An effective interferon-gamma-mediated inhibition of hepatitis C virus replication by natural killer cells is associated with spontaneous clearance of acute hepatitis C in human immunodeficiency virus-positive patients

Authors


  • Potential conflict of interest: Dr. Rockstroh consults for and advises Abbott, Boehringer Ingelheim, Bristol-Myers Squibb, Gilead, Merck, ViiV Healthcare, GlaxoSmithKline, and Janssen.

  • This work was supported by the German Research Foundation (DFG; SFB/TR 57), the H. W. and J. Hector Foundation (grant no.: M42), a grant from the BMBF (German Ministry for Science and Education; 01KI0791), a NEAT Gilead research grant (program no.: LSHP-CT-2006-037570), the DZIF TTU Hepatitis Project 8.3., and the German Center for Infection Research (DZIF).

Abstract

Hepatitis C virus (HCV) coinfection is an increasing health problem in human immunodeficiency virus-positive (HIV+) individuals. However, a considerable proportion of HIV+ patients manage to overcome acute hepatitis C (AHC) spontaneously. In the present study, we analyzed the role of natural killer (NK) cells in modulating the course of AHC in HIV+ patients. Twenty-seven HIV+ patients with AHC (self-limited course: n = 10; chronic course: n = 17), 12 HIV+ patients with chronic hepatitis C (CHC), 8 HIV monoinfected individuals, and 12 healthy controls were studied. NK cells were phenotypically analyzed by flow cytometry. Interferon-gamma (IFN-γ) secretion, degranulation (CD107a), and anti-HCV (= inhibition of HCV replication) activity of NK subpopulations were analyzed using the HuH7A2HCVreplicon cell system. NK cell frequency did not differ significantly between HIV+ patients with chronic and self-limited course of AHC. However, we found NK cells from patients with self-limiting infection to be significantly more effective in inhibiting HCV replication in vitro than NK cells from patients developing CHC. Of note, antiviral NK cell activity showed no significant correlation with NK cell degranulation, but was positively correlated with IFN-γ secretion, and blocking experiments confirmed an important role for IFN-γ in NK cell-mediated inhibition of HCV replication. Accordingly, NK cells from patients that spontaneously cleared the virus displayed a stronger IFN-γ secretion than those developing chronic infection. Finally, we observed high expression of NKG2D and NKp46, respectively, to be associated with self-limiting course of aHCV. Accordingly, we found that blocking of these NK cell receptors significantly impaired antiviral NK cell activity. Conclusion: Our data suggest a strong IFN-γ-mediated antiviral NK cell response to be associated with a self-limited course of AHC in HIV+ patients. (Hepatology 2014;59:814–827)

Abbreviations
Ab

antibody

AHC

acute hepatitis C

ALT

alanine aminotransferase

anti-HCV

HCV antibody

AST

aspartate aminotransferase

CHC

chronic hepatitis C

DAPI

4′,6-diamidino-2-phenylindole

ELISA

enzyme-linked immunosorbent assay

E:T

effector/target ratio

FACS

fluorescence-activated cell sorting

HAART

highly active antiretroviral therapy

HCV

hepatitis C virus

HIV+

human immunodeficiency virus positive

IFN-γ

interferon-gamma

IL

interleukin

KIR

killer immunglobuline-like receptor

NCRs

natural cytotoxicity receptors

NK

natural killer

NKRs

NK cell receptors

PBMCs

peripheral blood mononuclear cells

RFI

relative fluorescence intensity

rhIL

recombinant human interleukin

Approximately one third of the human immunodeficiency virus (HIV)-infected population in Europe and the United States is coinfected with the hepatitis C virus (HCV), and this is increasing as a result of an epidemic of acute HCV infection among HIV-infected homosexual men.[1, 2]

HCV coinfection in HIV-positive individuals is associated with significantly faster rates of fibrosis and higher associated liver morbidity and mortality than in hepatitis C alone.[3-5]

However, a considerable proportion of HIV+ patients manage to overcome acute hepatitis C (ACH) spontaneously, indicating that even HIV-infected individuals can eliminate HCV through their residual immune system. Accumulating evidence suggests that innate immunity, in particular, natural killer (NK) cells, plays an important role in this context.[6-8]

NK cells constitute a major component of the intrahepatic lymphocyte pool. In contrast to the peripheral blood, which contains approximately 5%-10% NK cells, intrahepatic lymphocytes comprise approximately 30% NK cells, and the percentage of intrahepatic NK cells may increase to >50% in liver diseases.[9]

