Potential conflict of interest: Nothing to report.
Liver fibrosis is a common scarring response to all forms of chronic liver injury and is always associated with inflammation that contributes to fibrogenesis. Although a variety of cell populations infiltrate the liver during inflammation, it is generically clear that CD8 T lymphocytes promote while natural killer (NK) cells inhibit liver fibrosis. However, the role of invariant natural killer T (iNKT) cells, which are abundant in the liver, in hepatic fibrogenesis, remains obscure. Here we show that iNKT-deficient mice are more susceptible to carbon tetrachloride (CCl4)-induced acute liver injury and inflammation. The protective effect of naturally activated iNKT in this model is likely mediated via suppression of the proinflammatory effect of activated hepatic stellate cells. Interestingly, strong activation of iNKT through injection of iNKT activator α-galactosylceramide (α-GalCer) accelerates CCl4-induced acute liver injury and fibrosis. In contrast, chronic CCl4 administration induces a similar degree of liver injury in iNKT-deficient and wild-type mice, and only a slightly higher grade of liver fibrosis in iNKT-deficient mice than wild-type mice 2 weeks but not 4 weeks after CCl4 injection, although iNKT cells are able to kill activated stellate cells. An insignificant role of iNKT in chronic liver injury and fibrosis may be attributable to hepatic iNKT cell depletion. Finally, chronic α-GalCer treatment had little effect on liver injury and fibrosis, which is attributable to iNKT tolerance after α-GalCer injection. Conclusion: Natural activation of hepatic iNKT cells inhibits, whereas strong activation of iNKT cells by α-GalCer accelerates CCl4-induced acute liver injury, inflammation, and fibrosis. During chronic liver injury, hepatic iNKT cells are depleted and play a role in inhibiting liver fibrosis in the early stage but not the late stage of fibrosis. (HEPATOLOGY 2009.)
Worldwide, alcohol drinking, hepatitis viral infection, and nonalcoholic steatohepatitis are the three major causes of chronic liver inflammation and injury, leading to liver fibrosis, cirrhosis, and hepatocellular carcinoma. Liver fibrosis is characterized by an accumulation of extracellular matrix proteins, which are mainly produced by activated hepatic stellate cells (HSCs).1–5 Increasing evidence suggests that the interaction of HSCs with inflammatory cells that are always associated with liver fibrosis plays an important role in fibrogenesis.1–5 Most notably, CD8 T cells have been shown to promote liver fibrosis via activation of HSCs,6 whereas natural killer (NK) cells inhibit liver fibrosis through killing of activated HSCs.7–10 However, the role of invariant natural killer T (iNKT) cells, which are abundant in the liver, in hepatic fibrogenesis is not clear.
NKT cells are a heterogeneous population of T lymphocytes that express markers of NK cells and T cell receptors.11, 12 These cells recognize endogenous lipid antigen isoglobotriaosylceramide and exogenous lipid antigens such as α-galactosylceramide (α-GalCer) by the nonclassical major histocompatibility class I–like molecule CD1.11, 12 CD1-dependent NKT cells can be broadly categorized into type I and type II NKT cells. Type I NKT cells, also known as “classical” NKT cells, iNKT cells, and Vα14NKT cells express the semi-invariant αβ T cell receptors encoded by Vα14 and Jα18 paired with a set of Vβ chains. Type I iNKT cells, which make up 90% to 95% of total NKT cells and recognize α-GalCer, are not detected in either Jα18−/− (iNKT-deficient) or CD1d−/− mice.11, 12 Type II NKT cells, also known as “nonclassical” NKT cells, express diverse T cell receptors and recognize sulfatide, but not α-GalCer. Type II NKT cells make up less than 5% of total NKT cells and are not detected in CD1d−/− mice, but can be detected in Jα18−/− mice.11, 12
