fax: (86) 551-360-6783
Impairment of liver regeneration correlates with activated hepatic NKT cells in HBV transgenic mice†
Article first published online: 23 MAY 2007
Copyright © 2007 American Association for the Study of Liver Diseases
Volume 45, Issue 6, pages 1400–1412, June 2007
How to Cite
Dong, Z., Zhang, J., Sun, R., Wei, H. and Tian, Z. (2007), Impairment of liver regeneration correlates with activated hepatic NKT cells in HBV transgenic mice. Hepatology, 45: 1400–1412. doi: 10.1002/hep.21597
Potential conflict of interest: Nothing to report.
- Issue published online: 29 MAY 2007
- Article first published online: 23 MAY 2007
- Manuscript Accepted: 12 JAN 2007
- Manuscript Received: 18 SEP 2006
- Natural Science Foundation of China. Grant Numbers: 30630059, 30671901, 30570819, 30571695
- Ministry of Science & Technology of China. Grant Numbers: 2006CB504300, 2006CB806504, 2004CB518807, 2003CB515501
A fraction of HBV carriers have a risk to develop liver cancer. Because liver possesses a strong regeneration capability, surgical resection of cancerous liver or transplantation with healthy liver is an alternate choice for HBV-caused hepatocarcinoma therapy. How HBV infection affects the regeneration of hepatectomized or transplanted liver remains elusive. We report that partial hepatectomy (PHx)-induced liver regeneration was reduced in HBV transgenic (HBV-tg) mice, a model of human HBV infection. PHx markedly triggered natural killer T (NKT) cell accumulation in the hepatectomized livers of HBV-tg mice, simultaneously with enhanced interferon gamma (IFN-γ) production and CD69 expression on hepatic NKT cells at the early stage of liver regeneration. The impairment of liver regeneration in HBV-tg mice was largely ameliorated by NKT cell depletion, but not by natural killer (NK) cell depletion. Blockage of CD1d-NKT cell interaction considerably alleviated NKT cell activation and their inhibitory effect on regenerating hepatocytes. Neutralization of IFN-γ enhanced bromodeoxyuridine incorporation in HBV-tg mice after PHx, and IFN-γ mainly induced hepatocyte cell cycle arrest. Adoptive transfer of NKT cells from regenerating HBV-tg liver, but not from normal mice, could inhibit liver regeneration in recipient mice. Conclusion: Activated NKT cells negatively regulate liver regeneration of HBV-tg mice in the PHx model. (HEPATOLOGY 2007.)
More than 350 million people are persistently infected with HBV and are at a risk of developing liver diseases, cirrhosis, and hepatocellular carcinoma. Because liver has a strong ability to regenerate to its original size and function, surgical removal of cancerous liver or transplantation with healthy liver have become two important alternate choices for hepatocarcinoma therapy, and the prognosis is to some extent dependent on the regeneration capability of hepatectomized or transplanted liver. Our previous studies have shown that murine cytomegalovirus (MCMV) infection or stimulation with the double-strand RNA virus mimic polyinosinic-polycytidylic acid could negatively regulate liver regeneration in normal mice in a natural killer (NK) cell- and IFN-γ-dependent manner.1 However, whether or how HBV infection impairs hepatectomy-induced liver regeneration is unknown.
Meanwhile, liver regeneration is a multistep process of tissue repairs following chemical-induced liver injury2 or partial hepatectomy (PHx). A large number of soluble cytokine and growth factors such as IL-6,3 hepatocyte growth factor,4 as well as epidermal growth factor,5 are involved in this complicated process. Because we have known that liver is abundant with a population of innate immune cells, and some investigations have demonstrated the critical involvement of these cells such as Kupffer cells,6 natural killer T (NKT) cells,7 and NK cells1 in the regulation of liver regeneration in normal mice. However, how liver innate immune cells influence the liver regeneration of HBV-infected carriers is never reported.
