Interleukin-15 prevents concanavalin A-induced liver injury in mice via NKT cell-dependent mechanism

Authors

  • Bofeng Li,

    1. Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei 230027, China
    Search for more papers by this author
  • Rui Sun,

    1. Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei 230027, China
    2. Section on Liver Biology, Laboratory of Physiologic Studies, National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, MD
    Search for more papers by this author
  • Haiming Wei,

    1. Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei 230027, China
    Search for more papers by this author
  • Bin Gao,

    1. Section on Liver Biology, Laboratory of Physiologic Studies, National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, MD
    Search for more papers by this author
  • Zhigang Tian

    Corresponding author
    1. Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei 230027, China
    2. Institute of Immunopharmacology and Immunotherapy, School of Pharmaceutical Sciences, Shandong University, Jinan 250002, China
    • School of Life Sciences, University of Science and Technology of China, 443 Huangshan Road, Hefei, Anhui 230027
    Search for more papers by this author
    • fax: 86-551-360-6783


  • Potential conflict of interest: Nothing to report.

Abstract

Administration of concanavalin A (Con A) induces a rapid and severe liver injury in mice. Natural killer T (NKT) cells are recognized to be the key effector cells, and a variety of cytokines [e.g., interleukin 4 (IL-4), IL-5, interferon gamma (IFN-γ), and tumor necrosis factor alpha (TNF-α)] have been shown to play vital roles in Con A–induced liver injury, whereas the role of IL-15, a critical cytokine in the development and homeostasis of NKT cells, remains obscure. In this study, pretreatment with IL-15 prevented mice from Con A–induced mortality, elevation of serum transaminase, liver necrosis, and hepatocyte apoptosis. Depletion of NKT cells abolished Con A–induced liver injury, which could be restored by adoptive transfer of purified NKT cells but not by that of in vivo or in vitro IL-15–treated hepatic NKT cells. Furthermore, transfer of wild-type NKT cells to CD1d−/− mice restored liver injury, whereas transfer of IL-15–treated NKT cells did not. IL-15 pretreatment decreased the NKT-derived IL-4, IL-5, and TNF-α production, thereby resulting in less infiltration of eosinophils, which play a critical role in Con A–induced liver injury. In conclusion, IL-15 protects against Con A–induced liver injury via an NKT cell–dependent mechanism by reducing their production of IL-4, IL-5, and infiltration of eosinophils. These findings suggest that IL-15 may be of therapeutic relevance in human autoimmune-related hepatitis. (HEPATOLOGY 2006;43:1211–1219.)

Hepatitis is a worldwide health care problem. Concanavalin A (Con A)–induced hepatitis is a well-established experimental murine model,1 being characterized by markedly increased serum levels of transaminase and simultaneous infiltration of T cells, eosinophils,2, 3 and Kupffer cells4 into the liver. The hepatic natural killer T (NKT) cells play essential roles in the Con A–induced liver injury by releasing a variety of cytokines, including interleukin 4 (IL-4), IL-5, interferon gamma (IFN-γ), and tumor necrosis factor alpha (TNF-α).5–11 Interestingly, many other cytokines are also involved in Con A–induced liver injury; for example, IL-6, IL-10, IL-11, and IL-22 show hepatoprotective effects,12–16 whereas IL-12 and IL-18 aggravate Con A–induced hepatitis.17, 18

IL-15 is a member of the 4α-helix bundle family of cytokines that interacts with a heterotrimeric receptor consisting of the β and γ subunits of the IL-2 receptor and a specific, high-affinity IL-15–binding subunit, IL-15Rα. IL-15 is produced by multiple cells such as macrophages, epithelial, and fibroblast cells.19, 20 More importantly, IL-15 is an indispensable cytokine for the proliferation, survival, and homeostasis of NKT cells.20, 21 However, the role of IL-15 in this NKT cell–mediated liver injury remains unclear. We now report that IL-15 pretreatment leads to an IL-15-dose–dependent protection of liver injury. The protective effect of IL-15 is via NKT cell–dependent manner by directly attenuating IL-4 and IL-5 production, resulting in less eosinophils infiltration into the liver.

