A Tim-3 ligand, galectin-9 (Gal-9), modulates various functions of innate and adaptive immune responses. In this study, we demonstrate that Gal-9 prolongs the survival of Meth-A tumor-bearing mice in a dose- and time-dependent manner. Although Gal-9 did not prolong the survival of tumor-bearing nude mice, transfer of naive spleen cells restored a prolonged Gal-9-induced survival in nude mice, indicating possible involvement of T cell-mediated immune responses in Gal-9-mediated antitumor activity. Gal-9 administration increased the number of IFN-γ-producing Tim-3+ CD8+ T cells with enhanced granzyme B and perforin expression, although it induced CD4+ T cell apoptosis. It simultaneously increased the number of Tim-3+CD86+ mature dendritic cells (DCs) in vivo and in vitro. Coculture of CD8+ T cells with DCs from Gal-9-treated mice increased the number of IFN-γ producing cells and IFN-γ production. Depletion of Tim-3+ DCs from DCs of Gal-9-treated tumor-bearing mice decreased the number of IFN-γ-producing CD8+ T cells. Such DC activity was significantly abrogated by Tim-3-Ig, suggesting that Gal-9 potentiates CD8+ T cell-mediated antitumor immunity via Gal-9-Tim-3 interactions between DCs and CD8+ T cells.

Galectin-9 (Gal-9),4 a β-galactoside-binding lectin, was first identified as an eosinophil chemoattractant and activation factor (1, 2, 3, 4). Subsequent studies in our laboratory revealed that Gal-9, similar to other galectins, modulates a variety of biological functions such as cell aggregation and adhesion, apoptosis of tumor cells, and others (5, 6). We showed that Gal-9 induced the apoptosis of activated but not resting human CD4+ T cells (7). We also reported that Gal-9 induced apoptosis of Th1 cells that expressed Tim-3 via Gal-9-Tim-3 interactions (where Tim is “T cell Ig- and mucin domain-containing molecule”), which resulted in the suppression of experimental autoimmune encephalitis (8). More recently, we showed that Gal-9 ameliorated another representative autoimmune model, collagen-induced arthritis, by inducing the apoptosis of synoviocytes (9), suppressing the generation of Th17 cells, and up-regulating the induction of regulatory T cells (Tregs) (10). These results suggest that Gal-9 induces immunotolerance in animals with exaggerated immune responses, including autoimmune diseases.

Nevertheless, the fact that Gal-9 can be a prognostic factor with antimetastatic potential in melanoma and breast cancer led us to postulate that Gal-9 exhibits antitumor activity in tumor-bearing hosts (11, 12). We recently found that Gal-9 induces the maturation of human dendritic cells (DCs) from immature DCs (iDCs) (13) and that it also stimulates innate immune cells, such as monocytes and DCs, to secrete low levels of TNF-α via Gal-9-Tim-3 interactions in both mouse and human model systems (14). These results led us to further hypothesize that Gal-9 ameliorates immune suppression in tumor-bearing hosts by promoting innate and adaptive immunity, specifically via DC maturation.

The purpose of the present study is to test the hypothesis that Gal-9 potentiates immune responses in the context of tumor-induced immune suppression by maturing Tim-3+ DCs and Tim-3+CD8+ T cells to produce IFN-γ via Gal-9-Tim-3 interactions between DCs and CD8 T+ cells.

Female BALB/c and BALB/c nude mice, 6- to 8-wk old, were purchased from Japan SLC and CLEA Japan, respectively. All animals were kept under standard conditions in a 12-h day/night rhythm with free access to food and water ad libitum and received humane care in accordance with international guidelines and national law. Study protocols were reviewed and approved by the Animal Care and Use Committee of Kagawa University (Kita-gun, Kagawa, Japan).

The medium used for cell culture experiments was RPMI 1640 (Sigma-Aldrich) supplemented with 10% FBS (JRH Biosciences), antibiotic solution (Sigma-Aldrich), and 2-ME (Invitrogen).

Meth-A cells (a methylcholanthrene-induced sarcoma of BALB/c origin; 5 × 105) were passaged in the peritoneal cavities of mice (15). Meth-A cells were then cultured for 1 wk in culture medium at 37°C in 5% CO2 to avoid the contamination of immune cells. Expression and purification of recombinant human stable Gal-9 have been previously described (16), and stable Gal-9 was used instead of wild-type Gal-9 throughout the present experiments. Meth-A cells (5 × 105) were inoculated into the peritoneal cavities of BALB/c mice, and 10, 30, or 100 μg of Gal-9 per mouse was administered daily from day 0 to day 14. In some experiments, we started the daily treatment of tumor-bearing mice with Gal-9 (100 μg/mouse) from days 0, 3, 7, and 10 and stopped it on day 14. Meth-A cells were inoculated into the peritoneal cavities of BALB/c nude mice identically as in the method described above for wild-type BALB/c mice. Further, CD4+ and CD8+ T cells (1.5 × 107) from naive BALB/c mice were transferred i.p. to BALB/c nude mice 1 day before the i.p. inoculation of Meth-A cells. Gal-9 (100 μg/mouse) was administrated daily from day 0 to day 14 postchallenge, and the survival rates of mice were monitored.

Peritoneal exudate cells (PECs) were obtained from Gal-9- or PBS-treated tumor-bearing mice. NK cells were depleted by positive immunoselection using MACS anti-DX5 beads (Miltenyi Biotech) according to the manufacturer’s instructions, followed by purifying mononuclear cells by density gradient centrifugation using Percoll (GE Healthcare). 51Cr-labeled Meth A cells (target cells) were added at varying E:T ratios. After a 6-h culture, supernatants were harvested and transferred to Luma plates (PerkinElmer). Released 51Cr was counted using a gamma counter (PerkinElmer). Spontaneous 51Cr release was determined by incubating the radiolabeled target cells in the absence of effector cells. Maximal 51Cr release was determined by incubating target cells in 2% Triton X-100. Cytotoxicity was calculated as follows: lysis (%) = ([cpm of test sample − cpm of spontaneous release]/[cpm of maximal release − cpm of spontaneous release]) × 100.

Ag-specific responses were enumerated by an IFN-γ ELISPOT assay (BD Biosciences) according to the manufacturer’s instructions. Spleen cells from PBS- or Gal-9-treated tumor-bearing mice were cocultured with Meth-A cells treated with 50 μg/ml mitomycin C (MMC) in a MultiScreen 96-well plate (Millipore) precoated with anti-IFN-γ Abs for 18 h at 37°C in 5% CO2. After culture, the membranes were thoroughly washed with distilled water and incubated with biotinylated anti-mouse IFN-γ for 1.5 h at 37°C. The IFN-γ spots were developed by a 3-amino-ethylcarbazone (AEC) substrate reagent (BD Biosciences) and counted with a dissecting microscope.

For quantification of cytokine production, spleen cells from PBS- or Gal-9-treated tumor-bearing mice were plated in anti-CD3 Ab-coated (2 μg/ml) plate at 3 × 105 cells/well for 3 days. The amounts of IL-4 and IFN-γ in the supernatants of cultured cells were measured with ELISA kits (R&D Systems) according to the manufacturer’s instructions. To evaluate the number of cytokine-producing cells, intracellular cytokine staining was performed. Spleen cells were cultured using the above conditions for 3 days and incubated with PMA (50 ng/ml; Sigma-Aldrich), ionomycin (1 μg/ml; Sigma-Aldrich), and monensin (2 μM; BioLegend) during the last 14 h. The cells were fixed and permeabilized with Cytofix/Cytoperm solution (BD Biosciences). The following Abs were used for staining: CD4-FITC (L3T4; eBioscience), CD8-FITC (Ly-2; eBioscience), IFN-γ-allophycocyanin (XMG1.2; eBioscience), IL-4-allophycocyanin (clone 11B11; eBioscience), Tim-1-PE (RMT1–4; BioLegend), and Tim-3-PE (8B.2C12; eBioscience). Stained cells were analyzed using a FACSCalibur flow cytometer (Becton Dickinson) and FlowJo software (Tree Star).

PECs and spleen cells were harvested from BALB/c mice 7 days after peritoneal Meth-A cell inoculation. Spleen cells were cultured in culture medium containing murine IL-2 (mIL-2) (1 ng/ml; Peprotech) for 5 days and incubated with monensin (2 μM; Biolegend) for 14 h. Spleen cells and PECs were fixed/permeabilized with Cytofix/Cytoperm solution (BD Biosciences). The following Abs were used: granzyme B-FITC (16G6; eBioscience) and perforin-PE (eBio MAK-D; eBioscience). The cells were analyzed by flow cytometry.

