| BackgroundThere are more than350million people worldwide and7.18%of population in China (nearly93million) suffered chronic infection of hepatitis B virus (HBV), which leads to increased risk of developing cirrhosis, hepatic decompensation and hepatocellular carcinoma. It’s still a great public health problem we should pay more attention to.HBV is transmitted mainly by blood such as perinatal, percutaneous, and sexual exposure. The risk of developing chronic HBV infection after acute exposure ranges from90%in newborns of HBeAg-positive mothers to25%to30%in infants and to less than5%in adults The natural history of chronic infenction of HBV acquired by perinatal exposure can be classified into four phases according virus-host interaction: immune tolerance, immune clearance, inactive carrier state, and reactivation. Among individuals with perinatally acquired HBV infection, a large percent of HBeAg-positive patients have high serum HBV-DNAbut normal alanine aminotransferase(ALT) levels who are considered to be in the "immune tolerant"(IT) phase. Many of these patients develop HBeAg-positive chronic hepatitis B with elevated ALT levels in the phase of immune clearance (chronic hepatitis B, CHB). The rate of clearance of HBeAg averages between8%and12%per Year. After spontaneous HBeAg seroconversion,67%to80%of carriers have low or undetectable HBV DNA and normal ALT levels with minimal or no necroinflammation on liver biopsy-the "inactive carrier state"(IC), a state of immune control of HBV. Mainly because of virus mutation for immune escape, the roughly10%to20%of inactive carriers may have reactivation of HBV replication and exacerbations of hepatitis to be HBeAg negative hepatitis B.Generally, HBeAg seroconversion signifying stable viral control, whether spontaneous or after antiviral therapy, reduces the risk of hepatic decompensation and improves survival. However, the rate of HBeAg seroconversion is low even in the chronic hepatitis B patients (CHB) receiving antiviral therapy, and the underline mechanism is not clear. A small number of studies have identified immune parameters that predict, and thus potentially contribute mechanistically, to HBeAg seroconversion. For example, serum IL-10and IL-12levels and associated genetic polymorphisms predict spontaneous HBeAg seroconversion; and serum IL-12and serum IL-21levels correlated with HBeAg seroconversion on antiviral therapy. There is currently no evidence that either innate or acquired cellular immune responses precede and/or predict HBeAg seroconversion. It’s found that the T cells responses to HBV were weak, narrow or barely detectable in CHB while it is vigorous, polyclonal and multi-specific in acute self-limited hepatitis B patients. And more functionally active HBV specific CD8+T cells were present in circulation and liver in IC patients compared to CHB patients; the IC patients or responders received IFN-a therapy can be expanded more a CD8+T cells response to HBV that was similar in strength and specificity to patients who had recovered from acute hepatitis than the non-responders. So the CD8+T cell immune hyporesponsiveness may also contribute to the failure control of HBV. Though host immunity mechanisms that potentially explain such T cell hyporesponsiveness to HBV were shown to be various, including impaired function of dendritic cells, high level of CD4+CD25+regulatory T cell, high expression of PD-1on HBV specific CD8+T cell and so on, the factors remain not completely clear and few were shown to closely relate to HBeAg seroconversion during antiviral treatment.Accumulated evidences show that immune regulation play an important role in the maintenance of immune tolerance or hyporesponsiveness to self and foreign antigens. Here we studied another regulatory T cells, CD4-CD8-T cells, named double negative T (DNT) cells. According to the T-cell receptor (TCR) αβ chain and γδ chain on the T cells, DNT cells are classified into two subsets, αβDNT cells and γδDNT cells. aPDNT cells have been shown to possess a unique features of cell surface markers, cytokines profile and inhibition to T cells in an antigen-specific, dose-dependent manner found in both mice and human. One recent study showed that in an HBeAg-TCR double transgenic mouse model, HBeAg-activated αβDNT cells can inhibit the cytokine production of HBeAg-specific and nonspecific effector cells. And γδDNT cells belong to the subsets of y8T cells, and most of γδT cells express neither CD4nor CD8. Though y8T cells play important role in protective immunity, growing evidences show that y8T cells were also crucial for immune tolerance, such as oral tolerance, ocular tolerance and hyporesponsiveness to tumor in murine. Similarly in human, γδT