Role of Hepatic Deposited Immunoglobulin G in the Pathogenesis of Liver Damage in Systemic Lupus Erythematosus

Abstract

The onset of hepatic disorders in patients with systemic lupus erythematosus (SLE) is frequent; however, the etiology and liver pathogenesis of SLE remain unknown. In the present study, the role of hepatic deposited immunoglobulin G (IgG) in SLE-derived liver damage was investigated. From a retrospective analysis of the medical records of 404 patients with lupus and from experimental studies on mice models, we found that liver dysfunction is common in SLE and liver damage with IgG deposition spontaneously develops in lupus-prone mice. Liver injury was recreated in mice by injecting IgG from lupus serum intrahepatically. The inflammation intensity in the liver decreased with IgG depletion and the lupus IgG-induced liver inflammation in FcγRIII-deficient mice was comparatively low; while, inflammation was increased in FcγRIIb-deficient mice. Macrophages, Kupffer cells, natural killer cells, and their products, but not lymphocytes, are required for the initiation of SLE-associated liver inflammation. Blocking IgG signaling using a spleen tyrosine kinase (Syk) inhibitor suppressed the liver damage. Our findings provided evidence of spontaneously established liver damage in SLE. They also suggested that hepatic-deposited lupus IgG is an important pathological factor in the development of liver injury and that hepatic inflammation is regulated by the Syk signaling pathway. Thus, Syk inhibition might promote the development of a therapeutic strategy to control liver damage in patients with SLE.

Keywords: hepatitis; immunoglobulin G deposition; innate immune cells; spleen tyrosine kinase; systemic lupus erythematosus.

Figure 1

Characterization of liver damage in lupus disease. (A) Clinical information in the cohort of 404 systemic lupus erythematosus (SLE) inpatients was reviewed and liver dysfunction (LD) was detected in 91 out of 404 SLE inpatients, while the other 313 SLE inpatients without LD served as the control group (normal liver function group). Serum biochemical markers of LD in 91 SLE patients are shown. (B) Histopathological examination of the liver and incidence of hepatic inflammation from lupus-prone mice MRL/lpr (n = 7, 22 weeks), B6 lpr (n = 10, 30 weeks), and B6 mice (n = 10, 30 weeks). Infiltration of inflammatory cells around the portal vein in these two mice models with different genetic backgrounds. The arrow indicates ballooning degeneration of hepatocytes that was observed in ~5% of the portal area. Original magnification, 100×. (C) Serum levels of liver enzymes in B6.MRL/lpr (B6 lpr) mice at various ages. ALT, alanine aminotransferase; AST, aspartate aminotransferase (n = 10). (D) Analysis of apoptosis of liver cells from B6 lpr mice (30 weeks) and age-matched control B6 mice. ***P < 0.001 (the result is representative of three independent experiments, n = 4 mice per group/experiment.) (E) Representative photomicrographs of the TUNEL in situ assay of liver sections of B6 lpr mice (30 weeks) with hematoxylin as a counter-stain. Original magnification, 400×.

Figure 2

Hepatic deposited lupus serum immunoglobulin G (IgG) induced liver inflammation in normal mice. (A) Immunohistochemical staining detected IgG deposition in the liver of MRL/lpr mice (25 weeks), B6 lpr mice (30 weeks), and C57BL/6 mice (30 weeks). Shown is a representative image of IgG deposition around the inflammatory portal area of a lupus-prone mice liver. Original magnification, 200× (B) Representative histopathologic photomicrograph of liver sections of C57BL/6 mice sacrificed 3 days after intrahepatic injection of serum (100 µL) from a lupus patient with liver disease, healthy control, lupus-prone MRL/lpr mice (35 and 5 weeka), and normal C57BL/6 mice or phosphate-buffered saline (PBS). H&E staining; original magnification, 200× (the result is representative of three independent experiments, n = 5 mice per group/experiment). (C) Kinetics of liver inflammation induced by intrahepatic injection of serum (100 µL) from a lupus patient with liver disease (left panel) and the severity of hepatic lesions triggered by various volumes of systemic lupus erythematosus (SLE)-serum 3 days after intrahepatic injection (right panel) in normal C57BL/6 mice (the result is representative of three independent experiments, n = 7 mice per group/experiment). (D) Intrahepatic IgG deposition on liver sections was examined by immunofluorescence 3 days after SLE-serum injection. Original magnification, 100×. (E) The severity of liver inflammation 3 days after intrahepatic injection of 100 µL SLE-serum with or without IgG depletion and 200 µg IgG from patients with SLE (s-IgG) or normal healthy individuals (n-IgG). **p < 0.01 (the result is representative of three independent experiments, n = 5 mice per group/experiment.) (F) Values of serum alanine aminotransferase and AST at various time points after intrahepatic injection of SLE-IgG or PBS in C57BL/6 mice. **p < 0.01, ***p < 0.001 (the result is representative of three independent experiments, n = 5 mice per group/experiment.).

