SEARCH

SEARCH BY CITATION

Keywords:

  • α-GalCer;
  • autoimmune hepatic inflammation;
  • cytotoxic T lymphocyte;
  • interferon-gamma;
  • natural killer T cell

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References
  10. Supporting Information

As natural killer (NK) T cells play an important role in the development of autoimmune diseases, they should have significant roles for the pathogenesis of autoimmune liver disease. Implication of the NK T cells in the generation of autoimmune-related hepatic inflammation was investigated using a novel mouse model. Immunization of mice with dendritic cells (DCs) loaded with hepatocyte-mimicking hepatocellular carcinoma cells (DC/Hepa1-6) induces cytotoxic T lymphocytes (CTL) capable of killing hepatocytes. Subsequent administration of interleukin (IL)-12, a potent interferon-gamma (IFN-γ) inducer, to the immunized mice generates autoimmune hepatic inflammation (AHI), as reported previously. Upon onset of the AHI response, the number of intrahepatic CD3+NK1·1+ NK T cells increased markedly, along with a decrease in the number of splenic NK T cells, augmented expression of CXCR6 on intrahepatic NK T cells and CXCL16 in hepatic tissue, suggesting that NK T cells were recruited into the inflamed liver. The NK T cells were strongly positive for CD69 and produced IFN-γ, but not IL-4. AHI activity was attenuated markedly in CD1d–/– NK T cell-deficient mice, indicating that NK T cells play a pivotal role in the development of AHI. Mice treated with DC/Hepa1-6 and alpha-galactosylceramide, a potent NK T cell activator, also exhibited similar hepatic inflammation, in which activated NK T cells producing IFN-γ and CD8+ T cells cytotoxic to hepatocytes were induced in liver-infiltrating mononuclear cells. Activated NK T cells producing IFN-γ potentiate DC-based AHI in the mouse model.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References
  10. Supporting Information

Autoimmune hepatitis (AIH) is an immune-mediated chronic inflammatory liver disease that predominantly affects women genetically predisposed to its development [1]. Although the pathogenic mechanism of AIH and aetiological agents involved are not known, it has been postulated that autoreactive T cells targeting at hepatocytes play a central role in the development of AIH [2]. Until recently, autoreactive CD4+ T cells were considered to be critical for disease development [3], but increasing evidence has shown that CD8+ T cells also play significant roles [4]. CD8+ T cells are observed mainly in areas of interface hepatitis of AIH, and CD4+ T cells are found in the central part of the portal tract [5]. Peripheral CD8+CD25+ lymphocytes are more prevalent in patients with acute-onset AIH than those with chronic AIH. After treatment with immunosuppressive therapies, the number of CD8+CD25+ T lymphocytes in the blood decrease in parallel with the serum aminotransferase level [6]. Although hepatic tissue in AIH exhibits abundant apoptosis of hepatocytes [7], and highly activated infiltrating T lymphocytes [8] may play an important role in the induction of hepatocyte apoptosis, the mechanism of establishment and progression of AIH remains unclear.

Natural killer (NK) T cells are innate immune cells that were described originally as expressing both T cell and NK cell phenotypes [9]. NK T cells are activated in a CD1d-dependent manner in response to glycolipid antigens such as α-galactosylceramide (α-GalCer) [10, 11], and rapidly produce large amounts of T helper type 1 (Th1) cytokine, interferon-gamma (IFN-γ) and Th2 cytokine, interleukin (IL)-4 [12]. The implication of NK T cells in autoimmune diseases such as multiple sclerosis [13-15] and diabetes [16-18] has been shown using well-established animal models. In mice, NK T cells represent up to 30% of T cells in the liver, where they reside within the sinusoids and appear to provide intravascular immune surveillance [19, 20], and may also be associated with the development of liver injury in the setting of hepatitis [21, 22]. In humans, it has not been clarified whether NK T cells are beneficial or harmful in the setting of liver disease [23]. The importance of NK T cells in the pathogenesis of autoimmune liver disease also remains unknown.