Liver lymphocytes are abundant in iNKT cells.13–17 For example, mouse liver lymphocytes contain approximately 30%–40% NKT cells, whereas peripheral blood lymphocytes contain less than 5% NKT cells.13, 14 Activation of iNKT cells by concanavalin A or α-GalCer induces acute hepatitis,17–19 suggesting that iNKT cell activation contributes to acute liver injury. Increasing evidence suggests that iNKT cells also contribute to the pathogenesis of a variety of liver disorders,16 including viral hepatitis,20, 21 alcoholic liver injury,22 primary biliary cirrhosis,23, 24 bile duct ligation–induced liver injury,25 and drug-induced liver injury.26, 27 However, the role of iNKT cells in chronic liver inflammation and fibrosis remains poorly understood.28 In this study, we investigated extensively the role of iNKT cells in hepatic inflammation, injury, and fibrosis induced by carbon tetrachloride (CCl4). Our findings suggest that natural activation of iNKT cells by endogenous lipid antigens plays a protective role, whereas strong activation of iNKT cells by exogenous lipid antigen α-GalCer plays a detrimental role in CCl4-induced acute liver injury, inflammation, and fibrosis. Moreover, chronic administration of CCl4 depletes hepatic iNKT and induces a higher grade of liver fibrosis in Jα18−/− (iNKT-deficient) mice than wild-type mice 2 weeks after administration, but a similar grade of liver fibrosis 4 weeks after injection. Interestingly, repeated treatment with α-GalCer had little effect on CCl4-induced chronic liver injury and fibrosis, which may be attributable to iNKT tolerance.
α-GalCer, α-galactosylceramide; α-SMA, alpha-smooth muscle actin; ALT, alanine aminotransferase; CCl4, carbon tetrachloride; FACS, fluorescence-activated cell sorting; HSC, hepatic stellate cell; IFN-γ, interferon gamma; IL, interleukin; iNKT, invariant natural killer T cells; MCP-1, monocyte chemotactic protein-1; NK, natural killer; NKG2D, natural-killer group 2, member D; NKT, natural killer T; PCR, polymerase chain reaction; SD, standard deviation; STAT1, signal transducer and activator of transcription protein 1; TNF-α, tumor necrosis factor alpha; TUNEL, terminal deoxynucleotidyl transferase-mediated nick-end labeling.
Materials and Methods
iNKT-deficient (Jα18−/−) mice on C57BL6 background were provided by Dr. Rachel Caspi (National Eye Institute [NEI], National Institutes of Health) with permission from Dr. Taniguchi (RIKEN Research Center for Allergy and Immunology, Japan). Mice lacking the Jα18 gene segment are devoid of Vα14 iNKT cells, but other lymphoid cell lineages are intact.29 Mice deficient in interferon-gamma (IFN-γ−/−) on C57BL6 background were purchased from the Jackson Laboratory (Bar Harbor, ME). Signal transducer and activator of transcription protein 1 (STAT1)-deficient mice (STAT1−/−) on C57BL6 background were described previously.30 All male mice were used in the current study and were housed in a specific pathogen-free facility and were cared for in accordance with National Institutes of Health guidelines and approved by the The National Institute on Alcohol Abuse and Alcoholism animal care and use committee.
Liver Injury Induced by CCl4.
For acute liver injury induced by CCl4, mice were injected intraperitoneally with a single dose of CCl4 (10% in olive oil, 2 mL/kg). For chronic liver injury, mice were injected intraperitoneally with CCl4 (10% in olive oil, 2 mL/kg, 3 times/week) for 2 or 4 weeks. Control groups were treated with vehicle (2 mL/kg of olive oil). After mice were sacrificed, liver tissues were frozen in liquid nitrogen or fixed in 10% buffered formalin and embedded in paraffin. There was no mortality in wild-type and Jα18−/− mice after acute or chronic CCl4 treatment.