Spleen cells were harvested from BALB/c mice 7 days after peritoneal Meth-A cell inoculation. Spleen cells from PBS- or Gal-9-treated tumor-bearing mice were pipetted into 96-well plates at 1 × 106 cells/well for 6 h. Spleen cells were fixed/permeabilized with Cytofix/Cytoperm solution (BD Biosciences). The following Abs were used: anti-Stat4-PE (pY693; BD Biosciences), anti-Stat6 Alexa Fluor 488 (pY641; BD Biosciences), and anti-Stat3-Alexa Fluor 647 (pY705; BD Biosciences). The cells were analyzed by flow cytometry.

CD4+ and CD8+ T cells were purified from naive BALB/c mice by magnetic cell sorting using CD4+ and CD8+ T cell isolation kits (Miltenyi Biotech) according to the manufacturer’s instructions. The purity of CD4+ and CD8+ T cells was >90% as assessed by flow cytometry. CD4+ or CD8+ T cells were cultured in anti-CD3 Ab-coated plates (2 μg/ml) at 3 × 105 cells/well for 24 h, followed by stimulation with Gal-9 at the indicated concentrations for 6 h. The proportion of apoptotic cells was determined by staining with propidium iodide (PI) and annexin V.

Purified CD4+ and CD8+ T cells from spleens of naive BALB/c mice were also used for lectin microarrays as previously described (17). Briefly, CMRA-labeled CD4+ and CD8+ T cells (2.5 × 106 cells/well) suspended in PBS/BSA were added to each well of a glass slide (100 μl/well) and incubated at 4°C for 1 h. After unbound cells were separated from the wells by gravity in cold PBS, cells bound to lectins immobilized on a glass slide were detected with an evanescent-field fluorescence scanner, SC-Profiler (Moritex), under Cy3 mode.

Spleen cells from tumor-bearing mice treated with PBS or Gal-9-for 7 days were washed in PBS with 0.5% FCS and 0.1% NaN3, and incubated with fluorochrome-labeled Abs as follows: anti-mouse Tim-3-PE (8B.2C12; eBioscience), CD11c-allophycocyanin (N418; eBioscience), I-A/I-E-PE (M5/114.15.2; eBioscience), and CD86-FITC (GL1; eBioscience). The cells were analyzed by flow cytometry.

iDCs were prepared from bone marrow precursors as previously described (18, 19). Briefly, bone marrow-derived cells obtained from femurs and tibias of BALB/c mice were cultured in medium supplemented with 20 ng/ml mouse GM-CSF (Peprotech). On days 2 and 4, nonadherent cells were removed, and adherent cells were cultured further. On day 6, nonadherent and loosely adherent iDCs were harvested and cultured in medium with varying concentrations of Gal-9 for another 24 or 48 h to assess the effects of Gal-9 on DC maturation from iDCs to mature DCs (mDCs). The cells were analyzed by flow cytometry.

T cells were depleted from spleen cells harvested from tumor-bearing mice treated with PBS or Gal-9 for 7 days by magnetic cell sorting using CD90 microbeads (Miltenyi Biotec) according to the manufacturer’s instructions. T cell-depleted splenocytes were used as APCs. After T cell depletion, CD3+ cells were <3% and I-A/I-E+ cells were >80%. In some experiments, Tim-3+ cells were depleted from isolated APCs by complement-dependent lysis using anti-Tim-3 mAbs as described previously (20). CD8+ T cells were also purified from spleens of naive BALB/c mice by magnetic cell sorting as described above. The generation of CTLs was performed in a 96-well plate in culture medium supplemented with anti-CD3 Ab (0.5 μg/ml) and mIL-2 (1 ng/ml). To inhibit Gal-9 activity, lactose (30 mM) was added to the culture medium. Sucrose (30 mM) was used as a control. Tim-3-Ig (100 ng/ml) was added to the culture medium to suppress the effects of Tim-3. Both APCs and CD8+ T cells were plated at 1.5 × 105 cells per well, cultured for 5 days, and incubated with PMA (50 ng/ml), ionomycin (1 μg/ml), and monensin (2 μM) for an additional 14 h. The cells were harvested and stained with anti-CD8 Ab and anti-IFN-γ Ab as described above. The levels of IFN-γ in the supernatant of the cultured cells were measured with ELISA kits (R&D Systems) according to the manufacturer’s instructions.

For statistical comparisons, nonparametric two-tailed Mann-Whitney U test, log-rank test, and one-way ANOVA were used. Comparisons between two groups were done by Mann-Whitney U test. All statistical analyses were done with Prizm 4 software (GraphPad Software).

We first designed experiments to determine whether Gal-9 exerts cell death-inducing activity in vitro against Meth-A cells (a murine fibrosarcoma cell line). Gal-9 apparently induced the cell death of Meth-A cells in a dose-dependent manner. Cell death was induced in almost all Meth-A cells by treatment with 10 μg/ml Gal-9 (supplemental figure 1).5 Such cell death was suppressed by lactose but not sucrose (data not shown). Thus, our next experiments were designed to clarify whether Gal-9 exhibits antitumor activity in vivo. Meth-A cells were inoculated into the peritoneal cavities of BALB/c mice. Intraperitoneal Gal-9 treatment was started at 1 h after the inoculation and continued every day until day 14. All of the control Meth-A-inoculated mice died by 16 days postchallenge, whereas Gal-9 treatment dramatically prolonged the survival of Meth-A-bearing mice in a dose-dependent manner (Fig. 1,A). Complete tumor rejection was observed in eight of 10 mice treated with 100 μg of Gal-9. Gal-9 treatment at 30 μg also resulted in significant survival prolongation, and complete rejection was observed in three of 10 mice (Fig. 1 A).

FIGURE 1.

Antitumor effect of Gal-9 in tumor-bearing mice is dependent upon T cells. BALB/c or BALB/c nude mice were inoculated with 5 × 105 Meth-A cells i.p., and the impact of Gal-9 treatment on tumor rejection was evaluated by monitoring the survival outcomes. Day 0 represents the day of tumor inoculation. A, Tumor-bearing mice were treated daily starting on day 0 with varying concentrations of Gal-9 (10, 30, and 100 μg/mouse; n = 10, each). B, Gal-9 treatment was started on various days after tumor inoculation (days 0, 3, 7, and 10; n = 10, each). C, Tumor-bearing nude mice were treated with 100 μg per mouse of Gal-9 after tumor inoculation (PBS, n = 7; Gal-9, n = 10). D, BALB/c nude mice were adoptively transferred with CD4+ and CD8+ T cells from naive BALB/c mice i.p. 1 day before tumor inoculation, followed by daily treatment with either PBS or 100 μg/mouse of Gal-9 (n = 10, each). ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001; NS, not significant.

FIGURE 1.

Antitumor effect of Gal-9 in tumor-bearing mice is dependent upon T cells. BALB/c or BALB/c nude mice were inoculated with 5 × 105 Meth-A cells i.p., and the impact of Gal-9 treatment on tumor rejection was evaluated by monitoring the survival outcomes. Day 0 represents the day of tumor inoculation. A, Tumor-bearing mice were treated daily starting on day 0 with varying concentrations of Gal-9 (10, 30, and 100 μg/mouse; n = 10, each). B, Gal-9 treatment was started on various days after tumor inoculation (days 0, 3, 7, and 10; n = 10, each). C, Tumor-bearing nude mice were treated with 100 μg per mouse of Gal-9 after tumor inoculation (PBS, n = 7; Gal-9, n = 10). D, BALB/c nude mice were adoptively transferred with CD4+ and CD8+ T cells from naive BALB/c mice i.p. 1 day before tumor inoculation, followed by daily treatment with either PBS or 100 μg/mouse of Gal-9 (n = 10, each). ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001; NS, not significant.

Close modal

To assess the therapeutic effects of Gal-9, daily treatment with Gal-9 was started on day 3, 7, or 10 postchallenge and stopped on day 18. Two of 10 mice treated with Gal-9 starting 3 days after Meth-A inoculation survived longer than 30 days (Fig. 1,B). Further, even when the treatment was started 10 days after Meth-A inoculation, Gal-9 treatment significantly prolonged the survival of Meth-A-bearing mice, although the difference was modest (Fig. 1 B). These results suggest that Gal-9 exhibits antitumor activity against Meth-A in vivo in dose- and time-dependent manners.