We have reported previously a mouse model of autoimmune hepatic inflammation (AHI) generated by immunization of C57BL/6 mice with dendritic cells (DC) loaded with well-differentiated hepatocellular carcinoma cells (Hepa1-6), followed by IL-12 administration [24, 25]. In this model, liver specific inflammation is mediated by hepatocyte-responsive autoreactive T cells. Our findings indicate that two independent steps are necessary for the development of autoimmune-mediated liver damage: one step concerns the induction of autoreactive T cells responsive to hepatocytes. Because of similar phenotypic expression between normal hepatocytes and Hepa1-6 cells, cytotoxic T lymphocytes (CTLs) recognizing shared antigen between them are induced by immunization of mice with DCs loaded with Hepa1-6 [24]. The other step is the modulation of the hepatic microenvironment to promote recruitment of autoreactive T cells into the liver and CTL response to hepatocytes, because the sole induction of autoreactive T cells cannot generate autoimmune hepatic injury in vivo. The key cytokine for this response is IFN-γ induced by IL-12, which provides the enhanced expression of major histocompatibility complex (MHC) class I, adhesion molecules and chemokines on hepatocytes [24]. In fact, treatment of the immunized mice with anti-IFN-γ monoclonal antibody or immunization of IFN-γ knock-out mice abolishes AHI activity [24]. Although, unlike human AIH, histological features of AHI are characterized by an acute inflammatory response located mainly in hepatic parenchyma, this model could contribute to analysis of the mechanism of the liver-specific autoimmune response [25].

In this study we show, using our mouse model, that intrahepatic NK T cells play a pivotal role in promoting the CTL-mediated autoimmune hepatic inflammation.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References
  10. Supporting Information

Animals

Eight-week-old female C57BL/6 wild-type (WT) mice were purchased from Sankyo Labo Service Co., Ltd (Tokyo, Japan). CD1d–/– mice (C57BL/6 genetic background) were kindly provided by Dr Shinsuke Taki (University of Shinshu, Japan). All animals were maintained in our facilities and received humane care according to the criteria outlined in the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences (NIH publication 86-23, revised 1985). Mice were used at the age of 8–10 weeks and were matched for sex and age.

Cell lines, cytokines and antibodies

Hepa1-6, a well-differentiated murine hepatocellular carcinoma cell line, was obtained from the American Type Culture Collection (Manassas, VA, USA). Recombinant murine granulocyte–macrophage colony-stimulating factor (GM-CSF) and recombinant IL-2, IL-4 and IL-12 were purchased from PeproTech (Rocky Hill, NJ, USA). Anti-CD8a monoclonal antibody (mAb) (clone 53-6·7) conjugated with fluorescein isothiocyanate (FITC) was purchased from BD Biosciences (San Diego, CA, USA). Anti-CD69 mAb (clone H1·2F3) conjugated with allophycocyanin (APC) and anti-CD62L mAb (clone MEL-14) conjugated with phycoerythrin (PE) were purchased from BioLegend (San Diego, CA, USA). Anti-CXCR6 mAb (clone 221002) conjugated with PE was purchased from R&D Systems, Inc. (Minneaoplis, MN, USA). Anti-IFN-γ (clone XMG1·2), anti-IL-4 (clone 11B11) conjugated with APC for intracellular cytokine staining were purchased from eBioscience (San Diego, CA, USA).

Treatment of mice

AHI was induced in mice as described previously [24, 25]. Briefly, bone marrow-derived DCs loaded with Hepa1-6 cells (DC/Hepa1-6) were generated by quick treatment of a mixture of the DCs and Hepa1-6 cells with 50% polyethylene glycol (Peg solution; Sigma-Aldrich, Inc., St Louis, MO, USA). DC/Hepa1-6 cells were injected subcutaneously into 8-week-old female C57BL/6 WT mice or CD1d–/– mice on days 1 and 14. Then, IL-12 (500 ng/mouse) was injected intraperitoneally on days 15, 17 and 19. The mice were sacrificed on day 21. To analyse the role of activated NK T cell in this mouse model, α-GalCer (KRN7000; Funakoshi Co., Ltd, Tokyo, Japan) was dissolved in 0·1 ml phosphate-buffered saline (PBS) and injected intraperitoneally (0·5 μg/mouse) instead of IL-12 to DC/Hepa1-6 pretreated and untreated WT mice on day 15. The mice were sacrificed 48 h after α-GalCer administration.