NKT cells are heterogeneous population of innate lymphocytes, present in virtually all lymphoid compartments.8 In contrast to their relatively low presence in peripheral lymphatic system, NKT cells are quite abundant in the liver, implying an important role of this cell type in liver biology. Like other innate immune cells, NKT cells can rapidly release remarkable quantities of immunomodulatory cytokines, such as IL-4 and IFN-γ, within minutes of stimulation after their T cell receptor (TCR) engagement with CD1d,9 one mechanism by which they are thought to participate in the liver injury triggered by concanavalin A (ConA).10, 11 NKT cell roles in liver injury have been extensively reported, but their roles in liver regeneration have only been revealed by a few studies. Earlier investigations found that repeated administration of IL-12 led to NKT cell accumulation in the liver and then inhibited liver regeneration in normal mice, which was further confirmed in NKT cell-deficient Ja281−/− mice.12 Paradoxically, in our study, results from another NKT cell-deficient CD1d−/− mouse showed that NKT cell did not take part in the regulation of liver regeneration,1 and even liver regeneration was up-regulated in CD1d−/− mouse.13 Additionally, when NKT cells were activated by specific agonist aGalcer or T cell stimulator ConA.13, 14 their effects on subsequent PHx-induced liver regeneration were diverse. ConA treatment could improve liver regeneration, mainly inducing NKT cell activation-induced death, implying the negative role of NKT in liver regeneration. However, aGalcer treatment accelerated liver regeneration, supporting that NKT cells improve liver regeneration.12 Taken together, those data involving NKT cell role in liver regeneration only come from artificial treatments or gene ablations in HBV-uninfected mice. To explore NKT cell role in disease conditions such as HBV infection, in this study we used an HBV transgenic (HBV-tg) mouse as a model for human HBV infection. We demonstrated that the impaired liver regeneration of HBV-tg mice was related with activated NKT cells and their IFN-γ production. Blockade of CD1d-NKT interaction restored the impaired liver regeneration in HBV-tg mice.
Materials and Methods
Male partial HBV-tg mice C57BL/6J-TgN (AlblHBV) 44Bri (H-2b, containing partial HBV genome including S, pre-S, and X gene) were purchased from VITALRIVER experiment animal company (Beijing), who purchased the mice from Jackson Lab (Bar Harbor, Maine) and bred them for us. Another male full-HBV-tg mouse was purchased from Infectious Disease Center of No. 458 Hospital (Guangzhou, China). The full-HBV-tg mouse lineage was initially produced on a BALB/c background. The transgene in these mice consists of 1.3 copies of the HBV adr complete genome. The full-HBV-tg mice express high level of HBsAg in their serum and have detectable HBV DNA in their serum.15 B6 mice or BALB/c mice were purchased from Shanghai Experimental Animal Center, Chinese Academy of Science (Shanghai, China) as respective control mice. All mice were fed and housed in a temperature-controlled animal facility with 12-hour day-night cycles. The full-HBV-tg mice were handled in the laboratory in accordance with Biological Safe Level 2. Animals were treated humanely, and all procedures were in compliance with the regulations of animal care of University of Science and Technology of China.
Two-thirds (approximately 70%) PHx was performed according to the method of Higgins and Andersen between 8:00 and 11:00 AM.16 There was no operative mortality. Two hours before they were killed, mice were intraperitoneally injected with bromodeoxyuridine (BrdU, 100 mg/kg body weight) to assess hepatocyte DNA synthesis. Groups of animals (n = 3-5 each group) were killed at 0, 12, 24, 48, and 72 hours after surgery. At the time of killing, livers were weighed and rapidly split into several pieces, some for formalin-fixation for immunofluorescence and immunohistochemistry as described below, the others for preparation of liver mononuclear cells (MNCs).
Cell Preparation, Transfer, and Depletion.
Murine livers were removed and passed through a 200-gauge stainless steel mesh and then suspended in RPMI 1640 medium containing 2% fetal bovine serum. After one washing, the cells were resuspended in 40% Percoll solution containing 100 U/ml heparin and were centrifuged at 500g for 10 minutes at room temperature. The pellet was re-suspended in a red blood cell lysis solution (155 mM NH4Cl, 10 mM KHCO3, 1 mM EDTA, and 170 mM Tris, pH 7.3). After incubation on the ice for 7 minutes, the cells were harvested by centrifugation and washed twice in Hank's balanced salt solution containing 5% fetal bovine serum before use. Hepatocyte isolation was followed as previously described.17
To deplete NK cells, CD4+ cells, or CD8+ cells in vivo, 100 μl of phosphate-buffered saline (PBS) containing anti-asialo GM-1 (AsGM-1; 50 μg, WAKO, Richmond, VA), anti-CD4 Ab (100 μg, ATCC, TIB 207), or anti-CD8 Ab (100 μg, ATCC TIB 105), respectively, was intravenously injected to mice 1 day before operation. For NK1.1+ cell (including NK and NKT cell) depletion, mice received two anti-NK1.1 injections (100 μg, ATCC, PK136) at days 3 and 1 before liver operation. The respective isotype antibodies were used as controls. The absences of respective cells were confirmed by flow cytometry.
For NKT cell purification, mice were injected with 50 μg AsGM-1 to deplete NK cells. After 1 day, liver lymphocytes were isolated from these NK-depleted mice, stained with FITC-conjugated anti-NK1.1 monoclonal antibody (MAb) (BD PharMingen), and incubated with anti-FITC Microbeads (Miltenyi Biotec, Auburn, CA) for 15 minutes at 4°C. NK1.1+ cells were enriched by positive MACS according to the manufacturer's protocol. Approximately 85% of the magnetic-activated cell sorting (MACS)–purified cells were NK1.1 and CD3 positive. Through trypan blue staining, viability of purified NKT cell for transfer assay is above 90%.
