Interleukin-22 but not interleukin-17 provides protection to hepatocytes during acute liver inflammation

Abstract

The cytokine interleukin-22 (IL-22) is primarily expressed by T helper 17 (Th17) CD4(+) T cells and is highly upregulated during chronic inflammatory diseases. IL-22 receptor expression is absent on immune cells, but is instead restricted to the tissues, providing signaling directionality from the immune system to the tissues. However, the role of IL-22 in inflammatory responses has been confounded by data suggesting both pro- and anti-inflammatory functions. Herein, we provide evidence that during inflammation, IL-22 played a protective role in preventing tissue injury. Hepatocytes from mice deficient in IL-22 were highly sensitive to the detrimental immune response associated with hepatitis. Additionally, IL-22-expressing Th17 cells provided protection during hepatitis in IL-22-deficient mice. On the other hand, interleukin-17 (IL-17), which is coexpressed with IL-22 and can induce similar cellular responses, had no observable role in liver inflammation. Our data suggest that IL-22 serves as a protective molecule to counteract the destructive nature of the immune response to limit tissue damage.

Figure 1. IL-22 deficient mice have no defects in innate or adaptive immunity to Listeria monocytogenes infection

IL-22 wild-type (+/+) or deficient (−/−) mice were injected with 1×10⁴ CFU rLM-ova. Bacterial loads in the (A) spleens and (B) livers of infected mice 3 days post-infection. Bars represent mean±SD of four mice per group. Seven days post-infection, (C) surface expression of CD44 and CD62L was quantitated by FACS on CD4+ T cells from uninfected mice and rLM-ova-infected mice. Shown is a FACS plot from one representative mouse, number represents the percentage of CD4 T cells that are CD44^high CD62L^low or CD44^low CD62L^high. (D) To analyze the antigen-specific response, T cells from the spleen or liver were stimulated with LLO_190–201 peptide or without peptide in the presence of GogiStop. IFNγ production was measured by intracellular cytokine staining. Percent shown is the percent of cells that are CD4+ IFNγ+. Plots are representative of four mice/group. Results are representative of two independent experiments.

Figure 2. Hepatocytes are highly responsive to IL-22

(A) IL-22R, IL-10Rα, IL-10Rβ expression in different cell subsets by real-time RT-PCR. Bars represent the gene/HPRT expression as determined by the relative quantification method (ΔΔC_T). ND=not detected. (B) FL38B cells or (C) bone-marrow derived macrophages were left untreated (Ø) or stimulated for 20 min with 50 ng/ml recombinant murine IL-10 or IL-22. Phosphorylation or total levels of Stat3 and β-actin in the cell lysates were determined by western blotting with the respective antibodies. (D) FL38B cells were either serum starved for 5 hrs (−FBS) or remained in media with serum (+FBS) and then stimulated with 50 ng/ml recombinant IL-22 for 0, 10 or 30 min. Phosphorylation or total levels of Akt and β-actin in the cell lysates was determined by western blotting with the respective antibodies. Experiment was performed four times with similar results.

Figure 3. IL-22 mRNA is upregulated in the liver during conA-mediated hepatitis

C57BL/6 mice were intravenously injected with the indicated dose of conA and 2 hrs later mRNA was harvested from the livers. mRNA was reverse transcribed, and IL-22 and IFNγ cDNA were quantitated by real-time PCR. Bars represent the mean±SD expression of the cytokine/HPRT using the ΔΔC_T method. Experiment was performed two times with similar results. ND=not detected.

Figure 4. IL-22 deficient mice are highly susceptible to conA-induced acute liver inflammation

IL-22 wild-type (+/+), heterozygous (+/−), and deficient mice (−/−) were intravenously injected with 10 μg/g conA or injected with PBS. (A) 18 hrs or 5 days post-injection mice were euthanized, and livers were harvested, fixed, sectioned and stained with H&E. Shown are representative sections from one of three mice per group. (B) Higher magnification of a lesion from an IL-22−/− mouse, 18 hrs post-conA. (C and D) ALT levels in the sera of mice injected 18 hrs previously with (C) PBS or (D) conA. Bars represent mean±SD of three mice/group. Experiment was performed three times with similar results.