Assay for serum transaminase levels

Serum alanine aminotransferase (ALT) levels were measured using the DriChem system (L3500V; Fuji Film Medical Co., Ltd, Tokyo, Japan), according to the manufacturer's instructions.

Histology

Liver tissue was fixed in 10% formalin for at least 24 h and paraffin-embedded. Sections of 2 μm thickness were stained with haematoxylin and eosin (H&E) to determine morphological changes. The numbers of inflammatory foci were determined as described previously [24].

Preparation of liver mononuclear cells

Hepatic mononuclear cells (MNCs) were isolated from murine liver, as described previously [24].

Quantitative reverse transcription–polymerase chain reaction (qRT–PCR)

Liver RNA extraction and messenger RNA (mRNA) quantification by real-time qRT–PCR were performed as described previously [25]. The expression levels of CXCL16 were normalized relative to that of glyceraldehyde-3-phosphate dehydrogenase (GAPDH).

Flow cytometry and intracellular cytokine staining

For cell-surface staining, after blocking with anti-FcR (clone 93; eBioscience), cells were incubated with various mAbs in darkness at 4°C for 30 min and examined by flow cytometry [fluorescence activated celll sorter (FACS)Calibur, BD Biosciences Immunocytometry Systems, San Jose, CA, USA]. For intracellular cytokine staining, isolated intrahepatic MNCs were stimulated with phorbol 12-myristate 13-acetate (PMA) (50 ng/ml; Sigma-Aldrich, Inc.) and ionomycin (1μg/ml; Sigma-Aldrich, Inc.) in the presence of Brefeldin A (10 μg/ml; Sigma-Aldrich, Inc.) for 5 h. After labelling the cell-surface antigens, cells were fixed and permeabilized using a Cytofix/Cytoperm plus kit (BD Biosciences) and then stained with anti-IFN-γ or IL-4 conjugated with APC. The stained cells were analysed by flow cytometry. The data were analysed using CellQuest Pro version 5·2 software (BD Biosciences Immunocytometry Systems, San Jose, CA, USA).

Isolation of hepatocytes and cytotoxicity assay

Hepatocytes were isolated as described previously [24]. A cytotoxicity assay against primary murine cultured hepatocytes was performed as described previously [24]. Briefly, isolated hepatocytes were seeded at 1 × 104 cells/well into 96-well collagen-coated plates (Iwaki, Asahi Techno Glass, Chiba, Japan). After overnight incubation, 4 × 105 effector, intrahepatic whole MNCs or CD8+ T cells or non-CD8+ T cells, which were isolated using a magnetic cell sorting system (CD8+ T Cell Isolation kit II; Miltenyi Biotec, Bergisch Gladbach, Germany), were co-cultured for 24 h. Aspartate aminotransferase (AST) activity of the culture supernatant was determined and percentage of cytotoxic activity was calculated as [(experimental AST release – spontaneous AST release)/(total AST release – spontaneous AST release) × 100], according to the formula described previously [24].

Statistical analysis

The significance of difference among the groups was analysed with Tukey's test for multiple group comparisons. Unpaired Student's t-test was used for comparison of means in two groups. Differences were considered to be significant at a P-value less than 0·05.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References
  10. Supporting Information

Accumulation of activated NK T cells producing IFN-γ in the AHI liver

As shown in Fig. 1a,i, the number of NK T cells increased significantly in the AHI liver generated by treatment of mice with DC/Hepa1-6 and IL-12. The number of NK T cells was higher in combined treatment with DC/Hepa1-6 and IL-12 than treatment with DC/Hepa1-6 or IL-12 alone (Supplementary Fig. S1A-a). The absolute number of intrahepatic CD69+ activated NK T cells was high in AHI (Fig. 1a,ii). Conversely, the number of NK T cells in the spleen was lower than in the untreated control at maximum hepatic inflammation (Fig. 1a,iii). Expression levels of CXCL16, a ligand of CXCR6, in hepatic tissue and the population of CXCR6+ intrahepatic NK T cells were elevated in AHI liver (Fig. 1b,i–iii). Although the population of IFN-γ-producing intrahepatic NK T cells increased at maximum hepatic inflammation (Fig. 1c,i,iii), the population of IL-4-producing intrahepatic NK T cells was not affected (Fig. 1c,ii,iii). These results suggest that activated NK T cells which show a Th1 phenotype might have been accumulated into the liver of AHI. Because of a vigorous increase in the number of intrahepatic T cells by generation of AHI, frequency of NK T cells in the intrahepatic MHCs was decreased in spite of an increase in absolute number of NK T cells (Supplementary Fig. S2).