For adoptive transfer of liver MNCs or purified NKT cells, recipient mice received two anti-NK1.1 injections at days 3 and 1 before liver operation. Recipient mice underwent a conventional PHx operation; 50 μl hepatic MNCs (5 × 106) or purified NKT cell (1 × 106) in PBS were injected into the spleen at a rate of 10 μl/s using a 29-gauge needle attached to a 1-ml syringe. Liver regeneration was assessed 48 hours after PHx.
Analyses of the Surface Phenotype and Intracellular Cytokine Expression by Flow Cytometry.
The MAbs used in this study included Cy5-anti-CD3e, phycoerythrin (PE)-anti-CD69, FITC-anti-CD4, Cy5-anti-CD8, and FITC- or PE-anti-NK1.1 MAb and PE-anti-CD1d (eBioscience, San Diego, CA). For intracellular cytokine staining, liver MNCs were incubated for 3 hours in the presence of Ionomycin (1 μg/ml, BD PharMingen), PMA (30 ng/ml, BD PharMingen), Monensin (1.7 μg/ml, BD PharMingen), and then stained with FITC-anti-NK1.1 MAb and Cy-5-anti-CD3e MAb. After fixation with fixation solution and permeabilization with permeabilization solution (eBioscience), intracellular cytokine staining was performed using PE-anti-IFN-γ MAb. Fc receptor (FcR) blocking antibody 2.4G2 was used to prevent nonspecific surface binding, whereas isotype antibody was used as control in the intracellular staining. Stained cells were acquired by FACScalibur and analyzed with WinMDI 2.9 software.
Histopathological Analysis and In Situ Cell Death Detection.
Hematoxylin-eosin (HE) staining was used to evaluate liver injury from formalin-fixed and paraffin-embedded liver tissue. Using terminal transferase deoxyuridine nick-end labeling (TUNEL), apoptotic cells on liver section were analyzed by an in situ cell death detection kit (TUNEL, Roche Diagnostics Ltd., Basel, Switzerland). The procedure was performed following the manufacturer's instructions.
Cell Cycle and Apoptosis Analysis.
Apoptotic HepG2 and HepG2.2.15 were determined by double staining with propidium iodide (PI) and Annexin V. Cell cycle analysis was performed with PI staining as previous described.18
Sections from formalin-fixed and paraffin-embedded liver tissue were stained with mouse monoclonal anti-BrdU antibody (Sigma, St. Louis, MO, 1:100 dilution), or mouse monoclonal antibody against proliferating cell nuclear antigen (PCNA) (Sigma, St. Louis, MO, 1:100 dilution), overnight at 4°C. Secondary antibody, peroxidase-labeled rabbit anti-mouse antibody (Dako Envision System Carpinteria, CA), was incubated at room temperature for 30 minutes and then stained with peroxidase substrate 3,3′-diamino-benzidine chromagen (Dako) and finally counterstained with hematoxylin or eosin. Two observers blinded to the animal's identity and treatment evaluated BrdU- or PCNA-positive hepatocytes. BrdU incorporation was determined by counting positively stained hepatocyte nuclei in three to six low-power (10×) microscope fields, and the mean was calculated. PCNA labeling was evaluated in a similar manner.
Sections from formalin-fixed and paraffin-embedded liver tissues were thawed onto glass slides, air dried, and fixed in acetone/methanol (1:1) at 4°C for 10 minutes. After being washed with PBS, the sections were blocked with 3% bovine serum albumin/PBS at room temperature for 30 minutes. Incubation was continued with rat anti-mouse PE-CD1d MAb (clone 1B1, Caltag, 1:100 dilution) and rat FITC-F4/80 MAb directed against murine macrophages (clone BM8, Dianova, Hamburg, Germany, 1:100 dilution) dissolved in 3% bovine serum albumin/PBS overnight at 4°C. After rinsing with PBS, the sections were coverslipped with 10% glycerol/PBS, pH 8.6, and examined by confocal laser-scanning microscopy (Axiovert 100 M, Carl Zeiss, Oberkochen, Germany).
Blocking CD1d-NKT Cell Interaction In Vivo.
Two hours before PHx, HBV-tg mice were intravenously injected with purified functional anti-CD1 MAb (100 μg in 100 μl PBS, 1B1, eBioscience) to block the interaction of NKT cells with CD1d as described previously,19 and IgG isotype MAb as controls.
Determination of IFN-γ Concentration and IL-12 in Liver and Serum.
