Negative Regulation of NK Cell Activities by Inhibitory Receptor CD94/NKG2A Leads to Altered NK Cell-Induced Modulation of Dendritic Cell Functions in Chronic Hepatitis C Virus Infection

Subjects

Ten healthy volunteers and 15 patients with chronic hepatitis C were enrolled in this study after informed consent had been obtained. The patients were positive for HCV RNA and were histologically diagnosed as having mild or moderate chronic hepatitis. They did not display any evidence of other types of liver diseases.

Isolation of PBL populations

PBL, obtained from healthy volunteers or patients with chronic hepatitis C infection, were isolated by Ficoll-Hypaque density centrifugation. Resting NK cells (CD56⁺CD3⁻; >90% purified) as well as naive CD4⁺ T cells (CD45RA⁺RO⁻; >95% purified) were further isolated using StemSep-negative selection systems (Stem Cell Technologies, Vancouver, Canada).

Generation of monocyte-derived DCs

Monocyte-derived DCs were generated from the peripheral venous blood of healthy volunteers as described previously (22, 23). In brief, PBMCs were centrifuged on a Percoll (Sigma-Aldrich, St. Louis, MO) gradient. The light density fraction floating on the middle layer, which contained highly purified monocytes, was seeded in 24-well culture plates at a density of 5.0 × 10⁵/well. After 45 min of incubation at 37°C, nonadherent cells were removed and the adherent cells were cultured in RPMI 1640 medium (Invitrogen Life Technologies, Gaithersburg, MD) containing 10% FBS, 10 U/ml penicillin/streptomycin, and 2 mM l-glutamine and supplemented with GM-CSF and IL-4 (10 ng/ml: PeproTech, Rocky Hill, NJ) for 6 days to generate DCs of an immature phenotype.

Cell lines

The Hep3B human hepatoma cell line was purchased from the American Type Culture Collection (Manassas, VA). The Huh7 human hepatoma cell line was a generous gift from Dr. M. Nanba (Institute for Molecular and Cellular Biology, Okayama University, Okayama, Japan). These hepatic cells were propagated in DMEM supplemented with 10% heat-inactivated FBS, antibiotics, and antimycotics (Invitrogen Life Technologies) in a humidified atmosphere of 5% CO₂ at 37°C. We also used the human chronic myeloid leukemia cell line K562 (American Type Culture Collection), propagated in RPMI 1640 supplemented with 10% heat-inactivated FBS, antibiotics, and antimycotics. Human nontransformed hepatocytes (PH) were purchased from Applied Cell Biology Research Institute (Kirkland, WA) and cultured in CS-C complete medium according to the manufacturer’s instructions.

NK cells, hepatic cells, and DC cocultures

Resting NK cells and DCs were seeded in the same wells of 24-well plates and cultured with human hepatic cells for 48 h (1 × 10⁵ of each type of cell/well). As a control, DCs were plated in 24-well culture plates. A Transwell insert was also used to prevent direct contact of the mixtures of NK cells, human hepatic cells, and DCs. In these Transwell systems, the soluble factors could freely pass through a microporous polycarbon membrane (0.4 μm). DCs were plated in an upper chamber, and NK cells and Hep3B cells were seeded in the bottom chamber at a ratio of 1:1 for 48 h. As a control, DCs (upper chamber) and NK cells alone (bottom chamber) were plated in 24-well culture plates.

Stimulation of DCs by coculture supernatants of NK/third-party cells

NK cells (1 × 10⁵ cells/well) were seeded in 24-well plates and stimulated with Hep3B cells or K562 cells (1 × 10⁵ cells/well) at a ratio of 1:1 for 24 h. In the case of nontransformed normal hepatocytes (PH), NK cells were stimulated with or without 50 ng/ml IL-2 for 24 h and then cocultured with PH (1 × 10⁵ cells/well) for 24 h. DCs were cultured in presence of the obtained supernatants of NK/hepatic cells or K562 for 24 h and were subjected to further analysis. In some experiments, anti-NKG2A mAb Z199 or isotype-matched control Ab was added during the cocultures of NK cells and Hep3B. Z199 mAb was previously confirmed to block NKG2A-mediated signal (24). The supernatant of NK/Hep3B cocultures was also treated with anti-IL-10 and TGFβ-neutralizing Ab (R&D Systems, Minneapolis, MN) and used for DC stimulation for 24 h.

