Skip to main content
NCBI home page
Clinical and Experimental Immunology logoLink to Clinical and Experimental Immunology
. 2015 Apr 12;180(3):499–508. doi: 10.1111/cei.12597

Activated natural killer cells accelerate liver damage in patients with chronic hepatitis B virus infection

Q Zheng *, Y Y Zhu *, J Chen *, Y B Ye †, J Y Li †, Y R Liu *, M L Hu ‡, Y C Zheng *, J J Jiang *,✉
PMCID: PMC4449778  PMID: 25639451

Abstract

Emerging evidence indicates that natural killer (NK) cells may contribute to liver injury in patients with hepatitis B virus (HBV) infection. Because HBV infection progresses through various disease phases, the cytolytic profiles of peripheral and intrahepatic NK cells in HBV-infected patients remain to be defined. In this study, we comprehensively characterized intrahepatic and peripheral NK cells in a cohort of HBV-infected individuals, and investigated their impact on liver pathogenesis during chronic HBV infection. The study population included 34 immune-clearance (IC) patients, 36 immune-tolerant (IT) carriers and 10 healthy subjects. We found that the activity of peripheral NK cells from IC patients was functionally elevated compared to IT carriers and controls, and NK cell activation was indicated by an increased expression of CD69, CD107a, interferon (IFN)-γ and tumour necrosis factor (TNF)-α. Further analysis showed that the increased activity of both peripheral and hepatic NK cells was correlated positively with liver injury, which was assessed by serum alanine aminotransferase levels (ALT) and the liver histological activity index (HAI). Interestingly, the frequency of peripheral NK cells was reduced in IC patients (especially those with higher HAI scores of 3–4), but there was a concomitant increase in hepatic NK cells. The functionally activated NK cells are enriched preferentially in the livers of IC patients and skew towards cytolytic activity that accelerates liver injury in chronic hepatitis B (CHB) patients.

Keywords: chronic hepatitis B, cytokine, innate immunity, liver injury, natural killer cells

Introduction

Chronic hepatitis B (CHB) is one of the most common infectious diseases, affecting more than 240 million people worldwide 1,2. Without therapeutic intervention, 15–40% of CHB patients will develop severe complications, including cirrhosis, liver failure or hepatocellular carcinoma (HCC) 2,3. The progression of chronic hepatitis B virus (HBV) infection results from the ineffective attempts by the host immune response to eliminate viral load, and the non-cytopathic and hepatotrophic characteristics of HBV contribute significantly to viral persistence 4. The immunopathogenesis of HBV liver disease is largely believed to be mediated by liver-infiltrating T lymphocytes 4–6. This viewpoint was challenged recently by observations that intrahepatic HBV-specific cytotoxic T lymphocytes (CTLs) are frequently present in the livers of patients without evidence of hepatic immunopathology 7. A larger number of non-viral specific lymphocytes is usually detected in the livers of CHB patients with hepatocellular damage 8, and natural killer (NK) cells were shown to exacerbate hepatocyte injury in a model of murine hepatitis virus strain 3 (MHV-3)-infected mice 9. These findings suggest that infiltrating NK cells may play a pivotal role in HBV-related liver immunopathogenesis.

Human NK cells are derived from CD34+ haematopoietic stem cells and constitute approximately 15% of the peripheral blood mononuclear cell fraction, representing a fundamental component of the innate immune system 10. NK cells are abundant in the liver, where they comprise up to 31% of the hepatic lymphocytic population 11. Unlike T lymphocytes, human NK cells generally do not express T cell antigen receptors (TCR) and the pan-T marker CD3, but NK cells co-express CD56 and NKp46 12. Because of this, NK cells can be classified further into subpopulations, including the more cytokine-responsive CD56bright (10%) and the more cytotoxic CD56dim (90%) subsets 13.

Thus far it has been fully documented that NK cells serve as a major innate immune component against viral infection. Especially in the early phase of viral infection, NK cells can either kill infected cells via perforin/granzyme or death receptor [e.g. Fas, tumour necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL)]-related pathways, or secret cytokines and chemokines [e.g. interferon (IFN)-γ, TNF-α] that have direct anti-viral activity or immunomodulatory effects. The cytolytic and proinflammatory activities of NK cells leads to the initial control of viral infections and efficient regulation of an adaptive immune response 14. Recent studies demonstrate clearly that activation of NK cells in response to viral infection also contributes to liver immunopathogenesis, both in rodent models and in patients with chronic HCV or HBV infection 8,9,15−17. In a cohort of immune-clearance (IC) patients, NK cells are activated and skew towards cytotoxicity, causing liver injury through the elevation of IL-12, IL-15 and IL-18 during active HBV infection 8. In addition, the activated NK cells are enriched preferentially in the livers of IC patients and mediate hepatocyte apoptosis by up-regulating TRAIL, Fas ligand (FasL) or NKG2D ligand 9,17. Although these studies have partially defined the role of NK cells in liver injury, the available data about NK cells in CHB patients are obtained primarily from the peripheral compartment 8. Detailed information regarding the correlation between hepatic NK cell activity and liver pathogenesis in CHB patients remains obscure.

