Introduction
Metabolic dysfunction-associated steatotic liver disease (MASLD), formerly designated as non-alcoholic fatty liver disease (NAFLD), has rapidly emerged as the leading etiology of chronic liver disease worldwide, currently affecting approximately 30% of the global population
[1,2]. MASLD encompasses a progressive spectrum of hepatic pathologies, initiating from simple steatosis and potentially advancing to metabolic dysfunction-associated steatohepatitis (MASH), advanced fibrosis, cirrhosis, and ultimately hepatocellular carcinoma (HCC)
[3]. The pathogenesis of MASH is fundamentally driven by lipotoxicity, chronic metabolic stress, and, crucially, persistent low-grade tissue inflammation
[4]. Intrahepatic macrophages, comprising both resident Kupffer cells and recruited monocyte-derived macrophages, serve as pivotal orchestrators of this inflammatory microenvironment
[5]. Their dynamic polarization into pro-inflammatory (M1-like) or anti-inflammatory (M2-like) phenotypes dictates the balance between hepatic injury and tissue repair, thereby functioning as a master regulator of MASLD progression
[6].
Interleukin-15 (IL-15) is a pleiotropic cytokine traditionally known for its essential roles in regulating the development, survival, and activation of natural killer (NK) cells and memory CD8+ T cells
[7]. Unlike most soluble cytokines, IL-15 predominantly functions via a unique mechanism termed “trans-presentation” wherein it binds with exceptionally high affinity to IL-15 receptor alpha (IL-15R
α) on the surface of a presenting cell to activate the heterotrimeric IL-15 receptor complex on adjacent responder cells
[8,9]. Emerging evidence has implicated the IL-15/IL-15R
α axis in chronic inflammatory and metabolic disorders, including obesity and insulin resistance
[10,11]. However, the precise role of the IL-15/IL-15R
α pathway in MASLD remains highly controversial, with conflicting historical studies reporting both protective and detrimental effects on hepatic steatosis and inflammation. This ongoing ambiguity stems largely from the lack of resolution regarding the tissue-specific and source-specific contributions of IL-15 signaling.
The gut-liver axis represents a crucial bidirectional anatomical and functional communication channel, whereby intestinal barrier integrity and gut-derived factors profoundly influence hepatic homeostasis
[12,13]. Components of the IL-15/IL-15R
α axis are expressed in both the liver and intestinal epithelium
[14,15], with IL15 and IL15RA transcripts detected across multiple immune, stromal, endothelial, and epithelial cell populations
[16]. Intestinal epithelial IL-15 trans-presentation has an established local role in intraepithelial lymphocyte biology
[17]; however, how these anatomically distinct hepatic and intestinal compartments cooperate or diverge in regulating metabolic inflammation remains incompletely characterized. Through public transcriptomic reanalysis of human MASLD cohorts (GSE246221), we observed associations between hepatic IL15 expression and macrophage-related transcripts across disease stages, providing context for the relevance of this axis without establishing causality. Nevertheless, whether local hepatic IL-15 signaling and distant intestinal epithelial IL-15/IL-15R
α pathways exert overlapping or distinct functions during MASLD progression has yet to be determined.
To address this fundamental knowledge gap, we systematically investigated the compartmentalized roles of IL-15 and IL-15Rα using tissue-specific conditional knockout mouse models. We generated hepatocyte-specific IL-15-deficient (Il15fl/fl; Alb-cre) mice, alongside intestinal epithelial cell (IEC)-specific IL-15-deficient (Il15fl/fl; Vil-cre) and IL-15Rα-deficient (Il15rafl/fl; Vil-cre) mice, and subjected them to a high-fat diet (HFD)-induced MASLD regimen. Strikingly, our findings reveal a remarkable spatial compartmentalization and divergent functionality of the IL-15 axis across the gut-liver axis. Hepatic IL-15 deficiency attenuated steatosis and liver injury without a detectable change in fibrosis, whereas intestinal epithelial IL-15Rα deficiency disrupts circulating IL-15/IL-15Rα complex homeostasis; this was associated with increased liver injury and fibrosis. Collectively, these findings indicate compartment-specific roles of the IL-15/IL-15Rα axis in diet-induced MASLD and provide a basis for further mechanistic testing.
Materials and methods
Animal ethics statement
All mice and experimental protocol in the study were approved by the Animal Care and Use Committee of the College of Animal Science and Technology at China Agricultural University, China (Aw70504202-1-1), and conducted in accordance with the guidelines.