Infections with viruses such as HCV result in activation of NK cells, which can kill virus-infected cells without previous immunization. In addition, NK cells produce proinflammatory cytokines, which induce an antiviral state of host cells and play a critical role in the recruitment of T cells to sites of inflammation. Moreover, NK cells enhance the interactions between antigen-presenting cells and T cells and directly modulate the function of both CD4+ and CD8+ T cells.[10, 11]

Function of NK cells is regulated by interactions of NK cell receptors (NKRs), which, in general, can be divided in activating and inhibitory NKRs, and their respective ligands.[12] Activating NKRs include NKG2C and NKG2D, and the “natural cytotoxicity receptors,” NKp30/44/46. Inhibitory receptors comprise NKG2A and members of the “killer immunglobuline-like” receptor family.

Immunogenetic studies indicate that NK cells may influence the outcome of acute HCV infection as well as immunopathogenesis in chronic hepatitis C (CHC).[13-16] Accordingly, in vitro studies suggest that NK cells are able to recognize and kill HCV-infected hepatocytes.[17, 18] Recent studies in HCV monoinfection indicated that NK cells are activated in the acute phase of HCV infection.[6-8] More important, these studies provided the first data indicating that NK cells might affect the natural course of AHC. However, it remained unclear whether and how NK cells display activity against HCV or rather indirectly modulate HCV antibody (anti-HCV) immune responses. Moreover, regulation of NK cell activity may be even more complex in HIV-coinfected patients because HIV monoinfection has been shown to be associated with significant alterations of the NK cell pool.[19]

In the present study, we show that during AHC, NK cells display a significant in vitro anti-HCV activity, even in HIV-coinfected patients. More important, we demonstrate, for the first time, that effective interferon-gamma (IFN-γ)-mediated inhibition of HCV by NK cells is associated with a self-limited course of acute HCV infection.

Patients and Methods

Patients

A total of 27 HIV+ patients with AHC, all from the Bonn/Cologne area in Germany, were studied, including 10 patients with self-limited course of HCV infection and 17 that subsequently developed CHC. As controls, 12 HIV+ patients with CHC and 8 individuals with HIV monoinfection were analyzed. All patients were HIV RNA negative under effective highly active antiretroviral therapy (HAART). As an additional control, we studied 12 healthy, HIV)/HCV donors (Table 1).

Table 1. Patient Characteristics
 HIV+/aHCVHIV+/cHCVHIV+Healthy Controls
Acute-> Self-LimitedAcute-> Chronic
  1. a

    Number (no.) of cases (no./total in %).

  2. b

    Mean (range).

  3. Abbreviations: MSM, men who have sex with men; NA, not analyzed.

Number101712812
Male Sexa10 (100%)17 (100%)11 (92%)6 (75%)7 (58%)
Age (years)b39.1 (31-46)41.6 (34-49)47.8 (38-58)51 (35-66)54 (21-54)
Risk Factors     
MSMa9 (90%)16 (94%)6 (50%)6 (75%)
Others/Unknowna1 (10%)1 (6%)6 (50%)2 (25%)
Clinical data     
ALT U/Lb742 (23-3,208)651 (26-2,262)70 (4-181)26 (16-40)n.a.
AST U/Lb493 (16-1,638)348 (24-2,006)49 (20-102)20 (16-24)n.a.
HIV Status     
HIV RNA Negative (under HAART)a10 (100%)17 (100%)12 (100%)8 (100%)
CD4 Cells/μLb647 (358-873)698 (300-1,377)620 (182-1,151)737 (421-1,199)n.a.
CD8 Cells/μLb1,205 (407-1,794)1,118 (220-2,775)631 (254-2,249)1,128 (366-1,940)n.a.
HCV Status     
HCV Load (x 106 IU/mL)b15.9 (<0.1-69)10.7 (<0.1-69)4.9 (0.8-26)
HCV Genotypes (%):     
1a8 (80)12 (71)10 (83)
2a0 (0)0 (0)0 (0)
3a0 (0)0 (0)0 (0)
4a0 (0)5 (29)2 (17)
Undetermined Genotypea2 (20%)0 (0 %)0 (0 %)

Informed consent was obtained from all patients. The study had been approved by the local ethics committee of the University of Bonn (Bonn, Germany).