figure

Figure 1. The dynamic statistics of natural killer (NK) T cells in the autoimmune hepatic inflammation (AHI) liver and spleen. (a) Absolute number of total NK T cells (i) and CD69+NK T cells (ii) in the liver and NK T cells in the spleen (iii) in AHI. The number was determined as [total number of mononuclear cells (MNCs) in the liver or spleen] × [the frequency of CD3+NK1·1+ cells or CD3+NK1·1+ CD69+cells] in each group [n = 5, mean ± standard deviation (s.d.), *P < 0·001]. (b) (i) Expression of CXCL16 in hepatic tissue. Levels of CXCL16 mRNA in each group were determined by quantitative reverse transcription–polymerase chain reaction (qRT–PCR). Bars indicate mean ± s.d., *P < 0·001. (ii) Population of intrahepatic CXCR6+NK T cells (n = 5, mean ± s.d., *P < 0·001). (iii) Representative flow cytometry of intrahepatic CXCR6+NK T cells in control and AHI. Grey-filled histograms; CXCR6+cells, dotted histogram; isotype control. The value shows the frequency of CXCR6+ cells and the numbers in parentheses shows mean fluorescence intensity (MFI) of CXCR6. (c) (i) Frequency of interferon (IFN)-γ+ NK T cells in each group (n = 5, mean ± s.d., *P < 0·001). (ii) Frequency of interleukin (IL)-4+ NK T cells in control and AHI. (iii) Representative flow cytometry of IFN-γ+ or IL-4+ intrahepatic NK T cells in each group. Grey-filled histograms; IFN-γ+ or IL-4+ cells, dotted histogram; isotype control. The value shows the frequency of IFN-γ- or IL-4-producing cells and the numbers in parentheses shows MFI of IFN-γ or IL-4. All experiments were repeated at least three times.

Download figure to PowerPoint

The activity of AHI was suppressed in the livers of CD1d–/– mice

When AHI was generated in NK T cell-deficient CD1d–/– mice, the inflammatory activity was markedly attenuated (Fig. 2a). The level of serum ALT, number of clusters of MHCs (inflammatory foci, IF) in hepatic tissue and absolute number of MNCs in the liver (Fig. 2b–d) were significantly lower in the AHI livers of CD1d–/– mice compared with those in AHI of WT mice. Although CD8+T cells were found in CD1d–/– mice, they were significantly fewer than in AHI liver of WT mice (Fig. 2e). These results demonstrate that NK T cells play a pivotal role in the establishment of AHI.

figure

Figure 2. Autoimmune hepatic inflammation (AHI) activity was attenuated in natural killer (NK) T cell-deficient CD1d–/– mice. AHI was generated in wild-type (WT) or CD1d–/– mice. (a) Histological change of AHI liver (haematoxylin and eosin staining, × 200). Arrows indicate inflammatory foci (IF). (i) Representative hepatic tissue of AHI in WT mice. (ii) Representative hepatic tissue of AHI in CD1d–/– mice. (b) Serum alanine aminotransferase (ALT) levels in WT and CD1d–/– mice (n = 5, mean ± standard deviation (s.d.), *P < 0·001). (c) Number of IF of AHI in WT and CD1d–/– mice (n = 5, mean ± s.d., *P < 0·001). (d) Absolute number of intrahepatic mononuclear cells (MNCs) in WT and CD1d–/– mice (n = 5, mean ± s.d., *P < 0·001) (E) Absolute number of intrahepatic CD8+T cells in WT and CD1d–/– mice (n = 5, mean ± s.d., *P < 0·001). All experiments were repeated at least three times.

