RAGE deficiency ameliorates autoimmune hepatitis involving inhibition of IL-6 production via suppressing protein Arid5a in mice

Activation of T cells and pro-inflammatory cytokines are essential for human autoimmune hepatitis. RAGE is one of the receptors for the inflammatory alarm molecule high mobility group box 1 (HMGB1), and it is involved in autoimmune hepatitis. However, the molecular mechanism of RAGE in the context of autoimmune hepatitis remains elusive. This study aimed to identify the function and mechanism of RAGE in autoimmune hepatitis. The role and underlying mechanisms of RAGE signaling-driven immune inflammatory response in ConA-induced experimental hepatitis were examined using the RAGE-deficient mice. We found that the RAGE deficiency protected the mouse from liver inflammatory injury caused by the ConA challenge. mRNA expression of VCAM-1, IL-6, and TNF-α within the livers is markedly decreased in RAGE-deficient mice compared to wild-type mice. In parallel, RAGE deficiency leads to reduced levels of the serum pro-inflammatory cytokines IL-6 and TNF-α as compared with wild-type control mice. RAGE-deficient mice exhibit increased hepatic NK cells and decreased CD4+ T cells compared with wild-type control mice. Notably, in vivo blockade of IL-6 in wild-type mice significantly protected mice from ConA-induced hepatic injury. Furthermore, RAGE deficiency impaired IL-6 production and was associated with decreased expression of Arid5a in liver tissues, a half-life IL-6 mRNA regulator. RAGE signaling is important in regulating the development of autoimmune hepatitis. Immune regulation of RAGE may represent a novel therapeutic strategy to prevent immune-mediated liver injury.


Introduction
RAGE (the receptor for advanced glycation end products) is a multiligand receptor that binds structurally diverse molecules, such as AGEs (advanced glycation end products), HMGB1 (high mobility group box 1), S100 family members, amyloid-beta and DNA [1,2].RAGE regulates a variety of cellular functions including inflammation, apoptosis, proliferation, and autophagy [3].Targeting RAGE signaling through the use of inhibitors and anti-RAGE antibodies can be a promising treatment strategy [4].Indeed, research by our team and others has demonstrated that the HMGB1/ RAGE signaling axis is crucial for the development of autoimmune hepatitis and liver damage [5][6][7][8].In a clinical setting, autoimmune hepatitis patients (AIH) displayed distinct profiles of serum EN-RAGE (extracellularly newly identified receptor for advanced glycation end products binding protein), sRAGE (soluble RAGE), or EN-RAGE/sRAGE compared to healthy controls.Furthermore, these three parameters exhibited potential as novel biomarkers for AIH diagnosis and prognosis evaluation [9].Inhibiting the signaling pathways for HMGB1/RAGE/NF-κB and HMGB1/ TLR4/NLRP3 alleviated hepatic inflammation caused by type II diabetes in mice [10].Additionally, earlier studies showed that RAGE is involved in hepatic ischemia/reperfusion (I/R) injury and liver cancer [11,12].However, the molecular mechanism by which RAGE participates in the pathogenesis of autoimmune hepatitis remains unclear.In the present study, we found that RAGE deficiency protects mice from ConA-induced hepatitis, and this protection is associated with the regulation of IL-6 production.Particularly, RAGE deficiency dramatically reduces the IL-6 production evoked by the ConA challenge via suppressing protein Arid5a, which stabilizes IL-6 mRNA (message RNA).

Mice
RAGE knockout mice with C57BL/6 background were a gift from Professor Fang Zheng in our department.6-8-weekold male wild-type mice weighing 18-22 g were used.They were bred in specific pathogen-free conditions, and all of the experiments were performed in accordance with the guidelines of the Tongji Medical College Animal Care and Use Committee.The study protocols were specifically reviewed and approved by this ethics committee.

ConA-induced hepatitis
ConA was dissolved in pyrogen-free phosphate-buffered saline (PBS) and intravenously (I.V.) administered to the mice at a dose of 15 mg/kg or 20 mg/kg body weight.Mice in the control group were injected with an equal volume of PBS.

