Introduction

Hepatitis B virus (HBV) is a noncytopathic enveloped virus that causes acute and chronic necroinflammatory liver diseases that affect more than 350 million people worldwide [1, 2]. Approximately 1 million people die annually from HBV-induced liver diseases [1, 2]. Antiviral-therapy-induced host immune responses against HBV play a crucial role in viral clearance [3, 4]; however, the pathogenesis and immunologic mechanisms of these responses remain unclear.

Invariant natural killer T (iNKT) cells express an invariant α-chain T-cell receptor (TCR; Vα24-Jα18 chain in humans and Vα14-Jα18 chain in mice), which is paired with one of three different β chains [5–7]. CD1d-restricted T cells are less diverse. Many use the same TCR α chain, but they also express natural killer (NK) cell receptors. The CD1d-restricted population is called iNKT cells, and one recognized ligand for CD1d molecules is α-galactosylceramide (α-GalCer) [5–8]. iNKT cells rapidly produce large quantities of both type 1 and type 2 helper-biased cytokines, including interleukin (IL)-4 and interferon (IFN)-γ [9]. It is thought that these cells primarily have a regulatory function, which enhances the function of NK cells, dendritic cells, and B cells, as well as that of conventional CD4+ and CD8+ T cells [9].

Research on chronic hepatitis B (CHB) infections has mainly focused on CD8+ T cells [10–12], but the role of iNKT cells has been neglected. Recently, iNKT cells were found to inhibit HBV replication in vivo directly and not rely on conventional CD4+ T and CD8+ T cells [13, 14]. In chronic HBV infections, iNKT cells exhibit functional decline. Unfortunately, the mechanism underlying the malfunction of iNKT cells in chronic HBV infections is not well understood. This functional decline could be related to the upregulation of surface programmed cell death 1 (PD-1) and T cell immunoglobulin mucin-3 (Tim-3) molecules [15–22]. PD-1 is a inhibitory receptor related to CD28, which was found to be upregulated on the surface of exhausted T cells in human immunodeficiency virus (HIV) [23–25], chronic lymphocytic choriomeningitis virus (LCMV), hepatitis C virus (HCV), and chronic hepatitis B virus infections [15, 24, 25]. PD-1 can bind to the B7-family ligands PDL1 (B7-H1) and PDL2 (B7-DC). Binding of PD-1 to PDL1 has been studied in detail and has been shown to transduce strong inhibitory signals to the iNKT cell [20]. The Tim-3 protein is specifically expressed on type 1 helper T (TH1) cells and negatively regulates TH1 cell responses. Some reports have suggested that Tim-3 is upregulated on effector CD8+ T cells in chronic virus infections [26]. Tim-3-expressing T cells have an inhibitory ability and dysfunctional cytokine responses, and blockade of Tim-3 improves the proliferative ability of HIV-specific T cells [27].

The aim of this project was to study the frequencies, functions, and surface expression levels of the co-stimulatory factors CD28, PD-1, and Tim-3 on iNKT cells in chronically HBV-infected patients. The low activities of iNKT cells by blocking the in vitro PD-1 and Tim-3 signaling pathways and/or the activation of CD28 signals were also investigated, which may improve the immune responses mediated by α-GalCer-activated iNKT cells in chronic HBV infection. The results indicated that the functional decline of iNKT cells was closely related to the decrease in CD28 expression and the increase of Tim-3 and PD-1 expression. In addition, clinical antiviral treatment with lamivudine could partially restore the immune function of iNKT cells in CHB patients.

Materials and methods

Subjects

Twenty-eight patients with CHB were enrolled in the study. Peripheral blood was taken from each patient, and none of the patients received interferon or nucleoside analogue drug treatment before sampling.

Fifteen patients who were pathologically confirmed to have chronic hepatitis B and stage I or II hepatocellular carcinoma (HCC) were enrolled. The clinical stage of tumor progression was determined according to the tumor-nodes-metastasis (TNM) classification system of the International Union Against Cancer (Edition 6). Of these 15 cases, five cases had received lamivudine orally as an anti-HBV treatment (LAM 100 mg/day) for 2–3 years, and the remaining 10 cases received no interferon or nucleoside analogue drug treatment before surgery and sampling. None of these patients received anticancer therapy before surgery and sampling. Fresh nontumor hepatic tissues, with approximate dimensions of 1.5 × 1.5 cm each, were obtained from the patients undergoing surgical resection from a location at least 3 cm distant from the tumor site.