Because T cell-mediated immune responses are required for sufficient antitumor activity (21), we did experiments in Meth-A-bearing BALB/c nude mice to test the contribution of T cells or to see whether Gal-9 directly kills tumor cells. Gal-9 treatment did not prolong the survival of Meth-A-bearing nude mice compared with PBS-treated mice (Fig. 1,C), in contrast to our results in wild-type BALB/c mice (Fig. 1,A). When either 1.5 × 107 CD4+ or CD8+ cells from naive spleen cells of normal mice were transferred into nude mice 1 day before Meth-A inoculation, Gal-9 treatment significantly prolonged the survival of those mice (CD4+, p < 0.0001; CD8+, p = 0.0045) (Fig. 1 D). These results suggest that although Gal-9 did not directly induce sufficient cell death in Meth-A cells in vivo, the involvement of both CD4+ and CD8+ T cell-mediated immune responses is critical for Gal-9-induced antitumor activity.

We next designed experiments to determine whether Gal-9 enhances the cytotoxic activity of spleen cells in tumor-bearing mice on day 7 after treatment. We harvested PECs from tumor-bearing mice treated with Gal-9 and found a significant increase in the number of activated CD8+ T cells (CD8+CD44+CD62L) in the peritoneums of Gal-9-treated tumor-bearing mice compared with CD8+ T cells in PBS-treated tumor-bearing mice on day 7 (Fig. 2,A). Granzyme B and perforin in PEC CD8+ T cells were measured as surrogate markers of cytotoxic activity (22, 23). The levels of granzyme B and perforin in PEC CD8+ T cells from Gal-9-treated tumor-bearing mice were significantly higher than those of PBS-treated tumor-bearing mice on day 7 posttreatment (Fig. 2,B and supplemental figure 2A). Indeed, NK cell-depleted PECs from Gal-9-treated tumor-bearing mice exhibited higher cytolytic activity against Meth-A cells than those from PBS-treated mice, suggesting that Gal-9 treatment results in enhanced cytotoxic T cell activity (Fig. 2 C).

FIGURE 2.

Gal-9 enhances cytotoxic T cell activity and granzyme B and perforin expression in tumor-bearing mice. PECs and spleens obtained from PBS- or Gal-9-treated (i.p., 100 μg/mouse) tumor-bearing mice were compared for the presence of infiltrating cytotoxic CD8+ T cells 7 days after challenge. A, Number of CD8+CD44+CD62L T cells in the peritoneal cavities of each group. Results are the mean ± SEM of five animals for each group. B, Percentage of granzyme B+ or perforin+ cells in the peritoneal cavities of each group (gated on CD8+ T cells). Results are the mean ± SEM of five animals for each group. C, Specific cytotoxicity was measured by 51Cr release assay. NK cell-depleted PECs were incubated for 6 h with Meth-A cells labeled with 51Cr. Percentage of specific lysis was as shown in Material and Methods. D, Spleen cells from PBS or Gal-9-treated tumor-bearing mice were cultured with 1 ng/ml mIL-2 for 5 days, and then the percentages of granzyme B+ and perforin+ CD8+ T cells were evaluated by flow cytometry. Results are the mean ± SEM of five animals for each group. E, Spleen cells from PBS- or Gal-9-treated tumor-bearing mice were cultured for 6 h and analyzed for intracellular expression of the indicated molecules. Histograms for the indicated molecules (solid lines) and isotype-matched controls (filled histograms) are shown. Numbers are mean fluorescence intensity for each molecule. ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001.

FIGURE 2.

Gal-9 enhances cytotoxic T cell activity and granzyme B and perforin expression in tumor-bearing mice. PECs and spleens obtained from PBS- or Gal-9-treated (i.p., 100 μg/mouse) tumor-bearing mice were compared for the presence of infiltrating cytotoxic CD8+ T cells 7 days after challenge. A, Number of CD8+CD44+CD62L T cells in the peritoneal cavities of each group. Results are the mean ± SEM of five animals for each group. B, Percentage of granzyme B+ or perforin+ cells in the peritoneal cavities of each group (gated on CD8+ T cells). Results are the mean ± SEM of five animals for each group. C, Specific cytotoxicity was measured by 51Cr release assay. NK cell-depleted PECs were incubated for 6 h with Meth-A cells labeled with 51Cr. Percentage of specific lysis was as shown in Material and Methods. D, Spleen cells from PBS or Gal-9-treated tumor-bearing mice were cultured with 1 ng/ml mIL-2 for 5 days, and then the percentages of granzyme B+ and perforin+ CD8+ T cells were evaluated by flow cytometry. Results are the mean ± SEM of five animals for each group. E, Spleen cells from PBS- or Gal-9-treated tumor-bearing mice were cultured for 6 h and analyzed for intracellular expression of the indicated molecules. Histograms for the indicated molecules (solid lines) and isotype-matched controls (filled histograms) are shown. Numbers are mean fluorescence intensity for each molecule. ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001.

Close modal

Next, we asked whether spleen CD8+ T cells also express granzyme B and perforin. Spleen cells were harvested on day 7 after Gal-9 treatment and cultured with mIL-2 for 5 days ex vivo. Spleen CD8+ T cells from Gal-9-treated mice expressed significantly higher levels of granzyme B and perforin compared with PBS-treated mice (Fig. 2 D and supplemental figure 2B), indicating that Gal-9 treatment may potentiate the expression of granzyme B and perforin in CD8+ T cells of tumor-bearing hosts to promote antitumor immunity.

To assess whether Gal-9-mediated cytotoxicity is associated with the activation of a STAT-signaling pathway, we compared the phosphorylation of STAT-3, STAT-4, and STAT-6 in spleen CD8+ T cells of Gal-9-treated tumor-bearing mice with those from PBS-treated tumor-bearing mice by flow cytometric analysis after incubation for 6 h. It was evident that STAT-4 in CD8+ T cells from Gal-9-treated tumor-bearing mice was activated (mean fluorescence intensity = 11.1 ± 0.37), whereas no change was detected in CD8+ T cells from PBS-treated mice (mean fluorescence intensity = 4.14 ± 0.12) (p = 0.0286) (Fig. 2,E). In contrast, the phosphorylation of STAT-3 and STAT-6 remained unchanged in the two groups (STAT-3: PBS = 2.12 ± 0.01, Gal-9 = 2.09 ± 0.01; STAT-6: PBS = 1.47 ± 0.02, Gal-9 = 1.63 ± 0.05) (Fig. 2 E). This suggested that Gal-9 treatment preferentially activated STAT-4, which is closely correlated with the generation of granzyme B and perforin (24), as well as IFN-γ (25).

CD8+ T cell-mediated cytotoxic activity and T cell-derived cytokines, such as IFN-γ and IL-4, may play critical roles in T cell-mediated antitumor immunity (26). Therefore, experiments were done to assess the effects of Gal-9 on IFN-γ and IL-4 production in tumor-bearing mice. We stimulated spleen cells with mitomycin C-treated Meth-A cells as a specific Ag and observed that the number of IFN-γ-secreting cells in the Meth-A-stimulated spleen cells from PBS-treated tumor-bearing mice was very low, suggesting that immunosuppression occurred in the tumor-bearing mice. In contrast, Meth-A stimulation significantly increased the number of IFN-γ-secreting cells in Gal-9-treated tumor-bearing mice (n = 5) compared with PBS-treated tumor-bearing mice (n = 3, p = 0.0357) (Fig. 3 A), suggesting an immunopotentiating role of Gal-9 in tumor-bearing mice. In the case of IL-4, the cytokine-secreting cells were not detectable even after Meth-A-stimulation (data not shown).

FIGURE 3.

Gal-9 up-regulates the Ag-specific responses against tumor cells. Analysis of IFN-γ and IL-4 production by spleen cells obtained from PBS- and Gal-9-treated tumor-bearing mice. A, Spleen cells obtained from PBS- or Gal-9-treated tumor-bearing mice on day 14 were stimulated with Meth-A cells treated with mitomycin C and analyzed by ELISPOT assay as described in Materials and Methods. Results are the mean ± SEM (PBS: n = 3; Gal-9: n = 5). B, Spleen cells from PBS- or Gal-9-treated tumor-bearing mice on day 7 were stimulated on anti-CD3 Ab-coated plates for 72 h. IFN-γ and IL-4 concentrations in the culture supernatant were determined by ELISA. Results are the mean ± SEM (n = 5, each). ∗, p < 0.05; ∗∗, p < 0.001; NS, not significant.

FIGURE 3.

Gal-9 up-regulates the Ag-specific responses against tumor cells. Analysis of IFN-γ and IL-4 production by spleen cells obtained from PBS- and Gal-9-treated tumor-bearing mice. A, Spleen cells obtained from PBS- or Gal-9-treated tumor-bearing mice on day 14 were stimulated with Meth-A cells treated with mitomycin C and analyzed by ELISPOT assay as described in Materials and Methods. Results are the mean ± SEM (PBS: n = 3; Gal-9: n = 5). B, Spleen cells from PBS- or Gal-9-treated tumor-bearing mice on day 7 were stimulated on anti-CD3 Ab-coated plates for 72 h. IFN-γ and IL-4 concentrations in the culture supernatant were determined by ELISA. Results are the mean ± SEM (n = 5, each). ∗, p < 0.05; ∗∗, p < 0.001; NS, not significant.