Abbreviations

Con A, concanavalin A; NKT, natural killer T; IL, interleukin; INF-γ, interferon gamma; TNF-α, tumor necrosis factor alpha; PBS, phosphate-buffered saline; mAb, monoclonal antibody; FITC, fluorescein isothiocyanate; RT-PCR, reverse transcription polymerase chain reaction; ALT, alanine aminotransferase; TUNEL, terminal deoxynucleotidyl transferase-mediated nick-end labeling; MNC, mononuclear cell; EPO, eosinophil peroxidase.

Materials and Methods

Reagents.

Con A was purchased from Sigma Chemical Co. (St. Louis, MO) and dissolved in phosphate-buffered saline (PBS) at a concentration of 2 mg/mL. Human recombinant IL-15 (Peprotech, London, UK) has a specific activity of 2 × 106 U/mg. Human IL-15 shares approximately 73% sequence identity with murine IL-15 and can be used by mice. The monoclonal antibodies (mAb) used in this study include fluorescein isothiocyanate (FITC)-conjugated anti-NK1.1 (eBioscience, San Diego, CA), PE-anti-IL-4, PE-anti-IL-5, PE-anti-IFN-γ, PE-anti-TNF-α (CALTAG, Burlingame, CA), PE-anti-NK1.1, PE-anti-CD122, PE-anti-CD132, and Cycrome-anti-CD3 (BD PharMingen, San Diego, CA), Biotin-anti-IL-15Rα (R&D Systems, Minneapolis, MN).

Animals.

Eight- to ten-week-old male C57BL/6 mice were purchased from the Shanghai Experimental Center, Chinese Academy of Science. CD1d−/− mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Mice were maintained under specific pathogen-free conditions in our animal facilities. All animals received care in compliance with the guidelines outlined in the Guide for the Care and Use of Laboratory Animals. In liver injury studies, mice received an intravenous Con A injection (15 μg/g body weight but 20 μg/g body weight for CD1d−/− mice). For survival study, Con A dosage increased to 25 μg/g body weight. In the IL-15 pretreatment group, 20 μg/mouse IL-15 was subcutaneously injected 6 hours before Con A administration. Sera were collected at indicated time points after Con A administration to measure the serum levels of alanine aminotransferase (ALT) and cytokines. Livers were also excised for the histological assay and terminal deoxynucleotidyl transferase-mediated nick-end labeling (TUNEL) assay.

Assay for Serum Transaminase Activity.

Serum samples from mice were obtained at different times. Serum ALT activities were determined by using the serum transaminase test kit (Rong Sheng, Shanghai) based on methods recommended by the International Federation of Clinical Chemistry.

Histology and TUNEL Assay.

Livers were fixed in 10% formalin, embedded in paraffin, cut into 5-μm sections, deparaffinized in xylene, and serially dehydrated in decreasing concentrations of ethanol. Sections were stained with hematoxylin-eosin and examined under light microscopy to detect lymphocyte infiltration and to evaluate tissue structure. The TUNEL assay was carried out using the In Situ Cell Death Detection Kit (Roche Diagnostics, Mannheim, Germany). Nuclear staining was evaluated under a light microscope. Immunohistochemistry was performed using a polyclonal Ab specific for the COOH terminus of CCR3 (Santa Cruz Biotechnology, Santa Cruz, CA). Staining of the primary antibody was performed with a horseradish peroxidase–labeled rabbit anti-goat Ig Ab and followed by DAB (diaminobenzidine) (Vector Laboratories, Burlingame, CA).

Preparation of Mouse Hepatocytes.

Primary mouse hepatocytes were isolated by collagenase perfusion.22 Briefly, mouse liver was perfused with HEPES buffer followed by perfusion of the liver with collagenase solution. The isolated hepatocytes were washed with cold PBS. The number and viability of isolated cells were determined by trypan blue exclusion. Trypan blue–negative hepatocytes were then stained extracellularly with Annexin-V as recommended by the manufacturer (BD Pharmingen) to determine the apoptosis cells or stained with IL-15R antibodies for flow cytometric analysis.

