinase; SDF-1α, stromal cell-derived factor 1α; PEC, peritoneal exudate cell; Med, medium.
- Received March 11, 2003.
- Accepted July 29, 2003.
- Copyright © 2003 by The American Association of Immunologists
References
In this issue