Figure 5. IL-22 deficient and wild-type mice have equal immune responses during conA-mediated hepatitis

IL-22 wild-type (+/+) and deficient (−/−) mice were intravenously injected with 10 μg/ml conA or left untreated and 6 hrs later the resulting immune response was analyzed. (A) Cytokine mRNA levels in the liver were semi-quantitated by real-time RT-PCR using the ΔΔC_T method. Bars represent mean±SD of four mice. (B) IFNγ and TNFα levels in the sera of mice were determined by ELISA. Bars represent mean±SD of six-eight mice per group. (C) Activation of CD4+ T cells from the livers and spleens was analyzed by surface expression of CD69 by FACS. Number represents the percentage of CD4 cells that are CD69^high. Shown are representative FACS plots of one mouse of three mice/group.

Figure 6. Hepatocyte damage in IL-22 deficient mice is dependent on inflammation and not inherent susceptibility of hepatocytes to apoptosis

IL-22 deficient (IL-22 −/−) or control wild-type mice (IL-22 +/+) were injected intraperitoneally with 1 μg/g mouse of anti-Fas Ab (clone Jo-2) or PBS and were euthanized 6 hrs later. (A) Representative liver sections stained with H&E. (B) ALT levels in the sera of Jo2 injected mice, mean±SD of four mice/group. Fold-induction of (C) IFNγ and (D) TNFα mRNA levels in the livers of Jo2 injected (Jo2) or conA injected mice (ConA) compared to untreated mice (Ø) as quantitated by real-time RT-PCR using the ΔΔC_T method. Experiment was performed three times with similar results.

Figure 7. IL-6 is not required for IL-22 induction during conA-mediated hepatitis

IL-6 wild-type (IL-6 +/+) or deficient mice (IL-6 −/−) were intravenously injected with 10 μg/g conA. 0, 2 or 4 hrs post-injection, liver lymphocytes were purified, RNA was harvested, reverse transcribed and IL-6 and IL-22 cDNA were quantitated by real-time PCR. Bars indicate mean±SD of the cytokine/HPRT ratio using the ΔΔC_T method in 3 mice per group. ND=not detected. Experiment was performed twice with similar results.

Figure 8. IL-22 and IL-17 have non-homologous roles in conA-mediated hepatitis

(A) C57BL/6 mice were injected with 10 μg/ml conA or PBS alone. 2 hrs later IL-17 mRNA fold-induction in the spleen compared to PBS mice was quantitated by real-time RT-PCR using the ΔΔC_T method. Bars indicate mean±SD. (B) IL-17A protein levels in the sera of C57BL/6 mice 6 hrs after conA injection. Bar indicates mean±SD. ND=not detected. (C and D) Wild-type (IL-22 +/+ IL-17 +/+), IL-22 deficient (IL-22 −/− IL-17 +/+), IL-17 deficient (IL-22 +/+ IL-17 −/−) or IL-22 IL-17 double deficient (IL-22 −/− IL-17 −/−) mice were intranvenously injected with 10 μg/g conA and 18 hrs later liver injury was examined. (B) H&E liver sections from the indicated mice. Shown are representative sections from one mouse of three per group. (C) ALT levels (mean±SD) in the sera of conA injected mice. (E) FL38B cells were stimulated for 20 min with 50 ng/ml of recombinant murine IL-22 or IL-17 or left unstimulated (Ø). Phosphorylation of Stat3 and Akt in the cell lysates was examined by western blotting with phospho-specific Abs as described in the Experimental Procedures.

Figure 9. Th17 differentiated IL-22 expressing cells protect hepatocytes during hepatitis

IL-22 wild-type (+/+) or deficient (−/−) naïve (CD44^low, CD62L^low, CD25-) CD4 T cells were differentiated in vitro with plate-bound α-CD3/α-CD28 Ab for 5 days under Th0 or Th17 conditions. (A) Cells were stimulated with PMA and ionomycin or remained unstimulated. 5 hr post-stimulation, mRNA was harvested, reverse transcribed, and cDNA was quantitated by real-time RT-PCR. Fold-induction over unstimulated cells was calculated by the ΔΔC_T method. ND=not detected. (B) 5×10⁶ IL-22 +/+ or IL-22−/− Th17 cells were injected into IL-22−/− mice, and then 6-12 hrs later, mice were injected with 10 μg/g conA. 6 and 18 hrs post-conA injection, ALT and AST levels in the sera were quantitated, mean±SD for 6 mice/group. A similar experiment yielded similar results.