Animal experimental design
To elucidate the role of hepatic-specific IL-15 in metabolic dysfunction-associated steatotic liver disease (MASLD), Il15fl/fl mice were crossed with Albumin-cre mice to generate hepatocyte-specific IL-15 knockout (Il15fl/fl; Alb-cre) mice. Cre-negative Il15fl/fl littermates served as genotype-matched controls.
To investigate the distinct functions of intestinal epithelial cell (IEC)-derived IL-15 and its receptor IL-15Rα, two IEC-specific knockout models were generated. Specifically, Il15fl/fl mice and Il15rafl/fl mice were respectively crossed with Villin-cre mice to obtain IEC-specific IL-15 knockout (Il15fl/fl; Vil-cre) and IL-15Rα knockout (Il15rafl/fl; Vil-cre) mice. Comparisons involving Il15rafl/fl; Vil-cre mice are exploratory between-cohort comparisons.
To induce the experimental MASLD model, all mouse groups were fed a high-fat diet (HFD).
Transcriptomic analysis
In accordance with the 2023 multi-society Delphi consensus statement on the new nomenclature for steatotic liver disease, the terms NAFLD and NASH from the original GEO dataset were updated to MASLD and MASH, respectively, throughout this study. Given the extensively reported high concordance (> 98%) between NAFLD and MASLD criteria in patient cohorts, historical NAFLD classifications were directly mapped to the MASLD framework to align with current clinical standards.
To systematically investigate the molecular mechanisms underlying hepatic steatosis, we conducted a reanalysis of the publicly available transcriptomic dataset GSE246221. Raw expression data for selected target genes were extracted, and hierarchical clustering heatmaps were generated using the pheatmap package in R. Furthermore, immune cell infiltration profiling was performed utilizing the CIBERSORTx analytical tool (
https://cibersortx.stanford.edu/). GSE246221 contains separate Homo sapiens and Mus musculus gene-count matrices
[18]. The human groups were healthy liver (n = 4), MASL/F1 (n = 6), MASH/F1 (n = 5), MASH/F2 (n = 5), MASH/F3 (n = 6), and cirrhosis/F4 (n = 6). The mouse component used male C57BL/6J mice treated with streptozotocin (40 mg/kg intraperitoneally for five consecutive days from seven weeks of age) and a 60%-fat diet from eight weeks; disease-stage labels were assigned by the source study
[18]. For each species, selected gene counts were standardized within each gene (row-wise z score). Rows are genes, columns are individual samples, and colors are row-wise z scores; the heatmaps are descriptive. Because CIBERSORTx estimates relative cell-type signatures from bulk tissue, the inferred M1 and M2 macrophage abundances were interpreted descriptively rather than as direct measurements of cell number or function.
RNA extraction and real-time quantitative PCR
Total RNA extraction from liver tissues was performed using TRNzol Universal Reagent (DP424; TIANGEN Biotech Co., Ltd., Beijing, China), followed by quality assessment via OD260 / OD280 and OD260 / OD230 ratios. First-strand cDNA was transcribed using the EVO M-MLV RT Premix for qPCR (AG11706; Accurate Biotechnology (Hunan) Co., Ltd., Changsha, China) according to the manufacturer's guidelines. Subsequent RT-qPCR assays were conducted employing the SYBR Green Premix Pro Taq HS qPCR Kit (AG11701; Accurate Biotechnology) and the primers listed in Table S2. Gene expression levels were normalized to β-actin and calculated using the 2−ΔΔCt method. Technical replicates were averaged before analysis; the number of independent biological animals is reported in the figure legends, and group comparisons were performed using the tests described under Statistical analysis.
Lipid content determination
Serum lipid concentrations, including triglyceride (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C) were determined using commercial assay kits (TG: A110-1-1, TC: A111-1-1, HDL-C: A112-1-1, LDL-C: A113-1-1, Nanjing Jiancheng Bioengineering Institute, Nanjing, China) following the manufacturer's protocols.
Hepatic histological assessment
Hepatic tissues were fixed in 4% neutral buffered formalin for a minimum of 24 hours before processing. Fixed samples were subsequently embedded in paraffin and sectioned at a thickness of 5 μm.