Download figure to PowerPoint

α-GalCer treatment of DC/Hepa1-6 immunized mice produced severe hepatic inflammation

Hepatic inflammation was induced when mice immunized with DC/Hepa1-6 were treated with α-GalCer instead of IL-12 (Fig. 3a,i). This was characterized by the emergence of abundant IF in hepatic lobules, similar to the AHI generated by treatment with DC/Hepa1-6 and IL-12. Importantly, the hepatic inflammatory activity generated by treatment with DC/Hepa1-6 and α-GalCer was greater than that induced by treatment with α-GalCer alone (Fig. 3a,i,ii,b,c). The population of IFN-γ-producing NK T cells were enhanced but the population of IL-4-producing NK T cells was not affected in hepatic inflammation induced by treatment with α-GalCer alone or DC/Hepa1-6 and α-GalCer (Fig. 3d,i,ii).

figure

Figure 3. Treatment of dendritic cell (DC)/Hepa1-6 pre-immunized mice with α-galactosylceramide (α-GalCer) generated marked hepatic inflammation with induction of interferon (IFN)-γ-producing natural killer (NK) T cells. (a) Histological changes in the liver (haematoxylin and eosin staining, × 100). Arrows indicate inflammatory foci (IF). (i) Hepatic tissue of DC/Hepa1-6 pre-immunized mice treated with α-GalCer. (ii) Hepatic tissue of mice treated with α-GalCer alone. (b) Serum alanine aminotransferase (ALT) levels of untreated control mice, mice treated with α-GalCer alone and DC/Hepa1-6 pre-immunized mice treated with α-GalCer [n = 5, mean ± standard deviation (s.d.), *P < 0·001]. (c) Numbers of IF in each group (n = 5, mean ± s.d., *P < 0·001). (d) (i) Frequency of IFN-γ+ intrahepatic NK T cells in each group (n = 5, mean ± s.d., *P < 0·001). (ii) Frequency of IL-4+ intrahepatic NK T cells in each group. All experiments were repeated at least three times.

Download figure to PowerPoint

Autoreactive CD8+ CTLs cytocidal to hepatocytes were induced by treatment of mice with DC/Hepa1-6 and α-GalCer

When DC/Hepa1-6 immunized mice were treated with α-GalCer and hepatic inflammation was generated, the population of CD8+ T cells in intrahepatic MNCs was increased (Fig. 4a,i). Among these CD8+ T cells, the percentage of CD62L active cells was increased (Fig. 4a,ii,iii). Notably, CD8+ T cells from the livers of mice treated with DC/Hepa1-6 and α-GalCer showed high cytotoxic activity against hepatocytes, while the cytotoxic activity of intrahepatic CD8+ T cells of mice treated with α-GalCer alone was lower (Fig. 4b). These results indicate that treatment of mice with DC/Hepa1-6 and α-GalCer could induce CTLs cytocidal to autologous hepatocytes, as for the treatment of mice with DC/Hepa1-6 and IL-12.

figure

Figure 4. Treatment of dendritic cell (DC)/Hepa1-6 pre-immunized mice with α-galactosylceramide (α-GalCer) generated active CD8+ T cells which were cytocidal to hepatocytes. (a) (i) Population of intrahepatic CD8+ T cells in each group. The mice were sacrificed 48 h after α-GalCer administration (17 days after the first DC/Hepa1-6 treatment). Intrahepatic mononuclear cells (MNCs) were isolated and the percentage of CD8+ T cells was determined by flow cytometry [n = 5, mean ± standard deviation (s.d.), *P < 0·001]. (ii) Representative flow cytometry of each group. Intrahepatic MNCs were isolated and analysed by flow cytometry using anti-CD8 and anti-CD62L monoclonal antibodies (mAbs). Upper left; control, upper right; DC/Hepa1-6, lower left; α-GalCer only, lower right; DC/Hepa1-6 + α-GalCer. Values in the upper and lower right indicate the frequency of cells in each area. (iii) Frequency of CD8+ CD62L activated T cells in total intrahepatic CD8+ T cells of each group (n = 5, mean ± s.d., *P < 0·001). (b) Cytotoxic activity of intrahepatic whole MNCs, CD8+ T cells and non-CD8+ T cells to autologous hepatocytes. Cytotoxic activity to primary cultured autologous hepatocytes was examined using aspartate aminotransferase (AST) release assay. Effector : target ratio was 40:1. All experiments were repeated at least three times.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References
  10. Supporting Information