A stock solution of α-GalCer (Alexis Biochemicals Corp., San Diego, CA) was diluted to 0.2 mg/mL in 0.5% polysorbate-20 and stored at −20°C. Mice were treated acutely with α-GalCer (2 μg/200 μL in phosphate-buffered saline per mouse) by intraperitoneal injection 3 hours before CCl4 administration. Chronic administration of α-GalCer (2 μg/200 μL in phosphate-buffered saline per mouse) was carried out by intraperitoneal injection once or twice per week. There was no mortality in mice treated with CCl4+α-GalCer.
The following methods are described in the supporting materials. Histology and immunohistochemistry, terminal deoxynucleotidyl transferase-mediated nick-end labeling (TUNEL) assay, Western blotting, real-time polymerase chain reaction (PCR), measurement of serum alanine aminotransferase (ALT) and serum cytokines, isolation and in vitro culture of HSC and Kupffer cells, isolation of mouse liver lymphocytes, liver NKT cells, and flow cytometric analysis, and cytotoxicity of liver lymphocytes and NKT cells against HSCs.
Data are expressed as means ± standard deviation (SD). To compare values obtained from three or more groups, one-factor analysis of variance was used, followed by Tukey's post hoc test. To compare values obtained from two groups, the Student t test was performed. Statistical significance was taken at the P < 0.05 level.
iNKT-Deficient (Jα18−/−) Mice Are More Susceptible to Acute Liver Injury and Inflammation Induced by CCl4.
After CCl4 treatment, the total number of mononuclear cells in the liver increased, with peak effect occurring 12 hours after administration. A similar increase was also observed in Jα18−/− mice (Fig. 1A). FACS analyses in Fig. 1B show that CCl4 treatment significantly decreased the percentage of NKT cells in liver lymphocytes. The total number of NKT cells increased slightly at 12 hours and then decreased significantly 24 hours after CCl4 treatment (Fig. 1C). Vehicle injection only caused a slight decrease in NKT cells (Fig. 1B). As expected, the number of NKT cells was very low in Jα18−/− mouse livers (Fig. 1C).
To determine whether down-regulation of hepatic NKT cells after CCl4 treatment is attributable to NKT cell death or loss of NKT markers, iNKT cell apoptosis was examined. As shown in Fig. 1D, hepatic iNKT cell apoptosis increased after injection of vehicle or CCl4, but was much higher in the CCl4 group than in the vehicle group. Moreover, expression of activation marker CD69 increased on hepatic NKT cells 12 hours after CCl4 treatment compared with vehicle group (Fig. 1E), suggesting that hepatic NKT cells are activated after CCl4 treatment.
Shown in Fig. 1F, serum ALT levels were much higher in Jα18−/− mice than in wild-type mice 12 hours after CCl4 treatment, but were comparable in both groups at 24 hours. TUNEL analyses showed that the number of apoptotic hepatocytes was greater in Jα18−/− mice than in wild-type mice 12 hours after CCl4 administration, but no difference was observed between these 2 groups at 24 hours (Fig. 1G and Supporting Fig. 1).
To examine hepatic inflammation after acute CCl4 injection, we measured the infiltration of neutrophils and monocytes into the liver by fluorescence-activated cell sorting (FACS) analyses of Gr-1 expression. It was reported that Gr-1high cells mainly represent neutrophils whereas Gr-1intermediate cells represent monocytes and eosinophils.31 As shown in Fig. 2A and 2B, infiltration of Gr-1high neutrophils increased after CCl4 treatment, which was higher in Jα18−/− mice than in wild-type mice 12 hours after injection but was comparable 24 hours after injection. CCl4 treatment also induced infiltration of Gr-1int monocytes, but such infiltration was lower in Jα18−/− mice compared with wild-type mice 12 hours after injection. Furthermore, immunohistochemistry staining confirmed the neutrophilic (myeloperoxidase-positive cells) infiltration after CCl4 treatment, which was significantly higher in Jα18−/− mice than in wild-type mice (Fig. 2C and Supporting Fig. 2A). Moreover, expression of hepatic chemokine (C-C motif) receptor 2 and CD68 (markers of monocytes/macrophages) was induced by CCl4 treatment; however, such induction was less evident in Jα18−/− mice than in wild-type mice (Fig. 2D), which is consistent with the findings in Fig. 2A showing that the number of Gr-1int monocytes was lower in Jα18−/− mice than in wild-type mice.