Serum IFN-γ concentration and IL-12 concentrations in whole-liver homogenates were determined by enzyme-linked immunosorbent assay using mouse IFN-γ or IL-12 enzyme-linked immunosorbent assay kit (Bender Med Systems, Vienna, Austria) following the manufacturer's instructions.
Differences between the groups were analyzed by Student t test, and P < 0.05 was considered statistically significant.
Liver Regeneration Is Delayed in HBV-tg Mice After PHx.
Our previous study has shown that MCMV infection or poly (I:C) injection could give rise to the delayed liver regeneration in wild-type (WT) mice via activating NK cells.1 Because MCMV infection or poly (I:C) injections are not real conditions in humans, to evaluate the effect of HBV infection on liver regeneration, we used HBV-tg mice, mimicking HBV carrier or chronic HBV hepatitis. HBV-tg mice were created with 70% PHx in comparison with WT mice. First, remnant livers showed a marked reduction in liver mass of partial HBV-tg mice through macroscopic evaluation if compared with littermate WT mice (Fig. 1A). Consistently, the percentages of remnant liver weight to body weight were substantially reduced both in 60-day-old and 120-day-old partial HBV-tg mice at the time points of 24, 48, and 72 hours after PHx (Fig. 1B). No significant difference was seen between these two strains of mice in the percentage of liver weight to body weight before PHx. Next, we analyzed hepatic PCNA labeling and BrdU incorporation of hepatocytes after PHx in WT mice and HBV-tg mice and found the positive staining of BrdU or PCNA peaked at 24 to 48 hours. However, the number of BrdU- and PCNA-positive hepatocytes per field in HBV-tg mice was much lower than that in WT mice at 48 and 72 hours (Fig. 1C, D), demonstrating liver regeneration of HBV-tg mice is inhibited in the PHx model. In long-term observations after 1 week, we found the percentages of liver weight to body weight in hepatectomized HBV mice were nearly similar to that in WT mice (data not shown), indicating that the liver regeneration in HBV-tg mice is just delayed.
However, when full-HBV-tg mice with BALB/c background were applied, we did not find that the percentage of liver weight to body weight 48 hours after PHx in these HBV-tg mice was lower than that in WT (See Supplementary Fig. 1: Supplementary material can be found on the HEPATOLOGY website (http://interscience.wiley.com/jpages/0270-9139/suppmat/index.html). In the following investigations, we chose C57BL/6J-TgN (AlblHBV) 44Bri mice as HBV infection model to explore the mechanism involving the impaired liver regeneration in the PHx model.
NKT Cells Negatively Regulate the Liver Regeneration of HBV-tg Mice
NKT Cells Number and Activation are Upregulated at the Early Liver Regeneration of HBV-tg Mice.
Flow cytometric analysis revealed that before PHx there were approximately 11% to 15% NKT cells (NK1.1+CD3+) in the liver of HBV-tg mice and 12% to 20% NKT cells in the liver of WT mice. PHx considerably elevated hepatic NKT cells to 20% to 50% of liver MNCs in HBV-tg mice, with the peak effect occurring at 12 hours, whereas NKT cell numbers were slightly elevated at 24 hours in WT mice (Fig. 2A). The absolute number of NKT cells per gram liver in HBV-tg mice was 2 to 3 times more than that in WT mice (Fig. 2B). To identify the subsets of elevated NKT cells induced by PHx in HBV-tg mice, liver lymphocytes were analyzed using three-color flow cytometry. The absolute numbers of hepatic CD4+ NKT cells and double negative (DN, CD4−CD8−) NKT cells in HBV-tg mice were obviously higher than that in WT mice after PHx, whereas CD8+ NKT cells showed no difference between two strains of mice (Fig. 2B). Next, we analyzed the expression of CD69, an early activation marker, on the NKT cells. Approximately 75% of hepatic NKT cells in HBV-tg mice were CD69 positive 24 hours after PHx, which was higher than that in WT mice (Fig. 2C). Absolute numbers of CD69-positive NKT cells in the regenerating liver of HBV-tg mice were also higher compared with WT (data not shown). These results collectively suggest that PHx dramatically upregulated hepatic NKT cell accumulation and activation in HBV-tg mice at the early stage of liver regeneration.
Depletion of NKT Cells, But Not NK Cells, Improves Liver Regeneration in HBV-tg Mice.
