Interleukin‐18 Binding Protein (IL‐18BP) Deficiency Affects Lymphocyte Activation and IL‐18 Expression in a Mouse Model of Liver Inflammation

The liver is a vital detoxification organ, continuously exposed to injury from xenobiotics, infections, and metabolic disturbances. It plays an important immunological role, with NK and NKT cells as major actors, and serves as the primary source of acute-phase proteins during inflammation [1]. Understanding the immune processes in acute hepatitis is essential for better characterizing the disease and helping recovery. A key player in inflammation regulation is interleukin (IL)-18. Produced mainly by innate immune cells, IL-18 is synthesized as proIL-18 in the cytosol, cleaved by caspase-1 upon inflammasome activation, and released. IL-18 activates NK and NKT cells and induces interferon-gamma (IFN-γ) production by T helper 1 cells [2]. Elevated IL-18 levels are linked to inflammatory disease [3] as well as liver pathologies, including acute hepatitis C virus (HCV) infection, metabolic-associated liver disease, and poor prognosis in hepatocarcinoma [4]. Mutations in the IL-18 binding protein (IL-18BP), a soluble receptor inhibiting the activity of IL-18, are associated in humans with fulminant hepatitis [5]. Studies in IL-18BP-deficient mice revealed that IL-18BP contributes to maintaining steady-state levels of circulating IL-18 [6] and that IL-18-induced IFN-γ upregulates IL-18BP, forming a negative feedback loop to resolve inflammation [3]. In this study, we investigated the IL-18/IL-18BP axis in acute mouse liver inflammation induced by concanavalin A (ConA) injection, a T cell- and macrophage-dependent liver injury model [7], or induced by coronavirus mouse hepatitis virus 3 (MHV-3), which mimics fulminant viral hepatitis [7]. Using these models, we compared wild-type (WT) and Il18bp-deficient (KO) mice. Both genotypes showed elevated mean ALT levels upon stimulation and weight loss starting at 60 h postinfection (hpi) (Figure S1). However, some MHV-3 injected mice exhibited ALT levels within the uninfected range at 72 hpi (gray dots in charts). Given the low dose of the virus used to avoid death before day 7, these animals might have cleared the virus more efficiently at the early stages. Focusing on immune-cell infiltration in the liver, we observed no significant differences in cell counts or population proportions between WT and KO mice (Figure 1A,B). We found that MHV-3-infected mice with normal ALT levels had a higher proportion of neutrophils (Figure S2A). Most strikingly, the percent of activated CD69+ lymphocytes was markedly reduced in the liver and spleen of MHV-3 infected KO mice (Figure 1C–F; Figure S2) compared with WT or ConA-treated mice. To investigate the cause of this lack of activation, we measured cytokine production by quantifying a panel of inflammatory cytokines using a bead-based immunoassay. We observe the distinct effect of MHV-3 and ConA challenges on cytokine levels in WT mice: the viral infection induced a high IL-18 level while ConA induced a greater rise in IL6, CXCL-1, and TNFα (Figure 1G,H; Figure S3) with a peak at 3–6 h. In Il18bp-deficient mice challenged with MHV-3 or ConA, circulating IL-18 levels remained at or below the limit of detection. These findings, also reported in other models, may result from a reduced half-life of IL-18 in the absence of IL-18BP, which would be both a chaperone and an inhibitor [6]. Additionally, a genetic regulatory mechanism could be at play. Indeed, Il18 mRNA levels were significantly higher in liver cells of WT compared with KO mice during MHV-3 infection (Figure 2A), but no differences between genotypes were found in the spleen or brain (not shown) nor after a chronic TLR9 stimulation [8]. During ConA-induced hepatitis, Il18 mRNA levels in the liver were reduced at 11 h but returned to baseline after 24 h (Figure 2B) in both genotypes. Since IL-18 injection worsens ConA-induced liver damage [9], a decrease in Il18 expression could be protective in this model by limiting local tissue damage. Our observations indicate that IL-18 regulation differs during hepatitis according to the type of challenge, which will need to be further explored. The role of IL-18BP in regulating Il18 mRNA expression warrants further investigation, especially in relation to the cellular sources of IL-18. In vitro experiments have shown that hepatic cell lines can constitutively produce Il18 mRNA ([5], [10]). To identify the in vivo sources of IL-18, we performed RNAscope in situ hybridization, using a duplex detection kit that enables sensitive detection of Il18 mRNA and cell-specific markers. Pecam-1 mRNA staining did not co-localize with Il-18 mRNA, ruling out liver sinusoidal endothelial cells as IL-18 producers (Figure S4). Il18 mRNAs were present in some hepatocytes (examples with yellow arrows) and also co-localized with Cd68 blue-stained macrophages (examples with red arrows) (Figure 2C–H). This study reveals that hepatocytes, alongside liver macrophages, are key producers of IL-18 during acute hepatitis, especially in response to viral infections like MHV-3. Hepatocytes may play a crucial compensatory role when macrophages are depleted by the virus. In MHV-3 infection and the absence of IL-18BP, IL-18 production by hepatocytes is impaired and lymphocyte activation is parallelly deficient. The consequences would be to prevent excessive inflammation in the absence of IL-18BP, as we did not observe increased liver symptom severity in KO mice. IL-18 plays various roles in liver pathophysiology, including during viral infection, cancer progression, fibrosis, and nonalcoholic steatohepatitis caused by nutritional stress. This might be facilitated by the production of IL-18 directly by hepatocytes, at the front line against external aggression. This work also suggests a regulatory loop between IL-18 and IL-18BP in acute hepatitis induced by viral infection. We thank Amélie Martin, Alanah Brown, and Avril McHugh for their technical assistance. We thank platforms: “Laboratoire de biochimie, CHU Rennes” and SFR Biosit—UMS 3480, US_S 018, France (H2P2, CYTOMETRI, and ARCHE). Funding was provided by INSERM, MESRI, University of Rennes, Région Bretagne, ANRS, and FRM. The authors declare no conflicts of interest. The peer review history for this article is available at https://publons.com/publon/10.1002/eji.202451579 Datasets are available from the corresponding author upon request. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.