Cell Depletion.

The anti-mouse-CD3, -CD4, -CD8, -NK1.1, and -anti-IL-4 antibodies were obtained from partially purified hybridoma culture supernatants using ammonium sulfate precipitation. The hybridoma cell lines were purchased from American Type Culture Collection (Rockville, MD). Anti-mouse-IL-5 (clone TRFK-5) was purchased from R&D systems. Mice were given two injections of the indicated mAb 48 and 36 hours before Con A injection to deplete specific immune cells and to block specific cytokines in vivo. To deplete NK cells, a dose of 50 μL of anti-ASGM1 antiserum (Wako Pure Chemical Industries, Ltd., Japan) diluted in 250 μL of pyrogen-free PBS was intravenously injected into mice 1 day before treatment. To deplete NK cells and NKT cells, mice were injected intraperitoneally with 0.5 mL PBS containing 250 μg anti-NK1.1 (mAb PK136, ATCC). Depletion of immune cells was confirmed by flow cytometry, and this protocol produced a >90% decrease in the number of the indicated cells.

Liver Mononuclear Cell Preparation.

Mouse livers were removed and pressed through a 200-gauge stainless steel mesh and then suspended in RPMI 1640 medium (Gibco, BRL) and centrifuged at 50g for 5 minutes. Supernatants containing hepatic mononuclear cells (MNCs) were collected, washed in PBS, and re-suspended in 40% Percoll (Sigma) in RPMI 1640 medium. The cell suspension was gently overlaid onto 70% Percoll and centrifuged for 20 minutes at 750g. MNCs were collected from the interphase, and washed twice in PBS. The degree of contamination by Kupffer cells and hepatocytes was minimal.14

Purification of NK1.1+CD3+T (NKT) Cells.

For purification of NKT cells, mice were injected with anti-ASGM1 to deplete NK cells (NK1.1+CD3). Twenty-four hours later, hepatic lymphocytes were isolated from these NK-depleted mice and stained with FITC-NK1.1 mAb and then incubated with anti-FITC Microbeads (Miltenyi Biotec, Auburn, CA) for 15 minutes at 4°C. NK1.1+ cells were enriched by positive magnetic cell sorting (MACS) according to the manufacturer's protocol. Approximately 90% of the magnetic cell sorting–purified cells were NK1.1 and CD3 positive.

Adoptive Transfer of Hepatic NKT Cells.

AntiNK1.1 antibody was injected into mice 48 and 36 hours before adoptive transfer, the time when antibody had depleted recipients' NK and NKT cells and did not affect the transferred NKT cells. Adoptive hepatic NKT cells transfer was performed as follows: 50 μL of purified hepatic NKT cells (1 × 106 cells) in PBS were injected into the lateral left lobe of the liver at a rate of 10 μL/sec using a 29-gauge needle attached to a 1-mL syringe, followed by intravenous injection of Con A. Liver injury was assessed 24 hours later by measuring serum ALT levels.6

Measurement of Serum Cytokine Levels.

Blood samples were obtained from retro-orbital puncture and immediately centrifuged at 4,000 rpm for 15 minutes. Serum samples were stored at −20°C until ready to be used. Serum levels of IL-4, IL-5, IFN-γ, and TNF-α were determined using ELISA kits from R&D systems.

Flow Cytometric Analysis.

MNCs were stained with antibodies diluted in PBS containing 2% fetal calf serum and 0.1% NaN3. For the intracellular cytokine assay, cells were fixed, permeabilized, and stained with intracellular IL-4, IL-5, IFN-γ, and TNF-α using a Cytofix/Cytoperm plus kit (BD PharMingen). Cells are analyzed by FACSCalibur flow cytometer with CellQuest software (Becton Dickinson, Franklin Lakes, NJ).

Reverse Transcription Polymerase Chain Reaction Analysis.