To examine whether NKT cells caused the decreased liver regeneration in HBV-tg mice, we generated NK1.1+ cell-depleted HBV-tg mice by treatment with anti-NK1.1 antibody. Anti-NK1.1 antibody could deplete nearly all NK cells and NKT cells, whereas anti-AsGM1 antibody only depleted NK cells. Two other cell-depleted mice were generated by CD4 and CD8 antibody treatment (Fig. 3A). The respective remaining cells are below 1%. Forty-eight hours after PHx, we found that depletion of NK1.1+ cells or CD4+ cells in WT mice could slightly improve liver regeneration. Compared with isotype control, HBV-tg mice depleted of NKT and NK cells were found to have dramatically elevated BrdU incorporation and PCNA labeling (Fig. 3B, C), whereas HBV-tg mice depleted of NK cells alone had nearly similar values as control isotype antibody. Depletion of CD4+ T cells, but not CD8+ T cells, also ameliorated the impaired liver regeneration in HBV-tg mice.
Adoptively Transferred NKT Cells from Regenerating HBV-tg Liver Inhibit Liver Regeneration of Recipient Mice.
To know the roles of NKT cells in HBV-tg mice in the decreased liver regeneration, several indicated hepatic lymphocytes were isolated and transferred to different recipient mice. Transfer with liver MNCs of hepatectomized HBV-tg mice but not WT mice could inhibit the liver regeneration of both recipient HBV-tg mice and WT B6 mice; however, if the MNCs were depleted of both NK and NKT cells with anti-NK1.1 MAb, the MNCs lost the inhibitory effects. Because of the lack of a specific MAb against NKT cells, we depleted NK cells from MNCs with anti-ASGM-1 to compare with NK1.1 depletion. The results demonstrated that NK cells-depleted MNCs revealed the same effect as whole MNCs, suggesting that the inhibitory effects of liver MNCs on hepatectomized HBV-tg mice possibly came from NKT cells, but not from NK cells, which was further confirmed by adoptive transfer with the purified NKT cells from hepatectomized liver of HBV-tg mice, but not from that of WT mice (Fig. 3D). In addition, when equal amounts of liver MNCs of hepatectomized HBV-tg mice were adoptively transferred to recipient HBV-tg mice, the inhibitory effect on liver regeneration was notably greater than that of recipient WT mice, and consistently, depletion of NK1.1+ cells in the transferred cells largely counteracted their inhibitory effects, indicating that HBV-tg mice are sensitive to adoptively transferred NKT cells from regenerating HBV-tg mice liver (Fig. 3D).
Elevated CD1d on Hepatocytes Contributes to NKT Cell Activation and Subsequent Impairment of Liver Regeneration in HBV-tg Mice
These data showed that PHx caused selective CD4+ and DN NKT cell accumulation and strong activation of NKT cells in livers of HBV-tg mice. Generally, CD4+ and DN NKT cells are CD1d-restricted.20, 21 We therefore examined the roles of CD1d in the involvement of NKT cells in impaired liver regeneration in HBV-tg mice. Immunofluorescence microscopy showed that hepatic CD1d expression (red) was at a low level in liver section of WT mice 24 hours after PHx, but greatly elevated in HBV-tg mice (Fig. 4A). CD1d-expressing F4/80+ macrophages/Kupffer cells (yellow) did not obviously increase in the sections of both WT mice and HBV-tg mice (Fig. 4A). To identify these CD1d-staining positive cells, primary hepatocytes were isolated for determination of CD1d expression using flow cytometry. Mean intensity of CD1d expression on hepatocytes in both WT and HBV-tg mice was relatively low before PHx; however, it dramatically increased 12 to 24 hours after PHx in HBV-tg mice, but not in WT mice (Fig. 4B). These data suggest that the CD1d increase during liver regeneration of HBV-tg mice was mostly due to hepatocytes, but not Kupffer cells and macrophages.
We then injected HBV-tg mice with anti-CD1d antibody to see whether blockage of CD1d recognition interferes with liver regeneration. Two hours after anti-CD1d antibody injection, immunofluorescence analyses showed that the treatment did not alter the absolute numbers of NKT cells, NK cells, or T cells in the livers (Supplementary Fig. 2), suggesting anti-CD1d antibody did not deplete any hepatic immune cells. Anti-CD1d antibody administration before PHx effectively improved the liver regeneration of HBV-tg mice, with elevated ratio of liver weight to body weight, PCNA labeling, and BrdU incorporation (Fig.4C,D), indicating that hepatic up-regulation of CD1d contributes to the impaired liver regeneration of HBV-tg mice.
NKT-Derived IFN-γ Plays a Critical Role in Inhibiting Liver Regeneration of HBV-tg Mice
IFN-γ Production by NKT Cell Is Enhanced in Hepatectomized HBV-tg Mice.