Biochemical and histological
Blood was collected 10 h after ConA injection, and serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) measurements were examined.Liver tissue was fixed in 4% paraformaldehyde and cut into 4-μm-thick sections for H&E staining.

Myeloperoxidase (MPO) assay
The liver tissues of experimental mice were measured using a myeloperoxidase (MPO) test kit (Nanjing Jiancheng Bioengineering Institution, China) according to the manufacturer's protocol.

ELISA
Blood was collected 10 h after ConA injection, and serum levels of IL-6, IFN-γ, and TNF-α were measured using an enzyme-linked immunosorbent assay kit (BioLegend, San Diego, CA) according to the manufacturer's protocol.

Quantitative real-time PCR
Total RNA was extracted from the liver tissue of the experimental mice using RNAiso Plus (TAKARA, Dalian, China), followed by cDNA synthesis using the Revert Aid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, USA).Subsequently, cDNA was used to measure the mRNA levels of IL-6, TNF-α and IFN-γ, using Fast Start Universal SYBR Green Master Mix (Roche Pharma, AG, Germany).β-actin was used as the normalization control.The 2 −△△CT method was used to calculate the relative mRNA levels.The primer sequences used were as follows: IL-6 forward: GAG ACT TCC ATC CAG TTG CC. reverse: AAG TGC ATC ATC GTT GTT CAT ACA .TNF-α forward: CAT CTT CTC AAA ATT CGA GTG ACA A. reverse: TGG GAG TAG ACA AGG TAC AACCC.

T cell proliferation in vitro
The spleen lymphocytes were labeled with CFSE (5 μM/ml, BioLegend) and stimulated with ConA (2 μg/ml) in vitro.The cells stained with CD3 were collected at 48 h and 72 h, respectively, for flow cytometry.

Statistical analysis
The results are shown as means ± standard deviation (SD).Statistical significance of differences was analyzed by a oneway ANOVA or Student's t test.Data were analyzed using GraphPad Prism 8 software.Values of p < 0.05 were considered significant.

Expression of RAGE increases in liver tissues and T cells by ConA challenge
We first examined the RAGE expression levels in liver tissues and T cells after injection of Con A. Immunochemical staining for RAGE in liver tissues showed that RAGE expression was significantly increased in ConA-challenged mice than that of PBS-injected control animals (Fig. 1A,  B).Also, RAGE expressed by CD4 + T cells isolated from hepatic mononuclear lymphocytes as well as spleens was markedly increased in the ConA-challenged group as compared with control mice with PBS injection (Fig. 1C).Interestingly, flow cytometric analysis showed that the ConA challenge or injection of PBS control did not affect the expression of RAGE by CD8 + T cells derived from either livers or spleens (Fig. 1D).

RAGE deficiency protects mice from ConA-induced hepatitis
Next, the survival rate of mice exposed to the ConA challenge was assessed in RAGE-deficient mice and control wild-type (WT) mice.Mice with RAGE deficiency survived significantly longer than wild-type (WT) control mice after the ConA challenge (Fig. 2A).In parallel, serum aminotransferase activities (ALT and AST) evoked by ConA injection in mice with RAGE deficiency were significantly lower than those of WT control mice (Fig. 2B).As shown in the liver and spleen samples, these organs derived from RAGEdeficient mice exhibited less damage after ConA treatment compared with those from wild-type control mice (Fig. 2C).Consistently, histological analysis revealed that the liver tissues generated from RAGE-deficient mice displayed less inflammation and necrosis as compared with those of WT control mice (Fig. 2D).

