Interleukin-22 ameliorates acute-on-chronic liver failure by reprogramming impaired regeneration pathways in mice

Abstract

Background & aims: Acute-on-chronic liver failure (ACLF) is a clinical syndrome defined by liver failure on pre-existing chronic liver disease. It is often associated with bacterial infection and high short-term mortality. Experimental models that fully reproduce ACLF are lacking, so too are effective pharmacological therapies for this condition.

Methods: To mimic ACLF conditions, we developed a severe liver injury model by combining chronic injury (chronic carbon tetrachloride [CCl₄] injection), acute hepatic insult (injection of a double dose of CCl₄), and bacterial infection (intraperitoneal injection of bacteria). Serum and liver samples from patients with ACLF or acute drug-induced liver injury (DILI) were used. Liver injury and regeneration were assessed to ascertain the potential benefits of interleukin-22 (IL-22Fc) administration.

Results: This severe liver injury model recapitulated some of the key features of clinical ACLF, including acute-on-chronic liver injury, bacterial infection, multi-organ injury, and high mortality. Liver regeneration in this model was severely impaired because of a shift from the activation of the pro-regenerative IL-6/STAT3 pathway to the anti-regenerative IFN-γ/STAT1 pathway. The impaired IL-6/STAT3 activation was due to the inability of Kupffer cells to produce IL-6; whereas the enhanced STAT1 activation was due to a strong innate immune response and subsequent production of IFN-γ. Compared to patients with DILI, patients with ACLF had higher levels of IFN-γ but lower liver regeneration. IL-22Fc treatment improved survival in ACLF mice by reversing the STAT1/STAT3 pathway imbalance and enhancing expression of many antibacterial genes in a manner involving the anti-apoptotic protein BCL2.

Conclusions: Acute-on-chronic liver injury or bacterial infection is associated with impaired liver regeneration due to a shift from a pro-regenerative to an anti-regenerative pathway. IL-22Fc therapy reverses this shift and attenuates bacterial infection, thus IL-22Fc may have therapeutic potential for ACLF treatment.

Lay summary: A mouse model combining chronic liver injury, acute hepatic insult, and bacterial infection recapitulates some of the key features of acute-on-chronic liver failure (ACLF) in patients. Both fibrosis and bacterial infection contribute to the impaired regenerative capacity of the liver in patients with ACLF. Herein, we show that IL-22Fc therapy improves ACLF by reprogramming impaired regenerative pathways and attenuating bacterial infection. Thus, it may have therapeutic potential for patients with ACLF.

Keywords: ACLF; Bacteria; IFN-γ; IL-6; STAT1; STAT3.

Figure 1.. Characterization of a mouse model of ACLF.

(A) Schematic timeline of the procedures. C57BL/6J mice were treated with 8-week CCl₄ to induce chronic liver injury, followed by additional challenges: Group 1, no additional challenges; Group 2, acute insult (a double dose of CCl₄ injection); Group 3, Klebsiella pneumoniae (K.P.) infection; Group 4, acute insult and K.P. infection. (B) Survival rates (n=8). (C) Serum ALT levels. (D) Total bilirubin (TBIL), white blood cells, neutrophil percentage, blood bacterial loads, and BUN were measured 72 hours post-acute CCl₄ injection. (E) Representative staining images from the group 4. (F) Hepatic and renal blood flow were determined 72 hours post-acute insult. One-way ANOVA was used for statistical evaluation (*P<0.05, **P<0.01, ***P<0.001, n=4-8 in panels C-E).

Figure 2.. Liver regeneration is inhibited in ACLF mice.

(A) Schematic timeline of CCl₄ and K.P. injection in C57BL/6J mice (the group 4 in Figure 1). (B) Serum ALT levels 24 hours post-acute insult (n=5). (C) BrdU was injected 2 hours before sacrifice. Representative BrdU immunostaining images are shown. The percentage of BrdU⁺ hepatocytes was calculated. One-way ANOVA was used for statistical evaluation (*P<0.05, **P<0.01, ***P<0.001, n=5-8 in each group).