Previous data have shown that IFN-γ production by NK cells impaired liver regeneration in WT mice.1 To address the issue, we detected serum IFN-γ concentration after PHx and found that serum IFN-γ in HBV-tg mice was slightly higher than that in WT mice (Fig. 5A), whereas the concentration decreased with NK1.1+ cell depletion rather than NK cell depletion. It was noted that 82.04% of hepatic NKT cells in HBV-tg mice expressed IFN-γ 24 hours after PHx, whereas this occurred in only 12.74% of NKT cells in WT mice (Fig. 5B). Systematical administration of anti-CD1d anybody to block CD1d-NKT interaction greatly inhibited serum IFN-γ levels in HBV-tg mice (Fig. 5C) and IFN-γ production by NKT cells (Fig. 5D). Therefore, CD1d-mediated NKT cell activation leads to more IFN-γ production in HBV-tg mice. We also detected IL-12 level in the liver lysate; no obvious difference was seen between WT and HBV-tg mice (data not shown), but neutralization of IL-12 could partially abrogate serum IFN-γ and IFN-γ secretion by NKT cells during liver regeneration in HBV-tg mice, indicating the physiological level of IL-12 is also necessary for NKT cell IFN-γ production and impaired liver regeneration in HBV-tg mice (Fig. 5C, D).
IFN-γ neutralization Improves Liver Regeneration of HBV-tg Mice.
To make clear the role of IFN-γ in impaired liver regeneration, HBV-tg mice were pretreated with IFN-γ-specific antibody 2 hours before PHx. IFN-γ neutralization could markedly improve the subsequent liver regeneration induced by PHx in HBV-tg mice, manifested by elevated ratios of remnant liver to body weight, BrdU incorporation, and PCNA labeling (Fig. 5E, F, and G).
Exogenous Administration with Recombinant IFN-γ Considerably Aggravates the Impairment of Liver Regeneration in HBV-tg Mice.
As indicated, HBV-tg mice are more sensitive to the adoptively transferred NKT cells from regenerating HBV-tg liver. We then wanted to determine whether HBV-tg mice were hypersensitive to IFN-γ secreted by NKT cells. Before PHx, both HBV-tg mice and WT received different doses of IFN-γ (0.5, 2, and 5 μg per mouse). As shown in Fig. 6, though liver regeneration in HBV-tg and WT mice was significantly inhibited at a 5-μg dose, consistent with previous data,1 live regeneration was not affected by administration of 0.5 or 2 μg in WT mice, but was obviously inhibited in HBV-tg mice (Fig. 6A). Therefore, HBV-tg mice are hypersensitive to IFN-γ at least in the PHx model.
IFN-γ Inhibits Liver Regeneration of HBV-tg Mice Through Negatively Controlling Cell Cycle Arrest, But Not Inducing Apoptosis of Hepatocyte.
We then explored whether IFN-γ induced the attenuated liver regeneration by inducing hepatocyte apoptosis, or by negatively controlling the cell cycle. Administration of IFN-γ even at a dose of 0.5 μg notably decreased PCNA labeling 48 hours after PHx in HBV-tg mice, but not in WT mice (Fig. 6B), indicating that cell proliferation is inhibited. Using in situ cell death analysis and pathological staining, no visible necrosis and apoptotic hepatocytes in the liver section were seen in either HBV-tg mice or WT mice (Fig. 6B), suggesting that apoptosis did not happen. As an injury model, mice were injected with 15 μg /g concanavalin A. Massive necrosis and brown apoptotic lesions were found in both HBV-tg and WT mice; however, manifestations on the liver sections of HBV-tg mice were more severe than those of WT mice (Fig. 6B). Furthermore, when the HBV gene-transfected human cell line HepG2.2.15 was stimulated with a series of human IFN-γ in vitro, the percentages of counted sub-G1 events did not exceed 2% of total counts (data not shown), the percentages of G0/G1 in all stages substantially went up in a dose-dependent manner, even at the low dose of IFN-γ (10 U/ ml), but increased in control HepG2 cells only at the high dose of IFN-γ (1,000 U/ml) (Fig. 6C), indicating that HBV gene-transfected cells are also sensitive to IFN-γ inhibition in vitro. Annexin V staining showed that no obvious change of apoptosis was found in both cell lines (Fig. 6C).