Flow Cytometry

For fluorescence-activated cell sorting (FACS) analysis, the following fluorochrome-labeled antibodies (Abs) were used: anti-CD3, anti-CD56, anti-CD57, anti-CD62L, anti-CD127, anti-CD158b, anti-CD158e, anti-CD161 and anti-NKp46 (BioLegend, Fell, Germany); anti-CD69, anti-NKG2A, anti-NKG2C, anti-NKG2D, and anti-IFN-γ were purchased from R&D Systems (Wiesbaden-Nordenstadt, Germany); anti-CD107a was purchased from BD Biosciences (Heidelberg, Germany); anti-NKp30 and anti-NKp44 were from Beckman Coulter (Krefeld, Germany); and eFluor670 was from eBioscience (Frankfurt, Germany). Samples were analyzed on a FACS Canto flow cytometer using the CellQuest Pro (BD Biosciences) and FlowJo 7.5 software packages (TreeStar Inc., Ashland, OR).

NK Cell Separation

NK cells were immunomagnetically separated from total peripheral blood mononuclear cells (PBMCs) by depletion of non-NK cells using MACS cell separation kits, following the manufacturer's recommendations (Miltenyi Biotec, Bergisch Gladbach, Germany). The purity of NK cells was >95%.

IFN-γ Secretion

NK cells were cultured in the presence or absence of recombinant human interleukin (rhIL)-12 (1 ng/mL; ebioscience, San Diego, CA) and rhIL-15 (10 ng/mL; ebioscience) for 18 hours. Then, brefeldin A (10 μg/mL; Sigma-Aldrich, St. Louis, MO) was added for another 4 hours, followed by intracellular staining with anti-IFN-γ and FACS analysis.

CD107a Degranulation and Killing Assay

Interleukin (IL)-2 (25 U/mL)-stimulated PBMCs were coincubated with HuH7A2HCVreplicon cells at an effector/target (E:T) ratio of 1:1 in the presence of anti-CD107a to assess degranulation, as described before.[36]

For the killing assay, isolated NK cells were cultured with IL-2 (25 U/mL) for 18 hours. Next, NK cells were coincubated with eFluor670-labeled HuH7A2HCVreplicon cells at different E:T ratios (1:1, 1:5, and 1:10). After 24 hours, cells were transfered to FACS tubes and 4′,6-diamidino-2-phenylindole (DAPI) was added before analysis. Specific lysis was calculated as (% DAPI + eFluor670 + [dead targets] − % spontaneous DAPI + eFluor670 + [dead targets])/(100 − % spontaneous DAPI + eFluor670 + [dead targets]).

HuH7A2HCVreplicon Cells

HuH7A2HCVreplicon cells[37] were kindly provided by V. Lohmann and R. Bartenschlager (University of Heidelberg, Heidelberg, Germany). Cells were grown in high-glucose Dulbecco's modified Eagle's medium (4.5 g/L) supplemented with glutamine (PAA Laboratories GmbH, Cölbe, Germany), 10% fetal calf serum, 1% nonessential amino acids (Biochrom AG, Berlin, Germany), and 1% penicillin/streptomycin (PAA). Blasticidin S hydrochloride (3 μg/mL) and G418 (1 mg/mL; PAA) were added to cells containing subgenomic replicons. HuH7A2HCVreplicon cells were passaged twice a week and were seeded at a dilution of 1:3.

Luciferase Assay

HuH7A2HCVreplicon cells (1 × 105) HuH7A2HCVreplicon cells were seeded in 48-well plates. After 3 hours, medium was removed and replicon cells were cocultured with sorted NK cells at different E:T ratios overnight in the presence of IL-2 (25 U/mL). In some experiments, the resulting supernatants were collected for determination of aspartate aminotransferase (AST) levels.

In addition, 1 × 105 HuH7A2HCVreplicon cells were cultured for 24 hours with the supernatants of the respective direct coculture experiments. The assay was performed using the Steady-Glo Luciferase Assay System (Promega, Mannheim, Germany) and measured with Tecan Infinite M200 (Tecan Group Ltd., Männedorf, Switzerland).

IFN-γ Enzyme-Linked Immunosorbent Assay

After coincubation of HuH7A2HCVreplicon cells with sorted NK cells (E:T ratio: 1:1) overnight supernatants were collected and IFN-γ concentrations were analyzed by enzyme-linked immunosorbent assay (ELISA), following the manufacturer's instructions (Diaclone, Sapphire Bioscience, Ann Arbor, MI).

Blocking Experiments

In blocking experiments, NK cells were preincubated with anti-NKp46, anti-NKG2D (R&D Systems), and anti-IFN-γ (10 μg/mL; BioLegend), respectively, or an isotype control.