RNA was extracted from liver tissue using Trizol Reagent (Invitrogen, Carlsbad, CA). One microgram RNA was reverse transcribed. The primer sequences used in polymerase chain reaction (PCR) were as follows: β-actin: sense, 5′-GGA CTC CTA TGT GGG TGG CGA GG-3′; antisense, 5′-GGG AGA GCA TGC CCT CGT AGA T-3′. IL-4: sense, 5′-ACA AAA ATC ACT TGA GAG AGA TCA T-3′; antisense, 5′-AGT AAT CCA TTT GCA TGA TGC TCT T'. IL-5: sense, 5′-GCC ATG GAG ATT CCC ATG AGC ACA-3′; antisense, 5′-GCC TTC CAT TGC CAA CTC TGT AC-3′ Eotaxin-1 (CCL11): sense, 5′-TCC ACA GCG CTT CTA TTC CT-3′; antisense, 5′- CTA TGG CTT TCA GGG TGC AT -3′. Eotaxin-2 (CCL24): sense, 5′- ACC CCA GCT TTG AAC TCT GA-3′; antisense, 5′- AAG GAC GTG CAG CAA GAT G-3′. RANETS (CCL5): sense, 5′- CCC CAT ATG GCT CGG ACA CCA-3′; antisense, 5′- CTA GCT CAT CTC CAA ATA GTT GAT-3′.

Assay of Hepatic EPO Activity.

Hepatic EPO (eosinophil peroxidase) activity was measured as described previously.23, 24 The enzyme activities of the hepatic tissues were calculated by subtracting the mean background optical density (OD) and expressed as change of OD490 per minute.

Statistical Analysis.

The statistical comparison between experimental and control groups was conducted by Student t test, and P < .05 was considered significant.

Results

IL-15 Prevents Mice From Con A–Induced Liver Injury.

IL-15 is a necessary cytokine for the survival and homeostasis of NKT cells. To know the role of IL-15 in Con A–induced liver injury, exogenous IL-15 was administered subcutaneously to mice. Fig. 1A illustrates that IL-15 pretreatment exerted a dose-dependent protection of liver injury. Pretreatment with 20 or 40 μg/mouse of IL-15 markedly inhibited Con A–induced elevation of serum ALT levels, whereas 10 μg/mouse of IL-15 only partly alleviated ALT elevations, and 2 μg/mouse IL-15 had no hepatoprotective effect. Next, serum ALT levels were measured at different times after Con A injection in IL-15– and PBS-treated groups. As shown in Fig. 1B, in the Con A–treated group, serum ALT started to increase as early as 6 hours after Con A injection, peaked at 24 hours, and declined thereafter. In the IL-15–treated group, significantly lower serum ALT levels were detected, particularly at 12 and 24 hours after Con A injection. Furthermore, we examined whether IL-15 protected mice from a lethal dose of Con A. As shown in Fig. 1C, 80% of IL-15–treated mice survived, whereas 100% of PBS-treated mice died within 12 hours after injection with a lethal dose of Con A.

Figure 1.

IL-15 administration exerts protective effects on Con A–induced hepatitis. (A) Various doses of lL-15 were administrated to C57BL/6 mice (5 mice per group) 6 hours before Con A injection (15 μg/g body weight). Sera were collected 24 hours after Con A injection. (B) PBS or IL-15 (20 μg/mouse) was administered subcutaneously to C57BL/6 mice 6 hours before intravenous injection of Con A (15 μg/g body weight). Serum ALT levels were measured at 0, 3, 6, 12, 24, and 48 hours after Con A injection. Values in panels A and B are shown as means ± SEM (n = 5) from 3 independent experiments. **P < .01 as comparison with Con A alone group. (C) Analysis of the effect of IL-15 pretreatment on survival in mice treated with Con A. C57BL/6 mice (n = 5) were injected with PBS or IL-15 (20 μg/mouse) 6 hours before administration with a lethal dose of Con A (25 μg/g body weight). Con A, concanavalin A; PBS, phosphate-buffered saline.

Liver histological studies were also used to determine the protective effect of IL-15 on Con A–induced liver injury. As shown in Fig. 2A, light microscopy examination showed massive necrosis in the livers of mice 24 hours after treatment with Con A alone, whereas mice pretreated with IL-15 showed minor damage. TUNEL assay showed significant hepatocyte apoptosis in the livers of mice treated with Con A, which was markedly prevented by IL-15 pretreatment (Fig. 2B). Similarly, hepatocytes isolated from IL-15–pretreated mice showed lower Annexin-V expression than that without IL-15 pretreatment, indicating that IL-15 can protect hepatocyte from apoptosis (Fig. 2C).