Close modal

We next used anti-CD3 stimulation for 3 days to further assess the enhanced T cell cytokine production due to the effects of Gal-9 treatment. ELISA analyses revealed that the spleen cells of Gal-9-treated tumor-bearing mice produced significantly higher levels of both IFN-γ and IL-4 compared with those of PBS-treated tumor-bearing mice (n = 5, each) (Fig. 3 B), indicating that Gal-9 may exhibit antitumor activity by enhancing both IFN-γ and IL-4 production.

Additional experiments were done to identify the cellular source of IFN-γ and IL-4. Spleen cells from PBS- or Gal-9-treated tumor-bearing mice were stimulated with anti-CD3 followed by incubation with PMA, ionomycin, and monensin for an additional 14 h. The proportions of intracellular IFN-γ- and IL-4-positive CD4+ and CD8+ T cells were assessed by flow cytometry. This analysis revealed that the up-regulation of IFN-γ occurred for CD8+ but not CD4+ T cells (Fig. 4,A). IFN-γ production in CD4+ T cells was slightly decreased by Gal-9 treatment (Fig. 4,A), possibly because the Tim-3+CD4+ T cells that are normally responsible for IFN-γ production were induced to apoptosis by Gal-9 treatment (8, 10). In contrast to IFN-γ, increases of IL-4-producing cells were detected in CD4+ but not in CD8+ T cells, although the increase was not significant (p > 0.05) (Fig. 4 B). These results led us to assume that there is little or no involvement of type 2 cytotoxic, IL-4-producing CD8+ T cells in Gal-9-mediated antitumor activity.

FIGURE 4.

The source of IFN-γ in tumor-bearing mice. Spleen cells from PBS- or Gal-9-treated tumor-bearing mice were harvested on day 7, stimulated in anti-CD3 Ab-coated plates for 5 days, and cytokine production was assessed by flow cytometry. Harvested spleen cells were restimulated with PMA and ionomycin in medium containing monensin for an additional 14 h as described in Materials and Methods. A, Percentages of IFN-γ-producing cells gated on CD8+ or CD4+ cells. Numbers shown are the percentages of IFN-γ-positive cells (upper) and negative cells (lower) in CD8+ or CD4+ T cells. B, Percentages of IL-4-positive CD8+ or CD4+ T cells. Numbers are the percentages of IL-4-positive cells (upper) and negative cells (lower) in CD8+ or CD4+ T cells. Results are the mean ± SEM (n = 4, each). ∗, p < 0.05, NS, not significant.

FIGURE 4.

The source of IFN-γ in tumor-bearing mice. Spleen cells from PBS- or Gal-9-treated tumor-bearing mice were harvested on day 7, stimulated in anti-CD3 Ab-coated plates for 5 days, and cytokine production was assessed by flow cytometry. Harvested spleen cells were restimulated with PMA and ionomycin in medium containing monensin for an additional 14 h as described in Materials and Methods. A, Percentages of IFN-γ-producing cells gated on CD8+ or CD4+ cells. Numbers shown are the percentages of IFN-γ-positive cells (upper) and negative cells (lower) in CD8+ or CD4+ T cells. B, Percentages of IL-4-positive CD8+ or CD4+ T cells. Numbers are the percentages of IL-4-positive cells (upper) and negative cells (lower) in CD8+ or CD4+ T cells. Results are the mean ± SEM (n = 4, each). ∗, p < 0.05, NS, not significant.

Close modal

We found that Gal-9 preferentially induces cell death, including apoptosis, in CD3-stimulated CD4+ T cells and to a lesser extent in CD8+ T cells in a dose-dependent manner (Fig. 5, A and B). It thus became intriguing to clarify the reason for such divergent effects of Gal-9 on CD4+ and CD8+ T cells. We hypothesized that CD4+ and CD8+ T cells express different sets of N- or O-glycans. Surface glycome profiling experiments using a lectin microarray revealed that CD8+ T cells were more evidently bound to lectins having affinities for O-glycans and polylactosamines (poly-LN) than CD4+ T cells (Fig. 5 C), thus indicating that CD8+ T cells evidently express more O-glycans and poly-LN on their surfaces than do CD4+ T cells.

FIGURE 5.

Divergent effects of Gal-9 treatment on CD4+ and CD8+ T cells. CD4+ and CD8+ T cells from naive BALB/c mice cells were cultured in anti-CD3 Ab-coated plates for 24 h, followed by stimulation with Gal-9 at the indicated concentrations for 6 h. A, Annexin V and PI staining in PBS or Gal-9-treated CD4+ and CD8+ T cells. One set of representative data from five independent experiments is shown. B, Percentages of annexin V+PI cells in Gal-9-treated CD4+ and CD8+ T cells. (Annexin V+PI cells represent early apoptotic cells.) Results are the mean ± SEM (n = 5). ∗∗∗, p < 0.001 C, Differential lectin binding profiling of CD4+ and CD8+ T cells. Freshly isolated CD4+ and CD8+ T cells (2.5 × 106 cells) were labeled with CMRA and allowed to bind to a lectin microarray. Bound cells were detected by an evanescent-field fluorescent scanner. Red rectangles indicate O-glycan binding lectins and yellow rectangles indicate poly-LN binding lectins. M; Marker.

FIGURE 5.

Divergent effects of Gal-9 treatment on CD4+ and CD8+ T cells. CD4+ and CD8+ T cells from naive BALB/c mice cells were cultured in anti-CD3 Ab-coated plates for 24 h, followed by stimulation with Gal-9 at the indicated concentrations for 6 h. A, Annexin V and PI staining in PBS or Gal-9-treated CD4+ and CD8+ T cells. One set of representative data from five independent experiments is shown. B, Percentages of annexin V+PI cells in Gal-9-treated CD4+ and CD8+ T cells. (Annexin V+PI cells represent early apoptotic cells.) Results are the mean ± SEM (n = 5). ∗∗∗, p < 0.001 C, Differential lectin binding profiling of CD4+ and CD8+ T cells. Freshly isolated CD4+ and CD8+ T cells (2.5 × 106 cells) were labeled with CMRA and allowed to bind to a lectin microarray. Bound cells were detected by an evanescent-field fluorescent scanner. Red rectangles indicate O-glycan binding lectins and yellow rectangles indicate poly-LN binding lectins. M; Marker.

Close modal

We found that Tim-3+ expression in CD3-stiumlated CD8+ T cells from Gal-9-treated tumor-bearing mice was significantly up-regulated compared with that in CD8+ T cells from PBS-treated mice (Fig. 6,A). Furthermore, the percentage of IFN-γ-producing cells for Tim-3+CD8+ T cells in Gal-9-treated tumor-bearing mice was significantly higher than that for Tim-3CD8+ T cells (Fig. 6 B). These results suggest that Gal-9 increases the number of Tim-3+CD8+ T cells and that the source of IFN-γ is naturally the Tim-3+CD8+ T cell.

FIGURE 6.

Gal-9 up-regulates Tim-3 expression on CD8+ T cells in vitro. Spleen cells from PBS or Gal-9-treated tumor-bearing mice were harvested on day 7 and then stimulated in anti-CD3 Ab-coated plates for 5 days. Tim-3 expression was evaluated by flow cytometry. A, Tim-3 expression was significantly induced in CD8+ T cells obtained from Gal-9-treated-mice. B, The frequency of IFN-γ-producing cells for Tim-3+CD8+ T cells was significantly higher than those for Tim-3CD8+ T cells. Results are the mean ± SEM (n = 4). ∗, p < 0.05.

FIGURE 6.

Gal-9 up-regulates Tim-3 expression on CD8+ T cells in vitro. Spleen cells from PBS or Gal-9-treated tumor-bearing mice were harvested on day 7 and then stimulated in anti-CD3 Ab-coated plates for 5 days. Tim-3 expression was evaluated by flow cytometry. A, Tim-3 expression was significantly induced in CD8+ T cells obtained from Gal-9-treated-mice. B, The frequency of IFN-γ-producing cells for Tim-3+CD8+ T cells was significantly higher than those for Tim-3CD8+ T cells. Results are the mean ± SEM (n = 4). ∗, p < 0.05.