Here, we comprehensively characterized the peripheral and hepatic NK cells in a cohort of hepatitis B e antigen (HBeAg)-positive individuals with chronic HBV infection and investigated the correlation of NK cell activity with serum alanine aminotransferase (ALT) levels and liver histological activity indices. We found that NK cells with hypercytolytic activity infiltrated the livers of IC patients and were correlated with liver injury during chronic HBV infection. Our findings may provide a strong rationale for further developing immunotherapeutic strategies for reducing viral load and subsequent hepatic damage.

Materials and methods

Study subjects

This study consecutively enrolled 70 patients with chronic HBV infection presenting to the First Affiliated Hospital of Fujian Medical University from September 2010 to January 2011. Inclusion criteria for the study were as follows: age (greater than 16 but less than 65 years); serum hepatitis B surface antigen (HBsAg)-positive; serum hepatitis B e antigen (HBeAg)-positive; serum hepatitis B e antibody (HBeAb)-negative; serum HBV DNA < 1000 copies/ml; and no immunosuppressive or anti-viral drug therapy within 6 months before sampling. Exclusion criteria were as follows: (1) concurrent hepatitis A virus (HAV), HCV, hepatitis D virus (HDV), hepatitis E virus (HEV), Epstein–Barr virus (EBV), cytomegalovirus (CMV) or human immunodeficiency virus (HIV) infection; (2) fatty liver disease (FLD); (3) alcoholic liver disease; (4) autoimmune liver disease; (5) drug-induced liver disease; (6) decompensated cirrhosis; (7) HCC or other malignancies; and (8) severe disease not correlated with malignancy (e.g. severe hypertension, cardiac disease, renal disease, psychosis, etc.).

In this study, the recruited patients were divided into two groups according to the phase of persistent HBV-infection: HBeAg(+) immune tolerance phase (IT; n = 36) and HBeAg(+) IC phase (n = 34). Minimal liver damage, normal ALT levels and active viral replication characterized the IT phase. The IC phase, which generally occurred following the IT phase, was characterized by increased liver injury and decreased viral replication 4,18. IT and IC patients were comparable with respect to age and sex, but ALT levels were significantly higher in IC patients (Table1).

Table 1.

Baseline characteristics of the study subjects

IT (n = 36) IC (n = 34) HCs (n = 10)
Age (years) 34 (20–55) 31 (17–46) 30 (25–38)
Sex (M/F) 26/10 22/12 6/4
ALT (IU/l) 28 (14–46) 149 (52–1733)a 18.5 (10–30)
HBV DNA (IU/ml) 2.11 × 105 (1.00 × 103−7.65 × 107) 4.06 × 105 (1.27 × 104−3.79 × 108) <1.0 × 103
Open in a new tab

Data are presented as mean and range.

P < 0·05 with respect to the immune-tolerant (IT) or healthy controls (HCs).

Ten age- and gender-matched healthy controls (HCs) were also enrolled. This study was approved by the ethics committee of Fujian Medical University and all participants (or their legal representatives) gave written informed consent before enrolment.

Blood samples were obtained from all enrolled subjects, and liver biopsy samples were collected from 24 patients with chronic HBV infection. The degree of hepatic inflammation was scored by an expert pathologist (Department of Pathology, Fujian Medical University, Fuzhou, China), according to the criteria described by Scheuer 19. We divided patients into two groups based on the score or grading of hepatic inflammation: patients with low to moderate hepatic inflammation (G1–2) and patients with significant hepatic inflammation (G3–4). Baseline characteristics of CHB patients who underwent percutaneous liver biopsies are shown in Table2.

Table 2.