Control fresh hepatic tissues were obtained from 10 patients diagnosed with hepatolithiasis or hepatic angiomatosis who did not have an HBV infection. Control peripheral blood samples were obtained from 16 healthy donors. Individuals with concurrent autoimmune disease or HIV, hepatitis C virus, or syphilis infections were excluded (Table 1).

Table 1 Baseline characteristics of CHB patients (n = 53) enrolled in the study, including the HBeAg (+) and HBeAb (+) patients
Full size table

All patients were from the Department of Infectious Diseases and the Department of Hepatobiliary Surgery of the First and the Second Affiliated Hospitals of Chongqing Medical University. The patients were enrolled between January 1, 2012, and December 31, 2013. This study received ethical approval from the Medical Ethics Committee of the Second Affiliated Hospital of Chongqing Medical University, and written informed consent was obtained from all participants.

Reagents

Allophycocyanin (APC), α-GalCer/human CD1d dimer, was purchased from Proimmune (USA). Anti-human Vα24-Jα18 TCR fluorescein isothiocyanate (FITC), anti-human CD279 (PD-1) FITC, phycoerythrin (PE)-conjugated anti-CD28 monoclonal antibody (mAb), anti-human IL-4 PE-Cy7, anti-human IFN-γ PerCP-Cy5.5, anti-human CD274 (PD-1; functional grade, purified), anti-human CD28 (functional grade, purified), and anti-human CD80 (B7-1; functional grade, purified) were purchased from eBioscience (San Diego, CA, USA). PE anti-human Tim-3 and anti-human Tim-3 (functional grade, purified) were purchased from Biolegend. The human IFN-γ enzyme-linked immunosorbent assay (ELISA) and human IL-4 ELISA were from R&D Systems (USA).

Cell isolation

Venous blood was sampled from healthy controls and patients. The venous blood was separated to obtain peripheral blood mononuclear cells (PBMCs) by Ficoll density gradient centrifugation. Hepatic mononuclear cells (MNCs) were isolated from the fresh nontumor hepatic tissues (at least 3 cm distant from the tumor site) from the patients who underwent surgical resection. Freshly obtained tissues were washed three times in sterile phosphate-buffered saline (PBS) before being cut into small pieces with scissors. The specimens were then homogenized in RPMI 1640 medium containing 10 % fetal calf serum (FCS) using a gentleMACSTM dissociator (Miltenyi Biotec, Germany). Dissociated cells were filtered through a 70-μm mesh and separated by Ficoll centrifugation. The fresh hepatic MNCs were washed two times in sterile PBS and resuspended in RPMI 1640 medium containing 10 % FCS and 1 % penicillin/streptomycin solution (penicillin, 10,000 units/ml; streptomycin, 10 mg/ml).

Flow cytometric analysis

The PD-1, Tim-3, and CD28 levels on iNKT cells were analyzed by flow cytometry (BD Bioscience). iNKT cells were assessed by Vα24 and α-galactosylceramide/CD1d tetramer markers. For intracellular cytokine detection in MNCs, cells were stained with mAbs against Vα24 and α-GalCer/CD1d tetramer markers, fixed with 4 % perfluoroalkoxy (PFA), and permeabilized with 0.1 % saponin/PBS. The fixed cells were then stained with PerCP-Cy5.5-anti-human IFN-γ and PE-Cy7-anti-human IL-4 and analyzed by flow cytometry as described above.