Flow cytometric analysis

DCs (5 × 10⁵/ml) were washed and resuspended in PBS containing 1% BSA and 0.05% NaN₃ and incubated with a series of mAbs at 4°C for 30 min. The following mAbs were used for immunofluorescent staining on DCs: anti-CD11c, anti-CD40, anti-CD86, anti-HLA-DR (BD Pharmingen, San Diego, CA), anti-CD83, and anti-CCR7 (Coulter Immunotech, Marseille, France). DCs were positively gated from triple cell cultures by CD11c expression.

For NK cell staining, the cells were washed and incubated at 4°C for 30 min in PC5-labeled CD56 mAb (Coulter Immunotech) and PE-labeled mAbs as follows: EB6 (KIR2DL1/2DS1), GL186 (KIR2DL2-3/2DS2), p50.3 (KIR2DS4), DX9 (KIR3DL1), HP-3B1 (CD94), Z199 (NKG2A), and 1D11(NKG2D). EB6, GL186, p50.3, HP3B1 and Z199 were purchased from Coulter Immunotech. DX9 was purchased from BD Pharmingen. 1D11 was kindly provided by Dr. V. Groh and Dr. T. Spies (Fred Hutchinson Cancer Research Institute, Seattle, WA) and was used as previously reported (25). The cells were then washed twice and fixed with 2% paraformaldehyde solution. The cells were analyzed by flow cytometric analysis using a FACScan system (BD Pharmingen), and data analysis was performed using CellQuest software.

The expression of HLA class I as well as HLA-E was examined on hepatic cells and K562 by using w6/32 or 3D12 (IgG1), respectively. 3D12 was kindly provided by Dr. E. Geraghty (Fred Hutchinson Cancer Research Institute) and was used as reported previously (26).

Measurements of cytokine production

The culture supernatants of IFN-γ, TNF-α, IL-10, and TGFβ were examined using ELISA kits according to the manufacturer’s instructions (IFN-γ, TNF-α, and IL-10; Endogen, Woburn, MA; TGFβ, R&D Systems).

Chemotactic assay

DC migration capacity was evaluated using the chemotaxis microchamber technique as described previously (8). In brief, human recombinant MIP-3β/CCL19 (R&D Systems) was diluted to 100 ng/ml RPMI 1640 medium and added to the bottom wells and then DCs (1 × 10⁵/well) were applied to the upper wells, which were separated from the bottom wells with 8-μm pore polycarbonate filters (Costar, Corning, NY). After 2 h of incubation at 37°C, the cells that migrated to the bottom chamber were collected and counted manually. As a control, the lower compartment of the control chambers contained medium alone.

Intracellular cytokine analysis

Intracellular staining was performed as described previously (22, 23). In brief, NK cells or CD4⁺ T cells were pretreated with 1 μl/ml GolgiPlug (BD Pharmingen) for 4 h at 37°C. At the end of the incubation period, NK cells and naive CD4⁺ T cells were stained with PC5-labeled CD56 mAb and CD4 mAb (Beckman Coulter) for 30 min, respectively. The cells were then fixed and permeabilized with Cytofix/Cytoperm buffer (BD Pharmingen) for 15 min at room temperature. Permeabilized cells were stained with FITC-labeled anti-IFN-γ mAb and PE-labeled anti-IL-10 mAb (mouse IgG1; BD Pharmingen) or isotype-matched control IgG. The stained cells were analyzed by flow cytometry.