Basic clinical data for the subgroup of immune-clearance (IC) patients with available liver biopsy samples

G1–2 (n = 10) G3–4 (n = 14)
Age (years) 31 (20–42) 32 (20–50)
Sex (M/F) 8/2 10/4
ALT (IU/l) 76 (52–233) 182 (72–652)
HBV DNA (IU/ml) 4.23 × 105 (1.27 × 103−3.79 × 107) 3.51 × 105 (1.63 × 104−7.65 × 108)
Open in a new tab

Cell surface staining and flow cytometric analysis

NK cells were identified in peripheral blood mononuclear cells (PBMCs) as CD3−/CD56+ 12. NK cells were identified phenotypically, as described previously, using a blood specimen incubated with the following antibodies (Becton Dickson Biosciences, Shanghai, China) 20: mouse anti-human CD3-fluorescein isothiocyanate (FITC) and mouse anti-human CD56-phycoerythrin (PE) monoclonal antibodies (mAbs). To test the NK activation status, cells were further labelled with mouse anti-human CD16-PE (Becton Dickson Biosciences) or mouse anti-human CD69-peridinin chlorophyll protein (PerCP) (Becton Dickinson Biosciences) 21. A PerCP-conjugated mouse immunoglobulin (Ig)G (Becton Dickinson) was used as an isotype control to detect non-specific binding. Next, red blood cells were lysed by fluorescence activated cell sorter (FACS) lysing solution, according to the manufacturer's instructions (Becton Dickinson). After washing with phosphate-buffered saline (PBS; Life Technologies, Shanghai, China), the harvested lymphocytes were fixed with 1% paraformaldehyde followed by flow cytometric analysis. FACS was performed using a two-laser BD FACSCalibur instrument equipped with CellQuest Pro software (Becton Dickinson). NK cell subpopulations were identified by gating on lymphocytes that were CD3−CD56+/CD16+ or CD3− CD56+/CD69+. Results were presented as the percentage of positive cells or the mean fluorescence intensity.

Intracellular cytokine staining

For TNF-α, IFN-γ, CD107a and perforin production, lymphocytes were stimulated directly with phorbol myristate acetate (PMA)/ionomycin or IL-12. Whole blood was cultured in complete medium (RPMI-1640 supplemented with 10% fetal bovine serum) with either (1) phorbol myristate acetate (PMA) (50 ng/ml; Sigma, St Louis, MO, USA) and ionomycin (1 ug/ml; Sigma) for 5 h or (2) human IL-12 (1000 ng/ml; PeproTech, Rocky Hill, NJ, USA) for 48 h. Unstimulated cells cultured in complete medium alone served as the negative control. Brefeldin A (Becton Dickinson) was added into the culture medium 1 h before the cells were harvested for phenotyping and intracellular cytokine staining. After washing, cells were collected and stained with the same antibodies against cell surface markers as described above. Next, FACS lysing solution was added to the culture medium as described above. Finally, the lymphocytes were washed, fixed and permeabilized by FACS permeabilization buffer (Becton Dickinson), according to the manufacturer's instructions. Cells were stained with mouse anti-human TNF-α, mouse anti-human IFN-γ, mouse anti-human CD107a or mouse anti-human perforin allophycocyanin (APC)-conjugated antibodies (Becton Dickinson). Mouse Ig isotype controls were included to assess non-specific binding. NK cell activity was measured by flow cytometry (Becton Dickson Biosciences) and expressed the level of cellular cytokine production (IFN-γ, TNF-α) and degranulation (perforin and CD107a).

Immunohistochemical staining

Immunohistochemical detection of CD56+ and CD3+ cells, as well as IFN-γ+ and CD107a+ levels in the liver tissues of HBV-infected individuals, was carried out as described with minor modifications 17,22. Briefly, formalin-fixed, paraffin-embedded liver tissues were cut into 4-μm sections. After deparaffinization and rehydration, sections were treated in 0·3% H2O2 to block endogenous peroxidase activity. Antigen retrieval was achieved via pressure cooking for 5 min in citrate buffer (pH 6·0). After a brief wash in water, sections were blocked with normal goat serum to suppress non-specific background staining. The sections were incubated with mouse monoclonal anti-human CD56+ (Zymed Laboratories/Life Technologies, Paisley, UK), anti-human CD3+ (Fuzhou Maixin Biotech, Fuzhou, China), anti-human IFN-γ+ (Beijing Biosynthesis Biotechnology, Beijing, China), and anti-human CD107a+ (Genentech, San Francisco, CA, USA) immunoglobulin for 60 min at 37 °C. Complete medium without primary antibody served as the negative control. Sections were washed again in PBS, and antibodies were detected using Elivision™ plus the polymer horseradish peroxidase (HRP) (mouse/rabbit) IHC kit (Fuzhou Maixin Biotech). Finally, sections were washed, stained with 3'3-diaminobenzidine tetrahydrochloride (DAB), dehydrated, placed into xylene and mounted in distrene–plasticizer–xylene (DPX) mountant (Sigma Aldrich, St Louis, Missouri, USA).

For quantitative analysis of hepatic CD3+ T cells and CD56+ NK cells, high-power fields (HPF, ×400) were used for counting positive cells in both portal and lobular areas. For semiquantitative analysis of IFN-γ+ and CD107a+ levels in liver tissues, low-power fields (LPF, ×200) were used for cellular assessment. The immunohistochemical staining intensity was analysed by Image-Pro Plus version 5·0 software (Rockville, MD, USA) and expressed as the integrated optical density (IOD).