AHI is generated in mice by induction of hepatocyte-responsive CTLs by immunization with DC/Hepa1-6 and accumulation of these autoreactive CTLs to the liver by enhanced expression of adhesion molecules, chemokine ligands and MHC in hepatic tissue by IFN-γ induced by IL-12 [24]. As IFN-γ makes hepatocyte vulnerable to CD8+CTL attack by up-regulation of MHC class I expression [3], IFN-γ is a key cytokine to generate AHI in this mouse model [24]. It is probable that activated NK T cells, accumulated into the liver upon onset of AHI, may contribute to the generation of the AHI as producers of IFN-γ. Because IL-12 is a potent IFN-γ inducer [26], it may stimulate various immune cell types to produce IFN-γ. Thus, the involvement of NK T cells as IFN-γ producers in the generation of AHI has not been clear. When CD1d–/– mice, which show NK T cell deficiency, were immunized with DC/Hepa1-6 and treated with IL-12, the resultant AHI was suppressed significantly, indicating that IFN-γ produced by NK T cells has an important role in development of AHI. Although the number of CD8+ T cells in the AHI liver was decreased significantly in CD1d–/– mice, CD8+ T cells existed in infiltrating cells of the AHI liver of CD1d–/– mice. This result suggests that IFN-γ produced by IL-12-activated immune cells other than NK T cells might have stimulated the CD8+T cell response.

The involvement of NK T cells in the generation of AHI was studied further by treatment of DC/Hepa1-6-immunized mice with α-GalCer. Because treatment with α-GalCer itself causes hepatic injury [27], the minimum dose of α-GalCer that induced hepatic injury (0·5 μg/mouse) was used to analyse the combined effect with DC/Hepa1-6 immunization and α-GalCer on generation of hepatic inflammation. The activity of hepatic inflammation was significantly higher in mice treated with DC/Hepa1-6 and α-GalCer than α-GalCer alone. CD8+ T cells from mice treated with DC/Hepa1-6 plus α-GalCer showed higher percentage of CD62 cells and significant cytotoxicity against primary cultured hepatocytes, but CD8+ T cells from mice treated with α-GalCer alone showed low cytotoxicity. These results suggest strongly that antigen-specific activated T cells capable of killing hepatocytes, which were induced by immunization with DC/Hepa1-6, had accumulated in the liver following modulation of the hepatic microenvironment by IFN-γ secreted from α-GalCer-activated NK T cells. Although infiltration of CD8+ T cells was seen in the liver of mice treated with DC/Hepa1-6 alone, inflammatory activity of the liver in such mice was somewhat low, as reported previously [24]. Accordingly, because of few numbers of intrahepatic CD8+ T cells of mice treated with DC/Hepa1-6 alone, it was extremely hard to collect them and examine their cytotoxic activity. Cytotoxic activity of CD8+ T cells in the AHI liver of CD1d–/– mice is important for interpretation of the role of NK T cells. According to our previous data, splenic CD8+ T cells from sole DC/Hepa1-6 immunized mice without IL-12 treatment could elicit significant cytotoxic activity to autologous hepatocytes in vitro. However, without the effect of IFN-γ, which was provided by systemic IL-12 treatment or activation of intrahepatic NK T cells, to increase the expression of MHC class I or several adhesion molecules in hepatic tissue, such CD8+ cytotoxic T cells could not recruit into the liver. Thus, it might be probable that CD8+ T cells in the AHI liver of CD1d–/– mice could show cytotoxic activity to hepatocytes, but because of lack of IFN-γ effect provided by activated NK T cells and less infiltration of CD8+CTLs into the liver, hepatic inflammatory activity was reduced in CD1d–/– mice.

Non-CD8+ T cells among the hepatic MHCs showed considerable cytotoxic activity to hepatocytes, although the activity was lower than that of CD8+ T cells. Under Th1 conditions rich in IFN-γ, activated intrahepatic bystander cells such as macrophages and NK or NK T cells might elicit non-specific cytotoxic activity to hepatocytes [28]. As intrahepatic NK T cells express FasL [29], it cannot be excluded that activated NK T cells might elicit direct hepatocyte injury, possibly through Fas-FasL interaction, and this point should be studied further.