Close modal

IL-4 is normally produced by Tim-1+CD4+ T cells (Th2) in allergic conditions (27) and also by type 2 cytotoxic CD8+ T cells (28). Although we found that spleen cells from Gal-9-treated tumor-bearing mice produce an increased level of IL-4 (Fig. 3,B), Gal-9 did not increase the percentage of IL-4-producing CD4+ T cells (Fig. 4 B). Moreover, we also found that the frequency of Tim-1+CD4+ T cells was not enhanced by Gal-9 treatment (supplemental figure 3A) and confirmed that the source of IL-4 is, as expected, Tim-1+CD4+ T cells in Gal-9-treated tumor-bearing mice (supplemental figure 3B). We thus suggest that although Gal-9 may not increase the number of IL-4-producing cells, it does enhance the amount of IL-4 produced by Tim-1+CD4+ T cells. However, additional studies are required to clarify the exact mechanisms.

From these results, it is quite likely that Gal-9 contributes to T cell-mediated antitumor immunity. There are at least two explanations for the Gal-9-mediated up-regulation of antitumor immunity; one is that Gal-9 decreases the number or function of regulatory T cells (Tregs) and myeloid-derived suppressor cells (MSCs) capable of down-regulating proinflammatory T cell immune responses, and another is that Gal-9 up-regulates DC maturation resulting in T cell activation. To address the first possibility, we next designed experiments to determine whether Gal-9 treatment affects Tregs. However, we found there was no significant difference in the percentages of Tregs between PBS and Gal-9-treated tumor-bearing mice (supplemental figure 4A). Thus, it is unlikely that Gal-9 treatment ameliorates tumor-induced immunosuppression by reducing Treg numbers. We also addressed whether MSCs are involved in the Gal-9-mediated antitumor activity and found that Gal-9 decreased the number of MSCs weakly but significantly (supplemental figure 4B). This suggested that the decreased number of MSCs by Gal-9 treatment was at least partly involved in Gal-9-mediated antitumor activity.

We previously reported that Gal-9 induces human DC maturation in vitro (13) and also activates human and mouse DCs to release low levels of TNF-α (14). Together, these results raise the possibility that Gal-9 induces DC maturation in tumor-bearing mice. We therefore assessed the effects of Gal-9 on DC maturation in tumor-bearing mice in vivo. When tumor-bearing mice were treated daily with Gal-9 for 7 days, the percentage of I-A/I-E+CD86+ cells in the spleens of Gal-9-treated tumor-bearing mice (n = 5) was higher than that of PBS-treated tumor-bearing mice (n = 5) (Fig. 7 A).

FIGURE 7.

Gal-9 induces DC maturation associated with the expression of Tim-3. Spleens obtained from PBS-treated or Gal-9-treated tumor-bearing mice were harvested on day 7 and the frequency of mDCs was assessed by flow cytometry. A, Percentages of I-A/I-E+ CD86+ cells from PBS- or Gal-9-treated tumor-bearing mice. B, CD86+cells in CD11c+ cells from PBS- or Gal-9-treated tumor-bearing mice. Results are the mean ± SEM (n = 5, each). C, Comparison of Tim-3 expression between iDC and mDCs in Gal-9-treated tumor-bearing mice. Histograms of Tim-3 (solid lines) and isotype-matched controls (filled histograms) are shown. D, Comparisons of the number of CD11c+CD86+Tim-3+ cells in spleens in PBS- and Gal-9-treated tumor-bearing mice. Data represent the mean ± SEM (n = 5, each). ∗, p < 0.05, ∗∗, p < 0.01.

FIGURE 7.

Gal-9 induces DC maturation associated with the expression of Tim-3. Spleens obtained from PBS-treated or Gal-9-treated tumor-bearing mice were harvested on day 7 and the frequency of mDCs was assessed by flow cytometry. A, Percentages of I-A/I-E+ CD86+ cells from PBS- or Gal-9-treated tumor-bearing mice. B, CD86+cells in CD11c+ cells from PBS- or Gal-9-treated tumor-bearing mice. Results are the mean ± SEM (n = 5, each). C, Comparison of Tim-3 expression between iDC and mDCs in Gal-9-treated tumor-bearing mice. Histograms of Tim-3 (solid lines) and isotype-matched controls (filled histograms) are shown. D, Comparisons of the number of CD11c+CD86+Tim-3+ cells in spleens in PBS- and Gal-9-treated tumor-bearing mice. Data represent the mean ± SEM (n = 5, each). ∗, p < 0.05, ∗∗, p < 0.01.

Close modal

Furthermore, the percentage of CD86+ cells for CD11c+ cells from the spleens of Gal-9-treated tumor-bearing mice (n = 5) was significantly higher than that of PBS-treated tumor-bearing mice (n = 5) (Fig. 7,B). We did additional experiments to investigate whether a different level of Tim-3 expression is observed between CD86CD11c+ (iDCs) and CD86+CD11c+ (mDCs), as we have described that Gal-9-Tim-3 interactions are involved in LPS-induced DC activation. We found that Tim-3 expression for mDCs was greater than for iDCs (Fig. 7,C). Moreover, Gal-9 treatment significantly increased the number of Tim-3+ mDCs in the spleens of tumor-bearing mice (n = 5) compared with those in PBS-treated mice (n = 5) (Fig. 7 D), indicating that Gal-9 may promote maturation of DCs with Tim-3 expression.

We also did in vitro experiments using iDCs generated from bone marrow-derived cells to assess the effects of Gal-9 on DC maturation. GM-CSF-induced iDCs were harvested on day 6 and shown to be >80% CD11c+. After washing, iDCs were further cultured with Gal-9 for an additional 24 or 48 h. Gal-9 induced the maturation of iDCs to mDCs in a dose-dependent manner as measured by CD86 expression of CD11c+ cells after 24 h culture (Fig. 8,A). However, Gal-9 failed to induce the up-regulation of Tim-3+ in mDCs after 24 h of stimulation (Fig. 8,B). When iDCs were cultured for an additional 48 h with Gal-9, the number of Tim-3+ mDCs was significantly enhanced (Fig. 8 B), suggesting that CD86 expression is followed by Tim-3 expression. Taken together, our results suggest that Gal-9 enhances T cell-mediated antitumor immunity by increasing not only Tm-3+CD8+ T cells but Tim-3+ mDCs as well.

FIGURE 8.

Gal-9 induces DC maturation in vitro. GM-CSF-induced iDCs were cultured for 24 or 48 h in varying concentrations of Gal-9 (1, 3, and 10 μg/ml). A, Percentages of CD86+ in CD11c+ cells at 24 h. B, Percentage of Tim-3+ cells in CD11c+ cells after 24 (upper) and 48 h (lower) of culture in Gal-9. Results are the mean ± SEM (n = 4, each). ∗, p < 0.05.

FIGURE 8.

Gal-9 induces DC maturation in vitro. GM-CSF-induced iDCs were cultured for 24 or 48 h in varying concentrations of Gal-9 (1, 3, and 10 μg/ml). A, Percentages of CD86+ in CD11c+ cells at 24 h. B, Percentage of Tim-3+ cells in CD11c+ cells after 24 (upper) and 48 h (lower) of culture in Gal-9. Results are the mean ± SEM (n = 4, each). ∗, p < 0.05.

Close modal

To determine whether DCs from Gal-9-treated tumor-bearing mice have an increased potential to activate CD8+ T cells compared with those from PBS-treated tumor-bearing mice, we compared the effects of DCs from PBS- and Gal-9-treated tumor-bearing mice on IFN-γ production in coculture experiments with naive CD8+ T cells stimulated with anti-CD3. T cell-depleted spleen cells were used as a source of DCs. Coculture of DCs from Gal-9-treated tumor-bearing mice significantly enhanced IFN-γ production by CD8+ T cells compared with DCs from PBS-treated mice (Fig. 9,A) and also increased the percentage of IFN-γ-producing CD8+ T cells during coculture (Fig. 9 B). These results suggest that DCs from Gal-9-treated mice may exhibit a higher potential to fully prime naive CD8+ T cells.

FIGURE 9.

Critical role for Tim-3-expressing DCs on CD8+ T cell activation. Naive CD8+ T cells were cocultured with DCs from PBS- or Gal-9-treated tumor-bearing mice and stimulated with CD3 and mIL-2. A, The levels of IFN-γ in culture supernatants of CD3-stimulated CD8+ T cells cocultured with DCs from PBS- or Gal-9-treated tumor-bearing mice. B, Percentages of IFN-γ-producing CD8+ T cells cocultured with DCs from PBS or Gal-9-treated tumor-bearing mice as measured by flow cytometry. C, Suppressive effect of lactose on IFN-γ production by CD8+ T cells. Naive CD8+ T cells and DCs from Gal-9-treated tumor-bearing mice were stimulated with CD3 and mIL-2 in the presence of 30 mM lactose (Lac). Sucrose (Suc) was used as a control. D, Percentages of IFN-γ-producing CD8+ T cells cocultured with DCs from Gal-9-treated tumor-bearing mice before and after Tim-3+ cell depletion. E, The effects of Tim-3-Ig on IFN-γ-production by CD8+ T cells cocultured with DCs from Gal-9-treated tumor-bearing mice. Cells were cultured in the presence or absence of 100 ng/ml Tim-3-Ig. Percentage of IFN-γ-producing CD8+ T cells was assessed by flow cytometry. Results are the mean ± SEM (n = 5). ∗, p < 0.05; ∗∗, p < 0.01.