Figure 2.

Pretreatment with IL-15 prevents liver necrosis and hepatocyte apoptosis in Con A–induced hepatitis. (A) Mice were treated with PBS or IL-15 (20 μg/mouse) and 6 hours later, followed by Con A injection (15 μg/g). Liver samples were collected 24 hours after Con A injection for hematoxylin-eosin staining (original magnification ×100 and ×200). N indicated necrosis area and red arrow indicated eosinophil. (B) Liver samples were collected 12 hours after Con A injection for TUNEL staining (original magnification ×100 and ×400). (C) Decrease of Annexin V–positive hepatocytes in the liver of IL-15–treated mice. Hepatocytes were isolated 12 hours after Con A injection and then stained with Annexin-V. Representative histograms for Con A- (thick dark line) and IL-15 plus Con A (faint broken line)–treated groups were shown overlaid on the PBS group (gray). Con A, concanavalin A; PBS, phosphate-buffered saline; HE, hematoxylin-eosin; TUNEL, terminal deoxynucleotidyl transferase-mediated nick-end labeling.

IL-15 Protects Against Con A–Induced Hepatitis via NKT Cell–Dependent Mechanism.

Because IL-15 receptors did not express on hepatocytes (Supplemental Fig. 1), IL-15 did not possibly exert any direct protective effects on hepatocytes.Therefore, we observed the change of liver lymphocytes and found that both absolute numbers and ratios of hepatic T, NKT, and NK cells did not significantly change after IL-15 pretreatment (Fig. 3A). We then explored the possible functional variation of lymphocyte populations in IL-15–treated mice. Various antibodies including anti-mouse-CD3, -CD4, -CD8, and -ASGM1, were used to eliminate the corresponding murine lymphocyte subpopulation. As already known1 and also shown in this study (data not shown), anti-mouse-CD3 or -CD4 antibodies abolished Con A–induced liver injury, mainly because of NKT cell depletion. Depletion of murine CD8+ T cells and ASGM1+ NK cells did not influence Con A–induced liver injury in either IL-15- and PBS-treated mice (Fig. 3B), indicating that the protective effect of IL-15 on Con A–induced hepatitis is not mediated via NK- or CD8+ T cell–dependent mechanisms.

Figure 3.

IL-15 protects against Con A–induced hepatitis via NKT cell-dependent mechanism. (A) C57BL/6 mice were treated with IL-15 (20 μg/mouse) or PBS 6 hours before Con A injection. After various time points, total hepatic MNCs were counted, and the lymphocyte populations were analyzed by FACS using anti-NK1.1 and anti-CD3 antibodies. (B) Cell depletion was conducted by using anti-mouse-CD8 and -ASGM1 antibodies, followed by IL-15 treatment or PBS. Six hours later, mice were treated with Con A (15 μ g/g). Sera were collected 24 hours after Con A injection for measuring ALT levels. (C) Depletion of NK and NKT cells by anti-NK1.1 (PK136) antibody treatment. (D) and (E) Various purified wild-type NKT cells (1 × 106 cells) were adoptively transferred into anti-NK1.1 Ab (PK136) treated mice or CD1d−/− mice, followed by intravenous injection with 15 μg/g or 20 μg/g Con A, respectively. Mice were killed 24 hours after Con A injection, and serum ALT levels were measured. These donor NKT cells were obtained from PBS or IL-15-treated wild-type mice or treated with IL-15 (2,000 ng/mL) in vitro for 4 hours. Values in panels A-E are shown as means ± SEM (n = 5) of three mice in each group from three independent experiments. **P < .01. Con A, concanavalin A; PBS, phosphate-buffered saline; NK, natural killer.