Furthermore, CCl4 treatment elevated serum and hepatic tumor necrosis factor alpha (TNF-α) and monocyte chemotactic protein-1 (MCP-1), such elevation was higher in Jα18−/− mice than in wild-type mice (Fig. 2E-F). In addition, CCl4-mediated induction of serum and hepatic levels of interleukin (IL) 6 was comparable between Jα18−/− and wild-type mice (Supporting Fig. 3). Hepatic levels of IL-4, IL-10, and IL-13 remained unchanged after CCl4 treatment, and were comparable between Jα18−/− and wild-type mice (Supporting Fig. 3).
CCl4 Metabolism Is Comparable Between Wild-Type and Jα18−/−Mice.
Because p450 CYP2E1-mediated CCl4 metabolism plays a key role in CCl4-induced liver injury, we wondered whether acceleration of liver injury in Jα18−/− mice was attributable to alterations in CCl4 metabolism in these mice. Expression of CYP2E1 was comparable in the livers from wild-type and Jα18−/− mice (Supporting Fig. 4). After the CCl4 challenge, expression of CYP2E1 decreased significantly in wild-type mice, as shown by western blot and immunohistochemical analyses (Supporting Fig. 4). A similar down-regulation was also observed in Jα18−/− mice, suggesting that CCl4 metabolism is similar in wild-type and Jα18−/− mice (Supporting Fig. 4).
HSCs from CCl4-Treated Jα18−/− Mice Produce Greater Levels of Proinflammatory Cytokines Than Those from Wild-Type Mice.
To understand why serum and hepatic cytokines were higher in Jα18−/− mice than in wild-type mice after CCl4 treatment, Kupffer cells and HSCs from these mice were isolated and cultured. As shown in Fig. 3A and 3B, HSCs from CCl4-treated Jα18−/− mice produced greater levels of TNF-α, IL-6, and MCP-1, and expressed higher levels of alpha-smooth muscle actin (α-SMA) and tissue inhibitor of metalloproteinase 1 but not transforming growth factor alpha compared with those from CCl4-treated wild-type mice. Kupffer cells from CCl4-treated Jα18−/− mice also produced greater levels of TNF-α, IL-6, and MCP-1 than those from CCl4-treated wild-type mice (Fig. 3C). Interestingly, the basal levels of TNF-α and IL-6 production by Kupffer cells were higher in Jα18−/− mice compared with wild-type mice (Fig. 3C). In contrast, production of IL-12, IFN-γ, and IL-10 by HSCs or Kupffer cells was comparable between Jα18−/− and wild-type mice (data not shown).
The findings that production of higher levels of cytokines by HSCs in Jα18−/− mice after CCl4 treatment suggest that iNKT cells may inhibit HSC activation. Previous studies have shown that NK cells were able to kill HSCs.7-9 This led us to test the hypothesis that NKT cells also may be able to kill HSCs. Shown in Fig. 3D, D0-HSCs were resistant to the cytotoxicity of liver lymphocytes (less than 1% cytotoxicity was observed), whereas D4-HSCs were susceptible to such killing. Liver lymphocytes from wild-type mice demonstrated 10% cytotoxicity against D4-HSCs but only 5% from corresponding liver lymphocytes from Jα18−/− mice that are devoid of iNKT cells, suggesting that NKT cells play an important role in killing activated HSCs (D4-HSCs). The cytotoxicity of liver lymphocytes against D4-HSCs was enhanced after acute α-GalCer treatment (Fig. 3E). Furthermore, purified liver NKT cells were able to kill D4-HSCs, which was significantly attenuated by treatment with an anti–natural-killer group 2, member D (NKG2D) antibody (Fig. 3F).
