Essential role of TNF family molecule LIGHT as a cytokine in the pathogenesis of hepatitis

Abstract

LIGHT is an important costimulatory molecule for T cell immunity. Recent studies have further implicated its role in innate immunity and inflammatory diseases, but its cellular and molecular mechanisms remain elusive. We report here that LIGHT is upregulated and functions as a proinflammatory cytokine in 2 independent experimental hepatitis models, induced by concanavalin A and Listeria monocytogenes. Molecular mutagenesis studies suggest that soluble LIGHT protein produced by cleavage from the cell membrane plays an important role in this effect through the interaction with the lymphotoxin-beta receptor (LTbetaR) but not herpes virus entry mediator. NK1.1+ T cells contribute to the production, but not the cleavage or effector functions, of soluble LIGHT. Importantly, treatment with a mAb that specifically interferes with the LIGHT-LTbetaR interaction protects mice from lethal hepatitis. Our studies thus identify a what we believe to be a novel function of soluble LIGHT in vivo and offer a potential target for therapeutic interventions in hepatic inflammatory diseases.

Figure 1. Pathogenic role of upregulated LIGHT in hepatitis.

(A) BALB/c mice were injected i.v. with 25 mg/kg ConA. At the indicated time points, the mice were sacrificed, and total RNA was extracted from liver and spleen. LIGHT, HVEM, LTβR, and GAPDH expression was examined by Northern blot analysis. (B and C) Wild-type (open circles) and LIGHT-deficient (filled circles) mice were injected i.v. with 30 mg/kg ConA. The survival (B) and serum ALT levels (C) of recipient mice were monitored. Sera were collected 18 hours after ConA injection, and the ALT levels were measured as described in Methods. SF, Sigma-Frankel. *p = 0.003.

Figure 2. Essential role of the soluble form of LIGHT in liver inflammation.

(A) B6 mice were injected i.v. with 30 mg/kg of ConA. At the indicated time points, serum was collected from the recipient mice and measured for soluble LIGHT concentration by LIGHT-specific ELISA. (B and C) B6 mice were injected with a sublethal dose of ConA (12.5 mg/kg) alone (filled circles, n = 13) or together with 20 μg plasmids encoding control pcDNA3.1 (open circles, n = 13), wild-type LIGHT (filled squares, n = 11), or LIGHTΔL (open squares, n = 10) by hydrodynamic injection technique. Survival of mice (B) and liver sections stained with H&E 18 hours after injection (C) were examined. N, necrotic area. *P = 0.3, **P = 0.033 between the groups by log-rank test. (D) BALB/c mice were injected i.p. with 50 μg of soluble LIGHT-flag fusion protein or control protein. One hour later, the mice were injected i.v. with 25 mg/kg ConA, and serum ALT levels were measured 6 hours later. One representative result from 3 independent experiments is shown as mean ± SD.

Figure 3. LTβR is necessary and sufficient for LIGHT-mediated hepatitis.

(A) B6 mice were injected with either empty vector or plasmid DNA encoding wild-type LIGHT or Y173F mutant by hydrodynamic method in combination with a sublethal dose of ConA (12.5 mg/kg). ALT levels were measured 18 hours after injection. One representative result from 2 independent experiments is shown as mean ± SD of 5 mice per group. (B) B6 mice were injected i.v. with 30 mg/kg of ConA together with 100 μg of control rat Ig (open circles, n = 12), anti-LTβR (filled circles, n = 10), or anti-HVEM mAb (filled squares, n = 9). The survival was monitored thereafter. (C) B6 mice were injected with either empty vector or plasmid DNA encoding LIGHT by hydrodynamic method in combination with a sublethal dose of ConA (12.5 mg/kg). The mice were treated i.p. with 100 μg of the indicated Abs 2 hours before the plasmid injections. After 18 hours, serum ALT levels were measured. One representative result from 2 independent experiments is shown as mean ± SD of 5 mice per group. hIg, control hamster Ig; rIg, control rat Ig.

Figure 4. Essential role of NKT cells in the production of soluble LIGHT in ConA-induced hepatitis.

(A and B) B6 mice were treated i.p. with 500 μg of either control mouse IgG or anti-NK1.1 mAb (PK136) on days 0 and 3 (A). Similarly, B6 mice were treated i.p. with 30 μg of control rabbit IgG or anti–asialo GM1 (ASGM1) on days 0 and 3 (B). On day 4, each group was injected i.v. with 30 mg/kg of ConA, and mouse sera were collected 1 hour later. The amounts of soluble LIGHT were measured by ELISA. *P = 0.02. (C) CD1d-deficient mice were injected with 20 μg of control vector or LIGHT-encoding plasmid by hydrodynamic injection in combination with a sublethal dose of ConA (12.5 mg/kg). Mouse sera were collected 18 hours later, and ALT levels were measured. **P = 0.002.

Figure 5. Blockade of LIGHT-LTβR interaction as a treatment for L. monocytogenes–induced hepatitis.

(A) B6 mice (open circles, n = 21) and LIGHT-deficient mice (filled triangles, n = 21) were injected i.p. with L. monocytogenes (2 × LD₅₀ per mouse), and their survival was monitored. P = 0.007 between the groups. (B) B6 mice were injected i.v. with 100 μg of control IgG (open circles, n = 10) or anti-LTβR mAb (filled circles, n = 11). At the same time, the mice were infected with L. monocytogenes (2 × LD₅₀ per mouse) by i.p. injection. Survival of mice was monitored thereafter. P = 0.03 between the groups. (C) As in B, B6 mice were infected with L. monocytogenes and treated with either anti-LTβR mAb or control IgG. Three days after infection, liver sections were prepared and stained with H&E.