Statistical analysis

Statistical analysis was performed by using SPSS version 19·0 software (Statistical Package of Social Sciences, Chicago, IL, USA). The between-group differences were analysed by χ2 test, analysis of variance (anova) or Mann–Whitney U-test, where appropriate. Multiple comparisons were made between different groups using the Kruskal–Wallis H non-parametric test. Correlations between variables were analysed using the Pearson correlation coefficient (r). For all tests, a two-sided P-value of less than 0·05 was considered to be statistically significant.

Results

NK cells accumulate selectively in the livers of IC patients

Analysis of NK cellular distribution by flow cytometry demonstrated that the proportion of circulating CD3−CD56+/CD16+ NK cells was significantly lower in chronic HBV-infected individuals compared with healthy subjects (P < 0·05) (Fig. 1a). In particular, fewer CD3−CD56+/CD16+ NK cells were present in peripheral blood samples from IC patients compared to IT patients (P < 0·05) (Fig. 1a). Chronic HBV-infected patients who underwent liver biopsy assessment showed further reductions in circulating NK cells, and patients with HAI scores of G3–4 had a lower percentage of NK cells than patients with scores of G1–2 (P > 0·05) (Fig. 1b).

Fig 1.

Fig 1

Open in a new tab

Flow cytometric analysis of peripheral CD3− CD56+/CD16+ natural killer (NK) cells in all enrolled subjects. (a) Pooled data show percentages of peripheral NK cells in immune-clearance (IC) patients (n = 34), immune-tolerant (IT) subjects (n = 36) and healthy control (HC) donors (n = 10). (b) Summarized data show the percentages of peripheral NK cells in partial IC patients with histological activity index (HAI) inflammation scores of G1–2 (n = 10) and those with HAI scores of G3–4 (n = 14). Each dot represents one individual. The horizontal bars indicate the median percentiles (*P < 0·05).

Liver biopsy assessment by immunohistochemistry was used to identify subsets of liver-infiltrating immune cells, specifically CD3+ T cells and hypercytolytic CD56+ NK cells 23. As shown in Fig. 2, low numbers of infiltrating CD3+ T cells and hypercytolytic CD56+ NK cells were identified in the liver biopsies with inflammation scores of G1–2. However, more CD3+ T cells and hypercytolytic CD56+ NK cells accumulated in livers with HAI scores of G3–G4. A quantitative analysis of liver-infiltrating CD3+ T cells and CD56+ NK cells further confirmed this observation and showed that a higher HAI score correlated with a significant increase in the number of liver-infiltrating immune cells (P < 0·05) (Fig. 2b). Although peripheral NK cells were reduced in IC patients (particularly those with a higher HAI score), our results indicate that NK cells were localized in liver tissues during the IC phase of infection.

Fig 2.

Fig 2

Open in a new tab

Immunohistochemical staining of CD56+ and CD3+cells in the liver sections obtained from patients with graded inflammation scores. Samples shown here were from partial immune-clearance (IC) patients with either low (G1–2, n = 10) or high (G3–4, n = 14) histological activity index (HAI) scores. (a) Hepatic CD56+and CD3+cells were stained brown and presented in both portal and lobular areas of livers (×400 magnification). (b) Quantitative analysis of CD56+ natural killer (NK) and CD3+ T cells in portal areas of liver tissues. Each bar graph indicates the mean ± standard deviation.

Peripheral NK cells are activated in IC patients

The CD69 expression levels were then measured to determine the activation state of peripheral blood NK cells. As shown in Fig. 3a, blood samples from IC patients showed a significant increase in the percentage of peripheral CD3−CD56+/CD16+ NK cells expressing CD69 compared to IT patients or HCs (P < 0·05). Moreover, patients with higher HAI scores (G3–4) had a higher percentage of peripheral CD3−CD56+/CD16+/CD69+ NK cells than those with lower HAI scores (P < 0·05) (Fig. 3b). Thus, NK cells are activated in IC patients, and an increase in activation state is correlated with a higher degree of liver inflammation.

Fig 3.

Fig 3

Open in a new tab

Flow cytometric analysis of CD69 expression on peripheral natural killer (NK) cells. CD3− CD56+/CD16 + NK cells were gated. (a) Pooled data show the percentage of peripheral NK cells expressing the CD69 activation marker from immune-clearance (IC), immune-tolerant (IT) patients and healthy controls (HCs). Boxes show the 10th, 25th, 75th and 90th percentiles and the median value (solid line) of each group (*P < 0·05). (b) Summarized data show the percentage of peripheral CD3−CD56+/CD16+cells expressing CD69 from partial IC patients who underwent percutaneous liver biopsies with either low (G1–2) or high (G3–4) HAI scores. Each dot represents one individual. The horizontal bars indicate the median percentiles (*P < 0·05).

