Bacteria engineered to produce IL-22 in intestine induce expression of REG3G to reduce ethanol-induced liver disease in mice

Abstract

Objective: Antimicrobial C-type lectin regenerating islet-derived 3 gamma (REG3G) is suppressed in the small intestine during chronic ethanol feeding. Our aim was to determine the mechanism that underlies REG3G suppression during experimental alcoholic liver disease.

Design: Interleukin 22 (IL-22) regulates expression of REG3G. Therefore, we investigated the role of IL-22 in mice subjected to chronic-binge ethanol feeding (NIAAA model).

Results: In a mouse model of alcoholic liver disease, we found that type 3 innate lymphoid cells produce lower levels of IL-22. Reduced IL-22 production was the result of ethanol-induced dysbiosis and lower intestinal levels of indole-3-acetic acid (IAA), a microbiota-derived ligand of the aryl hydrocarbon receptor (AHR), which regulates expression of IL-22. Importantly, faecal levels of IAA were also found to be lower in patients with alcoholic hepatitis compared with healthy controls. Supplementation to restore intestinal levels of IAA protected mice from ethanol-induced steatohepatitis by inducing intestinal expression of IL-22 and REG3G, which prevented translocation of bacteria to liver. We engineered Lactobacillus reuteri to produce IL-22 (L. reuteri/IL-22) and fed them to mice along with the ethanol diet; these mice had reduced liver damage, inflammation and bacterial translocation to the liver compared with mice fed an isogenic control strain and upregulated expression of REG3G in intestine. However, L. reuteri/IL-22 did not reduce ethanol-induced liver disease in Reg3g^-/- mice.

Conclusion: Ethanol-associated dysbiosis reduces levels of IAA and activation of the AHR to decrease expression of IL-22 in the intestine, leading to reduced expression of REG3G; this results in bacterial translocation to the liver and steatohepatitis. Bacteria engineered to produce IL-22 induce expression of REG3G to reduce ethanol-induced steatohepatitis.

Keywords: ILC; immune response; metabolome; microbiome.

Figure 1. Effects of the chronic–binge ethanol diet.

Plasma levels of ALT (a) and hepatic levels of triglycerides levels (b) in C57BL/6 mice following a control (isocaloric) or chronic–binge ethanol diet. Representative Oil red O staining images of liver tissue (c). Expression of Reg3g and Reg3b mRNA in jejunum epithelial cells (d, e). Expression of Il22 mRNA in jejunum lamina propria cells from C57BL/6 mice following a control or chronic–binge ethanol diet (f). Frequencies of IL22-expressing cells in the gated ILC3 population (CD3-RORγt⁺) after stimulation with IL23 for 4h in jejunum lamina propria cells from C57BL/6 mice after a control or chronic–binge ethanol diet (g). Gene expression levels are shown relative to those of control mice. Data presented are the mean ± SEM of n = 13–15 mice/group. Scale bars = 50 µm. *, *** indicate p<0.05, p<0.001, respectively.

Figure 2. Effects of the ethanol diet on tryptophan metabolism.

Expression of Il22 mRNA in jejunum lamina propria cells after stimulation with IL23 for 4h, co-incubated with an AHR inhibitor (10 uM CH-223191) or vehicle solution (a). Intestinal levels of indole-3-acetic acid (b), indole-3-sulfate (c), tryptamine (d), indole-3-aldehyde (e), indole-3-acetamide (f), and indole-3-lactic acid (g) of mice fed control (isocaloric) or chronic–binge ethanol diets. Levels of indole-3-acetic acid (h) and indole-3-lactic acid (i) in stool samples of alcoholic hepatitis patients (n=13) and healthy controls (n=17). Gene expression data are set relative to controls (a). Data in all panels are the mean ± SEM of n = 3 biological replicates (a), or n = 10 mice/group (b-f). *, **,**** indicates p<0.05, p<0.01, p<0.0001 respectively.

Figure 3. Effects of IAA supplementation.