FIGURE 9.

Critical role for Tim-3-expressing DCs on CD8+ T cell activation. Naive CD8+ T cells were cocultured with DCs from PBS- or Gal-9-treated tumor-bearing mice and stimulated with CD3 and mIL-2. A, The levels of IFN-γ in culture supernatants of CD3-stimulated CD8+ T cells cocultured with DCs from PBS- or Gal-9-treated tumor-bearing mice. B, Percentages of IFN-γ-producing CD8+ T cells cocultured with DCs from PBS or Gal-9-treated tumor-bearing mice as measured by flow cytometry. C, Suppressive effect of lactose on IFN-γ production by CD8+ T cells. Naive CD8+ T cells and DCs from Gal-9-treated tumor-bearing mice were stimulated with CD3 and mIL-2 in the presence of 30 mM lactose (Lac). Sucrose (Suc) was used as a control. D, Percentages of IFN-γ-producing CD8+ T cells cocultured with DCs from Gal-9-treated tumor-bearing mice before and after Tim-3+ cell depletion. E, The effects of Tim-3-Ig on IFN-γ-production by CD8+ T cells cocultured with DCs from Gal-9-treated tumor-bearing mice. Cells were cultured in the presence or absence of 100 ng/ml Tim-3-Ig. Percentage of IFN-γ-producing CD8+ T cells was assessed by flow cytometry. Results are the mean ± SEM (n = 5). ∗, p < 0.05; ∗∗, p < 0.01.

Close modal

Because lactose, a representative β-galactoside, can interrupt binding between Gal-9 and its ligand, we cultured CD8+ T cells with DCs from Gal-9-treated tumor-bearing mice in the presence of lactose. Lactose significantly suppressed IFN-γ-producing cells, although sucrose, a control disaccharide, did not (Fig. 9,C). This suggests that the lectin nature of Gal-9 is involved in the activation of IFN-γ production by CD8+ T cells. We also deleted Tim-3+ DCs to assess the effects of Tim-3 expression on DCs with regard to DC function, as up-regulation of Tim-3 expression is observed not only in DCs but also in CD8+ T cells. Removal of Tim-3+ DCs resulted in a decreased percentage of IFN-γ-producing CD8+ T cells (Fig. 9 D), suggesting a critical role for Tim-3+ DCs in eliciting IFN-γ production by CD8+ T cells. Therefore, additional experiments were done to clarify the involvement of Gal-9-Tim-3 interactions in the activation of CD8+ T cells during coculture with DCs.

To confirm the above possibility, we assessed the effects of a fusion protein, Tim-3-Ig, which inhibits the binding of Gal-9 with Tim-3 (8, 14). As expected, the percentage of IFN-γ-producing cells was significantly suppressed in the presence of Tim-3-Ig (Fig. 9 E). Collectively, the present results suggest that Gal-9-Tim-3 mediated interactions between DCs and CD8+ T cells play a critical role for potentiating IFN-γ production by Tim-3+CD8+ T cells.

In the present experiments, we have reported that Gal-9 treatment induces antitumor activity in a T cell-dependent manner, suggesting a possible immunopotentiating activity for Gal-9 in the context of tumor immunity. This model is inconsistent with our recent findings that Gal-9 suppresses hyperimmune conditions in autoimmune animal models by inducing the cell death of Tim-3+ Th1 and Th17 cells and increasing Treg generation (10), and with more recent findings by Wang et al. that the accumulation of Tim-3+CD8+ T cells in a skin graft model was suppressed by Gal-9 treatment (29). Thus, it is quite important to elucidate which mechanisms are involved in the Gal-9-mediated enhancement of antitumor immunity in tumor-bearing hosts.

Tumor cells can escape the attack of the host immune system through various mechanisms, including immune evasion, immunosuppression, or others (30). Progressive tumor growth seems to be at least partly ascribed to tumor cell-induced down-regulation of T cell-mediated immune responses (31). Indeed, an increased number or enhanced activity of Tregs (32, 33) and MSCs (34, 35) is observed in tumor-bearing mice and/or patients with cancer, and it has been thought to be the major cause of immunosuppression in tumor hosts. It is, however, unlikely that Gal-9 exhibits antitumor activity by decreasing Tregs, because Gal-9 treatment does not induce a decrease in Treg numbers. In contrast, a decreased functioning of MSCs may, at least in part, contribute to Gal-9-mediated antitumor activity, because Gal-9 decreased the frequency of Gr-1+CD11b+ cells, probably MSCs, in the spleens of Gal-9-treated tumor-bearing mice.

One strategy for combating tumors is to increase the number and activity of cytotoxic CD8+ T cells against tumor cells (36). Indeed, there is increasing evidence that inoculation of DCs or DC-derived exosomes is a useful therapy for some cancer patients (37, 38). We have shown that Gal-9 induces maturation of human immature DCs (13), as well as activating human and mouse Tim-3-expressing DCs to produce low levels of TNF-α (14). In the present experiments we have shown that Gal-9 increases the number of Tim-3+CD86+ mDCs in vivo and in vitro. However, the mechanisms by which Gal-9 can affect the maturation, number, and Tim-3 expression of DCs are still unclear, although it is imperative to elucidate them. Moreover, we have found that mature Tim-3+CD86+ DCs activate Tim-3+CD8+ T cells to produce IFN-γ and CD4+ T cells to produce IL-4. Although IFN-γ usually contributes to antitumor activity for many tumors (39, 40), IL-4 has been found to contribute antitumor activity for certain cell types such as Meth-A (41). Thus, it is conceivable that Gal-9 exhibits antitumor activity in Meth-A-bearing mice because Gal-9 enhances both IFN-γ and IL-4 production.

The fact that Gal-9 treatment also increases the number of Tim-3+CD8+ T cells in tumor-bearing mice is consistent with the findings that the increases of Tim-3+CD8+ T cells and Tim-3 ligand are observed in a mouse acute graft-vs-host disease model, although there is no direct evidence that the Tim-3 ligand in these experiments is Gal-9 (42). The exact mechanisms of Gal-9-induced up-regulation of Tim-3+ CD8+ T cells remain unclear, and future research should focus on this area.

Gal-9 induces the apoptosis of not only Tim-3-expressing Th1 cells (8) but also CD8 T cells (29) through Gal-9-Tim-3 interactions. In the present experiments we have, however, shown that Gal-9 does not induce cell death in activated CD8+ T cells, although the apoptosis of activated CD4+ T cells is evidently induced by Gal-9. These divergent effects of Gal-9 on CD4+ and CD8+ T cells may be explained by the findings that CD4+ and CD8+ T cells express different patterns of O-glycans and poly-LN, as CD8+ T cells express more O-glycans and poly-LN on their surfaces than do CD4+ T cells. Moreover, CD8+ T cells in a skin graft model (29) and CD8+ T cells in the present studies may have different glycosylation patterns, as stimuli promoting Th1, Th2, or Th-17 differentiation can differentially regulate the glycosylation patterns of Th cells and modulate their reactivities with Gal-1 (43). Therefore, the divergent effects of Gal-9 may arise from the different expression of glycans between CD4+ and CD8+ T cells.

More recently, Stowell et al. have shown that Gal-8 exhibits different binding, with an N-terminal carbohydrate recognition domain (CRD) of Gal-8 recognizing sulfated and sialylated glycans and a Gal-8 C-terminal CRD recognizing blood groups and poly-LN. This suggests that Gal-8 dimerization promotes functional bivalency for each CRD, which allows Gal-8 to signal phosphatidylserine exposure in leukocytes through C-terminal domain recognition of poly-LN (44). We have previously shown that Gal-9 has an entirely different binding affinity for galactosides compared with other galectins, and even the CRDs of N- and C-terminal Gal-9 exhibit different binding activities (45). Thus, the possibility cannot be excluded that both the different binding affinities of each Gal-9 CRD and the different glycosylation patterns of immune cells are associated with the divergent effects of Gal-9.