Proliferation assay

DCs were cultured with 10 μg/ml LPS or the supernatant of NK cells cultured with tumor cells (Hep3B and K562) for 48 h, and then were seeded in round-bottom 96-well plates for use as stimulator cells (10–0.33 × 10³/well). The DCs were cultured at graded numbers with the responder naive CD4⁺ cells (1.0 × 10⁵/well) for 72 h. The cocultured cells were pulsed with 1 μCi/well [³H]thymidine for 16 h of incubation and collected onto a glass fiber filter. [³H]Thymidine incorporation was quantified using a beta-plate liquid scintillation counter. The results were expressed as the mean cpm in triplicate cultures.

CD4⁺ T cell responses against recall Ag

DCs stimulated with the supernatant of the NK cell/Hep3B coculture medium for 24 h were given influenza Ag of the Texas 1/77 strain (Chemicon International, Temecula, CA) at 10 μg/ml (27) and were incubated overnight at 37°C. Autologous CD4⁺ T cells were stimulated by Ag-pulsed or unpulsed DCs for 48 h and subjected to proliferation or intracellular cytokine assay as described above.

Cytolytic assay

Target cells labeled with ⁵¹Cr were incubated in the NK cells for 4 h at various E:T ratios. The supernatants were obtained after the incubation and subjected to gamma counting. The maximum or spontaneous release was defined as counts from samples incubated with 5% Triton X-100 or medium alone, respectively. Cytolytic activity was calculated with the following formula: percent lysis = (experimental release − spontaneous release) × 100/(maximum release − spontaneous release). The spontaneous release in all assays was <20% of the maximum release.

Statistical analysis

Comparisons between groups were analyzed using the compared t test with Welch’s correction or Mann-Whitney U test. Differences were considered to be significant when the p < 0.05.

DCs gained matured phenotypes in the presence of N-NK and Hep3B hepatoma cells

To investigate whether NK cells could modulate DCs in the presence of human hepatic cells, monocyte-derived DCs were cultured with both NK cells and Hep3B human hepatoma cells for 48 h and analyzed for their phenotypes by flow cytometry. Unstimulated DCs displayed immature phenotypes, such as with moderate expression of costimulatory molecules (CD40 and CD86) as well as MHC class II (MHC II) and no expression of CD83 (Fig. 1⇓A). Neither NK cells nor Hep3B cells alone had any impact on DC phenotypes under our experimental conditions (Fig. 1⇓A and data not shown). DCs cultured with both N-NK and Hep3B cells exhibited the matured phenotype, exhibiting the elevated expression of CD40, CD86, MHC II, and CD83, at comparable levels to LPS-stimulated DCs (Fig. 1⇓). These results indicated that the coexistence of NK cells and Hep3B, but not NK cells alone, can induce maturation of DCs. In striking contrast, DCs cultured with both HCV-NK and Hep3B cells displayed little phenotypic changes of CD40, CD86, and CD83, whereas they showed up-regulation of MHC II expression.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Phenotypic change of DCs cocultured with N-NK and HCV-NK in the presence of third-party cells. NK cells (1 × 10⁵/well) derived from healthy volunteers (N) or chronic HCV-infected patients (HCV) were cocultured with allogeneic DCs from healthy donors at a ratio of 1:1 in the absence (NK) or presence (NK/Hep3B) of Hep3B cells (1 × 10⁵/well) for 48 h. As negative and positive controls, DCs were unstimulated (−) or were stimulated with 10 ng/ml LPS for 48 h, respectively. FACS analysis was performed to evaluate the expression of CD40, CDE86, MHC II, and CD83 on DCs after positively gating CD11c⁺CD56⁻ cell population (filled histograms). Open histograms represent the staining of control Ab. All experiments were performed at least three times and representative results (A) as well as the statistical analysis (B) are shown as the MFI of the staining cells. ∗, p < 0.05.