Statistical Analysis

Statistical analyses were performed using GraphPad Prism software (Version 5.0a; GraphPad Software Inc., San Diego, CA). A two-sided P value <0.05 was considered significant.

Results

Frequency and Phenotype of NK Cells in aHCV

First, we assessed the potential effect of acute HCV infection on the number and phenotype of NK cells. To this end, we compared NK cells in the circulating blood of HIV+ patients with AHC (HIV/aHCV) with those of HIV monoinfected individuals. As further controls, a group of HIV+ patients with chronic HCV infection (HIV/cHCV) and healthy individuals, respectively, were studied.

HIV+ patients with AHC displayed the lowest frequency of circulating CD3CD56+ NK cells (HIV/aHCV: 7.5% ± 0.8%; HIV, 12.4% ± 4.0%; HIV/cHCV: 10.3% ± 1.1%; healthy, 18.0% ± 1.9%), but increased number of CD56Bright NK cells. However, this was statistically significant only in comparison to healthy controls (P ≤ 0.001; data not shown).

When patients with chronic and self-limited course of AHC were analyzed separately, no statistically significant differences were observed with respect to frequency and composition of the peripheral NK cell pool (Fig. 1A).

Figure 1.

Distribution and phenotype of NK cell subsets in HIV+ patients with a self-limited and chronic course of AHC. (A) Comparison of percentages of total peripheral NK cells (left panel) or the proportions of CD56Dim (middle panel), and CD56Bright (right panel) NK cells from HIV+ patients with a self-limited (n = 10) and chronic (n = 17) course of HCV infection. (B) Exemplary illustration of expression of CD57, CD62L, CD127, and CD161, respectively, in HIV+ patients with a self-limited and chronic course of HCV infection. (C) Frequency of total NK cells and NK cell subsets (CD56Dim and CD56Bright) expressing CD57, CD62L, CD127, and CD161, respectively, in HIV+ patients with a self-limited (n = 10) and chronic (n = 17) course of HCV infection. All patients were studied in the acute phase of hepatitis C. Horizontal lines indicate the median percentages. n.s., not significant.

Analyzing surface expression of NK cell differentiation/maturation markers, we found patients who were able to clear the virus to display significantly higher numbers of NK cells expressing CD62L than patients that subsequently developed CHC (Fig. 1B,C). No such differences were found for expression of the other studied markers (CD57, CD127, and CD161).

Circulating NK Cells in AHC Are Activated

Next, we studied whether acute HCV infection was associated with changes in NK cell activation status and functional capacity.

Analyzing ex vivo functional activity of NK cells before coincubation with target cells, we observed significantly stronger degranulation in patients with AHC than in the control groups (Supporting Fig. 1A). Moreover, we found patients with AHC to display the highest expression of the NK cell activation marker, CD69 (Supporting Fig. 1B).

When patients with AHC were stratified according to outcome of HCV infection, we observed low spontaneous NK cell degranulation (Fig. 2A), but strong CD69 expression (Fig. 2B), to be associated with self-limited course of infection.

Figure 2.

NK cells are activated in the acute phase of HCV infection. (A) Comparison of spontaneous degranulation (CD107 expression) of NK cells obtained from patients with a self-limited (closed circles) and chronic course (open circles) of acute HCV infection, whereas CD69 expression is given in (B). Horizontal lines indicate the median percentages. n.s., not significant.

A Strong Anti-HCV NK Cell Activity Is Associated With Spontaneous Clearance of Acute HCV Infection

Next, we compared NK cell functions in the acute phase of hepatitis C in patients with different outcomes of HCV infection.

After coincubation of IL-2-stimulated NK cells with HuH7A2HCVreplicon cells, we observed significantly higher frequency of degranulating CD56Bright NK cells in patients with self-limiting hepatitis C than in those subsequently progressing toward CHC (Fig. 3A). More important, we found NK cells from patients with self-limiting infection to be significantly more effective in inhibiting HCV replication in vitro than NK cells from patients developing CHC (Fig. 3B). NK cell antiviral activity was dependent on the ratio of HuH7A2HCVreplicon cells, and a significant inhibition of viral replication was still detectable at an E:T ratio of 1:20 (Fig. 3C).

Figure 3.

