Exercise-Induced Irisin Decreases Inflammation and Improves NAFLD by Competitive Binding with MD2

Abstract

Non-alcoholic fatty liver disease (NAFLD) is a global clinical problem. The MD2-TLR4 pathway exacerbates NAFLD progression by promoting inflammation. Long-term exercise is considered to improve NAFLD but the underlying mechanism is still unclear. In this study, we examined the protective effect and molecular mechanism of exercise on high-fat diet (HFD)-induced liver injury. In an HFD-induced NAFLD mouse model, exercise training significantly decreased hepatic steatosis and fibrosis. Interestingly, exercise training blocked the binding of MD2-TLR4 and decreased the downstream inflammatory response. Irisin is a myokine that is highly expressed in response to exercise and exerts anti-inflammatory effects. We found that circulating irisin levels and muscle irisin expression were significantly increased in exercised mice, suggesting that irisin could mediate the effect of exercise on NAFLD. In vitro studies showed that irisin improved lipid metabolism, fibrosis, and inflammation in palmitic acid (PA)-stimulated AML12 cells. Moreover, binding assay results showed that irisin disturbed MD2-TLR4 complex formation by directly binding with MD2 but not TLR4, and interfered with the recognition of stimuli such as PA and lipopolysaccharide with MD2. Our study provides novel evidence that exercise-induced irisin inhibits inflammation via competitive binding with MD2 to improve NAFLD. Thus, irisin could be considered a potential therapy for NAFLD.

Keywords: MD2; NAFLD; exercise; inflammasome; irisin; myokine.

Figure 1

Liver gross appearance, body weight, liver weight, and liver enzymes in mice. (A) The NAFLD experimental C57BL/6 mouse model was generated according to the schematic timeline. (B) After sacrifice, the gross liver appearance was imaged. (C,D) The body weight (C), as well as the liver weight (D) was recorded. (E,F) Liver damage was assessed by measuring the hepatic levels of AST (E) and ALT (F). The data are presented as the mean ± SEM, n = 6 per group. # p < 0.05 vs. NCD group; * p < 0.05 vs. HFD group.

Figure 2

Exercise decreases steatosis and fibrosis in mouse livers. (A) Representative images of liver sections stained by H&E (200×, left panel) and Sirius Red (100×, right panel). (B) Histological scores of steatosis, lobular inflammation, and hepatocyte ballooning in H&E-stained livers. (C) NAFLD activity score was calculated by the histological score of steatosis, lobular inflammation, and hepatocyte ballooning. (D) The collagen areas seen by Sirius Red staining. (E,F) Protein levels of lipid metabolism markers FABP4 and PPARα in mouse liver tissues. Tubulin was used as the loading control. (G,H) Protein levels of fibrosis markers TGF-β1 and COL1 in mouse liver tissues. Tubulin was used as the loading control. (I) Relative mRNA levels of Fabp4, Ppara, Tgfb1, and Col1 in mouse liver tissues. The data are presented as the mean ± SEM, n = 6 per group. # p < 0.05 vs. NCD group; * p < 0.05 vs. HFD group.

Figure 3

Exercise decreases inflammation in mouse livers. (A) Representative F4/80 immunohistochemical staining of liver sections (200×). (B) Quantification of F4/80-positive area in immunohistochemical staining. (C) Cytokine levels of inflammation marker IL-1β in mouse liver tissues. (D,E) Protein levels of inflammation marker IL-6 in mouse liver tissues. Tubulin was used as the loading control. The data are presented as the mean ± SEM, n = 6 per group. # p < 0.05 vs. NCD group; * p < 0.05 vs. HFD group.

Figure 4

Exercise blocks MD2-TLR4 pathway activation in mouse livers. (A) MD2-TLR4 complex formation levels in mouse liver tissues detected by co-immunoprecipitation. (B) Protein levels of MAPK pathway and NF-κB pathway components, including p-ERK, p-JNK, p-p38, p-p65, and IκB-α. The corresponding unphosphorylated proteins and tubulin were used as the loading controls. (C) Relative mRNA levels of several pro-inflammatory markers Il6, Il1b, Tnf, Ccl2, Icam1, and Vcam1 in mouse liver tissues. The data are presented as the mean ± SEM, n = 6 per group. # p < 0.05 vs. NCD group; * p < 0.05 vs. HFD group.

Figure 5

Irisin changes in several tissues and circulating levels. Absolute irisin content in several skeletal muscles including the soleus (A), gastrocnemius (B), and quadriceps (C), circulating levels of irisin (D), and absolute hepatic irisin content (E) in mice were measured by ELISA. The data are presented as the mean ± SEM, n = 6 per group. # p < 0.05 vs. NCD group; * p < 0.05 vs. HFD group.

Figure 6

Irisin decreases steatosis and fibrosis in AML12 cells. (A–C) AML12 cells were pretreated with recombinant irisin (50 or 100 ng/mL) for 30 min followed by exposure to 200 μM PA for 36 h. (A) Representative images of AML12 stained by Oil Red O (400×). (B,C) Protein levels of FABP4 and PPARα in AML12 cells. (D,E) Protein levels of TGF-β1 and COL1 in AML12 cells. Tubulin was used as the loading control. (F) AML12 cells were pretreated with recombinant irisin (50 or 100 ng/mL) for 30 min followed by exposure to 200 μM PA for 12 h. Relative mRNA levels of Fabp4, Ppara, Tgfb1, and Col1 were detected. The data are presented as the mean ± SEM. # p < 0.05 vs. CON group; * p < 0.05 vs. PA group.

Figure 7

Irisin blocks NF-κB and MAPK pathways, and reduces inflammatory factors in AML12 cells. (A,B) AML12 cells were pretreated with recombinant irisin (50 or 100 ng/mL) for 30 min followed by exposure to 200 μM PA for 2 h. (A) MD2-TLR4 complex formation levels in AML12 cells detected by immunoprecipitation. (B) Protein levels of MAPK pathway and NF-κB pathway components, including p-ERK, p-JNK, p-p38, p-p65, and IκB-α. The corresponding unphosphorylated proteins and tubulin were used as loading controls. (C) AML12 cells were pretreated with recombinant irisin (50 or 100 ng/mL) for 30 min followed by exposure to 200 μM PA for 12 h. Relative mRNA levels of Il6, Il1b, Tnf, Ccl2, Icam1, and Vcam1 were detected. The data are presented as the mean ± SEM. # p < 0.05 vs. CON group; * p < 0.05 vs. PA group.

Figure 8

Irisin competitively binds to MD2 but not TLR4. (A,B) Immunoprecipitation analysis of the binding ability of recombinant irisin to MD2 (A) or TLR4 (B) in liver lysates. (C) ELISA analysis in the binding ability of recombinant irisin to MD2 or TLR4 in liver lysates. (D) Immunoprecipitation analysis in the binding ability of recombinant irisin to rhMD2. (E) ELISA analysis of the binding ability of recombinant irisin to rhMD2. (F) Surface plasmon resonance analysis between irisin with rhMD2. (G) ELISA analysis of the effect of recombinant irisin (0.1, 0.2, and 0.5 μg/mL) on the basal binding level of MD2-TLR4. (H,I) ELISA analysis of the competitive MD2 binding ability of recombinant irisin (0.1, 0.2, and 0.5 μg/mL) to PA or LPS. (J) Molecular docking of the dimeric irisin-MD2 complex. (K) ELISA analysis of irisin-MD2 binding levels in mouse liver tissue (n = 6 per group). The data are presented as the mean ± SEM. # p < 0.05 vs. CON or NCD group; * p < 0.05 vs. rhMD2 or HFD group.