NK cells from IC patients show enhanced cytokine and chemokine production

We further investigated TNF-α, IFN-γ, CD107a and perforin production by peripheral NK cells upon PMA/ionomycin or IL-12 stimulation. There was a significant increase in the production of TNF-α, IFN-γ, CD107a and perforin following PMA/ionomycin or IL-12 stimulation in IC patients compared to immune-tolerant (IT) patients (Fig. 4). We observed that the expression of CD107a on peripheral NK cells was similar among the three groups following PMA/ionomycin stimulation.

Fig 4.

Fig 4

Open in a new tab

Peripheral natural killer (NK) cells express the level of cellular cytokine [interferon (IFN)-γ, tumour necrosis factor (TNF)-α] and degranulation (perforin and CD107a) in immune-clearance (IC), immune-tolerant (IT) patients and healthy control (HC) subjects. CD3−CD56+/CD16+ NK cells were gated. (a) Representative dot-plots depict IFN-γ, TNF-α, perforin and CD107a production by peripheral NK cells among the three groups following phorbol myristate acetate (PMA)/ionomycin or interleukin (IL)-12 stimulation. Values in the upper right quadrants represent the percentage of NK cells expressing cytokine production and degranulation. (b) Pooled data show the percentage of peripheral NK cells expressing IFN-γ, TNF-α, perforin and CD107a among the three groups upon stimulation with PMA/ionomycin or IL-12. The median value (solid line) of each group (*P < 0·05).

Next, we compared the degranulation (CD107a expression), TNF-α and IFN-γ production of stimulated peripheral NK cells from patients with differing HAI scores. There was no statistically significant difference in the expression of perforin or TNF-α production among CHB patients, although data from patients with higher HAI scores (G3–4) exhibited a higher trend than data from the G1–2 group (data not shown). Upon PMA/ionomycin or IL-12 stimulation, there was a significant increase in the expression of both IFN-γ and CD107a by peripheral NK cells from patients with HAI scores of G3–4 compared to samples with lower scores (Fig. 5). In addition, liver biopsy specimens expressed detectable in-situ expression of CD107a and IFN-γ (Fig. 6). Both IFN-γ+ and CD107a+ expression levels were higher in liver tissues with high HAI scores (G3–4) compared to samples with lower scores (G1–2). Correlation analysis confirmed that expression of IFN-γ in situ correlated positively with levels of CD3+ T cells (r = 0·503, P = 0·012), and in-situ expression levels of CD107a correlated positively with levels of CD56+ NK cells (r = 0·682, P = 0·000). These results indicate that peripheral and hepatic NK cells from IC patients significantly up-regulate the expression of TNF-α, IFN-γ, CD107a and perforin, and expression levels are correlated with a higher HAI score.

Fig 5.

Fig 5

Open in a new tab

Flow cytometry analysis of natural killer (NK) cellular degranulation and interferon (IFN)-γ production in patients categorized by histological activity index (HAI). (a,b) Pooled data show the percentages of peripheral NK cells expressing IFN-γ (a) and CD107a (b) in two indicated groups upon stimulation with phorbol myristate acetate (PMA)/ionomycin or interleukin (IL)-12. CD3−CD56+CD16+ NK cells were gated. Each dot represents one individual. The horizontal bars indicate the median percentiles (*P < 0·05).

Fig 6.

Fig 6

Open in a new tab

Immunohistochemical analysis of interferon (IFN)-γ and CD107a, negative control in the liver tissues of chronic hepatitis B virus (HBV)-infected individuals categorized by histological activity index (HAI) score. (a) Representative in-situ immunohistochemical staining of liver sections from patients with either low (G1–2) or high (G3–4) HAI scores. Positively stained areas appear brown under × 200 magnification. (b) Semiquantitative analysis (expressed as integrated optical density, IOD) of IFN-γ or CD107a levels in livers of these subjects. Each dot represents one individual. The horizontal bars indicate the median percentiles (*P < 0·05).

