This study includes the following subjects (as summarized in Table S1 in Supplementary Material): (i) 30 treatment-naive patients with chronic hepatitis C. All patients tested positive for anti-HCV with a third-generation enzyme-linked immunosorbent assay (Kechuang and Xinhua, Shanghai, China). HCV RNA levels were quantified using a reverse transcription polymerase chain reaction (RT-PCR) assay (Qiagen, Shenzhen, China), with a lower limit of detection for 100 copies/ml. All patients were negative for antibodies against hepatitis B, hepatitis D, and human immunodeficiency virus; (ii) 25 HCV patients who achieved an EVR (defined as serum HCV RNA being undetectable, <100 copies/ml), at week 12 after initiation of Peg-IFN-α-2a (180 μg/week subcutaneously) and weight-based ribavirin (800–1,200 mg, depending on the HCV genotype); (iii) 20 HCV patients who achieved an SVR (defined as HCV RNA remains undetectable for at least 6 months after treatment discontinuation) following Peg-IFN-α and ribavirin therapy; (iv) 24 age- and sex-matched healthy volunteers. Informed consent was obtained from all participating subjects. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Research and Ethical Committee of Tangdu Hospital of the Fourth Military Medical University.

Peripheral blood mononuclear cells (PBMCs) were freshly isolated from peripheral blood by density gradient centrifugation using Ficoll-Hypaque (Sigma, St. Louis, MO, USA) and frozen or immediately studied as described below. If indicated, NK cells were further purified from PBMCs by negative selection using a human NK cell isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s instructions. The purity of NK cells was measured by flow cytometry after staining the cells with anti-CD3-Cy5.5/PerCP (Clone: UCHT1, BD Biosciences, San Jose, CA, USA) and anti-CD56-Pe-Cy7 (Clone: B159, BD Biosciences, San Jose, CA, USA), with the frequency of CD3⁻CD56⁺ cells above 90% and CD3⁺ cells less than 1% to be used for subsequent experiments.

Hepatitis C virus JFH-1 (genotype 2a) was produced as described previously (32). The JFH-1 expression construct (kindly provided by Dr. C. Rice, Rockefeller University, New York, NY, USA) was linearized, and full-length JFH-1 RNA was transcribed using a MEGAscript T7 in vitro transcription kit (Ambion, Austin, TX, USA) per manufacturer’s instructions. 5 × 10⁵ Huh7.5 cells (kindly provided by Dr. C. Rice) were transfected at 70–80% confluent in a six-well plate with 2 μg transcribed RNA using DMRIE-C reagent per company’s protocol (Invitrogen, Carlsbad, CA, USA). HCV antigen expression was examined at 48 h after transfection by immunofluorescence using HCV NS5 antibodies (ViroGen, Watertown, MA, USA) and the supernatant collected from HCV RNA-transfected Huh7.5 cells at 48 h was used to infect naive Huh7.5 cells to make HCV stocks. HCV titer was detected as previously described (33).

Peripheral blood mononuclear cells (0.5 × 10⁶ cells) from healthy subjects were infected by coculture with JFH-1/Huh7.5 cells at an effector-to-target (E:T) ratio of 10:1 or with HCV particles at a multiplicity of infection (MOI) of 10 for 48 h. Complete cell culture medium was used as negative control. After incubation, cells were stained for anti-CD3 (Clone: UCHT1), CD14 (Clone: M5E2), CD19 (Clone: HIB19), CD56 (Clone: B159), CD16 (Clone: 3G8), all from BD Biosciences, San Jose, CA, USA, and CD100 (Clone: A8, BioLegend, San Diego, CA, USA) monoclonal antibodies followed by flow cytometric analysis.

