Lymphocyte Activation Gene-3 (LAG-3) negatively regulates environmentally-induced autoimmunity

Abstract

Environmental factors including drugs, mineral oils and heavy metals such as lead, gold and mercury are triggers of autoimmune diseases in animal models or even in occupationally exposed humans. After exposure to subtoxic levels of mercury (Hg), genetically susceptible strains of mice develop an autoimmune disease characterized by the production of highly specific anti-nucleolar autoantibodies, hyperglobulinemia and nephritis. However, mice can be tolerized to the disease by a single low dose administration of Hg. Lymphocyte Activation Gene-3 (LAG-3) is a CD4-related, MHC-class II binding molecule expressed on activated T cells and NK cells which maintains lymphocyte homeostatic balance via various inhibitory mechanisms. In our model, administration of anti-LAG-3 monoclonal antibody broke tolerance to Hg resulting in autoantibody production and an increase in serum IgE level. In addition, LAG-3-deficient B6.SJL mice not only had increased susceptibility to Hg-induced autoimmunity but were also unresponsive to tolerance induction. Conversely, adoptive transfer of wild-type CD4(+) T cells was able to partially rescue LAG-3-deficient mice from the autoimmune disease. Further, in LAG-3-deficient mice, mercury elicited higher amounts of IL-6, IL-4 and IFN-γ, cytokines known to play a critical role in mercury-induced autoimmunity. Therefore, we conclude that LAG-3 exerts an important regulatory effect on autoimmunity elicited by a common environmental pollutant.

Figure 1. Exposure to Hg results in higher expression of LAG-3 on CD4⁺T cells and in an increased amount of soluble LAG-3 in serum.

(A) B6.SJL mice were left untreated or given 3 injections of HgCl₂ s.c. on days 0, 2 and 4. Splenocytes were harvested on day 8 and analyzed by flow cytometry. Figure depicts a representative dot plot of total splenocytes gated for LAG-3⁺ CD4⁺ T cells. (B) Results are expressed as frequency of LAG-3⁺ CD4⁺ T cells in each group (n = 3 or 4). (C) A.SW mice were challenged with 3 injections of HgCl₂ (30 µg/mouse) on days 0, 2 and 4. Concentration of soluble LAG-3 (sLAG-3) in serum was measured by ELISA (n = 4 or 5). Unpaired t test was used for statistical analysis where *indicates p<0.05.

Figure 2. LAG-3-deficient B6.SJL mice display an increased susceptibility to mercury-induced autoimmunity.

(A) Wild-type or Lag-3^−/− B6.SJL mice were challenged with 3 weekly s.c. injections of 30 µg HgCl₂/mouse during four weeks for a total of 12 injections. Serum Ig and autoantibodies were measured as described in the Material and Method section (n = 5 or 6). The levels of Ig depicted at week 0 represent the baseline levels in untreated LAG-3−/− and WT mice whereas the untreated mice in both groups had no detectable autoantibodies. (B) Mice from this experiment were sacrificed after 4 weeks. Their kidney sections were stained with goat anti-mouse IgG-FITC to detect IgG deposits (n = 5). Negative control depicts the staining of kidney section obtained from 3 week old untreated B6.SJL mice. The extent of kidney disease was scored depending on the amount and intensity of IgG staining as described in details in Material and Method section. To determine statistical significance between the groups of fig. 2A and 2B two way ANOVA was used where *indicates p<0.05, **indicates p<0.005 and ***indicates p<0.0001.

Figure 3. A short course of HgCl₂ injections is sufficient to induce mercury-induced autoimmunity in LAG-3-deficient B6.SJL mice.

(A) Wild-type or Lag-3^−/− B6.SJL mice were challenged with HgCl₂ (30 µg/mouse s.c.) on days 0, 2 and 4. Serum was analyzed by ELISA to determine IgG1 and IgE polyclonal antibodies and ANoA titers were measured by IF assay (n = 5 or 6). (B) Kidneys of mice from this experiment were harvested after 4 weeks, sectioned and stained with goat anti-mouse IgG-FITC to detect IgG deposit. The extent of kidney disease was scored depending on the amount and intensity of IgG staining as described in details in Material and Method section. To determine statistical significance between the groups of fig. 2A and 2B two way ANOVA was used where *indicates p<0.05, **indicates p<0.005 and ***indicates p<0.0001.

Figure 4. LAG-3-deficient mice cannot be tolerized to mercury.

Wild-type or LAG-3-deficient mice were given a tolerogenic dose of HgCl₂ (3 µg/mouse given i.p.) on day -7. Mice were then exposed to 3 injections of HgCl₂ (30 µg/mouse given s.c) on days 0, 2 and 4. Serum Ig were measured by ELISA and ANoA titers were determined by IF assay (n = 5 or 6). To determine statistical significance between the groups two way ANOVA was used where *indicates p<0.05.

Figure 5. Anti-LAG-3 monoclonal antibody breaks established tolerance against mercury-induced autoimmunity.

A.SW mice were tolerized to Hg and 7 days later they received a challenge dose of Hg in conjunction with anti-LAG-3 or control antibody as depicted in the Figure. Serum IgG1 and IgE levels were measured by ELISA and autoantibody titers were determined by IF assay (n = 5). The IgG and IgE concentrations depicted at wk -1 represent the baseline levels of antibodies in serum of untreated mice and no autoantibodies were detectable in untreated mice. To determine statistical significance between the groups two way ANOVA was used where *indicates p<0.05.

Figure 6. LAG-3-deficient mice produce higher amounts of cytokine following mercury treatment.

Wild-type and Lag-3^−/− mice were either given a challenge dose of HgCl₂ (30 µg/mouse s.c. on days 0, 2 and 4) or were left untreated. Splenocytes were harvested on day 8 and then stimulated in vitro with plate bound anti-CD3 and anti-CD28 antibodies. Supernatant were collected after 72 hrs to determine the levels of cytokine by ELISA (n = 3 or 4). Two way ANOVA was used to determine the statistical significance between the groups where *indicates p<0.05 and ***indicates p<0.0001.

Figure 7. Transfer of wild-type CD4⁺ T cells decreases autoimmunity in LAG-3-deficient mice.

CD4⁺ T cells were sorted using FACS and then transferred to LAG-3-deficient mice as depicted in the Figure. Autoantibody titers were measured by IF assay (n = 6). To determine statistical significance between the groups two way ANOVA was used where *indicates p<0.05.