These observed effects of DC maturation do not require direct cell-cell contact of DCs and NK/Hep3B cells, because the insertion of a Transwell membrane between DCs and cocultured NK/Hep3B cells induced DC maturation at levels similar to mixed cultures of DCs, NK cells, and Hep3B (Fig. 2⇓A). Next, we examined whether the supernatant of NK cells/Hep3B culture could affect the DC phenotypes. As shown in Fig. 2⇓, B and C, up-regulation of CD40, CD86, and MHC II was observed on DCs treated with the supernatant of N-NK/Hep3B cocultures. These results further confirmed that direct cell-to cell contact is not necessary for NK cells to regulate DC maturation upon human hepatic cells. In contrast, the supernatant of HCV-NK/Hep3B up-regulated MHC II expression on DCs, but had little effect on CD40 or CD86 expression. The differences of DC maturation status between N-NK and HCV-NK were similarly observed when NK cells were cocultured with another human hepatoma cell line, Huh7 (data not shown). In contrast, the maturation of DCs was similarly observed in the coculture of either N-NK or HCV-NK with K562 cells (Fig. 2⇓, B and C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Soluble factors produced from NK cells/Hep3B mixtures are responsible for promoting DC maturation. A, NK cells, in the presence or absence of Hep3B (NK/Hep3B or NK, respectively), were added to DC cultures either directly (NK/Hep3B-Mix) or to the bottom compartment separately through 0.4-μm pore size membrane (NK/Hep3B-Transwell) for 48 h. The expression of CD40, CD86, and MHC II on DCs was examined by flow cytometry. B and C, NK cells (1 × 10⁵) from healthy donors (N) and HCV-infected individuals (HCV) were cocultured in the absence (NK) or presence of Hep3B (NK/Hep3B) or K562 (NK/K562) at a ratio of 1:1 for 24 h. DCs from healthy donors were stimulated with or without ((−)) the supernatant obtained from the cocultured medium. The expression of CD40, CD86, and MHC II (filled histogram) were evaluated by FACS analysis. Open histograms represent the staining with control Ab. All experiments were performed at least five times and representative results (B) as well as the statistical analysis (C) are shown as MFI of the staining cells. ∗, p < 0.05. D and E, NK cells (NK, 1 × 10⁵) or those stimulated with IL-2 (50 ng/ml) for 24 h (IL-2 NK, 1 × 10⁵) were cocultured in the absence (NK) or presence of nontransformed normal hepatocytes (NK/PH) at a ratio of 1:1 for 24 h. DCs were then stimulated with the supernatant obtained from the cocultured medium. The expression of CD86 was evaluated by FACS analysis (filled histograms). Open histograms represent the staining with control Ab. All experiments were performed at least five times and representative results (D) as well as the statistical analysis (E) are shown as MFI of the staining cells. ∗, p < 0.05.

We examined whether primary-cultured, nontransformed human hepatocytes (PH) could modify NK cell activation of DCs. NK cells did not induce DC maturation either in the presence or absence of PH (Fig. 2⇑D). At the next step, NK cells were stimulated with IL-2 for 24 h and then cocultured with PH. The obtained supernatant of IL-2-stimulated HCV-NK alone up-regulated CD86 expression of DCs at levels similar to those of IL-2-stimulated N-NK alone. However, the maturation of DCs was clearly inhibited, especially in chronic HCV-infected patients when IL-2-stimulated NK cells were cocultured with PH (Fig. 2E⇑). Taken together, these results showed that HCV-NK reduced their ability to induce DC maturation when cultured with human hepatic cells.

DC functions were up-regulated by N-NK/Hep3B coculture supernatant, but not HCV-NK/Hep3B culture supernatant

One of the important aspects of DC activation is their acquisition of the migratory ability to regional lymph nodes as well as the stimulatory capacity of T cells (3). Therefore, we examined whether NK cells can modulate the DC chemotactic activity in the presence of human hepatic cells. Flow cytometric analysis revealed that DCs stimulated with the N-NK/Hep3B coculture supernatant showed significantly higher expression of CCR7 than those stimulated with the supernatant of NK cells alone (Fig. 3⇓, A and B). Furthermore, DCs stimulated with N-NK/Hep3B or N-NK/K562 supernatant gained the ability to migrate in response to CCL19, the specific ligand of CCR7 (Fig. 3⇓C). In contrast, CCR7 expression and the migratory activity of DCs were not induced by the treatment of HCV-NK/Hep3B culture supernatant, but were increased by the coculture supernatant of HCV-NK/K562 (Fig. 3⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