Concanavalin A (ConA) hepatitis is used widely as a mouse model of immune-related hepatitis, although the specific autoimmune response to hepatocytes has not been defined. Several studies have shown convincingly the involvement of NK T cells in the pathology of ConA hepatitis. Depletion of hepatic NK T cells in ConA-administered mice, or treatment of NK T cell-deficient mice with ConA, reduced the activity of ConA hepatitis significantly [30]. Moreover, in the interaction between CD8+ T cells responsive to ovalbumin in a MHC class I-restricted manner (OTI T cells) and hepatocytes expressing transferrin-membrane-bound ovalbumin (Tf-mO transferrin-mOVA) in Tf-mOVA mice, specific effector function to antigen was stimulated by co-activation of Vα14 NK T cells using α-GalCer [31]. Also, in an animal model of primary biliary cirrhosis, in which infection of mice with Novosphingobium aromaticivorans induces antibody against mitochondrial component and T cell-mediated autoimmunity, disease induction requires NK T cells [32]. These results suggest that NK T cells play pivotal roles for development of immune-related liver disease.

Our AHI model seems to be an artificial model generated by extreme treatment of mice. However, we demonstrated that in order to generate liver-specific autoimmune response, two independent factors, induction of hepatocyte-responsive CD8+ T cells by immunization with DC/Hepa1-6 and recruitment of such CD8+ T cells into the liver by modulation of hepatic environment with IFN-γ, were required. In humans, the similar mechanism for development of autoimmune hepatitis might be considered. Prior to onset of human autoimmune hepatitis, infection of hepatitis virus or drug-induced liver injury might contribute to induction of hepatocyte-responsive T lymphocytes just as does immunization with DC/Hepa1-6. Some promoting factors such as massive cytokine production from NK T cells for activation of these autoreactive T cells might be associated with onset of autoimmune hepatitis.

Although this study showed an important role for activated NK T cells in the generation of AHI, the implication of NK T cell activation in patients with autoimmune hepatitis remains obscure. Several potential physiological ligands for NK T cells have been reported [33, 34], but the specific endogenous ligand for NK T cell activation remains unknown [35]. Recently, several reports suggested the involvement of Toll-like receptors (TLRs) in the pathogenesis of autoimmune diabetes, inflammatory bowel diseases, multiple sclerosis and systemic lupus erythematosus [36, 37]. It may be possible that the inflammatory activity in human AIH is affected by activation of IFN-γ-producing NK T cells in the liver through TLR stimulation by intestinal microbial components. If so, the regulation of intrahepatic NK T cell activity might lead to the establishment of a new modality for controlling disease activity in human AIH.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References
  10. Supporting Information

This work was supported by Health Labor Science Research on Measures for intractable disease (Intractable Hepato-Biliary Disease Study Group) founded by the Ministry of Health, Labor and welfare of Japan.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
cei4664-sup-0001- fs1.TIF74K

Fig. S1. The dynamic statistics of natural killer (NK) T cells in the autoimmune hepatic inflammation (AHI) liver and spleen. (A) Absolute number of NK T cells in the liver (a) and spleen (b) in AHI. The number was determined as [total number of mononuclear cells (MNCs) in the liver or spleen] × [the frequency of CD3+NK1·1+ cells] in each group [n = 5, mean ± standard deviation (s.d.), *P < 0·001]. (B) (a) Population of intrahepatic CXCR6+ NK T cells (n = 5, mean ± s.d., *P < 0·001). (b) Expression of CXCL16 in hepatic tissue. Levels of CXCL16 mRNA in each group were determined by quantitative reverse transcription–polymerase chain reaction (qRT–PCR). Bars indicate mean ± s.d., *P < 0·001. (C) (a) Frequency of interferon (IFN)-γ+ NK T cells in each group (n = 5, mean ± s.d., *P < 0·001). (b) Frequency of interleukin (IL)-4+ NK T cells in each group. All experiments were repeated at least three times.

cei4664-sup-0002-fs2.TIF65K

Fig. S2. Frequency of natural killer (NK) T cells in the liver and spleen in autoimmune hepatic inflammation (AHI). Frequencies of NK T cells in intrahepatic major histocompatibility complexes (MHCs) and splenocytes were determined by flow cytometry (n = 5, mean ± standard deviation, *P < 0·001). Experiments were repeated at least three times.

Please note: Wiley Blackwell 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.