The activation status and degranulation capacity of NK cells correlate positively with liver injury

Both the proportion of activated (CD69+) peripheral NK cells and the degranulation of NK cells following in-vitro stimulation were higher in IC patients with high HAI scores (G3–4) compared to patients with lower scores (Figs 6). The correlation analysis illustrated further that the proportion of activated (CD69+) peripheral NK cells correlated positively with serum ALT levels, which served as a surrogate marker of liver injury (Fig. 7a). There was also a statistically significant positive correlation between the degranulation capacity (CD107a expression in response to various stimuli) of peripheral NK cells and serum ALT levels (Fig. 7b). However, no correlations were found between the percentage of peripheral CD3−CD56+/CD16+ NK cells and serum ALT levels in HBV-infected individuals (data not shown). Although PMA/ionomycin and IL-12 induction of cytokine (i.e. TNF-α, IFN-γ and perforin) expression was elevated in NK cells from IC patients with high HAI scores (G3–4) compared to patients with lower scores (G1–2), no relevant statistical correlations were found between cytokine production and serum ALT levels (data not shown). There were also no direct correlations between serum HBV levels and serum ALT levels (data not shown). Finally, neither NK cell activation status (CD69+ expression) nor cytokine and chemokine production (TNF-α, IFN-γ, CD107a and perforin) have direct correlations with serum HBV DNA levels (data not shown). Together, these data suggest that activated NK cells are correlated positively with HBV-related liver injury, and the cytolytic activity of NK cells contributes more towards accelerating liver disease than to viral control.

Fig 7.

Fig 7

Open in a new tab

Correlation analysis of CD69 or CD107a expression on peripheral natural killer (NK) cells and serum ALT levels. (a) CD69 and (b) CD107a expression. Results are expressed as Pearson correlation coefficients. Each dot represents one individual.

Discussion

This study has characterized comprehensively the immune status of NK cells at different stages of chronic HBV infection, providing insights into the role of NK cells in CHB. It demonstrates clearly that (1) NK cells are activated and skewed towards cytolytic activity in IC patients, especially those with HAI scores of G3–4; (2) NK cells with hypercytolytic activity are enriched preferentially in livers of IC patients and not in the peripheral blood; and (3) the elevated NK cytolytic activity contributes more towards accelerating liver injury than to HBV elimination in IC patients.

In accordance with previous reports of NK cells in chronic HBV infection 8,15,17, we found that expression of the CD69 early activation antigen on NK cells was largely increased in IC patients compared to IT/healthy control (HC) subjects (Fig. 3a). In addition, the expression levels of CD69 on freshly isolated peripheral NK cells were higher in HBV-infected individuals with HAI scores of G3–4 compared to scores of G1–2 (Fig. 3b). Furthermore, the proportion of peripheral activated (CD69+) NK cells correlated positively with serum ALT levels (but not with serum HBV DNA levels; Fig. 7a). Thus, these findings indicate that the activation status of NK cells was associated closely with liver necroinflammation and injury instead of viral control.

Subsequently, we analysed the associations between NK cells activity and liver injury. Consistent with previous reports that polarization of NK cells towards cytotoxicity contributes to liver injury and viral persistence 8,15,16, we found that peripheral NK cells from IC patients (especially those with an HAI score of G3–4) were prone to CD107a degranulation, which indicated cytotoxic potential 24. Correlation analysis illustrated further that the expression of CD107a on peripheral NK cells correlated positively with serum ALT levels, and this correlation was independent of serum HBV load (Fig. 7b). Moreover, IC patients with HAI scores of G3–4 displayed increased expression of CD107a in the liver (Fig. 6b) and up-regulation of hepatic CD107a was associated closely with an increase in the number of liver-infiltrating NK cells (Pearson's r = 0·682; P = 0·000). Taken together, these data suggest that polarization of NK cells towards cytotoxicity is responsible for chronic liver immunopathology.

In the present study, NK cell activation was accompanied by an increase in the expression of TNF-α, IFN-γ and perforin. Although previous reports have shown that NK-cell production of TNF-α and IFN-γ exacerbate liver injury 9,25, no direct correlations were found between the production of proinflammatory cytokines (TNF-α or IFN-γ) and serum ALT levels in this study. Expression of perforin by NK cells also showed no relevant statistical correlations with serum ALT levels. To our knowledge, this discrepancy may result from the types of in-vitro stimulation methods used in this study, which could lead to differing results 24. Future studies will investigate whether cytokine (TNF-α or IFN-γ) production and their proinflammatory functions can be influenced by differential expression of various NK cell receptor ligands, such as TRAIL, FasL or NKG2D.

We also investigated the cause of NK cellular activation in IC patients. NK cell activity is tightly regulated by various cytokines such as IFN-γ, IFN-α, IL-12, IL-15 and IL-18 8,12,26. In this study, we found that in-vitro exposure to IL-12 boosted NK degranulation and production of proinflammatory cytokines in IC patients, especially those with higher HAI scores. Activated NK cells were skewed towards a cytotoxic response profile and preferentially infiltrated the livers of IC patients, where hepatic IFN-γ expression was largely elevated (although not correlated with the presence of NK cells; data not shown). We found that the expression of IFN-γ in liver biopsies was associated closely with liver-infiltrating CD3+ T cells, as demonstrated by correlation analysis. In support of data from previous studies that examined regulation of NK cells 12,27, we propose that liver-infiltrating T cells may produce T helper type 1 (Th1) cytokines (i.e. IL-12, IFN-γ, TNF-α), leading to recruitment of peripheral NK cells into the liver and enhancing the function of NK cells.