Plasma levels of ALT (a) and hepatic triglycerides (b) in C57BL/6 mice on a control or chronic–binge ethanol diet, with IAA supplementation (20 mM/day) or vehicle. Representative Oil red O staining images of liver tissue (c). Expression of Il22 mRNA (d) and Reg3g mRNA (e) in ileum of C57BL/6 mice following the chronic–binge alcohol diet, with IAA supplementation or vehicle. Total bacteria in the liver, measured by qPCR for 16S rRNA, normalized to 18S rRNA (f). Gene expression levels are set relative to those of mice given vehicle (d, e). Fold change in hepatic bacteria are compared to mice on the control diet or isocaloric mice (f). Data in all panels are presented as mean ± SEM and are of n = 7 (vehicle/isocaloric), n = 12 (vehicle/ethanol), or n = 14 (IAA/ethanol) mice/group. Scale bars = 50 µm. *, **, *** indicates p<0.05, p<0.01, p<0.001, respectively.

Figure 4. Effects of antibiotics.

Plasma levels of ALT (a) and hepatic triglycerides (b) in C57BL/6 mice on a control (isocaloric) or chronic–binge ethanol diet, with or without antibiotics. Representative Oil red O staining images of liver tissue (c). Levels of Il22 mRNA in jejunum lamina propria cells from C57BL/6 mice after a control or chronic–binge ethanol diet, with or without antibiotics (d). Frequencies of IL22+ ILC3 (CD3–RORγt⁺) in jejunum lamina propria of C57BL/6 mice after a control or chronic–binge ethanol diet, with or without antibiotics (e). Gene expression levels are set relative to level of isocaloric-fed control mice. Data presented are the mean ± SEM of n = 10–12 mice/group (a, b) or representative of at least 3 independent stimulations of a minimum of 2 replicates (d, e). Scale bars = 50 µm. *, **,*** indicates p<0.05, p<0.01, and p<0.001, respectively.

Figure 5. L. reuteri engineered to produce IL22 reduce hepatic steatohepatitis and bacterial translocation to the liver.

Plasma levels of IL22 in C57BL/6 mice gavaged with PBS, L. reuteri, or L. reuteri/IL22 during the control (isocaloric) and chronic–binge ethanol diet (a). Mean plasma level of ALT in these groups of mice (b). Quantification and representative Oil red O staining images of liver tissues from ethanol-fed mice gavaged with PBS, L. reuteri, or L. reuteri/IL22 (c, d). Hepatic levels of Cxcl1 mRNA (e) and Cxcl2 mRNA (f) in mice from these groups. Expression of Reg3g mRNA in ileum from mice on the ethanol-diet given PBS, L. reuteri, or L. reuteri/IL22 (g). Total bacteria in livers of ethanol-fed mice, measured by qPCR for 16S rRNA, normalized to 18S rRNA (h). Levels are relative to mice given PBS on the control diet (e, f) or mice given PBS while on the ethanol diet (g, h). Data presented as mean ± SEM are of n = 6 (isocaloric groups) and n = 14–16 (ethanol diet) mice/group. Scale bars = 50 µm. * indicates p<0.05.

Figure 6. Effects of L. reuteri/IL22 in Reg3g^–/– mice.

Plasma level of ALT in Reg3g^–/– and wild type (Wt) littermates gavaged with PBS, L. reuteri, or L. reuteri/IL22 during chronic–binge ethanol diet (a). Hepatic triglycerides and representative Oil red O staining images of liver tissues (b, c). Hepatic expression of Cxcl1 mRNA (d) in the groups. Expression of Reg3g mRNA in ileum from Wt mice on chronic-binge ethanol diet (e). Total bacteria in the liver, measured by qPCR for 16S rRNA and normalized to 18S rRNA (f). Gene expression data are relative to Wt mice given PBS and are presented as mean ± SEM. Experiments performed in n = 51 Wt mice (PBS: 15; L. reuteri: 19; L. reuteri/IL22: 17) and n = 46 Reg3g^–/– mice (PBS: 14; L. reuteri: 16; L. reuteri/IL22: 16). Scale bars = 50 µm. *, **,*** indicates p<0.05, p<0.01, and p<0.001, respectively.

Figure 7. Model for the effects of alcohol on IL22 and REG3G expression.

In healthy conditions, bacterial catabolism of tryptophan into indoles such as IAA promotes AHR-dependent production of IL22 by ILC3s, which maintains expression of REG3G. Chronic alcohol consumption leads to intestinal dysbiosis, which is associated with reduced levels of IAA in the gut. Decreased IAA availability reduces AHR-mediated production of IL22 production by ILC3. This reduces expression of REG3G, allowing bacteria to translocate to the liver, where they promote progression of alcoholic liver disease. (Figure made using Servier Medical Art; http://smart.servier.com)