Effective NK cell-mediated inhibition of HCV replication is associated with spontaneous clearance of acute HCV infection. (A) Degranulation of IL-2-stimulated NK cells obtained from patients with a self-limited (closed circles) and chronic course (open circles) of acute HCV infection after coincubation with HuH7A2HCVreplicon cells. To study NK cell-mediated inhibition of HCV replication, NK cells were cocultured with HuH7A2HCVreplicon cells at an E:T ratio of 1:1. Then, luciferase activity was measured. Graph shows inhibition of luciferase activity (LU) in relation to HuH7A2HCVreplicon cells cultured alone in patients with a self-limited (closed circles) and chronic course (open circles) of acute HCV infection (B). (C) Comparison of NK cell-mediated inhibition of HCV replication in patients with a self-limited (closed circles; n = 5) and chronic course (open circles; n = 5) of AHC at different E:T ratios. (D) Correlation between frequency of CD107a+NK cells and in vitro inhibition of HCV replication. (E) AST levels in the supernatants for each coculture, whereas (F) illustrates specific lysis of HuH7A2HCVreplicon cells at different E:T ratios. All patients were studied in the acute phase of hepatitis C. *P ≤ 0.05; **P ≤ 0.01.

However, antiviral NK cell activity showed an inverse correlation with ex vivo degranulation (Supporting Fig. 2) and did not correlate with CD107a expression after coincubation with HuH7A2HCVreplicon cells (Fig. 3D), suggesting that noncytolytic mechanisms may be more important in the context of anti-HCV NK cell activity. Parallel analysis of AST levels, as a marker of hepatocyte lysis, in coincubation experiments confirmed this hypothesis because we did not observe any significant increase of AST, even at a high E:T ratio of 1:1 (Fig. 3E). This observation is in agreement with the results from the lysis assay (Fig. 3F), also showing only minimal specific killing of HuH7A2HCVreplicon cells at different E:T ratios.

Antiviral NK Cell Activity Is IFN-γ Dependent

Given the strong antiviral activity of IFN-γ, we therefore studied the potential role of this cytokine in NK cell-mediated inhibition of HCV replication. To this end, we first analyzed ex vivo IFN-γ secretion of unstimulated circulating NK cells. We observed significantly higher spontaneous secretion of IFN-γ in patients who cleared HCV infection than in those subsequently developing CHC (Supporting Fig. 3).

Moreover, IL-12-/IL-15-stimulated CD56Bright NK cells from resolvers displayed a significantly higher capacity to secrete IFN-γ, as compared to NK cells from patients with a chronic course of infection (Fig. 4A).

Figure 4.

IFN-γ is importantly involved in NK cell-mediated blocking of HCV replication. To assess the NK cell capacity to secrete IFN-γ, purified circulating NK cells were stimulated with IL-12 (1 ng/mL) and IL-15 (10 ng/mL) and then analyzed for IFN-γ expression by flow cytometry (A). (B) Correlation between frequency of IFN-γ+NK cells after cytokine stimulation and NK cell-mediated in vitro inhibition of HCV replication (% inhibition luciferase activity [LU]). In further experiments, NK cells from patients with self-limited (closed circles) and chronic course (open circles) of acute HCV infection were cocultured with HuH7A2HCVreplicon cells. Resulting supernatants were harvested and analyzed for IFN-γ by ELISA (C) or their capacity to block HCV replication in vitro (D). (E) Effect of IFN-γ blockade on anti-HCV activity of NK cells obtained from patients with spontaneous clearance of AHC and compares anti-HCV NK cell activity in patients with a self-limited (closed circles) and chronic course (open circles) of acute HCV infection in the presence of an IFN-γ-neutralizing Ab. All patients were studied in the acute phase of hepatitis C.

Of note, IFN-γ secretion was positively correlated with NK cell-mediated inhibition of HCV replication, supporting a role for IFN-γ in anti-HCV NK cell functions (Fig. 4B).

Thus, we next studied IFN-γ concentration after coincubation of NK cells with HuH7A2HCVreplicon cells. We found significantly higher levels of IFN-γ in supernatants of NK cells from patients with self-limiting course of AHC, as compared to patients developing chronic HCV infection (Fig. 4C). Accordingly, supernatants of NK cells from patients who were able to spontaneously clear the virus were significantly more effective in blocking viral replication than supernatants of NK cells isolated from patients who established CHC (Fig. 4D). In addition, we observed that coincubation with an IFN-γ-specific Ab significantly reduced anti-HCV activity of NK cells from patients with a self-limiting course of AHC, which further supported a role for IFN-γ in NK cell-mediated anti-HCV activity (Fig. 4E). Interestingly, we found that NK cells from patients who were able to clear the virus and from those that progressed toward chronic infection did not differ significantly with respect to in vitro inhibition of HCV replication after blocking of IFN-γ (Fig. 4E).