Next, to investigate whether IL-15 protects against Con A–induced liver injury via modulation of NKT cells, adoptive transfers of in vivo or in vitro IL-15–treated hepatic NKT cells were performed in NK and NKT cell–depleted mice. Treatment with anti-NK1.1 (PK136) antibody completely depleted NK and NKT cells (Fig. 3C) and abolished Con A–induced elevation of serum ALT levels (Fig. 3D), but did not influence the transferred NKT cells thereafter (Supplemental Fig. 2). Adoptive transfer of hepatic NKT cells isolated from PBS-treated mice restored Con A–induced hepatitis, whereas adoptive transfer of hepatic NKT cells from IL-15–treated mice failed to restore Con A–induced liver injury, indicating that IL-15 inhibits the ability of NKT cells to restore Con A–induced liver injury. Furthermore, adoptive transfer of in vitro IL-15–treated hepatic NKT cells also prevented NKT-depleted mice from Con A–induced liver injury, suggesting that IL-15 directly acts on NKT cells (Fig. 3D). To further confirm these results, we used NKT cell-deficient CD1d−/− mice as recipients. Adoptive transfer of wild-type NKT cells treated with IL-15 in vitro or from in vivo IL-15–treated mice into CD1d−/− mice also suppressed Con A–induced hepatitis (Fig. 3E). Taken together, these results indicated that IL-15 protected against Con A–induced hepatitis via an NKT cell–dependent mechanism.

IL-15 Inhibits Cytokine Productions of NKT Cells in Con A–Induced Hepatitis.

NKT cells rapidly secrete a large amount of IL-4, IL-5, IFN-γ, and TNF-α, all of which have been shown to play essential roles in Con A–induced liver injury.3, 7–12 We next wanted to know whether IL-15 influences these cytokines' production. As shown in Fig. 4A-D, IL-15 pretreatment reduced serum levels of IL-4 by 30% and 50% at 2 and 3 hours after Con A injection and reduced serum IL-5 at 3, 6, and 12 hours after Con A injection. Serum TNF-α production also decreased in IL-15–treated mice although the serum levels of IFN-γ did not alter. We then used flow cytometry to analyze intracellular cytokines produced by hepatic NKT, T, and NK cells. As shown in Fig. 4E, IL-4, IL-5, IFN-γ, and TNF-α were mainly produced by NKT cells, and partially produced by T and NK cells. IL-15 pretreatment greatly suppressed NKT-derived IL-4, IL-5, and TNF-α production, whereas IFN-γ production of NKT cells was not obviously influenced.

Figure 4.

IL-15 pretreatment decreases NKT-derived cytokine production in Con A–induced hepatitis. Mice were treated subcutaneously with PBS or IL-15 for 6 hours, followed by administration with Con A. Serum samples were collected at various times post–Con A injection, and serum levels of IL-4 (A), IL-5 (B), IFN-γ (C), and TNF-α (D) were measured by ELISA. Values in panels A-D are shown as means ± SEM (n = 5). **P < .01, *P < .05 as comparison with Con A group. (E) For flow cytometric analysis, 1 hour after Con A injection, liver MNCs were prepared and examined for various intracellular cytokines production. T (CD3+NK1.1), NKT (CD3+NK1.1+) and NK (CD3NK1.1+) cells were then examined for intracellular IL-4, IL-5 IFN-γ and TNF-α by flow cytometry after triple staining with anti-NK1.1, anti-CD3, and the indicated intracellular antibodies. Con A, concanavalin A; PBS, phosphate-buffered saline; NK, natural killer.

IL-15 Pretreatment Reduces Infiltration of Eosinophils in Con A–Induced Liver Injury.