Activation of iNKT Cells by a Single Dose of α-GalCer Synergistically Enhances CCl4-Induced Acute Liver Injury and Fibrosis: Dependent on IFN-γ/STAT1.
It has been reported that activation of iNKT cells by α-GalCer, an iNKT activator, results in mild liver injury.19 Here we examined the effects of α-GalCer on CCl4-induced liver injury. As shown in Fig. 4A, at the 12-hour time point, treatment with α-GalCer or CCl4 alone yielded mild elevations in serum ALT levels to 200 IU/L and 1200 IU/L, respectively, whereas cotreatment with α-GalCer and CCl4 elevated synergistically serum ALT levels up to 6000 IU/L. Such synergistic effect was not observed in Jα18−/− mice. At the 24-hour time point, treatment with α-GalCer did not further enhance CCl4 -induced elevation of serum ALT levels in wild-type and Jα18−/− mice. Moreover, α-GalCer pretreatment enhanced significantly liver fibrosis 72 hours after CCl4 injection, as demonstrated by enhancing α-SMA immunostaining and messenger RNA in the liver (Fig. 4B).
To understand the mechanisms underlying α-GalCer acceleration of CCl4-induced acute liver injury, serum cytokines were measured. Treatment with α-GalCer elevated a variety of serum cytokines, including TNF-α, MCP-1, IFN-γ, and IL-6 (Fig. 4C). Injection with CCl4 by itself elevated TNF-α, MCP-1, and IL-6. Cotreatment with CCl4 and α-GalCer induced synergistically elevation of TNF-α but not other cytokines. Because α-GalCer injection induced high levels of serum IFN-γ, we wondered whether IFN-γ and its downstream signal STAT1 contributed to α-GalCer acceleration of CCl4-induced liver injury. As shown in Fig. 4D, CCl4 treatment induced a similar grade of liver injury in wild-type, IFN-γ−/−, and STAT1−/− mice. α-GalCer treatment enhanced synergistically CCl4-mediated liver injury in wild-type mice but not in IFN-γ−/− and STAT1−/− mice. α-GalCer injection alone induced mild liver injury (ALT reached 300 IU/L) in wild-type mice but not in IFN-γ−/− and STAT1−/− mice (data not shown). Moreover, induction of TNF-α, MCP-1, but not IL-6, by α-GalCer was diminished in IFN-γ−/− mice (Fig. 4E). Finally, α-GalCer treatment significantly increased the number of hepatocyte apoptosis in CCl4-treated mice, which was partially diminished in IFN-γ−/− mice (Fig. 4F).
The data showed that acute treatment with CCl4 resulted in iNKT depletion in the liver. Next we examined the effects of chronic CCl4 treatment on hepatic NKT cells. As shown in Fig. 5A, normal C57BL6 mouse liver contains approximately 37% NK1.1+CD3+ and 32% CD3+CD1d/αGalCer+ cells. In the livers of 2-week or 4-week CCl4-treated mice, the percentage of NKT (NK1.1+CD3+ or CD3+CD1d/αGalCer+) cells decreased significantly. Interestingly, vehicle injection also slightly reduced the percentage of NKT cells but increased the total number of NKT cells 2 weeks after injection (Figs. 5A, B). The total number of NKT cells in the liver was markedly decreased 2 and 4 weeks after CCl4 injection (Fig. 5B). In contrast, vehicle and CCl4 both increased the total number of NK cells (Fig. 5B). Figure 5C shows that the percentage of apoptotic liver NKT cells increased significantly from 2-week or 4-week CCl4-treated mice, suggesting that NKT depletion was attributable to apoptosis after chronic CCl4 treatment. Finally, Fig. 5D shows that the percentage of NKT (NK1.1+CD3+) in the spleen increased slightly after chronic CCl4 treatment.
iNKT Cells Play a Role in Inhibiting Liver Fibrosis in the Early Stage But Not Late Stage of Liver Fibrosis Induced by Chronic CCl4 Treatment.