NK/Hep3B coculture supernatants can regulate chemotactic activity of DCs. NK cells from healthy (N) and HCV-infected (HCV) individuals were cultured with Hep3B or K562 at a ratio of 1:1 for 24 h. DCs from healthy donors were stimulated with the obtained supernatant for 24 h and were evaluated for their expression of CCR7 (filled histograms). Open histograms represent the staining with control Ab. All experiments were performed five times and representative results (A) as well as the statistical analysis (B) are shown as MFI of the staining cells. ∗, p < 0.05. C, CCL19 (100 ng/ml) were loaded in the lower compartment and DCs stimulated as described above were seeded in the upper compartment of Boyden microchambers. After 2 h, the cells that migrated to the lower chambers were collected and counted. These examinations were performed with triplicate cultures, and the results are presented as a migration index (cell number in CCD19-contained medium/cell numbers in medium alone). Similar results were obtained in three independent experiments. ∗, p < 0.05.

We next evaluated whether the DCs exposed to the NK/Hep3B supernatant could stimulate naive CD4⁺ T cells from allogeneic donors. When pretreated with the supernatant of either N-NK/Hep3B or N-NK/K562 mixtures, DCs were capable of stimulating proliferation and IFN-γ production of allogeneic CD4⁺ T cells (Fig. 4⇓, A and B). In contrast, the allostimulatory capacity of DCs was severely impaired when pretreated with the supernatant of HCV-NK/Hep3B coculture, but not when pretreated with HCV-NK/K562 supernatant. We further examined for differences between N-NK and HCV-NK cultured with Hep3B cells with respect to DC-induced activation of Ag-specific CD4⁺ T cells. DCs endowed autologous CD4⁺ T cells with the ability to proliferate and secrete IFN-γ in the presence of the influenza A Ag when cultured in the supernatant of N-NK/Hep3B mixtures. However, the addition of HCV-NK/Hep3B supernatant did not induce DCs to stimulate the influenza A Ag-dependent autologous CD4⁺ T cell responses (Fig. 4⇓, C and D).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

N-NK and HCV-NK differ in their ability to trigger DC-mediated induction of CD4⁺ T cell responses. A and B, DCs were cultured with the supernatant of N-NK (N) or HCV-NK (HCV)/Hep3B coculture medium for 24 h and then used for the stimulation of allogeneic CD4⁺ T cells for 48 h. C and D, DCs cultured with N-NK (N) or HCV-NK (HCV)/Hep3B coculture supernatant were pulsed with or without influenza A Ag (10 μg/ml) and used for the stimulation of autologous CD4⁺ T cells for 48 h. In these settings, DCs did not induce autologous CD4⁺ T cell responses without recall Ag. CD4⁺ T cells (1 × 10⁵/well) were cultured with graded numbers of DCs and proliferation was examined by [³H]thymidine corporation (A and C), and intracellular expression of IFN-γ was determined by FACS analysis (B and D). The number in the right upper quadrant represents the percentages of CD4-positive cells expressing IFN-γ (B and D). Similar results were obtained in two independent experiments and representative data are shown.