For morphological evaluation, sections were stained with hematoxylin and eosin (H&E) by Wuhan Servicebio Technology Co., Ltd. (Wuhan, China). Histopathological changes, including hepatic steatosis scores and the NAFLD activity score (NAS), were evaluated under bright-field microscopy based on the H&E-stained sections. Steatosis was graded 0 (< 5%), 1 (5-33%), 2 (> 33-66%), or 3 (> 66%); lobular inflammation was graded 0 (none), 1 (< 2 foci/200 × field), 2 (2-4 foci/200 × field), or 3 (> 4 foci/200 × field); ballooning was graded 0 (none), 1 (few cells), or 2 (many/prominent cells); and NAS was their unweighted sum (0-8)
[19].
To visualize collagen deposition and assess hepatic fibrosis, parallel paraffin sections were stained with Picro-Sirius Red (PSR). All stained sections were examined under a bright-field microscope at 200× magnification. Quantitative analysis of liver fibrosis was performed by calculating the percentage of the PSR-positive area relative to the total tissue area using ImageJ software.
Statistical analysis
Data are expressed as mean ± standard error of the mean (SEM). Statistical analysis was performed using SPSS 23.0 (IBM Corp., Armonk, USA). For multi-group comparisons, one-way ANOVA followed by Duncan's multiple-range test or the Kruskal-Wallis test followed by Dunn's multiple-comparison test was used. Pearson correlation assessed expression associations, and two-sided unpaired Student's t-tests were employed. Statistical significance thresholds were defined as follows: P < 0.01 (significant), P < 0.05 (significant), 0.05 ≤ P < 0.10 (exploratory trend), and P ≥ 0.10 (not significant). Graphs were generated using GraphPad Prism 10 (GraphPad Software, San Diego, USA). Groups sharing at least one letter are not significantly different, whereas groups with no letter in common differ at P < 0.05. Tests involving the unmatched Il15rafl/fl; Vil-cre cohort quantify exploratory between-cohort differences rather than genotype-specific effects.
Results
Hepatic IL15 expression is dynamically regulated during MASLD progression and correlates with macrophage infiltration
To investigate the role of the IL-15 signaling axis in human fatty liver disease, we evaluated the hepatic transcriptome profiles across distinct pathological stages of metabolic dysfunction-associated steatotic liver disease (MASLD), ranging from healthy controls to cirrhosis. Our analysis revealed a stage-specific, dynamic expression pattern for
IL15. Specifically, hepatic
IL15 mRNA levels progressively increased from the healthy state to simple steatosis (MASL), peaking significantly during the early-to-moderate stages of steatohepatitis (MASH F1 and MASH F2) (
Figure 1A). Although its expression exhibited a slight decline in advanced fibrosis (MASH F3) and cirrhosis, it remained dysregulated compared to the baseline. In contrast, the hepatic expression of its specific receptor,
IL15RA, remained relatively constitutive and stable across all disease stages (
Figure 1B).
To further contextualize these changes within the hepatic microenvironment, we mapped the expression of
IL15 alongside genes regulating critical biological functions, including lipid metabolism, inflammation, fibrosis, and immune cell infiltration (
Figure 1C). The transcriptomic heatmap demonstrated a disease-driven reprogramming. As the disease progressed from MASL toward advanced MASH and cirrhosis, we observed a coordinated upregulation of gene clusters driving inflammation (
TNF,
IL1B,
CCL2), extracellular matrix (ECM) remodeling (
TIMP1,
MMP9), and fibrogenesis (
COL1A1,
COL1A2,
ACTA2). Concurrently, the expression of pan-macrophage markers (
ADGRE1,
CD68), as well as specific markers for both M1 (
CD86,
NOS2) and M2 (
ARG1,
CD163,
MRC1) polarization, were markedly enriched in the MASH cohorts, mirroring the inflammatory cascade.
Given the spatial and temporal co-emergence of
IL15 and macrophage signatures, we performed Pearson correlation analyses to delineate their relationship. In these bulk-tissue data, hepatic
IL15 expression showed modest positive associations with the M1 macrophage marker
CD86 (r = 0.3632,
P = 0.0410;
Figure 1D) and the M2 macrophage marker
CD163 (r = 0.3892,
P = 0.0277;
Figure 1E). Together, these bulk-tissue associations indicate that hepatic IL15 expression tracks with macrophage-related transcripts, although these data do not establish a functional or causal relationship.