The role of iNKT cells in CCl4-induced chronic liver injury and fibrosis was examined in wild-type and Jα18−/− mice. As shown in Fig. 6A, serum levels of ALT were similar in Jα18−/− and wild-type mice 2 and 4 weeks after CCl4 injection. Figure 6B shows that chronic CCl4 injection induced slightly higher levels of collagen deposition (Sirius red staining) and HSC activation (α-SMA staining) in Jα18−/− mice than in wild-type mice 2 weeks after injection but a similar grade of liver fibrosis 4 weeks after injection. Serum levels of TNF-α, MCP-1, and IL-6, as well as hepatic levels of TNF-α, MCP-1, IL-6, IL-12, and chemokine (C-C motif) receptor 2 were elevated similarly between these two groups after CCl4 treatment (data not shown).
Chronic Treatment with α-GalCer Has Little Effect on CCl4-Induced Chronic Liver Injury and Fibrosis.
The effects of chronic treatment with the iNKT activator α-GalCer on CCl4-induced liver injury and fibrosis are shown in Fig. 7. Surprisingly, serum levels of ALT, liver fibrosis grade, and serum cytokine levels (in other words, IL-6, MCP-1, and TNF-α) were comparable between groups treated with CCl4 alone or with CCl4 plus α-GalCer injection (Fig. 7A-D). Serum levels of IL-4 and IL-10 were under the detectable limit in these groups. Hepatic expression of Th2 cytokines such as IL-4, IL-10, and IL-13 was comparable between CCl4 and CCl4 plus α-GalCer group (Supporting Fig. 5). Chronic α-GalCer injection alone had little effect on liver injury and fibrosis (data not shown).
In this paper, we extensively investigated the role of iNKT cells in acute and chronic liver injury, inflammation, and fibrosis. Our findings indicate that (1) natural activation of iNKT cells inhibits CCl4-induced acute liver injury, whereas strong iNKT activation by the exogenous ligand α-GalCer accelerates CCl4-induced acute liver injury and fibrosis; (2) acute and chronic CCl4 treatment induces hepatic iNKT cell depletion; (3) iNKT cells play a role in inhibiting liver fibrosis at the early stage but not late stage; (4) repeated injection of exogenous iNKT ligand α-GalCer has little effect on chronic CCl4-induced liver injury and fibrosis. We have integrated these findings into a model (summarized in Fig. 8) depicting the complex role of iNKT in CCl4-induced acute and chronic liver injury, inflammation, and fibrosis.
Natural Activation of iNKT Cells Inhibits, Whereas Strong Activation of iNKT Cells by α-GalCer Accelerates CCl4-Induced Acute Liver Injury and Inflammation.
Although Jα18−/− mice are resistant to concanavalin A-induced and α-GalCer–induced acute liver injury and inflammation,17–19 we showed that Jα18−/− mice were more susceptible to CCl4-induced liver injury and inflammation compared with wild-type mice. This suggests that natural activation of iNKT cells plays an anti-inflammatory role in CCl4-induced liver injury. In contrast, strong activation of iNKT by α-GalCer markedly accelerated CCl4-induced acute liver injury. Reasons for the existence of opposing roles by naturally activated iNKT and the strongly activated iNKT by α-GalCer in CCl4-induced liver injury is not fully understood. We speculate that natural activation of iNKT after CCl4 treatment occurs locally and weakly in the liver, thereby inhibiting inflammation, whereas, in contrast, α-GalCer–mediated iNKT activation is systemic and strong, resulting in the stimulation of inflammatory responses.