Cell culture

MNCs were isolated as described above. Suspended enriched MNCs were cultured in RPMI-1640 medium (containing 15 % FCS, 100 U of penicillin–streptomycin per mL, 0.1 mM minimum essential medium (MEM) with nonessential amino acids, 1 mM sodium pyruvate, 5.5 mM 2-mercaptoethanol [2-ME], and 10 mM HEPES buffer) and α-GalCer (200 ng/mL) at a concentration of 1 × 106 cells/ml. The cells were cultured in flat-bottomed 96-well plates (0.2 × 106 per 96-well plate with 200 μL/well). Briefly, these cells were divided into five groups and given different treatments. Group I was specified as the blank controls. In group II, anti-human Tim-3 (functional grade, purified; 0.5 μg/well) was added to block the Tim-3 signaling pathway. Group III was given anti-human CD274 (B7-H1) (1 μg/well) to block the PD-1 signaling pathways. Group IV was costimulated with anti-human CD28 (1 μg/well) and anti-human CD80 (B7.1) (1 μg/well) mAb. Group V was given a mixture including anti-human Tim-3 mAb, anti-human CD274 mAb, and anti-human CD28/80 mAb. Cells were cultured in a 5 % CO2 incubator for 72 h. The supernatant was collected and subjected to ELISA to measure the cytokines. Golgi Plug™ (1 μL/mL) was added in the next step, and the incubation was continued for 4 h until the analysis.

ELISA analysis

The IL-4 and IFN-γ levels in the cell supernatants were measured with human-IL-4- and IFN-γ-specific ELISA kits (RayBiotech, Inc., Norcross, GA, USA), according to the manufacturers’ instructions. Briefly, a standard curve was established by adding 100 µl of standard-concentration solutions to the wells of the pre-coated ELISA plate. Blank wells and sample wells were prepared with 100 µl of the appropriate samples. The reaction plate was sealed with film and allowed to incubate at 37 °C for 90 minutes. After incubation, the liquid was discarded, and the plate was washed four times with washing buffer. To each well, 100 µl of biotinylated antibody was added, and the plate was again incubated at 37 °C for 60 minutes. The plate was washed again, as before, and 100 µl of HRP-conjugated streptavidin was added to each well. After incubation for 30 minutes (37 °C), 100 µl of TMB was added and allowed to react for 20 minutes in the dark at 37 °C. Finally, the OD450 was determined on an ELx800NB plate reader (BioTek USA, Winooski, VT, USA). The sample concentrations were calculated using a standard curve. Data are shown as the mean ± SD, and statistical significance was defined as p < 0.05.

Statistical analysis

All data are shown as the mean ± standard deviation (SD). All statistical analysis was performed with the SPSS version 13.0 software (SPSS, Inc., Chicago, IL, USA). Multigroup comparisons were carried out by analysis of variance (ANOVA). The relationship between two quantitative indices was analyzed using Pearson’s correlation coefficient. Statistical significance was set at p < 0.05.

Results

iNKT cells in CHB patients and healthy controls

The frequency of isolating iNKT cells was determined by flow cytometry by applying fluorescence-activated cell sorting (FACS). iNKT cells were defined by FACS analysis as cells that were positive for both α-GalCer/CD1d tetramers and Vα24-Jα18 TCR. The iNKT cell frequency was significantly lower in CHB patients than in the control groups (PBMCs, 0.055 ± 0.012 % vs. 0.008 ± 0.015 %; hepatic MNCs, 1.42 ± 0.53 % vs. 0.53 ± 0.25 % [CHB patients without clinical anti-HBV treatment], 1.42 ± 0.53 % vs. 0.51 ± 0.09 % [CHB patients with clinical anti-HBV treatment] p < 0.05). It is important to note that, due to the limited amount of some specimens, no significant difference in the percentage of iNKT cells was found between CHB W-anti-HBV and CHB W/O-anti-HBV (p > 0.05; Fig. 1).

Fig. 1
figure 1

iNKT cell frequencies in MNCs from different groups, as analyzed by flow cytometry. iNKT cells (shown in the right upper quadrant) were defined as APC α-GalCer/CD1d tetramers and Vα24-Jα18 TCR FITC T cells. (A) The numbers indicate the percentage of iNKT cells in the isolated PBMC population. (B) Graph comparing the frequency of iNKT cells from liver MNCs between CHB with (w) anti-HBV treatment, CHB without (w/o) anti-HBV treatment and normal controls. *, p < 0.05, as determined by ANOVA. MNCs, mononuclear cells; PBMCs, peripheral blood mononuclear cells