Figure 3

Liver inflammation triggered by intrahepatic deposited lupus immunoglobulin G (IgG) depends on macrophages and TNF-α. (A) Immunohistochemistry detected Kupffer cells (F4/80+), dendritic cells (CD11c+), T cells (CD3+), and B cells (CD20+) in liver sections from lupus-prone mice. The results were from three independent experiments. Original magnification, 200× (B) Severity of liver inflammation and serum alanine aminotransferase (ALT) level 48 h after intrahepatic injection of systemic lupus erythematosus (SLE)-IgG (200 µg) in RAG-1^–/– and wild-type mice (the result is representative of three independent experiments, n = 5 mice per group/experiment). (C) Severity of liver inflammation and serum ALT level 48 h after intrahepatic injection of SLE-IgG (200 µg) in macrophage-depleted mice with clodronate liposomes (Clod.)-pretreated and control mice. **p < 0.01 (the result is representative of three independent experiments, n = 5 mice per group/experiment). (D) Representative histopathologic photomicrograph showing the severity of liver inflammation and serum ALT level 48 h after intrahepatic injection of SLE IgG in wild-type and TNF-α deficient mice. *p < 0.05, ***p < 0.001 (the result is representative of three independent experiments, n = 5 mice per group/experiment).

Figure 4

Synergistic effect of macrophage and natural killer (NK) cells on hepatic apoptosis induced by lupus immunoglobulin G (IgG). (A) Western blot showing the levels of TNF-α and IFN-γ in the liver from C57BL/6 mice with intrahepatic injection of the same dose of systemic lupus erythematosus (SLE) and normal serum, SLE, and normal IgG. (B) Flow cytometry analysis of NK cell activation in the liver from B6 lpr (25 w) and age-matched control mice using anti-NK1.1, anti-CD3e and anti CD69 antibodies. The result is representative of three independent experiments, n = 6 mice per group/experiment. (C) Flow cytometry analysis of NK cell activation in the liver from C57BL/6 mice with intrahepatic injection of SLE-serum (treated for 72 h), SLE-IgG (treated for 48 h), or phosphate-buffered saline (treated 48 h) using anti-NK1.1, anti-CD3e and anti CD69 antibodies (the result is representative of three independent experiments, n = 5 mice per group/experiment). (D) Severity of liver inflammation 48 h after intrahepatic injection of SLE-IgG in C57BL/6 mice with or without NK depletion using anti-ASGM1 antibody treatment. *p < 0.05 (the result is representative of three independent experiments, n = 5 mice per group/experiment.) (E) Severity of liver inflammation 48 h after intrahepatic injection of SLE-IgG (200 µg) in wild type and TNF-α deficient mice with or without NK cell depletion, respectively (***p < 0.001, *p < 0.05 the result is representative of three independent experiments, n = 5 mice per group/experiment). (F) Representative images of liver inflammation triggered by intrahepatic injection of recombined murine IFN-γ (500 ng/mouse) and/or TNF-α (200 ng/mouse) by HE staining. Original manifestation, 100× (the result is representative of three independent experiments, n = 7 mice per group/experiment). (G) Analysis of apoptotic hepatocytes 48 h after intrahepatic injection of SLE-IgG (200 μg/mouse) in mice with macrophage or NK cell depletion (the results are representative of three independent experiments, n = 4 mice per group/experiment).

Figure 5

Spleen tyrosine kinase (Syk) inhibitor suppressed liver inflammation induced by deposited lupus immunoglobulin G (IgG). (A) Severity of liver inflammation and serum alanine aminotransferase (ALT) level 48 h after intrahepatic injection of systemic lupus erythematosus (SLE)-IgG in mice with FcγRIIb-deficient (FcγRIIb^−/−) and FcγRIII-deficient (FcγRIII^−/−). *p < 0.05 (the result is representative of three independent experiments, n = 6 mice per group/experiment). (B) Representative image of Syk and p-Syk expression in inflamed liver sections from lupus-prone mice (aged 29 weeks) by immunohistochemistry and immunofluorescence, respectively. (C) Western Blot detected phosphorylation of Syk (p-Syk) or Syk in BMDMs stimulated by various doses of SLE-IgG for 30 min or stimulated by SLE-IgG (100 µg/ml) for the indicated time points as shown. BMDMs (2 × 10⁶ cells) were isolated from wild-type (WT) mice. The results were from three independent experiments. (D) Western blot analyzing p-Syk or Syk in BMDMs stimulated by 100 µg/ml SLE-IgG in presence or absence of Syk inhibitor R406 (2 µM) for 30 min. BMDMs (2 × 10⁶ cells) were isolated from WT or FcγR^−/− mice. The results are from three independent experiments. (E) WT-derived BMDMs (2 × 10⁶ cells) were stimulated with 100 µg/ml SLE-IgG in the presence or absence of Syk inhibitor R406 (2 µM) for 30 min. Cell lysates were subjected to immunoblotting analysis to detect the designated proteins or their phosphorylated cognates. (F) The severity of liver inflammation, serum ALT level, and analysis of apoptotic hepatocytes in C57BL/6 mice treated with or without Syk inhibitor R406 (10 mg/kg body) sacrificed 48 h after intrahepatic injection of SLE-IgG (200 µg) from a patient with lupus. *p < 0.05, **p < 0.01, ***p < 0.001 (the result is representative of three independent experiments, n = 7 mice per group/experiment).