As shown in Fig. 1, expression of CD69, an activation marker, is elevated on liver NKT cells after CCl4 treatment, suggesting that iNKT cells in the liver are activated during CCl4-induced acute liver injury. Additionally, the fact that the number of hepatic iNKT cells decreased after acute CCl4 injection also indirectly suggests activation of iNKT cells because iNKT cells die after activation (a typical activation-induced death).32, 33 Moreover, serum levels of IFN-γ (a major cytokine produced by activated iNKT cells) were elevated only slightly after CCl4 injection (data not shown), indicating that iNKT cell activation after CCl4 treatment may occur weakly in the liver. Moreover, the number of infiltrated neutrophils was significantly higher after acute CCl4 treatment in Jα18−/− mice than in wild-type mice. Taken together, these findings suggest that iNKT cells are activated and play an important role in inhibiting neutrophil infiltration in acute CCl4-induced liver injury. Interestingly, the anti-neutrophil inflammatory response of iNKT cells was also recently reported in another mouse model of cholestatic liver injury.34 However, the mechanisms by which iNKT is naturally activated after CCl4 injection and contributes to antiinflammatory effects are not clear. It has been reported that HSCs are liver-resident antigen-presenting cells that can present lipid antigens to induce iNKT cell activation.35 Thus, hepatic iNKT activation could be caused by HSC-presenting lipid antigens released from damaged hepatocytes after CCl4 treatment. Furthermore, we provide evidence suggesting that the antiinflammatory effect of natural activation of iNKT after CCl4 injection is mediated, at least in part, by inhibition of HSC activation. First, HSCs from Jα18−/− mice produce greater TNF-α and IL-6 than wild-type mice during CCl4-induced liver injury, suggesting that iNKT deficiency increases the proinflammatory effect of HSCs. Second, in vitro cytotoxicity assays showed that iNKT cells can directly kill early-activated HSCs, but not quiescent HSCs, through an NKG2D-dependent mechanism, similar to NK cell killing of activated HSCs.7 Third, iNKT cells may inhibit HSC activation through production of IFN-γ; a cytokine that has been shown to inhibit HSC proliferation and activation.30, 36 Last, activated HSCs have been shown to participate in liver inflammation.4, 37, 38
In contrast to weak natural iNKT cell activation, injection of α-GalCer caused strong and systemic iNKT activation as evidenced by markedly elevated serum cytokines, including IFN-γ (Fig. 4). Further studies suggest that elevation of IFN-γ contributes to α-GalCer acceleration of CCl4-induced acute liver injury because the acceleration was completely abolished in IFN-γ −/− mice. Because IFN-γ is able to induce hepatocyte apoptosis via a STAT1-dependent mechanism,39 thus it is plausible that IFN-γ production after α-GalCer treatment can increase the susceptibility of hepatocyte apoptosis during CCl4-induced liver injury. Indeed, the number of apoptotic hepatocytes was much greater in the α-GalCer plus CCl4 group than in the group treated with CCl4 alone.
iNKT Cells Are Depleted and Play a Minor Role in CCl4-Induced Chronic Liver Injury and Inflammation.
Although CCl4-induced acute liver injury was accelerated in Jα18−/− mice compared with wild-type mice, CCl4-induced chronic liver injury and inflammation were comparable between these two groups, suggesting that iNKT cells play a minor role in chronic liver injury in this model. This may occur because hepatic iNKT cells were depleted during chronic CCl4 treatment. The mechanism underlying iNKT cell depletion during CCl4-induced liver injury remains obscure. It was reported that endoplasmic reticulum stress decreases CD1d protein expression on hepatocytes, resulting in down-regulation of NKT cells in murine fatty livers.40 Thus, the endoplasmic reticulum stress caused by CCl4 injection also may contribute to hepatic NKT cell depletion during CCl4-induced liver injury. Moreover, depletion of hepatic NKT cells was also observed in a variety of liver injury models induced by concanavalin A, polyriboinosinic polyribocytidylic acid, α-GalCer, etc.,17–19 which may be caused by either activation-induced NKT cell death or loss of cell markers such as NK1.1, or a combination of both mechanisms.32, 33 Three lines of evidence from our studies suggest that depletion of hepatic iNKT cells after CCl4 is mainly mediated via activation-induced NKT cell death. First, the number of apoptotic NKT cells in the liver was significantly increased after acute and chronic CCl4 treatment. Second, depletion of NKT cells was observed in both analyses using NK.1.1/CD3 markers and CD1 tetramer marker. Third, expression of Vα14 messenger RNA, a marker of iNKT cells, was down-regulated after CCl4 treatment (data not shown).