Full size image

Functional and phenotypic analysis of iNKT cells

The expression levels of PD-1, Tim-3, and CD28 on iNKT cells were compared between patients with CHB and healthy controls. In addition, comparisons were made between IL-4+ iNKT and IFN-γ+ iNKT between the two groups. After lymphocyte gating based on forward and side scatter properties, the expression levels of Tim-3+ iNKT, IL-4/PD-1+ iNKT, and IFN-γ/CD28+ iNKT were analyzed for the indicated cell markers. Compared to iNKT cells from healthy controls, those from CHB patients expressed lower levels of CD28 (PBMCs, 46.2 ± 10.6 % vs. 90.1 ± 6.3 %; hepatic MNCs [CHB patients without anti-HBV treatment], 45.3 ± 3.5 % vs. 87.9 ± 9.2 %, [CHB patients with anti-HBV treatment], 67.7 ± 7.2 % vs. 87.9 ± 9.2 %; p < 0.05). In contrast, compared to iNKT cells from healthy controls, those from CHB patients expressed significantly higher levels of Tim-3 and PD-1 (PBMCs, 89.4 ± 10.1 % vs. 23.3 ± 12.6 % and 84.5 ± 9.7 % vs. 35.5 ± 5.5 %, respectively; hepatic MNCs [CHB patients without anti-HBV treatment], 89.3 ± 8.1 % vs. 37.6 ± 4.8 % and 89.1 ± 7.7 % vs. 35.2 ± 9.1 %, respectively; [CHB patients with anti-HBV treatment], 68.9 ± 7.9 % vs. 37.6 ± 4.8 % and 72.4 ± 5.2 % vs. 35.2 ± 9.1 %, respectively; p < 0.05). The IL-4 and IFN-γ levels and number of iNKT cells from CHB patients were much lower than those of the healthy controls (p < 0.05). It is noteworthy that for the CHB patients who received lamivudine as an anti-HBV treatment, the production of both IL-4 and IFN-γ by iNKT cells were increased, but the Tim-3+ expression was slightly reduced in comparison to patients without antiviral treatment (Fig. 2).

Fig. 2
figure 2figure 2

Functional and phenotypic analysis of iNKT cells before cultivation. lymphocyte gating based on side-scatter (SSC) and forward-scatter (FSC) properties. Graph comparing the frequencies of (A/D) Tim-3+iNKT, (B/E) IL-4/PD-1+iNKT, and (C/F) IFN-γ/CD28+iNKT between CHB patients and normal controls. Panels A, B, and C show data from PBMCs, and panels D, E, and F show data from liver MNCs. Data are shown as the mean ± SD. *, p < 0.05

Full size image

Tim-3/Tim-L3 and/or PD-1/PDL1 blockade and/or CD28/CD80 activation enhances the function of α-GalCer-activated iNKT cells in vitro

We designed a method for expanding iNKT cells from PBMCs and liver MNCs with α-GalCer ex vivo. These cells were divided into five groups and given different treatments, as described in “Materials and methods”. The blank controls (group I) showed no change in the frequency of iNKT cells (p > 0.05), and the IL-4/IFN-γ +iNKT levels were not significantly different before and after cell culture (p > 0.05). The PD-1/Tim-3/CD28 expression levels of the iNKT cells after cell culture and after activation by α-GalCer were examined (Fig. 3). In the other four groups, the secretion of IL-4 and IFN-γ increased by different degrees, and IFN-γ secretion increased most obviously. It is important to note that the effect in group IV was significant (Tables 2 and 3). Finally, supernatant fluids were collected and subjected to ELISA to measure IL-4 and IFN-γ cytokines. The IL-4 and IFN-γ levels determined by ELISA and flow cytometry (FCM) technology were identical (Fig. 4)

Fig. 3
figure 3figure 3

In vitro expansion of human iNKT cells. Human PBMCs and liver MNCs were cultured in the presence of α-GalCer at a dose of 200 ng/ml for 72 h. (A/E) iNKT cells were identified by the presence of α-GalCer/CD1d tetramers with Vα24-Jα18 TCR. (B/F) Tim-3+iNKT, (C/G) IL-4/PD-1+iNKT, and (D/H) IFN-γ/CD28 +iNKT were analyzed for their expression of the indicated cell surface markers. Panels A, B, C, and D show data from PBMCs, and panels E, F, G, and H show data from liver MNCs. Data are shown as the mean ± SD, *, p < 0.05

Full size image
Table 2 PBMCs
Full size table
Table 3 Liver MNCs
Full size table
Fig. 4
figure 4