Hep3B cells activated N-NK cells, but not HCV-NK cells

We evaluated whether the differences between N-NK and HCV-NK in their ability to modulate DC functions are associated with the NK cell effector functions. N-NK and HCV-NK efficiently killed K562 target cells at similar levels, but the cytolytic activity of HCV-NK against Hep3B cells was much lower than that of N-NK (Fig. 5⇓A). In addition, the cell surface expressions of activation markers CD25 and CD69 increased when N-NK were cultured with K562 or Hep3B cells. In contrast, these enhancements were observed when HCV-NK were cultured with K562 cells, but not when they were cultured with Hep3B cells (Fig. 5⇓B). We next examined the cytokine production profiles of NK cells cultured with tumor cells. Intracellular cytokine staining revealed that N-NK produced IFN-γ when stimulated with Hep3B or K562. In contrast, HCV-NK did not efficiently produce IFN-γ upon stimulation with Hep3B, although they responded to K562 cells. IL-10 production was much higher in HCV-NK than in N-NK when cultured with Hep3B cells (Fig. 5⇓C). ELISA data confirmed that N-NK produced significantly higher levels of IFN-γ as well as TNF-α than HCV-NK when cultured with Hep3B cells. The levels of IL-10 production from NK cells were higher in HCV-infected patients than those in healthy donors. Similarly, HCV-NK produced much higher amounts of TGFβ than N-NK (Fig. 5⇓D). The different patterns of these cytokine productions between N-NK and HCV-NK were not observed when NK cells were cocultured with K562 cells (Fig. 5⇓, A–C). These results indicate that HCV-NK, upon stimulation of Hep3B cells, show a decrease in NK effector functions, such as cytolytic activity and production of proinflammatory cytokines, correlating well with their decreased ability to activate DCs. Furthermore, HCV-NK, in response to Hep3B cells, produced higher levels of immunoregulatory cytokines IL-10 and TGFβ than N-NK.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Human hepatic cells can activate N-NK, but not HCV-NK. A, Cytolytic ability of NK cells against Hep3B or K562 cells. NK cells isolated from healthy volunteers (N) or HCV-infected patients (HCV) were cultured with ⁵¹Cr-labeled Hep3B or K562 at the indicated ratios for 4 h and then subjected to the standard ⁵¹Cr-releasing assay. Similar results were obtained for three independent donors, and the data are presented as a composite examined for each group. ∗, p < 0.05. B, The change of CD25 and CD69 expression on NK cells. Resting NK cells from healthy volunteers (N) or HCV-infected patients (HCV) were cultured with or without Hep3B or K562 for 24 h and analyzed for the activation markers CD25 and CD69 by flow cytometry. ∗, p < 0.05. C, Intracellular IFN-γ and IL-10 expression of NK cells. NK cells from healthy volunteers (N) or HCV-infected patients (HCV) were cultured with or without Hep3B or K562 for 24 h, and IFN-γ and IL-10 expression were determined by flow cytometry. The number in the right upper quadrant represents the percentage of CD56-positive cells expressing IFN-γ or IL-10. D, NK cells generated from healthy (N) and HCV-infected (HCV) individuals were cultured with or without Hep3B or K562 for 24 h (n = 3 for each group). IFN-γ, TNF-α, IL-10, and TGFβ were measured in each culture supernatant by ELISA. ∗, p < 0.05.

Increased frequency of NK cells expressing the inhibitory receptor CD94/NKG2A in HCV-infected individuals

NK cell activities have been reported to be regulated by the interaction of NK cell inhibitory and activating receptors and their ligands (10). Thus, we next examined the expression of NK cell receptors on N-NK as well as HCV-NK. As shown in Fig. 6⇓, flow cytometric analysis revealed that all of the KIRs tested and activating receptor NKG2D were expressed on HCV-NK at levels similar to those on N-NK. The expression of C-type lectin receptor CD94 and NKG2A were significantly elevated in HCV-NK compared with N-NK. These results demonstrated that CD94 and NKG2A are molecules expressed at much high levels in HCV-NK. It has been reported that human NK cells were composed of CD56^bright and CD56^dim subsets and that CD56^bright NK cells express CD94/NKG2A more than CD56^dim subsets (28). The percentages of the CD56^bright subset of HCV-NK were similar to those of N-NK (data not shown), indicating that the higher expression levels of CD94/NKG2A of HCV-NK were not due to expansion of the CD56^bright subpopulation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Profiles of NK receptor expression on NK cells. NK cells isolated from normal volunteers (n = 10; N) or from HCV-infected patients (n = 15; HCV) were stained with Abs for various KIRs, CD94, NKG2A, and NKG2D as described in Materials and Methods. The receptor expression of CD56⁺ cells was examined by flow cytometry and was expressed as a percentage of the positive cell rate. ∗, p < 0.05.