CIBERSORTx deconvolution reveals a pro-inflammatory macrophage polarization shift during human MASLD progression
To characterize the cellular orchestration underlying the transcriptomic changes, we employed the CIBERSORTx deconvolution algorithm on the human bulk RNA-seq dataset (GSE246221) to infer the relative abundance of 22 distinct infiltrating immune cell subsets. The stacked area landscape (
Figure 2A) illustrated a profound and dynamic reshaping of the hepatic immune repertoire as the disease advanced from a healthy state towards cirrhosis.
Given the associations between
IL15 expression and macrophage-related transcripts, we specifically focused on the dynamics of the monocyte-macrophage lineage. Interestingly, the relative intrahepatic abundance of circulating monocytes remained broadly stable across all disease stages without statistically significant fluctuations (
Figure 2B). However, the resident and differentiated macrophage pools underwent a dramatic phenotypic remodeling. The proportion of unpolarized (M0) macrophages exhibited a marked expansion specifically during the early-to-moderate stages of steatohepatitis (MASH F1 and F2) compared to healthy controls (
Figure 2C).
More importantly, disease progression was characterized by a reciprocal shift in macrophage functional phenotypes. The abundance of pro-inflammatory M1 macrophages demonstrated a striking, stepwise accumulation, increasing significantly from the healthy state to MASL, and maintaining high levels through the advanced MASH and cirrhotic stages (
Figure 2D). Conversely, the proportion of anti-inflammatory and tissue-reparative M2 macrophages was progressively diminished, reaching its lowest significant point in the advanced fibrosis (MASH F3) cohort (
Figure 2E). Collectively, these
in silico deconvolution results suggest that the relative abundance of M1- and M2-associated macrophage signatures shifts during MASLD progression, without establishing a functional polarization state or causal role.
Murine hepatic Il15 expression specifically correlates with M1 macrophage polarization during experimental MASH progression
To examine the IL-15 signaling axis in a controlled experimental setting, we reanalyzed the Mus musculus component of GSE246221, a published STZ/HFD disease-course dataset spanning healthy liver, MASH-associated fibrosis, and hepatocellular carcinoma (HCC)
[18]. Consistent with the human data, hepatic
Il15 expression exhibited a disease-stage-dependent dynamic pattern.
Il15 mRNA levels progressively increased and peaked significantly at the MASH Fibrosis stage compared to the healthy and MASL groups, before showing slight fluctuations in the advanced fibrosis and HCC stages (
Figure 3A). Interestingly, unlike the constitutive expression observed in the human cohort, the expression of its receptor,
Il15ra, was dynamically downregulated during the progression of steatosis and fibrosis (
Figure 3B).
We next evaluated the global remodeling of the hepatic microenvironment in the murine model. The heatmap (
Figure 3C) confirmed the successful recapitulation of MASH pathogenesis. The transition towards MASH Fibrosis and HCC was accompanied by a robust and coordinated upregulation of gene signatures associated with inflammation (
Tnf,
Il1b,
Ccl2), extracellular matrix remodeling (
Timp1,
Mmp9), and fibrogenesis (
Col1a1,
Col1a2,
Acta2), alongside a concomitant dysregulation of lipid metabolism networks (
Ppara,
Acaca). Crucially, markers of macrophage infiltration and polarization were profoundly altered. While pan-macrophage (
Adgre1,
Cd68) and M1-specific markers (
Cd86,
Nos2) were strongly enriched in advanced disease stages, the expression patterns of M2 markers (
Arg1,
Cd163,
Mrc1) were more heterogeneous.
To precisely define the relationship between IL-15 and macrophage subsets in the murine liver, we conducted correlation analyses. Strikingly, hepatic
Il15 expression showed a significant moderate-to-strong positive correlation with the M1 macrophage marker
Cd86 (r = 0.6430,
P < 0.0001;
Figure 3D). In sharp contrast to the human cohort,
Il15 showed no significant correlation with the M2 macrophage marker
Cd163 (r = -0.1800,
P = 0.3008;
Figure 3E). Taken together, these
in vivo data not only corroborate the involvement of the IL-15 system in MASH-associated fibrogenesis but also highlight a specific, potentially causal link between IL-15 and the pro-inflammatory (M1) polarization of hepatic macrophages.