Diverse Roles of iNKT Cells in Liver Fibrosis.
In contrast to NK cells, which have been shown to play an important role in inhibiting liver fibrosis,7–10 iNKT cells play a less important role in regulating liver fibrosis because of iNKT cell depletion and tolerance. As shown in Fig. 5, chronic CCl4 treatment caused marked depletion of hepatic iNKT cells. Thus, chronic CCl4-treated wild-type mice were very similar to Jα18−/− mice, whereas iNKT cells in the liver were depleted in both groups. This also may explain why chronic CCl4 treatment only induced slightly greater liver fibrosis in Jα18−/− mice than in wild-type mice at an early stage (2-week treatment) but not at a later stage (4-week treatment), although NKT cells are able to kill HSCs. In contrast to iNKT cell depletion, the total number of NK cells was not decreased, but rather increased after CCl4 treatment (Figs. 1C; 5B). The cytotoxicity of hepatic NK cells against activated HSCs was also increased after 2 week CCl4 treatment (Jeong, Park, Gao, unpublished data). These findings suggest that NK cells could compensate the depletion of iNKT to inhibit liver fibrosis during chronic CCl4 treatment.
Chronic treatment with the NK cell activator polyriboinosinic polyribocytidylic acid markedly inhibited liver fibrosis as demonstrated previously,7, 9, 41 whereas chronic treatment with the iNKT activator, α-GalCer, had little effect on chronic liver injury and fibrosis (Fig. 7). The obvious reason for the unresponsiveness of α-GalCer in this model is attributable to the lack of hepatic iNKT cells during CCl4-induced chronic liver injury. An additional mechanism is likely, because of long-term iNKT cell anergy and tolerance after α-GalCer stimulation.42, 43 Interestingly, in contrast to α-GalCer, a naturally occurring glycolipid, β-glucosylceramide, has been shown to ameliorate liver fibrosis via modulation of NKT and CD8 lymphocyte distribution.44 However, it is not clear whether repeated β-glucosylceramide treatment also causes iNKT cell anergy.
Contrary to an insignificant role of chronic α-GalCer treatment on chronic liver fibrosis, a single injection of α-GalCer markedly enhanced CCl4-induced acute liver fibrosis (Fig. 4), although such an injection enhanced iNKT cell killing of activated HSCs and elevated serum IFN-γ levels that inhibited liver fibrosis (Figs. 3E; 4). Because a single dose of α-GalCer markedly enhanced CCl4-induced liver injury (Fig. 4), we speculated that α-GalCer injection may inhibit liver fibrosis through production of IFN-γ and induction of iNKT cell killing of HSCs, but also may accelerate liver fibrosis by increasing liver injury. Acceleration of liver fibrosis by increased liver injury may dominate over the inhibitory effect of α-GalCer on liver fibrosis, leading to stimulatory effects of a single α-GalCer injection on liver fibrosis induced by acute CCl4 treatment.
In summary, our findings suggest that iNKT cells may play a protective or detrimental role in CCl4-induced acute liver injury, depending on the degree of iNKT cell activation. During chronic liver injury, iNKT cells are depleted, playing a role in inhibiting the early stage but not late stage of liver fibrosis. The roles of iNKT cells in human liver injury and fibrosis remain unknown. De Lalla et al.45 reported that iNKT cells increase in chronically infected livers and produce profibrotic cytokines such as IL-4 and IL-13, suggesting that iNKT cells may contribute to the progression of liver fibrosis in patients with chronic hepatitis B viral infection.