Inhibition of NKG2A during culture with HLA-E-bearing Hep3B cells restored the activation status of HCV-NK

We next analyzed the expression of HLA class I and HLA-E, which function as ligands of KIRs and CD94/NKG2A, respectively (10, 29), on human hepatic cells or K562. As shown in Fig. 7⇓, all human hepatic cells expressed HLA-E as well as HLA class I molecules on their cell surface, and the levels of HLA-E expression varied among them. Neither HLA-E nor HLA class I was detected on K562, consistent with previous reports (30, 31). The above findings raise the possibility that the interaction of CD94/NKG2A and HLA-E may attenuate NK cell effector functions against HLA-E-bearing human hepatic cells in patients with HCV infection.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Surface expression of HLA-E and HLA class I in Hep3B, Huh7, primary hepatocyte, and K562 was assessed by flow cytometry (filled histograms). Peripheral blood leukocytes (PBL) were used as positive control for HLA-E and HLA class I expression. Open histograms show isotype control staining.

Although the formation of a heterodimer complex is necessary to express NKG2A and CD94 at cell surface levels to exert their biological effects, CD94 does not trigger any functions, and NKG2 families are critical for exerting their biological effects (32). We next investigated the cytolytic activity of NK cells with masking Ab of NKG2A during coculture of NK cells and Hep3B cells. Addition of anti-NKG2A Ab, which can specifically block CD94/NKG2A heterodimer-HLA-E interaction (24), resulted in a marked increase in Hep3B cytolysis of HCV-NK, but to a lesser extent for N-NK. In contrast, the NKG2A blockade had no effect on N-NK or HCV-NK cytolysis of HLA-E-negative K562 cells (Fig. 8⇓A). We also examined the effect of anti-NKG2A Ab on cytokine production from NK cells. The addition of anti-NKG2A Ab during NK cells/Hep3B cocultures increased production of proinflammatory cytokines, such as IFN-γ and TNF-α Again, the blockade of the NKG2A-mediated signal resulted in a substantial decrease in IL-10 and TGFβ production from HCV-NK, but to a lesser extent for N-NK (Fig. 8⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

Blockade of NKG2A enhances NK cell activities. A, NK cells isolated from healthy volunteers (N) or HCV-infected patients (HCV) were cultured with ⁵¹Cr-labeled Hep3B or K562 in the presence of the masking Ab of NKG2A or isotype control IgG (20 μg/ml) at the indicated ratios in 96-well culture plates for 4 h, and then were subjected to standard ⁵¹Cr-releasing assay. Similar results were obtained for three independent donors, and the data are presented as a composite examined for each groups. ∗, p < 0.05. B, NK cells isolated from healthy (N) or HCV-infected (HCV) individuals were cultured with Hep3B in the presence of masking Ab of NKG2A or isotype control IgG (20 μg/ml) for 24 h (n = 3 for each group). IFN-γ, TNF-α, IL-10, and TGFβ were measured in each culture supernatant by ELISA. ∗, p < 0.05.

Blockade of NKG2A signals enhanced the ability of NK cells to activate DCs

To investigate the involvement of the NKG2A-mediated signal in NK cell activation of DCs, NK cells and Hep3B cells were cultured in the presence of anti-NKG2A Ab for 24 h. Next, DCs were stimulated with the obtained NK/Hep3B supernatant for 24 h. FACS analysis showed that the expression levels of CD86, CD40, CD83, and CCR7 were elevated on DCs when anti-NKG2A Ab was added during the cocultures of NK cells and Hep3B; the extent was greater in HCV-NK than in N-NK. The NKG2A blockade during IL-2-stimulated NK cells and nontransformed hepatocytes also restored DC maturation, indicating that the ability of normal hepatocytes to inhibit NK cell activation of DCs relied mainly upon the NKG2A-mediated inhibitory signal (Fig. 9⇓A and data not shown). Moreover, the blockade of NKG2A during HCV-NK/Hep3B coculture restored the ability of the supernatant-treated DCs to stimulate allogeneic CD4⁺ T cells, whereas the NKG2A blockade slightly enhanced proliferation of CD4⁺ T cells by N-NK/Hep3B supernatant (Fig. 9⇓B). We also examined whether any differences existed between anti-NKG2A Ab and control IgG treatment in DC-stimulated cytokine production from CD4⁺ T cells. Intracellular cytokine staining showed that CD4⁺ T cells can produce both IFN-γ and IL-10 when stimulated by DCs that had been exposed to N-NK/Hep3B coculture supernatant. DCs pretreated with the HCV-NK/Hep3B supernatant generated CD4⁺ T cells, preferentially producing IL-10. Blockade of NKG2A during the coculture of NK cells and Hep3B inhibited IL-10 production and promoted IFN-γ production from CD4⁺ T cells (Fig. 8⇑C). Taken together, these findings identified high expression levels of NK inhibitory receptor NKG2A as the critical factor for suppressing the ability of NK cells to activate DCs.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 9.