Validation of liver-specific IL15 and intestinal epithelium-specific IL15Rα conditional knockout mice models
The levels of IL-15/IL-15R
α complex in the liver and serum of
Il15fl/fl; Alb-cre and
Il15rafl/fl; Vil-cre mice were measured by ELISA. The results showed that, compared with
Il15fl/fl mice,
Il15fl/fl; Alb-cre mice exhibited decreased IL-15/IL-15R
α complex levels in the liver (
P = 0.071) (
Figure 4A), but no change in serum levels (
P > 0.05) (
Figure 4B). Compared with
Il15fl/fl and
Il15fl/fl; Vil-cre mice,
Il15rafl/fl; Vil-cre mice showed no change in hepatic IL-15/IL-15R
α complex levels (
P > 0.05) (
Figure 4C), whereas serum levels were significantly reduced (
P < 0.01) (
Figure 4D). The hepatic
P = 0.071 value is a non-significant downward trend, and residual whole-liver signal may arise from non-hepatocyte sources. These measurements are consistent with reduced hepatic IL-15/IL-15R
α levels in the liver-specific model and with reduced circulating complex in the intestinal IL-15R
α knockout model; however, cell-specific recombination and direct ligand expression require confirmation.
Effects of liver-specific IL15 and intestinal epithelium-specific IL15Rα signaling on fatty liver and lipid metabolism
Starting from week 8 of the experiment, the body weight gain rate of
Il15fl/fl; Alb-cre mice was significantly lower than that of
Il15fl/fl mice (
P < 0.05) (
Figure 5A). No significant differences were observed between the two groups in glucose clearance (
Figure 5B) or liver index (
Figure 5C) (
P > 0.05), whereas the degree of hepatocyte steatosis and NAFLD activity score were significantly lower in
Il15fl/fl; Alb-cre mice than in
Il15fl/fl mice (
P < 0.05) (
Figure 5D-5F). Compared with
Il15fl/fl mice,
Il15fl/fl; Alb-cre mice exhibited significantly reduced serum HDL-C levels and ALT activity (
P < 0.05), with a trend toward lower serum TG (
P = 0.064) (
Figure 5G-I), and significantly increased hepatic mRNA expression of
Prkaa1 and
Ppara (
P < 0.05) (
Figure 5J). These results indicate that hepatocyte-specific IL-15 deficiency was associated with lower body-weight gain, hepatic steatosis, histological activity, and serum ALT in HFD-fed mice, although the lower HDL-C level should be interpreted cautiously.
Il15rafl/fl;
Vil-cre mice showed a trend toward higher serum ALT activity (
P = 0.083) and significantly higher AST activity (
P < 0.05) compared with
Il15fl/fl mice (
Figure 5I), indicating impaired liver function in
Il15rafl/fl;
Vil-cre mice.
Effects of liver-specific IL15 and intestinal epithelium-specific IL15Rα signaling on liver fibrosis
Quantitative analysis of liver Sirius red staining showed no significant difference in the degree of liver fibrosis between
Il15fl/fl and
Il15fl/fl; Alb-cre mice (
P > 0.05) (
Figure 6A). Similarly, there were no significant differences in the hepatic expression of fibrosis markers (
Col1a2,
Pdgfrb, and
Col1a1) between the two groups (
P > 0.05) (
Figure 6B), providing no evidence of a detectable fibrosis difference between these genotype-matched groups under the tested conditions. In the three-group comparison,, compared with
Il15fl/fl mice,
Il15rafl/fl; Vil-cre mice exhibited significantly increased liver fibrosis (
P < 0.05) (
Figure 6A), and showed a trend toward higher mRNA level of the fibrosis marker
Col1a2 (
P = 0.069) (
Figure 6B). The
Il15rafl/fl; Vil-cre comparisons are exploratory between-cohort observations.