Defect in RAGE signaling inhibits the infiltration of neutrophil and the production of pro-inflammatory cytokines
Previously, we demonstrated the implication of HMGB1 in ConA-induced acute hepatitis [5].We now sought to address the role of RAGE, one of the receptors for HMGB1 in the inflammatory response during ConAinduced hepatitis.As shown in Fig. 3, RAGE deficiency resulted in a significant decrease of inflammation-related adhesion molecule VCAM-1 (vascular cell adhesion molecule-1, VCAM-1) expression in liver tissues upon ConA challenge, while another adhesion molecule ICAM-1(intercellular adhesion molecule-1, ICAM-1) did not change between RAGE-deficient mice and WT controls after ConA treatment (Fig. 3A).Moreover, in contrast to controls, the infiltration of neutrophils in RAGE-deficient liver was dramatically attenuated upon administration of ConA (Fig. 3B).Quantitative RT-PCR analysis of IL-6 in the liver after Con A administration revealed a remarkable decrease in mRNA levels of IL-6 after RAGE deficiency, while mRNA levels of IFN-γ and TNF-α did not reach the significant change between RAGE-deficient mice and WT controls (Fig. 3C).Interestingly, serum pro-inflammatory cytokines IL-6 and TNF-α protein levels remarkably decreased in RAGE-deficient mice as compared with WT control animals (Fig. 3D).

RAGE deficiency changes the proportions of leukocytes in the liver challenged by ConA
Resident leukocytes in the liver play a critical role in autoimmune hepatitis, therefore, the effect of RAGE deficiency on the development of leukocytes in the liver was examined.It was found that RAGE deficiency led to hepatic T cells significantly lower than those of WT control mice, while the proportion of NKT cells was comparable between RAGE-deficient mice and WT control mice upon ConA injection (Fig. 4A).In contrast, NK cells markedly increased in RAGE-deficient mice as compared with WT control mice after the ConA challenge (Fig. 4A).Moreover, mice challenged with ConA, in the hepatic T cell subset, mice with RAGE deficiency exhibited significantly fewer CD4 + T cells than that of WT control mice.However, hepatic CD8 + T cells show no difference between RAGE defect mice and control mice (Fig. 4B).

T cells without RAGE affect T cell subsets activation and proliferation evoked by ConA challenge
To examine the effect of RAGE deficiency on T cell activation and proliferation stimulation by ConA, flow cytometry staining analysis (using CD69 as an activation marker) revealed that RAGE deficiency did not affect the activation of T cells in the liver upon ConA stimulation (Fig. 5A).Interestingly, as compared with WT control, defect in RAGE significantly inhibited the activation of spleen CD4 + T cells while did not affect the activation of spleen CD8 + T cells by Con A stimulation (Fig. 5B).CFSE staining analysis revealed that RAGE deficiency decreased proliferation of spleen T cells as compared with WT T cells (Fig. 5C).

The defect of RAGE inhibits the differentiation of Th17 cells in the liver
To further define the influence of RAGE deficiency on the CD4 + T cell subsets differentiation, the cytokines profiles produced by T cells were evaluated by flow cytometry using intracellular cytokines staining.It was found that  6A-D).However, RAGE deficiency significantly inhibited the proportion of Th17 cells with the capacity for IL-17 production (Fig. 6E, F).

Blockade of IL-6 alleviates hepatitis in RAGE competent mice challenged with ConA
To confirm that IL-6 was a critical pro-inflammatory cytokine impaired by RAGE deficiency, the anti-IL-6 antibody was used to block the effect of IL-6-caused hepatitis in vivo.Blockade of IL-6 markedly reduced serum levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) in WT control mice to the levels comparable with RAGE-deficient mice (Fig. 7A).Consistently, the pathological analysis revealed that blockade of IL-6 significantly ameliorated hepatitis induced by ConA injection in wild-type mice (Fig. 7B).Collectively, these results confirmed our hypothesis that IL-6 was pivotal in the effect of RAGE on Con A-induced hepatitis.