NKG2D and NKp46 Modulate Antiviral NK Cell Activity and May Affect Outcome of AHC

NK cell activity is tightly regulated by the integration of signals delivered by a diverse array of NK cell receptors. To clarify whether more effective NK cell anti-HCV activity in resolvers was related to differential NKR expression, we next analyzed surface expression of NK cell receptors on NK cells directly ex vivo.

Regarding expression of NKp30, NKp44, and CD158b, respectively, no significant differences were found between patients with a chronic versus self-limited course of infection (data not shown).

In contrast, we observed a significantly higher frequency of the NK cells expressing NKG2A, but lower numbers of NKG2C+ and CD158e+ NK cells, respectively, in patients with a self-limiting course of AHC than in patients who subsequently developed chronic HCV infection (Fig. 5). However, neither expression of CD158e nor expression of NKG2A showed any significant correlation with antiviral NK cell functions (Fig. 5B,C), and frequency of NKG2C-positive NK cells was negatively correlated with antiviral NK cell activity (Fig. 5E), suggesting that these NK cell receptors are not directly involved in inhibition of HCV replication.

Figure 5.

Expression of NKG2A, NKG2C, and CD158e differs between patients with a self-limited and chronic course of AHC. (A) Freshly isolated PBMCs were stained with anti-CD3, anti-CD56, and anti-CD158e and NKG2A and NKG2C, respectively. CD3CD56+ NK cells were then gated for quantification of the respective NK cell receptor. (B) Frequency of circulating CD158e+ NK cells in patients with a self-limited (closed circles) and chronic course (open circles) of acute HCV infection, and the correlation between CD158e expression and NK cell anti-HCV activity is given. (C) Frequency of NKG2A+ NK cells in patients with a self-limited (closed circles) and chronic course (open circles) of acute HCV infection, and the correlation between NKG2A expression and NK cell anti-HCV activity is illustrated. (D) Frequency of NKG2C-expressing NK cells in AHC, whereas (E) illustrates the correlation between frequency of circulating NKG2C+ NK cells and in vitro inhibition of HCV replication (right diagram) and the proportion of IFN-γ+ NK cells (left diagram) in HIV+ patients with acute hepatitis C.

Regarding NKp46, we found higher expression of this NK cell receptor on CD56Bright NK cells in patients who spontaneously cleared HCV than in those who subsequently developed CHC (Fig. 6A,B). Of note, both frequency of total NKp46+ NK cells as well as density of NKp46 expression on CD56Bright NK cells were positively correlated with antiviral function of NK cells (Fig. 6C,D), and blocking experiments indicated a functional role for NKp46 (Fig. 6E).

Figure 6.

Role of NKp46 in anti-HCV activity of NK cells from HIV+ patients with AHC. PBMCs were stained with anti-CD3, anti-CD56, and anti-NKp46 (A). (B) Frequency of NKp46-positive NK cells (left graph) as well as the density of NKp46 surface expression (RFI; right graph) in patients with a self-limited (closed circles) and chronic course (open circles) of acute HCV infection. (C) Correlation between frequency of circulating NKp46-expressing NK cells and in vitro inhibition of HCV replication (left diagram) and proportion of IFN-γ+ NK cells (right diagram) in HIV+ patients with AHC, whereas the correlation between the density of NKp46 expression (RFI) on CD56Bright NK cells and in vitro inhibition of HCV replication is depicted in (D). (E) Effect of a neutralizing, NKp46-specific Ab on anti-HCV activity of NK cells. Column represents mean and standard error of the mean from four independent experiments. *P < 0.05. All patients were studied in the acute phase of hepatitis C.

Finally, we found frequency of NKG2D-positive NK cells as well as density of NKG2D surface expression relative fluorescence intensity (RFI; Fig. 7A,B) to be significantly higher in resolvers than in patients with chronic evolution of hepatitis. Importantly, both parameters showed a positive correlation with IFN-γ production as well as inhibition of HCV replication (Fig. 7C,D). Functional experiments confirmed a role of NKG2D in anti-HCV NK cell activity, because preincubation with a NKG2D-specific Ab significantly reduced NK cell-mediated inhibition of HCV replication (Fig. 7E).

Figure 7.