Furthermore, we analysed the correlations between NK cell activity and control of viral load. Interestingly, neither NK cell activation status (CD69+ expression) nor activity (TNF-α, IFN-γ, CD107a and perforin production) correlated with serum HBV DNA levels in this study. Our results indicate that although NK cell activity is enhanced in IC patients, it is not sufficient to eliminate virus infection. The well-established notion that HBV eradication is mediated mainly by an HBV-specific T cell response which is hampered during chronic HBV infection 27. In accordance with recent reports, activated NK cells cytolytically eliminate activated CD4+ T cells that affect CD8+ T cell function and exhaustion 28. Natural killer cell activation enhances immune pathology and promotes chronic infection by limiting CD8+ T cell immunity 29. In this study, we also found that activated NK cell function accelerated liver damage, but not sufficient to achieve viral clearance. Further study is needed to investigate whether NK cell function contributes to HBV-specific CD8+ T cell responses according to viral persistence in chronic HBV patients.

In summary, this study shows that NK cells are activated, skewed towards cytolytic activity, move from the periphery to infiltrate the liver and exacerbate hepatic injury without viral clearance during chronic HBV infection. Further studies are required to explore the best protocols that block NK cell-mediated liver injury by suppressing NK cytotoxicity as well as harnessing its non-cytolytic anti-viral effects.

Acknowledgments

The authors would like to thank L. Y. Gao and L. H. Chen (Fujian Medical University, Fuzhou, Fujian, China) for helpful analysis of liver pathology samples. This study was supported by the Fujian Province Natural Science Fund (grant no. 2013J01304).

Disclosures

The authors have no conflicts of interest to declare.