Cytotoxic T lymphocytes have been considered as one of the critical effector cells in the pathogenesis of HBV infection.22, 23 In HBV-tg mice, hepatitis B surface antigen (HBsAg)-specific cytotoxic T lymphocyte plays a major role in the spread of liver necrosis mechanisms.24 In addition, HBx protein could inhibit liver cell proliferation in a paracrine manner in HBx transgenic mice.25 Apart from these reports, few studies have focused on the effect of HBV infection on liver cell proliferation in humans and mice. With many innate immune cells abundant in liver, especially NKT cells, the roles of NKT cells thus came to light in the defense of virus infection26, 27 and in the pathogenesis of liver diseases.28 Although NKT or NK cells have widely reported to attack hepatocytes in ConA-induced liver injury28 or poly (I:C)-induced liver injury,29 respectively, how these cells play their roles in liver regeneration, as a repairing process following liver injury, is still unclear. Our data provide evidence supporting the negative regulation of innate immunity, but not direct cytotoxicity, on liver regeneration in HBV carriers. This finding has an important clinical relevance in HBV-related immunopathogenesis. When surgical resection of cancerous liver or transplantation with new healthy liver was performed in HBV patients with hepatocarcinoma, the negative effect of HBV infection on liver regeneration will be taken into account. Preventive interference with activated NKT cells in the early stage of hepatectomy may be a possible approach for immune-mediated impaired liver regeneration in HBV patients.
In our study, we observed inconsistent results in two kinds of HBV-tg mice. We did not find the impairment of liver regeneration in full-HBV-tg mice with BALB/c background containing the complete virus particle. Presumably, NKT cell functions in different genetic backgrounds show some divergences,30 and possibly are similar to NK cell functions in BALB/c strain; NKT cell function in full HBV-tg mice may be weaker than that in the B6 strain. Therefore, backcross of the full-HBV mice to B6 background is a critical answer to the enigmatic question. Conversely, two mice have some distinguished differences in the introduced HBV gene, which determine their distinct pathological changes. Our used partial HBV-tg mice have obvious HBsAg intracellular retention in endoplasmic reticulum, a very similar outcome to human HBV chronic infection. Generally, HBsAg retention in hepatocytes is closely associated with glass-like denaturalization, even liver fibrosis, in humans. However, our full-HBV-tg do not have any HBV protein deposit in hepatocyte, and conversely, existence of contagious HBV Dane particles is detected in the liver of full-HBV-tg mice. Our data hint that not all HBV infection would lead to impaired liver regeneration, which probably relies on HBV existence. When more HBV protein is deposited in hepatocytes, it will limit liver regeneration, mainly by inducing NKT cell activation, whereas complete HBV particles do not contribute to NKT cell activation in the PHx model. Therefore, demonstrating the divergent phenomena in two HBV-tg mice will hold benefits for the explaining the question of which kind of HBV patients easily develop impairment of liver regeneration.
Previously, we revealed that NK cells negatively regulated liver regeneration in normal mice after hepatectomy or acute MCMV infection1; however, we still do not know whether the liver regeneration is impaired in the condition of HBV infection, and which cells will control this process. Interestingly, in this study, we found that NKT cells, but not NK cells, negatively controlled the liver regeneration after hepatectomy in HBV-tg mice. Acute MCMV infection is different from HBV infection, as has been extensively verified. We used partial HBV transgenic mice to explore liver regeneration. In contrast to NK cells in WT mice, we found that NKT cells are major players in negatively regulating liver regeneration in HBV-tg mice. Why were NK cells not responsible for the impairment of liver regeneration in HBV-tg mice? The difference may be attributable to the distinct feature of two pathogens on hepatic NK cell function. As reported, MCMV infection often causes NK cell accumulation and activation.31 Conversely, NK cell functions such as cytokine production and cytotoxicity were obviously attenuated in chronic HBV infection.32 Thus, NK cells' ability to negatively regulate liver regeneration would not be found in HBV-tg mice.
Liver NKT cells are a highly heterogeneous population. Originally, NKT cells were classically considered to be either CD4+ or DN, and CD1d-restricted (e.g., their development was dependent on CD1d).33 In this study, we found that the number and activation of CD4+ NKT cells were increased after PHx in HBV-tg mice, and blockage of interaction between CD1d and NKT cells could counteract NKT cell activation by reduction of IFN-γ production and improved impaired liver regeneration in HBV-tg mice. Thus, these populations can be considered as CD1d-reactive NKT cells. Because CD1d−/− HBV-tg mice will be greatly helpful in further evaluation of the exact roles of NKT cells in the development of impaired liver regeneration in HBV-tg mice, we are raising the hybrid mice in our setting for further study.
NKT cells are usually activated by antigen-loading CD1d on APCs and soluble cytokines such as IL-12.34, 35 We also observed that elevation of CD1d expression and an increase in the physiological level of IL-12 played necessary roles in NKT cell activation in HBV-tg mice. CD1d-expressing cells in the liver mainly refer to macrophages/Kupffer cells and hepatocytes.6, 36, 37 In our study, we observed that HBV-infected hepatocytes plus surgery-initiated liver regeneration, as a kind of “stress” triggering, induced overexpression of CD1d in HBV-infected hepatocytes, but not in macrophages/Kupffer cells, which contributes to NKT cell activation followed by releasing inflammatory cytokines (Fig. 5). The roles of macrophages/Kupffer cells in NKT cell activation are possibly to produce IL-12, as previously reported by us.29 This finding indicates that the virus-infected hepatocytes may exist as a “stress-inducible” target cell recognized by NKT cells through CD1d molecules. The phenomenon also has implications of wide general biomedical importance in virus-infected tissue repair (such as gut, respiratory tract, pancreas etc.).