Figure 3.. The IL-6/STAT3 pathway is attenuated while the IFN-γ/STAT1 pathway is enhanced in fibrotic livers post-acute injury.

(A) Schematic timeline of the fibrotic and control mice with a double dose of CCl₄ (0.4 ml/kg) (without K.P infection). (B, C) Serum IL-6 and IFN-γ, and relative mRNA expressions of Il-6 and Ifng, and their downstream target genes in the liver were determined. (D) Western blot analyses of IL-6 and IFN-γ downstream signaling pathways. (E) Kupffer cells were isolated from control and fibrotic livers and stimulated with LPS. The IL-6 mRNA and protein (in the supernatant) were measured. Student’s t-test was used for statistical evaluation (*P<0.05, **P<0.01, ***P<0.001, n= 3-5 in panel B-E).

Figure 4.. The IFN-γ/STAT1 pathway is strongly activated in fibrotic livers post bacterial infection.

(A) Schematic timeline of CCl₄ and K.P. injection. (B, C) Serum IL-6 and IFN-γ, and relative mRNA expressions of Il-6 and Ifng, and their downstream target genes in the liver were determined. (D) Liver extracts were subjected to western blot analysis. Relative protein expression was quantitated. Student’s t-test was used for statistical evaluation (*P<0.05, **P<0.01, ***P<0.001, n=5 in each group).

Figure 5.. Serum IL-6 and IFN-γ levels, and liver regeneration in DILI and ACLF patients.

(A) Serum IL-6 and IFN-γ in DILI and ACLF patients (INR<1.5 or ≥1.5). (B) Immunohistochemistry staining of cell proliferating biomarker PCNA or Ki67 in DILI, HBV-ACLF and alcohol-ACLF patients. The percentage of PCNA⁺ or Ki67⁺ hepatocytes was calculated by randomly counting 10 images (10x) per sample. Kruskal-Wallis test for non-parametric variables was used for panel A, and one-way ANOVA was used for pane B statistical evaluation (*P<0.05, **P<0.01, ***P<0.001).

Figure 6.. IL-22Fc therapy ameliorates ACLF in mice by reprograming the impaired regeneration pathways and inducing anti-bacterial proteins.

(A) Schematic experiment timeline. Survival rates were determined and Mantel-Cox log-rank test was used for statistical analysis (**P<0.01). (B) Serum ALT and comparison of the ALT levels at 72 hours post-acute insult are shown. Blood bacterial loads are shown in the right. (C) Representative BrdU immunostaining images are shown, the percentage of BrdU⁺ hepatocytes was calculated. (D) Western blot analysis. (E) Hepatic and renal microcirculation were determined 72 hours post-acute insult with representative images shown on top and quantification graph below. (F) Mouse AML12 hepatocytes were stimulated with IFN-γ, IL-22Fc, or both, followed by western blot analysis. (G) Heat map of the relative hepatic expression of anti-bacterial genes in IL-22Fc and IgG2-treated mice. Student’s t-test was used for statistical evaluation (*P<0.05, **P<0.01, n=8 in each group in panels B-C, n=5 in each group in panels D-E).

Figure 7.. Bcl2^Hep-TG mice have greater survival rate and liver regeneration compared to wild-type (WT) following acute liver injury and bacterial infection.

(A) Schematic diagram of CCl₄ and K.P. injection procedures. Bcl2^Hep-TG and WT (Bcl2^Flox-STOP) mice were injected with a single dose of CCl₄ and 24 hours later injection of K.P. (B) Serum ALT. (C) Survival rate was determined and Mantel-Cox log-rank test was used for statistical analysis (*P<0.05). (D) Representative images and quantification of BrdU staining in mice post the acute insult. (E) Bacterial loads in the blood were determined post the acute insult. (F) Heat map of the relative hepatic expression of anti-bacterial genes following acute insult and bacterial infection. Student’s t-test was used for statistical evaluation (*P<0.05, **P<0.01, n=8 in each group in panels B, D, E).