Functional analysis of iNKT cells after culture. PBMCs and liver MNCs were isolated from different control groups and coactivated with α-GalCer/anti-Tim-3/anti-PDL1/anti-CD28/CD80 for 72 h. Concentrations of IL-4+iNKT and IFN-γ+iNKT cytokines in the culture supernatants were measured by ELISA

Full size image

Discussion

Patients with chronic hepatitis B infections usually exhibit decreased Th1 responses and increased Th2 responses. iNKT cells have unique intermediary roles between the innate and acquired immune systems [28]. Both Tim-3 and PD-1 are important negative immune regulation factors that play essential roles in the immune response to HBV infection. However, whether the presence of Tim-3 or PD-1 molecules on iNKT cells plays a key role in HBV infection remains unclear. In this study, we demonstrated that the frequency of iNKT cells was significantly decreased in CHB patients, and high levels of Tim-3 and PD-1 inhibitory receptors were detected on these cells. However, the expression of the costimulant CD28 was much lower in CHB patients than in healthy controls. The iNKT cells had an impaired capability to secrete IL-4 and IFN-γ cytokines in CHB patients. Our study showed that the increase in Tim-3/PD-1 and the low level of CD28 were closely related to the immune impairment of iNKT cells. When the iNKT cells were activated with α-GalCer for 72 h, we found the Tim-3, PD-1, and CD28 expression levels increased, but there were no changes in IFN-γ and IL-4 production. These experiments suggest that α-GalCer not only activates iNKT cells but also induces anergy.

Previous reports have shown that the Tim-3/galectin-9 pathway and Tim-3L, one of the ligands for Tim-3, modulate IFN-γ production. This pathway is particularly important for the immune response to virus infection [29, 30]. Research has shown that IFN-γ cytokines can inhibit HBV-DNA replication [31]. In addition, α-GalCer can induce iNKT cell anergy. In the present study, when the iNKTs from CHB patients were coactivated with α-GalCer/anti-Tim-3/anti-PDL1/anti-CD28/CD80 for 72 h, IFN-γ and IL-4 levels were significantly higher than they were before activation. In the anti-CD28/CD80 treatment group, the effect was particularly obvious. The data also showed that antiviral therapy with lamivudine resulted in a decrease in the expression of the inhibitory factors PD-1 and Tim-3, and an increase in the expression of the positive costimulatory factor CD28 on iNKTs, as compared to the levels before antiviral treatment. However, the PD-1 and Tim-3 expression levels were still higher than those of healthy controls, whereas the CD28 expression level remained relatively low. These results indicated that the Tim-3/Tim-3L, PD-1/PDL1, and CD28/CD80 pathways could regulate the functions of iNKT cells and have some complementary effects. These experiments form an important theoretical basis for the clinical inhibition of HBV-DNA replication.

As mentioned above, using coactivated iNKT cells from different treatment groups, we collected supernatant fluid and performed ELISA to measure IFN-γ and IL-4 cytokines. The results were consistent with the FACS results. These results further showed that through our different processing methods, the functions of iNKT cells could be partially restored. The data also indicated that antiviral therapy with lamivudine could partly restore the anergic T cell activities in CHB patients (although perhaps transiently) when compared to patients without antiviral therapy. However, from a therapeutic perspective, the detailed mechanism still requires further study. Unfortunately, due to clinical and ethical constraints, few liver tissue specimens were collected after lamivudine antiviral therapy, and we did not obtain reliable statistical significance. Thus, increasing the number of samples may be necessary to obtain reliable statistical results in future studies for validation.

In conclusion, we have highlighted for the first time the role of Tim-3 on iNKT cells during HBV infection. In addition, we partly restored the function of iNKT cells by blocking the Tim-3/Tim-3L and PD-1/PDL1 pathways and/or by exciting the CD28/CD80 pathway. Our data showed that the anergic T cell activities in CHB patients were closely related to the high Tim-3 and PD-1 expression levels and low CD28 expression levels. Upregulation of the CD28/CD80 signal in α-GalCer-activated iNKT cells is more effective in increasing IFN-γ and IL-4 secretion by iNKT cells in CHB patients than the downregulation of the Tim-3/Tim-3L and PD-1/PDL1 signals.