Gal-9 can potentiate the expression of granzyme B and perforin and IFN-γ production by Tim-3+CD8+ T cells, and it activates STAT-4 but not STAT-3 or STAT-6. It has been shown that IL-27 directly activates not only STAT-4 but also STAT-1, STAT-2, STAT-3, and STAT-5 in CD3-stimuated CD8+ T cells, and it enhances antitumor activity by up-regulation of granzyme B, perforin, and IFN-γ production (46). Thus, STAT-4 appears to play a more critical role in the generation of cytotoxic CD8+ T cells, although further studies are required to verify this. Furthermore, it is well known that CD44 is expressed not only on activated CD4+ but also on activated CD8+ T cells (47), and we have shown that Gal-9 binds to CD44 on T cells and suppresses the binding of hyaluronic acid to CD44 (48). Thus, it cannot be excluded that CD44 is also involved in the function of Gal-9 together with Tim-3, although this remains to be clarified.

The removal of Tim-3+ DCs resulted in decreased IFN-γ production from CD8+ T cells, raising the hypothesis for the significance of Gal-9-Tim-3 interactions in CD8+ T cell activation. Similar galectin-mediated crosslinking has been demonstrated in marine sponges, where galectins mediate crosslinking between aggregating factors expressed on sponge cells necessary for cell aggregation (49). From the present results, Gal-9 may enhance antitumor immunity via Gal-9-Tim-3 interactions between Tim-3+ DCs and Tim-3+ CD8+ T cells in tumor-induced immune suppressive conditions, although Gal-9 induces the cell death of Tim-3+ Th1, Th17 and CD8+ T cells in hyperimmune conditions (10).

In total, the possibility can be raised that Gal-9 exhibits a pivotal “thermostat”-like function to maintain immunological homeostasis; it exhibits immunopotentiating activity during a state of immunosuppression and an immunosuppressive function in the context of a hyperimmune condition.

We thank the staffs at Division of Animal Experiment and Radioisotope Research, Life Science Research Center, Institute of Research Promotion, Kagawa University.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This study was supported in part by a grant from the Japanese Ministry of Education, Culture, Sports, Science, and Technology.

4

Abbreviations used in this paper: Gal-9, Galectin-9; CRD, carbohydrate recognition domain; DC, dendritic cell; iDC, immature DC; mDC, mature dendritic cell; mIL-2, murine IL-2; MSC, myeloid-derived suppressor cell; PEC, peritoneal exudate cell; PI, propidium iodide; poly-LN, polylactosamine; Tim, T cell Ig- and mucin domain-containing molecule; Treg, regulatory T cell.

5

The online version of this article contains supplemental material.