Decreased expression of ARID5A protein expression without RAGE results in the reduction of IL-6 production in the liver
Finally, to explore the molecular mechanism by which RAGE deficiency results in reduced IL-6 production in the liver, which alleviates hepatitis in mice induced by ConA challenge (Fig. 8A), macrophages and T cells in the liver were examined to produce IL-6 by intracellular cytokine staining.As shown in Fig. 8B and C, hepatic Kupffer cells and T cells (both CD4 + and CD8 + T cells) separated from mice with RAGE deficiency exhibited significantly decreased production of IL-6 as compared with those cells generated from wild-type control mice upon ConA challenge.Due to the AT-rich interactive domain-containing protein 5a (Arid5a) plays a critical role in regulating the half-life of Interleukin-6 (IL-6) mRNA [13].Western blotting analysis revealed that hepatic tissues with RAGE deficiency displayed a significantly decreased expression of Arid5a than that of control wild-type mice after the ConA challenge (Fig. 8D).Furthermore, inflammation-related signaling pathways molecules expression such as STAT3 with phosphorylation (p-STAT3) and JAK2 with phosphorylation (p-JAK2) in hepatic tissues come from mice with RAGE deficiency significantly lower as compared with of wild-type control mice (Fig. 8E).

Discussion
In the present study, we investigated the effects and mechanisms of RAGE deficiency on the pathogenesis of hepatitis in a mouse autoimmune hepatitis model.We demonstrated that deficiency of RAGE in the mice resulted in ameliorated hepatitis.This protection of ConA-induced hepatitis in RAGE-deficient mice was caused by decreased IL-6 secretion by RAGE-deficient hepatic Kupffer cells and T cells.
The biological function of HMGB1 is mediated by multiple receptors, including the receptor for advanced glycation end products (RAGE) and Toll-like receptors (TLRs), which are expressed in different hepatic cells [14].Previous studies have shown that innate immune cells such as macrophages, dendritic cells, and neutrophils express RAGE [1,15,16].In this study, we observed that mice injected with ConA induced the expression of RAGE in inflammatory liver tissues.Interestingly, ConA stimulation particularly promoted the expression of RAGE on CD4 + T cells rather than CD8 + T cells isolated from both the liver and spleen.
We previously have demonstrated that pro-inflammatory cytokines TNF-α and IFN-γ play important roles in detrimental inflammation in hepatitis caused by the ConA challenge [5,17].Now, we showed that RAGE deficiency impaired the expression of IL-6 by hepatic Kupffer cells and CD4 + T cells to attenuate hepatitis.To our knowledge, this is the first time to report that RAGE deficiency reduces the secretion of proinflammatory IL-6 by hepatic immune cells.Indeed, previous studies have suggested that IL-6 participates in the pathogenesis of inflammation and tumor in the liver, and hepatitis viral B has the capability of stimulating IL-6 secretion by hepatocytes [18,19].In addition, RAGE deficiency downregulates the expression of adhesion molecule VCAM-1, and decreases the infiltration of neutrophils to alleviate autoimmune hepatitis caused by the ConA challenge.
In addition to aberrant Th1/Th2 T cell responses in the liver involving hepatic inflammation, Th17 cells were a recently discovered subtype of CD4 + T-helper cells.TGF-β and IL-6, two cytokines abundantly present in the injured liver, promote Th17 cell differentiation to participate in hepatic inflammation [20].In this study, we observed that RAGE deficiency not only inhibits the production of IL-6 by hepatic Kupffer cells and CD4 + T cells, but the decreased IL-6 also impaired the generation of hepatic Th17 cells, which further ameliorated autoimmune hepatitis by ConA challenge.Importantly, in vivo blockade of IL-6 using an What is the mechanism by which RAGE deficiency inhibits the secretion of IL-6 by hepatic immune cells?To answer this question, we checked the expression of Arid5a as a unique RNA binding protein in liver tissues, which stabilizes IL-6 mRNA through binding to the 3' untranslated region of IL-6 mRNA [21].We found that RAGE deficiency impaired the hepatic expression of Arid5a and as a result, inhibited the elevation of IL-6 serum level in ConA-treated mice.In-line with our finding, it has been shown that Arid5a not only regulates the expression of IL-6 and generation of Th17 but is also involved in inflammatory diseases [22,23].In parallel, our investigation was supported by Weinhage et al.'s study [24].Moreover, in the present study, we provided a novel molecular explanation that pro-inflammatory cytokine IL-6 plays a critical role in the RAGE signaling axis in the pathogenesis of autoimmune hepatitis.
In conclusion, we have presented strong evidence to support that RAGE signaling plays a crucial role in regulating the secretion of IL-6 by hepatic immune cells in the context of ConA-induced autoimmune hepatitis.This is a new finding that demonstrates a regulatory role of RAGE in autoimmune hepatitis, providing a reference value for the basic research and clinical treatment of autoimmune hepatitis.