Effects of liver-specific IL15 and intestinal epithelium-specific IL15Rα signaling on markers of macrophages and cytotoxic T cells
Macrophages play an important role in MASLD, and IL15 has been reported to regulate macrophages. Therefore, we first examined the expression of macrophage markers in the liver. The results showed that, compared with
Il15fl/fl mice,
Il15fl/fl; Alb-cre mice had no significant change in hepatic
F4/80 mRNA expression (
Figure 7A), but exhibited
higher Cd86 transcript levels and the M2 macrophage markers
Arg1,
Cd163, and
Cd206 (
Figures 7B-7C), showing altered expression of selected macrophage-related transcripts. In addition, we examined the cytotoxic T cell marker
Cd8a, and found no significant difference in hepatic
Cd8a mRNA levels between the two groups (
P > 0.05) (
Figure 7D). Collectively, these data suggest that liver-specific IL15 promotes MASLD progression, potentially by modulating macrophage activity. Compared with
Il15fl/fl mice,
Il15rafl/fl; Vil-cre mice showed significantly increased hepatic
F4/80 mRNA levels (
P < 0.05) (
Figure 7E), no significant differences in the expression of M1 or M2 macrophage markers (
P > 0.05) (
Figures 7F-7G), and significantly decreased
Cd8a mRNA levels (
P < 0.05) (
Figure 7H). Within the three-group comparison shown in
Figure 7G, Cd163 was higher in
Il15fl/fl;
Vil-cre mice than in both
Il15fl/fl and
Il15rafl/fl;
Vil-cre mice, whereas the
Il15rafl/fl;
Vil-cre group did not differ significantly from
Il15fl/fl.
Discussion
This study examined compartment-specific IL-15/IL-15R
α signaling along the gut-liver axis in MASLD. Using tissue-specific knockout mice, our data support distinct roles: hepatocyte-derived IL-15 was associated with steatohepatitis, whereas loss of intestinal epithelial IL-15R
α was associated with exacerbated HFD-induced liver fibrosis. These observations help reconcile previous reports describing both protective and detrimental effects of IL-15 in fatty liver disease
[20,21] and underscore the importance of cellular origin when interpreting IL-15 biology.
Our data indicate that hepatocyte-specific IL-15 deficiency alleviated HFD-induced steatosis and liver injury and reduced the histological activity score. The phenotype was associated with concurrent upregulation of
Cd86 and selected M2 markers without a change in total macrophage abundance, indicating a mixed macrophage-marker profile, rather than a uniform polarization state, which requires functional validation. Human and murine transcriptomic analyses showed a positive association between
IL15 expression and
CD86, consistent with a macrophage-associated response that warrants further investigation
[22,23]. The concurrent upregulation of
Prkaa1 and
Ppara in IL-15-deficient livers may reflect altered lipid-handling programs; functional studies are needed before attributing this response to suppressed lipid oxidation
[24,25].
In sharp contrast, intestinal epithelium-specific IL-15R
α knockout did not alter steatosis but significantly aggravated hepatic fibrosis and elevated serum transaminases. This was accompanied by a marked reduction in circulating IL-15/IL-15R
α complexes, increased hepatic
F4/80 (pan-macrophage transcript), and reduced
Cd8a transcript levels; Cd8a may also be expressed by non-CD8 immune cells. IEC-mediated IL-15 trans-presentation has an established local role in intraepithelial lymphocyte biology
[17]. The concurrent reduction in circulating IL-15/IL-15R
α complexes and hepatic Cd8a expression raises the possibility that intestinal epithelial IL-15R
α may also influence systemic IL-15-related immune homeostasis
[26,27]. Hepatic CD8
+ T-cell subsets may regulate fibrosis in a context-dependent manner. This association is consistent with, but does not establish, a role for intestinal epithelial IL-15R
α in hepatic fibrogenesis through systemic immune regulation involving CD8
+ cells; however, this study measured hepatic Cd8a transcript rather than peripheral CD8
+ cell abundance. The proposed link therefore requires flow-cytometric or functional validation
[28,29]; greater confidence in the fibrosis phenotype would also benefit from Masson's trichrome or biochemical collagen validation.
Our reanalysis of human MASLD transcriptomic data (GSE246221) and CIBERSORTx deconvolution further contextualizes these observations. IL15 expression progressively increases from healthy liver to early MASH, correlating positively with M1 macrophage markers, and M1 macrophages accumulate with disease severity while M2 macrophages decline. These public-cohort associations provide context for the murine findings but do not establish causality or translational efficacy.
Collectively, these findings support a compartment-specific working model in which hepatocyte-derived IL-15 is associated with hepatic steatosis and injury through incompletely defined macrophage- and lipid-handling-related mechanisms, whereas loss of intestinal epithelial IL-15Rα is associated with reduced circulating IL-15/IL-15Rα complexes and exacerbated hepatic fibrosis. These observations indicate that the cellular origin and context of IL-15 signaling are important; however, direct immune-cell measurements and functional rescue experiments are needed before proposing selective therapeutic targeting.
Open Access
PDF
(7852KB)