The importance of CD1d molecule in NKT activation has been extensively reported, but no study has focused on HBV-infected hepatocytes. CD1d could selectively activate hepatic NK cells to eliminate experimentally disseminated hepatoma cells in murine liver38; overexpression of CD1d by keratinocytes in psoriasis could lead to NKT cell activation and releasing of IFN-γ39; transgenic CD1d mice had stronger NKT cell function in nonobese diabetic mice, alleviating insulin-dependent diabetes mellitus pathogenesis40; high levels of pro-inflammatory hepatic CD1d-reactive NKT cells are seen in HCV infection, and up-regulated CD1d was also found on infected hepatic cells in HCV-infected donors.41 In the current study, we found that blocking interaction of NKT cells with CD1d significantly inhibited IFN-γ production by NKT cells (Fig. 5). Therefore, CD1d upregulation in the liver is an initial step for activation of hepatic CD1d-reactive NKT in HBV-tg mice. In contrast to major histocompatibility complex I upregulation on virus-infected cells, which is associated with conventional T cell activation,42 CD1d upregulation will help NKT cell activation.34, 40 Different from major histocompatibility complex I-loaded peptide, antigens presented by CD1d are considered to be lipid. Lipid on hepatocytes may be altered to be suitable for CD1d presentation; or certain molecules for CD1d-dependent presentation, such as TAP1, are abnormally activated because of HBV protein retention in hepatocytes, which will be helpful in explaining hepatic NKT cell activation in HBV-tg mice. Recently, however, HSV-mediated downregulation of CD1d recycling has been reported, implying that virus can escape from NKT cell immunosurveillance via interfering with the CD1d-presenting pathway.43 How HBV virus infection increased CD1d levels during PHx remains to be determined. Further investigations should expand to analysis of CD1d gene expression, CD1d protein synthesis, and CD1d recycling in the subcellular organelle.
NKT cells exert their regulatory roles mainly by two pathways: direct cytotoxicity and releasing cytokines such as IFN-γ and IL-4. Previous studies showed that the IFN-γ produced by NKT cells or NK cells exerts its suppressive response in liver regeneration.1 However, how NKT negatively modulates liver growth through IFN-γ is not known. NKT cells can induce hepatocyte apoptosis by the surface FasL and TRAIL expression.28, 44, 45 IFN-γ also is a critical cytokine for massive necrosis and apoptosis of hepatocyte in ConA-induced liver injury.46 In this study, we found that although NKT cells produced more IFN-γ in HBV-tg at the early stage of PHx, serum levels of IFN-γ elevated only slightly. We did not find any necrotic lesions or apoptotic cells in HBV-tg mice liver sections after PHx. HBV-tg mice were hypersensitive to low doses of exogenous supplemental IFN-γ via arrest of the cell cycle of the HBV gene-transfected hepatocytes, rather than directly induced apoptosis. In conclusion, in addition to overproduction of IFN-γ by more sensitively activated hepatic NKT cells, the impaired liver regeneration in HBV-tg mice also may be closely relevant to hepatocyte itself, which is more sensitive to exogenous stimulus, such as IFN-γ47, 48 this mechanisms will be helpful for the elucidation of HBV-associated fulminant hepatitis in human beings.
- 16Experimental pathology of the liver. I: restoration of the liver of the white rat following partial surgical removal. Arch Pathol 1931; 12: 186–202., .
Supplementary material for this article can be found on the H EPATOLOGY website ( http://interscience.wiley.com/jpages/0270-9139/suppmat/index.html ).
|jws-hep.21597.fig1.tif||271K||Figure S1. Liver regeneration is not impaired in full-HBV-tg mice after PHx. 48 h after full-HBV-tg mice and BALB/c mice were subject to PHx, the ratio of liver weight to body weight was calculated.|
|jws-hep.21597.fig2.tif||294K||Figure S2. Anti-CD1d treatment does not affect hepatic immune cell numbers. 2 h after partial HBV-tg mice were received with 100 μg anti-CD1 mAb or isotype antibody control in 100 μL PBS, liver MNCs were isolated for two-color flow cytometry (CD3 and NK1.1). The absolute numbers of NK (CD3 -NK1.1 + ), NKT (CD3 +NK1.1 + ) and T (CD3 +NK1.1 - ) cells were calculated by multiplying their respective percentages with total liver MNCs number.|
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.