Fig. 2
Fig. 2 RAGE deficiency protects mice from ConA-induced hepatitis.A Survival was monitored for 36 h after ConA (20 μg/g body weight) injection.n = 10 per group.B Serum levels of ALT and AST in WT mice and RAGE-deficient mice after intravenous injection of ConA (15 μg/g body weight).n = 3 or 6 per group.C Liver and spleen of WT mice and RAGE −/− mice displayed different pathologi-

Fig. 3 Fig. 4
Fig. 3 RAGE deficiency attenuates the hepatic inflammation response.A Western blot of liver tissue of VCAM-1 and ICAM-1 in WT mice and RAGE −/− mice.Representative blots from three independent experiments are shown.B Immunohistochemistry and myeloperoxidase activity were used to detect neutrophil infiltration in liver tissue in WT mice and RAGE −/− mice after intravenous injection of ConA.n = 3 or 6 per group.C Hepatic mRNA levels of IL-6, IFN-γ and TNF-α in WT and RAGE −/− mice following ConA injection were measured by real-time quantitative PCR.n = 3 or 6 per group.D Serum levels of IL-6, IFN-γ and TNF-α in WT mice and RAGE −/− mice after intravenous injection of ConA were measured by ELISA.n = 3 or 6 per group.Mice treated with PBS were the control group.Data are presented as means ± SD of three to six individual mice per group.Similar results were obtained for three independent experiments.NS: no significant; *p < 0.05; **p < 0.01; ***p < 0.001 vs Control group mice ◂

Fig. 5 Fig. 6 Fig. 7
Fig. 5 deficiency affects T cell activation and proliferation following ConA injection.A RAGE deficiency does not affect T cell activation in the liver.The right bar chart shows the percentage of CD69 in liver CD4 + T cells and CD8 + T cells from the WT and RAGE −/− mice treated with PBS or ConA for 10 h.n = 3 or 6 per group.B RAGE deficiency affects CD4 + T cell activation in the spleen.The right bar chart shows the percentage of CD69 in spleen CD4 + T cells and CD8 + T cells from the WT and RAGE −/− mice

Fig. 8
Fig. 8 RAGE deficiency reduces the expression of ARID5A protein in the liver, which in turn reduces the expression of IL-6.Hepatic mononuclear from the WT and RAGE −/− mice were isolated after ConA injection A Intracellular staining of IL-6 in liver DC was performed.The right bar chart shows the percentage of IL-6 + cells in liver CD11C + cells from the WT and RAGE −/− mice treated with PBS or ConA for 10 h.B Intracellular staining of IL-6 in liver KCs was performed.The right bar chart shows the percentage of IL-6 + cells in liver F4/80 + cells from the WT and RAGE −/− mice treated with PBS or ConA for 10 h.n = 3-6 per group.C Intracellular staining of IL-6 in liver CD3 + T cells, CD4 + T cells and CD8 + T cells were performed.The bar charts below show the percentage of IL-6 + cells in liver T cells from the WT and RAGE −/− mice treated with PBS or ConA for 10 h.n = 3 or 5 per group.D Western blot of Arid5a in liver tissue of WT mice and RAGE −/− mice.Representative blots from three independent experiments are shown.E Western blots of the total proteins and phosphorylated proteins of JAK2 and STAT3 in liver tissue of WT mice and RAGE −/− mice.Representative blots from three independent experiments are shown.Data are presented as means ± SD of three to six individual mice per group.Similar results were obtained for three independent experiments.NS: no significant; *p < 0.05; **p < 0.01; *** p < 0.001; **** p < 0.0001 vs Control group mice ◂