NKG2D is involved in anti-HCV activity of NK cells from HIV+ patients with AHC. PBMCs were stained with anti-CD3, anti-CD56, and anti-NKG2D (A). CD3CD56+ NK cells were then gated for quantification of NKG2D-expressing NK cells and density of NKG2D surface expression (B) in patients with a self-limited (closed circles) and chronic course (open circles) of acute HCV infection. (C) Correlation between frequency of circulating NKG2D+ NK cells and in vitro inhibition of HCV replication (left diagram) and proportion of IFN-γ+ NK cells in HIV+ patients with AHC (right diagram), whereas the correlation between surface density of NKG2D (RFI) and NK cell activity is given in (D). (E) Preincubation with anti-NKG2D significantly blocks anti-HCV activity of NK cells. Column represents mean and standard error of the mean from four independent experiments. *P < 0.05. All patients were studied in the acute phase of hepatitis C.

Discussion

A substantial proportion of de novo HCV-infected patients are able to spontaneously eliminate the virus, even in the presence of HIV coinfection, but the factors determining the natural course of AHC are only incompletely defined.

Increasing data indicate an important role for NK cells,[6-8, 13-16, 20-22] and recent studies suggested that distinct populations of NK cells may be involved in the early control and clearance of HCV infection.[6-8]

Here, we show that during the acute phase of HCV infection, NK cells effectively block HCV replication in vitro. More important, we provide first evidence that anti-HCV NK cell activity during the acute phase of hepatitis C is associated with different outcomes of the infection, because we found NK cells from patients clearing the virus to be significantly more effective in blocking viral replication, as compared to NK cells obtained from patients subsequently progressing to CHC.

This is in contrast to previous studies in HCV+/HIV patients[6-8] that observed significant alterations of NK cell phenotype and activity, but did not find any significant associations between outcome of acute HCV infection and NK cell functions. However, these studies used the hematopoietic K562 and/or 221 cell lines as targets,[6-8] which might not be ideal for studying NK cell activity against a hepatotropic virus such as HCV. To overcome this obstacle, we used the well-established HuH7A2HCVreplicon cell system.

Noncytolytic effector functions are considered critical for antiviral activity of HCV-specific CD8+ T cells.[23] In line with these data, we recently demonstrated a role for IFN-γ secretion in NK cell-mediated inhibition of viral replication in CHC.[20] In the current study, we present several lines of evidence indicating that IFN-γ secretion by NK cells may also play an important role in modulating the outcome of acute HCV infection.

First, we show that NK cells from resolvers display a significantly stronger IFN-γ secretion after coincubation with HuH7A2HCVreplicon cells than patients with a chronic course of infection. Second, we found the frequency of IFN- γ+ NK cells to be positively correlated with antiviral activity of NK cells, whereas no such correlation was observed for NK cell degranulation. Third, we demonstrated that supernatants of NK cells effectively block HCV replication, indicating that anti-HCV NK cell activity is not dependent on direct cell-cell contact. Fourth, we could show that blocking of IFN- γ with a specific Ab significantly reduced the antiviral function of NK cells. Of note, blocking of IFN- γ neutralized the differences in anti-HCV activity between NK cells from patients who were going to clear the virus and those developing CHC.

Taken together, these data suggest that a robust IFN-γ secretion enables NK cell to effectively fight HCV, thereby favoring the self-limited course of AHC.

This resembles findings in studies on HCV-specific CD8+ T cells showing that during AHC, the first decline of HCV RNA correlates with the initiation of IFN-γ production of virus-specific CD8+ T lymphocytes[24] and the appearance of IFN-γ-producing CD8+ T cells in the liver.[25, 26] Of note, intrahepatic accumulation of IFN-γ-secreting HCV-specific CD8+ T cells has also been observed in the absence of significant liver disease, which further supports the noncytopathic nature of this pathway.[26]

On the other hand, we found patients with AHC to display significantly higher levels of alanine aminotransferase (ALT) and AST, respectively, compared to patients with chronic HCV infection, indicating the presence of cytolytic effector functions. However, in our in vitro studies, a strong NK cell-mediated inhibition of HCV replication was observed in the presence of only minimal cytolysis of HuH7A2HCVreplicon cells and mild AST elevations, even at the highest E:T ratio, suggesting a limited role of NK cell cytotoxicity in AHC. Accordingly, patients with a self-limiting and chronic course of acute HCV infection did not differ significantly with respect to serum levels of AST and ALT, respectively.

However, NK cells have been suggested to be more activated in the intrahepatic compartment, and cytotoxic activity of liver-infiltrating has been proposed to be associated with liver cell damage in CHC.[27] Of note, CD56Bright NK cells, which represent the dominant intrahepatic NK cell subset, displayed a higher degranulation in patients with self-limited infection, as compared to patients developing CHC. Thus, we cannot completely exclude a role for intrahepatic CD56Bright NK cells in liver inflammation during the acute phase of HCV infection.