Recently, Louis et al.3 reported that depletion of eosinophils completely prevented Con A–induced liver damage,3 and IL-4 and IL-5 played an essential role in recruiting eosinophils into liver.2, 3 Here, we observed that when anti-IL-4 or anti-IL-5 Ab was used to deplete corresponding cytokines, the ALT levels and infiltration of eosinophils into the liver were greatly decreased (Fig. 5A). As already shown in Fig. 4E, IL-4 and IL-5 were mainly secreted by hepatic NKT cells and IL-15 pretreatment completely suppressed IL-5 production and significantly decreased IL-4 production. We further demonstrated that IL-15 pretreatment inhibited the IL-4 and IL-5 productions of purified NKT cells by reverse transcription polymerase chain reaction (RT-PCR) analysis (Fig. 5B). We then measured the EPO activity to confirm the eosinophils infiltration. As shown in Fig. 5C, Con A injection significantly elevated EPO activity in the livers of PBS-treated mice but not in that of IL-15–treated mice. We also observed that IL-15 pretreatment reduced the expression of eotaxin-1 (CCL11) and eotaxin-2 (CCL24) in liver, two important chemokines playing an important role in eosinophils recruitment (Fig. 5B). Because eotaxin is the ligand of CCR3, the main receptor expressed on eosinophils' surface,25–27 we then used immunohistochemical staining to detect eosinophils in liver. As shown in Fig. 5D, Con A–treated mice had massive liver necrosis, which was surrounded by brown CCR3+ cells (eosinophils). IL-15 pretreatment greatly reduced eosinophil infiltration into the liver. Additionally, the numbers of eosinophils per field in Fig. 5D were almost parallel to those of Fig. 5C. Therefore, down-regulating NKT-derived-IL-4 and IL-5 production in liver by IL-15 treatment is an important mechanism responsible for IL-15 to attenuate eosinophil infiltration.

Figure 5.

IL-15 pretreatment reduces infiltration of eosinophils in Con A–induced liver injury. (A) Mice were injected with anti-mouse-IL-4 and -IL-5 Abs or control-IgG. Sera and livers were collected 16 hours after Con A injection. ALT levels and EPO activities were measured. Values are shown as means ± SEM (n = 3). **P < .01. (B) Mice were killed 1 hour after Con A injection. RNA samples from Purified NKT cells or liver tissue were used to test expressions of IL-4, IL-5, eotaxin-1, eotaxin-2, and RANTES. RT-PCR analysis was used with corresponding primers. (C) Mice were treated with PBS or IL-15 6 hours before Con A injection. Livers were collected 16 hours after Con A injection. EPO activities were measured. (D) Liver samples were collected 16 hours after Con A injection for anti-CCR3 staining (original magnification ×200). N, necrosis area. Ten fields were randomly selected and the numbers of eosinophil were counted. Con-A, concanavalin A; ALT, alanine aminotransferase; RT-PCR, reverse transcription polymerase chain reaction; PBS, phosphate-buffered saline.

Discussion

Traditionally, IL-15 is thought to be a proinflammatory factor in autoimmune diseases such as rheumatoid arthritis (RA), ulcerative colitis (UC) and psoriasis.28, 29 IL-15 contributes to the pathogenesis of these autoimmune diseases through (1) recruiting and activating T cells to the inflammatory tissue; (2) leading to TNF-α and IFN-γ secretion by autoreactive T cells; and (3) protecting autoreactive T cells from apoptosis.30, 31 On the contrary, in Con A model, IL-15 administration neither recruited more lymphocytes into liver nor protected NKT cells from apoptosis (Fig. 3A). Moreover, IL-15 down-regulated serum TNF-α production (Fig. 4D) and did not influence IFN-γ production (Fig. 4C). Meanwhile, although IL-15 can induce autoreactive T cells to attack self organ, it is usually a long-term inflammation process. In Con A–induced fulminant hepatitis, the process was too short because ALT level peaked at 24 hours and declined to normal level at 48 hours. Why IL-15 plays a different role in chronic autoimmune diseases, such as RA, UV and psoriasis, and fulminant hepatitis, such as Con A model, needs further investigation.