References

  1. Wright TL. Introduction to chronic hepatitis B infection. Am J Gastroenterol. 2006;101:S1–6. doi: 10.1111/j.1572-0241.2006.00469.x. [DOI] [PubMed] [Google Scholar]
  2. World Health Organization (WHO) Hepatitis B: Fact Sheet no. 204. Geneva: WHO; 2008. Available at: http://www.who.int/mediacentre/factsheets/fs204/en/ (accessed 22 July 2013) [Google Scholar]
  3. Lok AS. Chronic hepatitis B. N Engl J Med. 2002;346:1682–3. doi: 10.1056/NEJM200205303462202. [DOI] [PubMed] [Google Scholar]
  4. Rehermann B, Nascimbeni M. Immunology of hepatitis B virus and hepatitis C virus infection. Nat Rev Immunol. 2005;5:215–29. doi: 10.1038/nri1573. [DOI] [PubMed] [Google Scholar]
  5. Chang JJ, Lewin SR. Immunopathogenesis of hepatitis B virus infection. Immunol Cell Biol. 2007;85:16–23. doi: 10.1038/sj.icb.7100009. [DOI] [PubMed] [Google Scholar]
  6. Chisari FV, Ferrari C. Hepatitis B virus immunopathogenesis. Annu Rev Immunol. 1995;13:29–60. doi: 10.1146/annurev.iy.13.040195.000333. [DOI] [PubMed] [Google Scholar]
  7. Maini MK, Boni C, Lee CK, et al. The role of virus-specific CD8(+) cells in liver damage and viral control during persistent hepatitis B virus infection. J Exp Med. 2000;191:1269–80. doi: 10.1084/jem.191.8.1269. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Zhang Z, Zhang S, Zou Z, et al. Hypercytolytic activity of hepatic natural killer cells correlates with liver injury in chronic hepatitis B patients. Hepatology. 2011;53:73–85. doi: 10.1002/hep.23977. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Zou Y, Chen T, Han M, et al. Increased killing of liver NK cells by Fas/Fas ligand and NKG2D/NKG2D ligand contributes to hepatocyte necrosis in virus-induced liver failure. J Immunol. 2010;184:466–75. doi: 10.4049/jimmunol.0900687. [DOI] [PubMed] [Google Scholar]
  10. Miller JS, Verfaillie C, McGlave P. The generation of human natural killer cells from CD34+/DR− primitive progenitors in long-term bone marrow culture. Blood. 1992;80:2182–7. [PubMed] [Google Scholar]
  11. Racanelli V, Rehermann B. The liver as an immunological organ. Heptology. 2006;43:S54–62. doi: 10.1002/hep.21060. [DOI] [PubMed] [Google Scholar]
  12. Caligiuri MA. Human natural killer cells. Blood. 2008;112:461–9. doi: 10.1182/blood-2007-09-077438. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Notas G, Kisseleva T, Brenner D. NK and NKT cells in liver injury and fibrosis. Clin Immunol. 2009;130:16–26. doi: 10.1016/j.clim.2008.08.008. [DOI] [PubMed] [Google Scholar]
  14. Andoniou CE, Andrews DM, Degli-Esposti MA. Natural killer cells in viral infection: more than just killers. Immunol Rev. 2006;214:239–50. doi: 10.1111/j.1600-065X.2006.00465.x. [DOI] [PubMed] [Google Scholar]
  15. Oliviero B, Varchetta S, Paudice E, et al. Natural killer cell functional dichotomy in chronic hepatitis B and chronic hepatitis C virus infections. Gastroenterology. 2009;137:1151–60. doi: 10.1053/j.gastro.2009.05.047. [DOI] [PubMed] [Google Scholar]
  16. Ahlenstiel G, Titerence RH, Koh C, et al. Natural killer cells are polarized toward cytotoxicity in chronic hepatitis C in an interferon-alfa-dependent manner. Gastroenterology. 2010;138:325–35. doi: 10.1053/j.gastro.2009.08.066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dunn C, Brunetto M, Reynolds G, et al. Cytokines induced during chronic hepatitis B virus infection promote a pathway for NK cell-mediated liver damage. J Exp Med. 2007;204:667–80. doi: 10.1084/jem.20061287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Ganem D, Prince AM. Hepatitis B virus infection – natural history and clinical consequences. N Engl J Med. 2004;350:1118–29. doi: 10.1056/NEJMra031087. [DOI] [PubMed] [Google Scholar]
  19. Scheuer PJ. Classification of chronic viral hepatitis: a need for reassessment. J Hepatol. 1991;13:372–74. doi: 10.1016/0168-8278(91)90084-o. [DOI] [PubMed] [Google Scholar]
  20. Mania A, Kaczmarek M, Kemnitz P, et al. Alterations in NK cell phenotype in relation to liver steatosis in children with chronic hepatitis C. Inflammation. 2013;36:1004–12. doi: 10.1007/s10753-013-9632-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Varchetta S, Oliviero B, Francesca Donato M, et al. Prospective study of natural killer cell phenotype in recurrent hepatitis C virus infection following liver transplantation. J Hepatol. 2009;50:314–22. doi: 10.1016/j.jhep.2008.10.018. [DOI] [PubMed] [Google Scholar]
  22. Zhang Z, Chen D, Yao J, et al. Increased infiltration of intrahepatic DC subsets closely correlate with viral control and liver injury in immune active pediatric patients with chronic hepatitis B. Clin Immunol. 2007;122:173–180. doi: 10.1016/j.clim.2006.09.006. [DOI] [PubMed] [Google Scholar]
  23. Lopez-Vergès S, Milush JM, Pandey S, et al. CD57 defines a functionally distinct population of mature NK cells in the human CD56dimCD16+ NK-cell subset. Blood. 2010;116:3865–74. doi: 10.1182/blood-2010-04-282301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Rehermann B, Naoumov NV. Immunological techniques in viral hepatitis. J Hepatol. 2007;46:508–20. doi: 10.1016/j.jhep.2007.01.002. [DOI] [PubMed] [Google Scholar]
  25. Chen Y, Sun R, Jiang W, Wei H, Tian Z. Liver-specific HBsAg transgenic mice are oversensitive to Poly(I:C)-induced liver injury in NK cell- and IFN-gamma-dependent manner. J Hepatol. 2007;47:183–90. doi: 10.1016/j.jhep.2007.02.020. [DOI] [PubMed] [Google Scholar]
  26. Vivier E, Tomasello E, Baratin M, Walzer T, Ugolini S. Functions of natural killer cells. Nat Immunol. 2008;9:503–10. doi: 10.1038/ni1582. [DOI] [PubMed] [Google Scholar]
  27. Shimizu Y. T cell immunopathogenesis and immunotherapeutic strategies for chronic hepatitis B virus infection. World J Gastroenterol. 2012;18:2443–51. doi: 10.3748/wjg.v18.i20.2443. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Waggoner SN, Cornberg M, Selin LK, Welsh RM. Natural killer cells act as rheostats modulating antiviral T cells. Nature. 2011;481:394–8. doi: 10.1038/nature10624. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lang PA, Lang KS, Xu HC, et al. Natural killer cell activation enhances immune pathology and promotes chronic infection by limiting CD8+ T-cell immunity. Proc Natl Acad Sci USA. 2012;24:1210–5. doi: 10.1073/pnas.1118834109. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Clinical and Experimental Immunology are provided here courtesy of British Society for Immunology

RESOURCES