1
Matsumoto, R., H. Matsumoto, M. Seki, M. Hata, Y. Asano, S. Kanegasaki, R. L. Stevens, M. Hirashima.
1998
. Human ecalectin, a variant of human galectin-9, is a novel eosinophil chemoattractant produced by T lymphocytes.
J. Biol. Chem.
273
:
16976
-16984.
2
Matsushita, N., N. Nishi, M. Seki, R. Matsumoto, I. Kuwabara, F. T. Liu, Y. Hata, T. Nakamura, M. Hirashima.
2000
. Requirement of divalent galactoside-binding activity of ecalectin/galectin-9 for eosinophil chemoattraction.
J. Biol. Chem.
275
:
8355
-8360.
3
Matsumoto, R., M. Hirashima, H. Kita, G. J. Gleich.
2002
. Biological activities of ecalectin: a novel eosinophil-activating factor.
J. Immunol.
168
:
1961
-1967.
4
Saita, N., E. Goto, T. Yamamoto, I. Cho, K. Tsumori, H. Kohrogi, K. Maruo, T. Ono, M. Takeya, Y. Kashio, et al
2002
. Association of galectin-9 with eosinophil apoptosis.
Int. Arch. Allergy Immunol.
128
:
42
-50.
5
Asakura, H., Y. Kashio, K. Nakamura, M. Seki, S. Dai, Y. Shirato, M. J. Abedin, N. Yoshida, N. Nishi, T. Imaizumi, et al
2002
. Selective eosinophil adhesion to fibroblast via IFN-γ-induced galectin-9.
J. Immunol.
169
:
5912
-5918.
6
Hirashima, M., Y. Kashio, N. Nishi, A. Yamauchi, T. A. Imaizumi, T. Kageshita, N. Saita, T. Nakamura.
2004
. Galectin-9 in physiological and pathological conditions.
Glycoconj. J.
19
:
593
-600.
7
Kashio, Y., K. Nakamura, M. J. Abedin, M. Seki, N. Nishi, N. Yoshida, T. Nakamura, M. Hirashima.
2003
. Galectin-9 induces apoptosis through the calcium-calpain-caspase-1 pathway.
J. Immunol.
170
:
3631
-3636.
8
Zhu, C., A. C. Anderson, A. Schubart, H. Xiong, J. Imitola, S. J. Khoury, X. X. Zheng, T. B. Strom, V. K. Kuchroo.
2005
. The Tim-3 ligand galectin-9 negatively regulates T helper type 1 immunity.
Nat. Immunol.
6
:
1245
-1252.
9
Seki, M., K. M. Sakata, S. Oomizu, T. Arikawa, A. Sakata, M. Ueno, A. Nobumoto, T. Niki, N. Saita, K. Ito, et al
2007
. Beneficial effect of galectin 9 on rheumatoid arthritis by induction of apoptosis of synovial fibroblasts.
Arthritis Rheum.
56
:
3968
-3976.
10
Seki, M., S. Oomizu, K. M. Sakata, A. Sakata, T. Arikawa, K. Watanabe, K. Ito, K. Takeshita, T. Niki, N. Saita, et al
2008
. Galectin-9 suppresses the generation of Th17, promotes the induction of regulatory T cells, and regulates experimental autoimmune arthritis.
Clin. Immunol.
127
:
78
-88.
11
Kageshita, T., Y. Kashio, A. Yamauchi, M. Seki, M. J. Abedin, N. Nishi, H. Shoji, T. Nakamura, T. Ono, M. Hirashima.
2002
. Possible role of galectin-9 in cell aggregation and apoptosis of human melanoma cell lines and its clinical significance.
Int. J. Cancer
99
:
809
-816.
12
Irie, A., A. Yamauchi, K. Kontani, M. Kihara, D. Liu, Y. Shirato, M. Seki, N. Nishi, T. Nakamura, H. Yokomise, M. Hirashima.
2005
. Galectin-9 as a prognostic factor with antimetastatic potential in breast cancer.
Clin. Cancer Res.
11
:
2962
-2968.
13
Dai, S. Y., R. Nakagawa, A. Itoh, H. Murakami, Y. Kashio, H. Abe, S. Katoh, K. Kontani, M. Kihara, S. L. Zhang, et al
2005
. Galectin-9 induces maturation of human monocyte-derived dendritic cells.
J. Immunol.
175
:
2974
-2981.
14
Anderson, A. C., D. E. Anderson, L. Bregoli, W. D. Hastings, N. Kassam, C. Lei, R. Chandwaskar, J. Karman, E. W. Su, M. Hirashima, et al
2007
. Promotion of tissue inflammation by the immune receptor Tim-3 expressed on innate immune cells.
Science
318
:
1141
-1143.
15
Noguchi, Y., E. C. Richards, Y. T. Chen, L. J. Old.
1995
. Influence of interleukin 12 on p53 peptide vaccination against established Meth A sarcoma.
Proc. Natl. Acad. Sci. USA
92
:
2219
-2223.
16
Nishi, N., A. Itoh, A. Fujiyama, N. Yoshida, S. Araya, M. Hirashima, H. Shoji, T. Nakamura.
2005
. Development of highly stable galectins: truncation of the linker peptide confers protease-resistance on tandem-repeat type galectins.
FEBS Lett.
579
:
2058
-2064.
17
Tateno, H., N. Uchiyama, A. Kuno, A. Togayachi, T. Sato, H. Narimatsu, J. Hirabayashi.
2007
. A novel strategy for mammalian cell surface glycome profiling using lectin microarray.
Glycobiology
17
:
1138
-1146.
18
Inaba, K., M. Inaba, N. Romani, H. Aya, M. Deguchi, S. Ikehara, S. Muramatsu, R. M. Steinman.
1992
. Generation of large numbers of dendritic cells from mouse bone marrow cultures supplemented with granulocyte/macrophage colony-stimulating factor.
J. Exp. Med.
176
:
1693
-1702.
19
Haase, C., T. N. Jorgensen, B. K. Michelsen.
2002
. Both exogenous and endogenous interleukin-10 affects the maturation of bone-marrow-derived dendritic cells in vitro and strongly influences T-cell priming in vivo.
Immunology
107
:
489
-499.
20
Nishimura, T., K. Iwakabe, M. Sekimoto, Y. Ohmi, T. Yahata, M. Nakui, T. Sato, S. Habu, H. Tashiro, M. Sato, A. Ohta.
1999
. Distinct role of antigen-specific T helper type 1 (Th1) and Th2 cells in tumor eradication in vivo.
J. Exp. Med.
190
:
617
-627.
21
Smyth, M. J., D. I. Godfrey, J. A. Trapani.
2001
. A fresh look at tumor immunosurveillance and immunotherapy.
Nat. Immunol.
2
:
293
-299.
22
Masson, F., T. Calzascia, W. Di Berardino-Besson, N. de Tribolet, P. Y. Dietrich, P. R. Walker.
2007
. Brain microenvironment promotes the final functional maturation of tumor-specific effector CD8+ T cells.
J. Immunol.
179
:
845
-853.
23
Yang, Z. Z., A. J. Novak, S. C. Ziesmer, T. E. Witzig, S. M. Ansell.
2006
. Attenuation of CD8+ T-cell function by CD4+CD25+ regulatory T cells in B-cell non-Hodgkin’s lymphoma.
Cancer Res.
66
:
10145
-10152.
24
Curtsinger, J. M., J. O. Valenzuela, P. Agarwal, D. Lins, M. F. Mescher.
2005
. Type I IFNs provide a third signal to CD8 T cells to stimulate clonal expansion and differentiation.
J. Immunol.
174
:
4465
-4469.
25
Li, Q., C. Eppolito, K. Odunsi, P. A. Shrikant.
2006
. IL-12-programmed long-term CD8+ T cell responses require STAT4.
J. Immunol.
177
:
7618
-7625.
26
Cronin, S. J., J. M. Penninger.
2007
. From T-cell activation signals to signaling control of anti-cancer immunity.
Immunol. Rev.
220
:
151
-168.
27
de Souza, A. J., T. B. Oriss, J. O'Malley, K. A. Ray, L. P. Kane.
2005
. T cell Ig and mucin 1 (TIM-1) is expressed on in vivo-activated T cells and provides a costimulatory signal for T cell activation.
Proc. Natl. Acad. Sci. USA
102
:
17113
-17118.
28
Boni, A., G. Iezzi, E. Degl'Innocenti, M. Grioni, E. Jachetti, A. Camporeale, M. Bellone.
2006
. Prolonged exposure of dendritic cells to maturation stimuli favors the induction of type-2 cytotoxic T lymphocytes.
Eur. J. Immunol.
36
:
3157
-3166.
29
Wang, F., W. He, H. Zhou, J. Yuan, K. Wu, L. Xu, Z. K. Chen.
2007
. The Tim-3 ligand galectin-9 negatively regulates CD8+ alloreactive T cell and prolongs survival of skin graft.
Cell. Immunol.
250
:
68
-74.
30
Zou, W..
2005
. Immunosuppressive networks in the tumour environment and their therapeutic relevance.
Nat. Rev. Cancer
5
:
263
-274.
31
Kim, R., M. Emi, K. Tanabe, K. Arihiro.
2006
. Tumor-driven evolution of immunosuppressive networks during malignant progression.
Cancer Res.
66
:
5527
-5536.
32
Curiel, T. J., G. Coukos, L. Zou, X. Alvarez, P. Cheng, P. Mottram, M. Evdemon-Hogan, J. R. Conejo-Garcia, L. Zhang, M. Burow, et al
2004
. Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival.
Nat. Med.
10
:
942
-949.
33
Somasundaram, R., L. Jacob, R. Swoboda, L. Caputo, H. Song, S. Basak, D. Monos, D. Peritt, F. Marincola, D. Cai, et al
2002
. Inhibition of cytolytic T lymphocyte proliferation by autologous CD4+/CD25+ regulatory T cells in a colorectal carcinoma patient is mediated by transforming growth factor-β.
Cancer Res.
62
:
5267
-5272.
34
Serafini, P., I. Borrello, V. Bronte.
2006
. Myeloid suppressor cells in cancer: recruitment, phenotype, properties, and mechanisms of immune suppression.
Semin. Cancer Biol.
16
:
53
-65.
35
Yang, R., Z. Cai, Y. Zhang, W. H. Yutzy, IV, K. F. Roby, R. B. Roden.
2006
. CD80 in immune suppression by mouse ovarian carcinoma-associated Gr-1+CD11b+ myeloid cells.
Cancer Res.
66
:
6807
-6815.
36
Yan, X., B. D. Johnson, R. J. Orentas.
2004
. Murine CD8 lymphocyte expansion in vitro by artificial antigen-presenting cells expressing CD137L (4-1BBL) is superior to CD28, and CD137L expressed on neuroblastoma expands CD8 tumour-reactive effector cells in vivo.
Immunology
112
:
105
-116.
37
Chaput, N., N. E. Schartz, F. Andre, J. Taieb, S. Novault, P. Bonnaventure, N. Aubert, J. Bernard, F. Lemonnier, M. Merad, et al
2004
. Exosomes as potent cell-free peptide-based vaccine. II. Exosomes in CpG adjuvants efficiently prime naive Tc1 lymphocytes leading to tumor rejection.
J. Immunol.
172
:
2137
-2146.
38
Zitvogel, L., A. Regnault, A. Lozier, J. Wolfers, C. Flament, D. Tenza, P. Ricciardi-Castagnoli, G. Raposo, S. Amigorena.
1998
. Eradication of established murine tumors using a novel cell-free vaccine: dendritic cell-derived exosomes.
Nat. Med.
4
:
594
-600.
39
Ferris, R. L., T. L. Whiteside, S. Ferrone.
2006
. Immune escape associated with functional defects in antigen-processing machinery in head and neck cancer.
Clin. Cancer Res.
12
:
3890
-3895.
40
Rudge, G., S. P. Barrett, B. Scott, I. R. van Driel.
2007
. Infiltration of a mesothelioma by IFN-γ-producing cells and tumor rejection after depletion of regulatory T cells.
J. Immunol.
178
:
4089
-4096.
41
Hiraoka, S., N. Takeuchi, Y. Bian, H. Nakahara, M. Kogo, K. Dunussi-Joannopoulos, S. Wolf, S. Ono, H. Fujiwara.
2005
. B7.2-Ig fusion proteins enhance IL-4-dependent differentiation of tumor-sensitized CD8+ cytotoxic T lymphocyte precursors.
Int. Immunol.
17
:
1071
-1079.
42
Oikawa, T., Y. Kamimura, H. Akiba, H. Yagita, K. Okumura, H. Takahashi, M. Zeniya, H. Tajiri, M. Azuma.
2006
. Preferential involvement of Tim-3 in the regulation of hepatic CD8+ T cells in murine acute graft-versus-host disease.
J. Immunol.
177
:
4281
-4287.
43
Toscano, M. A., G. A. Bianco, J. M. Ilarregui, D. O. Croci, J. Correale, J. D. Hernandez, N. W. Zwirner, F. Poirier, E. M. Riley, L. G. Baum, G. A. Rabinovich.
2007
. Differential glycosylation of TH1, TH2 and TH-17 effector cells selectively regulates susceptibility to cell death.
Nat. Immunol.
8
:
825
-834.
44
Stowell, S. R., C. M. Arthur, K. A. Slanina, J. R. Horton, D. F. Smith, R. D. Cummings.
2008
. Dimeric Galectin-8 induces phosphatidylserine exposure in leukocytes through polylactosamine recognition by the C-terminal domain.
J. Biol. Chem.
283
:
20547
-20559.
45
Cebo, C., G. Vergoten, J. P. Zanetta.
2002
. Lectin activities of cytokines: functions and putative carbohydrate-recognition domains.
Biochim. Biophys. Acta
1572
:
422
-434.
46
Morishima, N., T. Owaki, M. Asakawa, S. Kamiya, J. Mizuguchi, T. Yoshimoto.
2005
. Augmentation of effector CD8+ T cell generation with enhanced granzyme B expression by IL-27.
J. Immunol.
175
:
1686
-1693.
47
Lesley, J., N. Howes, A. Perschl, R. Hyman.
1994
. Hyaluronan binding function of CD44 is transiently activated on T cells during an in vivo immune response.
J. Exp. Med.
180
:
383
-387.
48
Katoh, S., N. Ishii, A. Nobumoto, K. Takeshita, S. Y. Dai, R. Shinonaga, T. Niki, N. Nishi, A. Tominaga, A. Yamauchi, M. Hirashima.
2007
. Galectin-9 inhibits CD44-hyaluronan interaction and suppresses a murine model of allergic asthma.
Am. J. Respir. Crit. Care Med.
176
:
27
-35.
49
Pfeifer, K., M. Haasemann, V. Gamulin, H. Bretting, F. Fahrenholz, W. E. Muller.
1993
. S-type lectins occur also in invertebrates: high conservation of the carbohydrate recognition domain in the lectin genes from the marine sponge Geodia cydonium.
Glycobiology
3
:
179
-184.