Recently, Bulfone-Paul et al. reported that human IL-15-murine IgG2b fusion protein completely prevented mice from anti-Fas antibody–induced serious liver injury,32 the first report about IL-15 protective effect on liver injury. They focused on that IL-15 exerted direct anti-apoptotic effects on activated T cells from many types of tissues, including liver, spleen, and thymus. However, in their study, they did not survey the inflammatory process (e.g., the injury interaction between leukocytes and hepatocytes) and how IL-15 protected mice from the fulminant hepatitis. We consider that the underlying mechanisms of Fas-induced murine hepatitis are totally different from Con A–induced hepatitis. The former model is through a clear target molecule, a death receptor that may directly induce the apoptosis of a variety of cells (including hepatocytes) in many organs (such as liver); the latter is through an extremely complicated and unclear cellular immunologic process. Because IL-15Rα, IL-15Rβ (CD122) and IL-15Rγ (CD132) were not expressed on hepatocytes from our study (Supplemental Fig. 1), we concluded that IL-15 would not directly act on hepatocytes. Because NKT cells are recognized to play vital roles in Con A–induced liver injury and IL-15 is an indispensable cytokine in the development and homeostasis of NKT cells, we thought IL-15 might directly influence NKT cells' function. Now, we demonstrated the hepatoprotective effect of IL-15 was via NKT cell-dependent mechanism. Because Fas-induced hepatitis is not possibly mediated by inflammatory leukocytes such as NKT cells and eosinophils, the actual protective mechanisms of IL-15 on apoptosis of hepatocyte and liver injury in this model are not known.

In fact, except for IL-15, many other cytokines such as IL-6, IL-10, and IL-22 were shown to have a protective effect in Con A–induced liver injury through different mechanisms.13, 14, 16 IL-10 reduced the Con A–induced hepatotoxicity by diminishing the production of proinflammatory cytokines IL-12, IFN-γ, and TNF-α13; IL-22 was a survival factor for hepatocytes and protected mice from Con A–induced liver injury via STAT3 activation; IL-6, similarly to IL-15, alleviated the Con A–induced liver injury through an NKT-dependent manner, as reported in our previous publication.14 However, adoptive transfer of in vitro IL-15–treated NKT cells attenuated liver injury (Fig. 3D-E), whereas transfer of in vitro IL-6-treated–NKT cells did not. The reason is that IL-6 influenced NKT cells in CD4+ T cell-dependent manners, whereas IL-15 directly acted on NKT cells, as indicated in this article.

NKT cells have the unique potential to rapidly secrete large amounts of cytokines, which might be important for the initiation and regulation of some immune responses and autoimmune diseases.33 In the Con A model, NKT–derived IL-5 might improve eosinophil maturation and infiltration, and NKT-derived IL-4 might upregulate expression of eotaxin-1 (CCL11), eotaxin-2 (CCL24), and RANTES (CCL5),2, 34 which are ligands of eosinophil receptor CCR3.25–27 Fig. 5A showed liver hepatitis was completely inhibited in anti-IL-4 Ab-treated mice while partly inhibited in IL-5–depleted mice because IL-4 also induced an increase in the expression of granzyme B and FasL on NKT cells.7 Moreover, we and others35, 36 found NKT cells treated with IL-15 failed to produce IL-5. These best explain why IL-15–treated NKT cells may exert negative regulatory effects on eosinophils. The negative regulatory effects of IL-15 on eosinophils were also observed in asthma, which is related to decreased production of IL-5.37 However, the molecular mechanism regarding how IL-15 reduces the production of IL-4 and IL-5 is not clear.

Previous studies reported that TNF-α played an important role in the Con A model by neutralizing TNF-α or using TNFR knockout mice.9–12 We also found IL-15 pretreatment could not only reduce serum TNF-α level but also suppress the hepatic NKT-derived TNF-α production (Fig. 4D-E). However, NKT cells are not the only main source for secreting TNF-α; Kupffer cells are possibly another important source of TNF-α in liver.4 We still do not know whether IL-15 directly inhibits TNF-α production by Kupffer cells.

In summary, we report that IL-15 pretreatment prevented Con A–induced mortality, elevation of serum transaminase, and liver necrosis, clearly indicating that IL-15 is able to protect mice from Con A–induced liver injury. The protective effects of IL-15 on Con A–induced liver injury were mediated via an NKT cell–dependent mechanism by inhibiting their IL-4, IL-5, and TNF-α productions and therefore reducing eosinophil infiltration into the liver. IL-15 is potentially of therapeutic relevance in treating human autoimmune-related hepatitis.

Ancillary