We also observed the effect of IFN-α on CD100 and plexin-B1/B2 expression. PBMCs (0.5 × 10⁶ cells) were incubated in 1 ml complete medium supplemented with IFN-α-2a (at a concentration ranging from 0.01 to 1,000 ng/ml) (Roche Bioscience, Hillview Avenue Palo Alto, CA, USA) in a 24-well round-bottom plate (Corning, One Riverfront Plaza, NY, USA.) for 2, 6, 12, 24, and 48 h, respectively, and Huh7.5 cells were stimulated with IFN-α-2a (10 ng/ml) for 48 h. Complete cell culture medium was used as controls. After incubation, cells were stained with monoclonal antibodies as described above for flow cytometric analysis.

For phenotypic analysis, PBMCs isolated from HCV patients and healthy subjects were stained with CD3-Cy5.5/PerCP (Clone: UCHT1), CD14-Cy5.5/PerCP (Clone: M5E2), CD19-Cy5.5/PerCP (Clone: HIB19), CD16-Cy7/APC (Clone: 3G8), CD56-PeCy7 (Clone: B159), CD69-FITC (Clone: FN50), TRAIL-PE (Clone: S35-934), all from BD Biosciences, San Jose, CA, USA, CD100-FITC (Clone: A8, BioLegend, San Diego, CA, USA) or isotype matched controls (BD Biosciences, San Jose, CA, USA). Plexin-B1/B2 expression on K562 (ATCC, Manassas, VA, USA) and Huh7.5 cells (used as target cells) were measured using RT-PCR (primers were listed in Table S2 in Supplementary Material) and flow cytometry with plexin B1 and plexin B2 antibody (Clone: #559830&#537223, R&D Systems Inc., Minneapolis, MN, USA). Immunostained cells were analyzed on a multicolor Arial II (BD Biosciences, San Jose, CA, USA) and FlowJo Version 7.6.2 software.

For analysis of cytokine production and degranulation, purified NK cells were resuspended at 10⁵ cells/ml and stimulated with either (i) K562, Huh7.5 or JFH-1/Huh7.5 cells at an E:T ratio of 1:1; (ii) K562, Huh7.5 or JFH-1/Huh7.5 cells at an E:T ratio of 1:1 in the presence of IFN-α (10 ng/ml); (iii) K562, Huh7.5 or JFH-1/Huh7.5 cells preincubated for 2 h with soluble CD100 (sCD100) (PeproTech, Rocky Hill, NJ, USA) or antiplexin B1/2 antibodies mix (Clone: sc28372&sc373969, Santa Cruz, CA, USA) to block availability of plexin-B1/B2; (iv) complete cell culture medium without target cells and cytokines as control. After 8 h incubation at 37°C, Brefeldin A (10 ng/ml) (eBioscience, San Diego, CA, USA) and monensin (0.67 μl/ml) (BD Biosciences, San Jose, CA, USA) were added for another 4 h prior to harvesting the cells. Finally, NK cells were washed, stained with anti-CD3, CD16, CD56, and CD100, and analyzed by flow cytometry.

RNA was extracted using the RNeasy mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Quantification of messenger RNA (mRNA) levels of plexin-B1 and plexin-B2 were performed by real-time PCR. GAPDH served as the internal control. Primer sequences used for real-time PCR analysis were listed in Table S2 in Supplementary Material.

Statistical analyses were performed with GraphPad Prism Version 5.0 (GraphPad Software Inc., San Diego, CA, USA). The Shapiro–Wilk test was used to test the normal distribution of quantitative variables. When quantitative variables were normally distributed, the results were presented as mean ± SD, otherwise median and interquartile range (25th–75th percentile) were reported; differences between two or three groups were evaluated by parametric or non-parametric tests according to data distribution. Paired Student’s t-test was used to assess paired variables. Correlations between variables were evaluated with the Spearman rank correlation test. p < 0.05 was considered as statistically significant.