We and others recently showed that the activating NK cell receptor, NKp46, might be involved in antiviral NK cell activity in patients chronically infected with HCV and demonstrated intrahepatic frequency of NKp46-expressing NK cells to be inversely correlated with HCV loads.[20, 28] In line with these data, we found that during AHC, the frequency of NKp46+ NK cells was positively correlated with in vitro inhibition of HCV replication. Furthermore, surface expression of NKp46 was higher in patients that were able to clear the virus, as compared to those that became chronically infected, although this effect was statistically significant only in the CD56Bright NK cell subset. Furthermore, blocking of NKp46 significantly reduced the capability of circulating NK cells to block HCV replication, suggesting a role of NKp46 in the regulation of anti-HCV immune responses also in the acute phase of HCV infection.

However, the strongest association between NKR expression and outcome of AHC was found for the activating NKG2D receptor with significantly higher surface expression and frequency of NKG2D+ NK cells in patients with a self-limited course of HCV infection. In addition, there was a strong positive correlation between NKG2D expression and IFN-γ secretion as well as inhibition of HCV replication. Accordingly, we found that blocking of NKG2D significantly reduced antiviral NK cell activity, supporting an important role of NKG2D in NK cell-mediated immune responses against HCV. Of note, major histocompatibility complex class I chain-related molecules, the natural ligands for NKG2D, have been shown to be induced on HCV-infected hepatocytes.[29] Thus, it is conceivable that NK cells can identify and kill HCV-infected hepatocytes in a NKG2D-dependent fashion. Moreover, in chronic HCV monoinfection, Sène et al. recently identified HCV NS5A-induced down-regulation of NKG2D, resulting in impaired NK cell cytolytic activity and IFN-γ secretion, as a mechanism potentially involved in viral escape.[29] However, reports on NKG2D in CHC are controversial, and thus further studies are needed to clearly delineate the role of this NK cell receptor in HCV infection.[29-32]

Both patient subgroups (resolvers and chronics) did not differ significantly with respect to age, gender distribution, route of HCV transmission, CD4+ T cell counts, viral loads, or comorbidities. Patients with chronic course of HCV infection comprised a higher number of individuals with a HCV genotype 4 infection, but a lower number of individuals carrying an IL28B C/C genotype. However, none of these parameters was associated with NK cell phenotype and functions, respectively (data not shown). Thus, it remained unclear why patients with a self-limited and chronic course of AHC displayed the observed differences in NK cell phenotype and function.

Another important issue relates to the question of whether our results are specific for HIV/HCV coinfection or can be extrapolated to patients infected with HCV alone. In line with previous reports, we observed HIV monoinfection to be associated with significant phenotypic and functional alterations of the NK cell compartment. Thus, some of our observations may, at least in part, be attributable to HIV coinfection, despite the fact that all our HIV- patients were HIV RNA negative under efficient HAART and displayed robust CD4+ T cell counts above 400 cells/μL. Indeed, increased expression of NKG2D and/or NKp46 as a potential predictor of self-limited course of AHC is somewhat in contrast to reports in HCV monoinfected patients[6-8] and therefore may represent an HIV-associated phenomenon. Accordingly, dys-regulated expression of NKG2D and NKp46 has been proposed as a mechanism potentially involved in impaired NK cell functions in HIV RNA+ patients,[33, 34] and effective antiretroviral therapy has been shown to be associated with normalization of NK cell receptor expression.[34] Thus, higher frequencies of NKG2D and/or NKp46-expressing NK cells in resolvers may characterize HIV patients that are more likely to clear HCV as the result of a better reconstitution of NK cell functions under HAART. However, data obtained in HCV monoinfection are still conflicting. For instance, Alter et al. found low expression of NKG2D to be associated with spontaneous viral clearance,[6] whereas Amadei et al. did not.[7] Moreover, functional data obtained in monoinfected patients support a role for these two receptors in modulating anti-HCV immune responses irrespective of HIV coinfection.[20, 28, 29, 35]

Taken together, our data indicate that during the acute phase of hepatitis C, a robust IFN-γ-mediated inhibition of HCV replication by NK cells is associated with spontaneous clearance of HCV infection, underpinning the important role of NK cells.

Acknowledgment

The authors extend their grateful thanks to Claudia Zwank for her perfect technical assistance and to all the participants.

Ancillary

Advertisement