NKG2A-mediated signal suppresses NK cell activation of DCs. NK cells (1 × 10⁵) from healthy donors (N) as well as HCV-infected individuals (HCV) were cocultured with Hep3B at a ratio of 1:1 in the presence of anti-NKG2A or control IgG (20 μg/ml) for 24 h. DCs (1 × 10⁵) were stimulated with the supernatant obtained from the cocultured medium for 24 h. A, The expressions of CD40, CD86, and CCR7 were evaluated by FACS analysis. All experiments were performed five times and statistical analysis is shown. ∗, p < 0.05. B, Allogeneic naive CD4⁺ T cells (1 × 10⁵/well) were cultured for 72 h by graded numbers of DCs pretreated as described above and then the proliferation of CD4⁺ T cells was examined by [³H]thymidine corporation. C, DCs were stimulated with the supernatant of N-NK (N) or HCV-NK (HCV)/Hep3B coculture in the presence of the masking Ab of NKG2A or control IgG (20 μg/ml) as described above, and after washing three times, DCs (1 × 10⁵/well) were cocultured with allogeneic naive CD4⁺ T cells (5 × 10⁵/well) for 48 h. After positive gating as CD4⁺ cells to exclude the DC population, intracellular expression of IFN-γ and IL-10 in CD4⁺ T cells was determined by FACS analysis. The number in each quadrant represents the percentage of cell numbers. Similar results were obtained in three independent experiments, and representative data are shown.

Immunosuppressive cytokines, IL-10 and TGFβ, act as downstream mediators of NKG2A-induced inhibition of NK cell activation of DCs

IL-10 and TGFβ are known to act as suppressive factors of DC activation to induce the anergic states (33, 34). We examined whether IL-10 and TGFβ, produced in large amounts by HCV-NK, are responsible for the impaired DC activation. Toward this goal, the culture supernatant of HCV-NK/Hep3B cells was treated with neutralizing Abs of IL-10 and TGFβ and was used for the stimulation of DCs. The blockade of IL-10/TGFβ clearly enhanced the expression of CD40, CD86, CD83, and CCR7 on DCs. In contrast, when DCs were pretreated with the supernatant of the anti-NKG2A Ab-added culture, the IL-10/TGFβ blockade had little additional effects on DC maturation (Fig. 10⇓). These results indicated that the IL-10/TGFβ production, triggered by the interaction of NKG2A and its ligand, could directly interfere with DC activation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 10.

IL-10 and TGFβ are responsible for NKG2A-mediated inhibition of HCV-NK activation of DCs. NK cells (1 × 10⁵) from HCV-infected individuals (HCV-NK) were cocultured with Hep3B at a ratio of 1:1 for 24 h. In some experiments, NK cells were cocultured with Hep3B in the presence of anti-NKG2A Ab (20 μg/ml; anti-NKG2A). DCs (1 × 10⁵) were stimulated with the culture supernatant in the presence or absence of both anti-IL-10 (20 μg/ml) and anti-TGFβ-neutralizing Ab (30 μg/ml) for 24 h. The expressions of CD40, CD86, CD83, and CCR7 were evaluated by FACS analysis. All experiments were performed at least three times and representative results (A) as well as the statistical analysis of CD40 and CD86 (B) are shown as the MFI of the stained cells. ∗, p < 0.05.