Based on CD56 and CD16 expression, NK cells can be divided into CD56^bright (CD56⁺⁺CD16^±) and CD56^dim (CD56⁺CD16⁺) subsets (34). CD3/14/19⁻CD56⁺ NK cells and their CD56^bright and CD56^dim subsets were identified in PBMCs using multicolor flow cytometry by immune-staining for CD3, CD14, CD19, CD56, and CD16 (Figure 1A). To identify the specific effect of HCV infection on NK cells, we conducted an ex vivo cross-sectional analysis to investigate the CD100, CD69, and TRAIL expression on NK cells from HCV patients versus healthy subjects. As shown in Figure 1B, the percentage of CD100⁺ NK cells was slightly decreased in patients with chronic HCV infection compared to healthy subjects, though this did not achieve a significant difference. Similarly, no significant changes were observed in the frequency of CD69⁺ NK and TRAIL⁺ NK cells between HCV patients and healthy subjects, though these phenotypic markers were slightly upregulated in the setting of HCV infection. These data suggest that NK cells were not successfully activated during chronic HCV infection.

To investigate whether CD100 was implicated in liver injury and disease progression, we observed the correlation between the measured molecules with ALT and HCV-RNA levels. We found that CD100⁺ NK as well as TRAIL⁺ NK cells positively correlated with the serum ALT levels, but inversely correlated with the HCV-RNA titers (Figures 1C,D). Moreover, the level of CD100 was found to be related to the TRAIL expression on CD56^bright NK cells (Figure 1E); however, no correlations were found between the CD69 expression on NK cells and the ALT or HCV-RNA levels.

To determine whether IFN-α can activate NK cells in vivo and upregulate CD100 expression, we examined CD100, CD69 and TRAIL expressions by NK cells before and after IFN-α treatment in chronically HCV-infected patients who achieved EVR and SVR, respectively. We found that the proportion of CD100⁺, CD69⁺ and TRAIL⁺ NK cells was significantly upregulated in patients with EVR, and returned to normal in patients with SVR after IFN-α treatment (Figure 2A). Of note, the expression of CD100 was increased in both CD56^bright and CD56^dim NK subsets, whereas CD69 expression mainly increased in CD56^dim NKs, and TRAIL expression primarily increased in CD56^bright NK subsets, respectively (Figure 2A). Furthermore, we also examined the relationship between CD100 or TRAIL expression and HCV RNA titer in 6 patients at the very early phase of Peg-IFN-α and ribavirin treatment. PBMCs from these patients were collected on days 0 and 7 of treatment, and HCV-RNA in serum was quantified. At day 0 of treatment, about 74.3% of CD56^bright NKs and 52.1% of CD56^dim NKs were CD100⁺ cells. After 7 days of treatment, however, the percentage of CD100⁺ CD56^bright NKs and CD100⁺ CD56^dim NKs increased to 89.1 and 74.5%, respectively. Similarly, TRAIL⁺ cells were also increased from 22.1 to 58.7% in CD56^bright NK subset (Figure 2B). Importantly, the upregulation of CD100 on both NK subsets and TRAIL on CD56^bright NK cells were inversely correlated with the HCV-RNA decline during the early phase of antiviral treatment (Figure 2C). These results indicated that IFN-α treatment efficiently activated NK cells, and the increased CD100 and TRAIL expressions on NK cells correlated with the HCV control.

To further elucidate the role of HCV in regulation of CD100 expression and to mimic the in vivo setting of HCV infection, we employed a cell coculture system by incubating healthy PBMCs with Huh7.5 hepatocytes transfected with HCV JFH-1 strain in vitro. NS5 protein was detected in HCV-JFH-1 transfected Huh7.5 cells by immunofluorescent staining (Figure S1A in Supplementary Material). Additionally, Huh7.5 cells can be infected by the supernatant of JFH-1-transfected Huh7.5 cells (Figure S1B in Supplementary Material), suggesting that HCV particles are secreted from the HCV mRNA-transfected hepatocytes into the culture media. We incubated healthy PBMCs with medium alone vs. medium containing HCV virions (MOI = 10) or Huh7.5 vs. JFH-1-infected Huh7.5 cells for 48 h, followed by immune-staining and flow cytometric analysis. As shown in Figure 3A, CD100 expression on NK cells was not affected by the HCV particles; but it was remarkably decreased on PBMCs cocultured with JFH-1-infected Huh7.5 compared with those incubated with Huh7.5 cells without HCV infection (Figure 3B).

