Identification of potential cytokine pathways for therapeutic intervention in murine primary biliary cirrhosis

Abstract

Primary biliary cirrhosis (PBC) is considered a model autoimmune disease, with the most highly directed and specific autoantibody in both murine and human autoimmunity, the anti-mitochondrial autoantibody (AMA). However, therapeutic advances in this disease have lagged behind. Herein we have taken advantage of our unique model of murine PBC in which mice immunized with 2-octynoic acid coupled to BSA (2OA-BSA), a compound identified by quantitative structure activity relationships (QSAR) of human AMA binding, develop an intense inflammatory cholangitis with striking similarities to humans with PBC. In particular, we have constructed several unique gene-deleted mice, including mice deleted of IL-12p40, IL-12p35, IFN-γ, IL-23p19, IL-17A, IL-17F and IL-22, immunized these animals with 2OA-BSA and followed the natural history of immunopathology to identify key pathways that might provide clues for successful therapy. Our data indicate that whereas both IL-12/Th1 and IL-23/Th17 are involved in cholangitis, it is the IL-12/Th1 signaling pathway that elicits pathology. In fact, deletion of IFN-γ prevents disease and suppresses autoantibodies. Importantly, deletion of the Th17 cytokines IL-17A and IL-22, but not IL-17F, reduces biliary damage; IL-17A-knockout mice have reduced levels of anti-mitochondrial antibody. We further demonstrate that the production of IFN-γ is significantly decreased in the liver of IL-23p19(-/-), IL-17A(-/-) and IL-22(-/-) mice compared with controls. However, the ability of T cells to produce IFN-γ was not affected in Th17 cytokine-deficient mice. Our data indicate that a deficient Th17 pathway suppresses the accumulation of IFN-γ producing cells in liver during the early phase of cholangitis. In conclusion, whereas IFN-γ has a pivotal role in the early events involved in the pathogenesis of autoimmune cholangitis induced by 2OA-BSA, the IL-23/Th17 pathway potentiates the effects of IL-12/IFN-γ-mediated immunopathology.

Figure 1. IL-12p40 is required for 2OA-BSA-induced cholangitis.

(A) H&E staining of representative liver from WT, IL-12p40^−/− (p40^−/−), IL-23^−/− (p19^−/−) and IL-12p35^−/− (p35^−/−) mice 8 weeks after 2OA-BSA immunization. Portal inflammatory changes with interlobular bile duct damage (red arrow) were observed in the liver of WT, p19^−/− and p35^−/− mice but not p40^−/− mice; normal bile ducts in p40^−/− mice (blue arrow). (B) Portal inflammation and bile duct damage were examined in individual animals. The pathological score of portal inflammation and biliary cell damage were evaluated in WT (n = 17), p19^−/− (n = 14), p35^−/− (n = 3) and p40^−/− (n = 5) mice. *p<0.05, **p<0.01, ***p<0.001. (C) Immunohistochemical analysis of the liver of WT mice and IL-23p19^−/− mice at 8 weeks after 2OA-BSA immunization using mAbs to CD4 and CD8.

Figure 2. Flow cytometric analysis of the lymphocytic infiltration in liver from IL-23p19^−/−, IL-12p35^−/− and control WT mice.

HMNCs were isolated from liver samples at 8 weeks after the first immunization. The data are expressed as the cell number per gram of tissue. **p<0.01, ***p<0.001.

Figure 3. Inflammatory cytokine production in mice immunized with 2OA-BSA.

(A) Serum levels of IFN-γ and IL-17A in B6 and IL-23p19^−/− mice 2 and 4 weeks after initial 2OA-BSA immunization. (B) Production of IFN-γ and IL-17A in supernatant fluids of cultured splenic and hepatic MNCs (n = 4) with anti-CD3/CD28 mAbs for 3 days (C) inflammatory cytokines in extracted liver protein from WT mice (n = 8) and IL-23p19^−/− (n = 8) mice. *p<0.05, **p<0.01, ***p<0.001.

Figure 4. Pathological changes in liver of mice immunized with 2OA-BSA.

(A) Representative H&E staining profiles of liver from IL-17A^−/−, IL-17F^−/− and IL-22^−/− mice 8 weeks after immunization. Portal inflammatory changes with interlobular bile duct damage (red arrow) were observed. IL-17F^−/− demonstrating more epithelioid granulomas (red arrowhead) in liver compared to IL-17A^−/− and IL-22^−/− mice. (B) Scoring of portal inflammation and bile duct damage in liver from WT (n = 25), IL-17A^−/− (n = 19), IL-17F^−/− (n = 13) and IL-22^−/− (n = 14) mice.

Figure 5. Inflammatory cytokine production in mice immunized with 2OA-BSA.

(A) Serum levels of IL-17A, IFN-γ, TNF-α and IL-6 in WT, IL-17A^−/−, IL-17F^−/− and IL-22^−/− mice, respectively, at 2 weeks following 2OA-BSA immunization. (B) Production of IFN-γ in supernatant fluids of cultured splenic and hepatic MNCs (n = 4) with anti-CD3/CD28 mAbs at 3 days. (C) The level of inflammatory cytokines in extracted liver protein from WT, IL-17A^−/−, IL-17F^−/− and IL-22^−/− mice, respectively. Each group n = 8. *p<0.05, **p<0.01, ***p<0.001.

Figure 6. Pathological changes in the liver of IFN-γ^−/− mice immunized with 2OA-BSA.

(A) Representative H&E stained of liver sections from IFN-γ^−/− mice compared with WT mice 8 weeks after 2OA-BSA immunization. Portal inflammatory changes with interlobular bile duct damage (red arrow) were observed in WT mice; normal bile ducts in IFN-γ^−/− mice (blue arrows). (B) Scoring of portal inflammation and bile duct damage in sections of liver from WT and IFN-γ^−/− mice (each group, n = 16).

Figure 7. Genetic deletion of IFN-γ results in decreased lymphocytic infiltration of the liver.

MNCs were isolated from spleen and liver samples of WT and IFN-γ^−/− mice 8 weeks after 2OA-BSA immunization. The number of MNCs and cell subpopulations were analyzed by standard flow cytometry. The data are expressed as the cell number per gram of tissue. Each group, n = 16; * p<0.05; ***p<0.001; ****p<0.0001.

Figure 8. Comparative levels of PDC-E2-specific autoantibodies in cytokine deficient mice induced by 2OA-BSA immunization; samples were collected at serial time points.

The fold changes of AMA OD values of each group are compared to the data of WT mice (n = 8–16/group). AMA values of IL-17A^−/− and IFN-γ mice compared to other groups, *p<0.05.