Galectin-9 activates and expands human T-helper 1 cells

Abstract

Galectin-9 (Gal-9) is known for induction of apoptosis in IFN-γ and IL-17 producing T-cells and amelioration of autoimmunity in murine models. On the other hand, Gal-9 induced IFN-γ positive T-cells in a sarcoma mouse model and in food allergy, suggesting that Gal-9 can have diametric effects on T-cell immunity. Here, we aimed to delineate the immunomodulatory effect of Gal-9 on human resting and ex vivo activated peripheral blood lymphocytes. Treatment of resting lymphocytes with low concentrations of Gal-9 (5-30 nM) induced apoptosis in ∼60% of T-cells after 1 day, but activated the surviving T-cells. These viable T-cells started to expand after 4 days with up to 6 cell divisions by day 7 and an associated shift from naïve towards central memory and IFN-γ producing phenotype. In the presence of T-cell activation signals (anti-CD3/IL-2) Gal-9 did not induce T-cell expansion, but shifted the CD4/CD8 balance towards a CD4-dominated T-cell response. Thus, Gal-9 activates resting T-cells in the absence of typical T-cell activating signals and promotes their transition to a TH1/C1 phenotype. In the presence of T-cell activating signals T-cell immunity is directed towards a CD4-driven response by Gal-9. Thus, Gal-9 may specifically enhance reactive immunological memory.

Figure 1. Gal-9 triggers TIM-3-independent cell death and PBMC expansion.

A. Resting PBMCs were incubated with anti-TIM3-APC or isotype control antibody or B. with streptavidin-Alexa488, biotinylated Gal-9+streptavidin-Alexa488 in the presence or absence of α-lactose. Cell surface staining was evaluated by flow cytometry C. Resting PBMCs (n = 3) were treated with medium or 15 nM of recombinant Gal-9 for up to 7 days. Cell death was determined by Annexin-V staining. D. PBMCs (n = 8) treated as in (C) were analyzed for cell number. E. Resting PBMCs (n = 8) were treated as in (C) in the presence of α-lactose or sucrose, and analyzed for cell number. F. Resting PBMCs (n = 3) were treated with medium or 15 nM of recombinant Gal-9, Gal-1, Gal-2, Gal-3, or Gal-8 for up to 7 days, after which cell number was determined. All graphs represent mean +/− SD.

Figure 2. Recombinant and Gal-9(s) dose dependently activate T-cells.

A. Resting PBMCs (n = 5) were incubated in medium for up to 7 days. Distribution of cell populations was analyzed every day by flow cytometry. B. Similar as (A), but in the presence of 15 nM Gal-9. C. Resting PBMCs (n = 5) treated with medium or 15 nM of recombinant Gal-9 for up to 7 days were analyzed for T-cell activation by staining for activation marker CD25. D. Resting PBMCs were treated as in (C) in the presence of α-lactose, after which CD25 expression was analyzed at 1 day. E. Resting PBMCs (n = 5) were treated as in (C) and analyzed for expression of TIM-3. F. PBMCs or monocyte-depleted PBMCs were treated with Gal-9 for 7 days, and analyzed for CD25 expression G. Resting PBMCs (n = 6) were treated with a concentration range of recombinant Gal-9 or physiologically occurring isoform Gal-9(S) for 7 days and analyzed for cell density. H. Resting PBMCs (n = 6) were treated as in G and CD25 expression was determined. All graphs represent mean +/– SD.

Figure 3. Galectin-9 treatment expands CD4⁺ T-cells.

A. Representative plot of 6 independent experiments of resting PBMCs stained with CFSE and subsequently incubated in medium for up to 7 days. At Day 7, PBMCs were harvested, stained with the T-cell marker CD3, and CFSE peak pattern was analyzed within the CD3⁺ cells by flow cytometry. B. Representative plot of 6 independent experiments of resting PBMCs as treated in A, but in the presence of 15 nM Gal-9. C. Analysis of (A) and (B) showing percentage of CD3⁺ T-cells in the respective peak of all independent experiments (mean +/− SEM). D. Analysis of (B) showing the number of CFSE peaks of all independent experiments. E. Representative plots of 3 independent experiments of resting PBMCs stained with CFSE and subsequently incubated in medium or 15 nM Gal-9 (+/− lactose) for up to 7 days. At Day 7, PBMCs were harvested, stained with the T-cell marker CD4, and CFSE peak pattern was analyzed within the CD-3⁺ cells by flow cytometry. F. As in (E) but stained for the T-cell maker CD8. G. Resting PBMCs (n = 4) were treated for up to 7 days with medium or Gal-9 and analyzed for CD4 and CD8 distribution. All graphs represent mean +/− SD unless stated otherwise.

Figure 4. Gal-9 treatment of resting PBMCs shifts T-cells towards central memory and T-helper 1 phenotype.

A−C. PBMCs were treated for 7 days with Gal-9 or Gal-9(S) after which the percentage of naïve (A), central memory (B), and effector memory (C) was evaluated by flow cytometry. D−G. Resting PBMCs were treated for 7 days with Gal-9 or Gal-9(S), after which T-cell cytokine production was analyzed by flow cytometry as described in M&M. The percentage of IL-2 (n = 11) (D), IFN-γ (E), IL-17 (F) and IL-4 (G) was determined. H. Supernatants of medium and Gal-9 (15 nM) treated PBMCs was harvested at day7 after which the amount of secreted IFN-γ and IL-17 was determined with ELISA. I. PBMCs (n = 11) were treated for 7 days with Gal-9 or Gal-9(S) after which the percentage of regulatory T-cells was determined. Unless indicated otherwise; (n = 12).

Figure 5. Gal-9 treatment of TCR-activated T-cells reverses CD8/CD4 distribution and shifts T-cells towards a central memory phenotype.

A−B. T-cells were activated by anti-CD3/IL-2, or additionally with 15 nM Gal-9 or Gal-9(S). After 7 days the percentage of CD4 (A) and CD8 (B) were determined by flow cytometry. C. T-cells were activated as in (A) and percentage of T-cells with central memory phenotype was determined. D−G. T-cells were activated as in A, after which T-cell cytokine production was analyzed by flow cytometry as described in M&M. The percentage of IL-2 (n = 11) (D), IFN-γ (E), IL-17 (F) and IL-4 (G) was determined. H. T-cells were activated as in (A), after which the percentage of regulatory T-cells was determined. Unless indicated otherwise (n = 12).