Next, we examined the effect of IFN-α on CD100 expression on NK cells, and found that CD100⁺ NK cells were significantly upregulated, in a dose- and time-dependent manner, in PBMCs stimulated with IFN-α (Figure 3C). CD100 expression began to rise at 6 h and peaked at 12 h following IFN-α treatment. As shown in Figure 3D, the same effect was observed in different NK subsets in PBMCs derived different donors incubated with IFN-α for 12 h. We also examined whether IFN-α can affect the expression of CD100 receptors, plexin-B1/B2, on Huh7.5 cells. Indeed, plexin-B1 and plexin-B2 expressions were upregulated by IFN-α, which reached a maximum effect at 24 h, following IFN-α treatment (Figures 3E,F). The CD72, another important CD100 receptor, was checked as well. However, it cannot be detected on Huh7.5 cells or HCV-infected/IFN-α treated Huh7.5 cells. These results indicated that IFN-α treatment can enhance CD100 and plexin-B1/B2 expressions on NK cells, which may influence signal transduction between NK cells and target cells.

Natural killer cell activation typically results in cytotoxicity/degranulation and release of antiviral cytokines, including IFN-γ. Several lines of evidence have shown that NK cells from HCV-infected patients display a functional dichotomy, characterized by normal or enhanced cytotoxicity and reduced production of IFN-γ (8–10, 35, 36). To investigate whether HCV and IFN-α could affect NK cell killing activity to target cells in the in vitro system, we cocultured purified NK cells with K562, Huh7.5 or HCV JFH-1-infected Huh7.5 (JFH-1/Huh7.5) cells in the presence or absence of IFN-α. As shown in Figure S1C in Supplementary Material, HCV had a strong effect on CD107a expression, especially on CD56^dim subsets, but not on IFN-γ production (Figure S1D in Supplementary Material). In response to K562 cells, IFN-α treatment effectively activated NK cells through increased degranulation on both NK subsets and enhanced IFN-γ production on CD56^dim NK subset (Figures 4A,B). Correspondingly, similar results were observed (increased CD107 and IFN-γ expression in NK cells, particularly in CD56^dim subsets) upon coculturing NK cells with Huh7.5 (Figures 4C,D) or JFH-1/Huh7.5 cells (Figures 4E,F).

Previous studies have reported that degranulation and IFN-γ production of NK cells could be induced via activation of the IFN-α/β receptor (37, 38), it was not clear however whether other molecules were involved in this process, especially in response to target cells. To explore the mechanisms underlying NK cell functions in response to target cells, we studied the potential involvement of IFN-α, CD100, and plexin-B1/B2 in NK degranulation and IFN-γ production. Since we have shown that IFN-α is involved in NK regulation, we hypothesized that CD100 and plexin-B1/B2 interactions have functional consequences on the NK killing process. To test this hypothesis, we blocked the availability of plexin-B1/B2 by preincubating sCD100 with K562 cells. As shown in Figure 5A, the optimal blockade was observed at a concentration of 50 ng/ml, with a plateau being observed thereafter. Interestingly, it was found that the expression of CD107a on NK cells declined, whereas CD100 expression upregulated with increasing the sCD100 concentrations. This phenomenon might imply a negative loop between CD100–plexin-B interaction and CD100 expression. CD100–plexin B ligation could trigger Erk signaling within the cells (20). Thus, we believe that the ligations of CD100–plexin-B downregulate CD100 expression on NK cells through the MAPK signal transduction, whereas the interaction blockage by sCD100 rescues CD100 expression in NK cells. The plexin B1/2 antibodies were also used for CD100/plexin B interaction blockade according to previous studies (39). As expected, CD107a expression decreased in total NK as well as in the two NK subsets, particularly in the presence IFN-α after blocking plexin-B1/B2 availability by sCD100 or plexin B1/2 antibodies (Figure 5B); this was also true for IFN-γ production (Figure 5C). Taken together, these results revealed an important role of CD100–plexin-B1/B2 interactions in NK degranulation and IFN-γ production in response to target cells.