cells were also found to possess regulatory potential. So both αβDNT cells and γδDNT cells may involved in the immune hyporesponsiveness to HBV in CHBObjectTo study the correlation between DNT cells and chronic HBV infection and the role of DNT cells in chronic HBV infection.Methods1. Study Subjects. This study was conducted according to the ethical guidelines of the1975Declaration of Helsinki, and was approved by the Ethical Committee of Nanfang Hospital. Written informed consent was obtained from all subjects.The age, serum ALT, serum AST and serum HBV-DNA levels of the three groups of subjects with a chronic HBV infection in the cross-sectional study, defined according to AASLD criteria. All these subjects were HBsAg-positive on two occasions at least6months apart, and were recruited from the outpatients department of Nanfang Hospital. The IT group (n=22) were HBeAg-positive, had normal ALT levels (ALT<40IU/L) on at least2occasions during at least one year of follow up, and had high serum HBV DNA levels (>1×107copies/mL). The IC group (n=25) were HBeAg-negative, anti-HBe-positive, had serum ALT levels≤40IU/L and serum HBV-DNA<10,000copies/mL. All patients with CHB were HBeAg-positive, had serum ALT>40IU/L on at least2occasions, and had a serum HBV-DNA>100,000copies/mL. The healthy control (HC) group (n=32) were recruited from the staff and students of the Southern Medical University, were negative for HBsAg, HBeAg and anti-HBe and had serum ALT<40IU/L. None of the HC subjects had a history of liver disease.The patients for the longitudinal study (n=51) were recruited from the outpatients department of Nanfang Hospital, and were participating in either the phase IV, multi-center, open-label clinical trial of telbivudine (n=21, LDT600mg/d, trial CLDT600ACN07T) or the Efficacy Optimization Research of Telbivudine Therapy trial (EFFORT, n=30, trial NCT00962533). All patients were HBsAg-positive and had been HBeAg-positive for more than6months. The criteria for inclusion in the trials were a serum ALT>40IU/L and serum HBV-DNA>100,000copies/mL. Twenty mL of heparinized blood were taken at baseline,12weeks (12W),24weeks (24W) and52weeks (52W) during LDT treatment. The patients were classified as responders (n=20) if they achieved HBeAg seroconversion within52 weeks of starting treatment, and were classified as non-responders (n=31) if they did not achieve HBeAg seroconversion. The age, gender, baseline ALT, AST and HBV-DNA levels, and HBV genotypes of these two groups were compared. All these subjects were negative for the anti-hepatitis C virus (HCV), anti-hepatitis E virus IgM, anti-hepatitis A virus IgM and anti-human immunodeficiency virus antibodies.2. Isolation of Peripheral Blood Mononuclear Cells (PBMC) and Flow Cytometric Analysis. PBMC were isolated by Ficoll-Hypaque density gradient centrifugation, frozen in fetal bovine serum containing10%dimethyl sulfoxide and stored in liquid nitrogen until further use. After thawing, cells were stained with a selection from the following fluorochrome-conjugated antibodies:anti-CD3-APC (UCTH1), anti-TCRγδ-FITC (11F2), anti-CD4-PE, anti-CD8-PE and anti-TCRαβ-PE-Cy7(IP26). Flow cytometric analysis was performed using a FACS CanTo Ⅱ flow cytometer. A minimum of30,000lymphocytes were collected and analyzed using BD FACSDiva software.3. Isolation of Liver-infiltrating Lymphocytes (LIL). Liver biopsy tissue was obtained from patients for routine histology, and LIL were isolated from the residual liver biopsy fragment not need for diagnosis. Briefly, liver tissue was gently sliced into1-2mm3pieces, washed twice with10%FBS-RPMI1640(R10) to remove contaminating blood and digested in RPMI1640-containing collagenase IV (500μg/mL), DNase I (20μg/mL), FBS (2%) and BSA (0.6%) at37℃for30min. After a brief vortexing, the suspension was passed through a70-mm-nylon mesh filter to remove cell clumps and undissociated tissue. The filtered suspension was centrifuged and the cell pellet was washed twice in phosphate buffered saline before staining for phenotypic analysis4. Intracellular Cytokine Staining (ICS). A panel of2618-mer peptides overlapping by10residues and covering the full HBV core open reading frame (Core) was obtained from Sigma-Aldrich. These peptides were dissolved in dimethyl sulfoxide, pooled and named the core peptides pool. A single18-mer HIV gag peptide was used as the negative control. Intracellular cytokine staining was performed as follow:Half million cells in100μL R10medium supplemented with stimulation ("blank","anti-CD3(OKT,5μg/mL) and anti-CD28(1μg/mL) antibodies","peptides (3μg/mL), anti-CD49d (1μg/mL) and anti-CD28(1μg/mL)" or "50ng/mL, Phorbol-12-myristate-13-acetate (PMA) plus lumol/L calcium ionomycin") for1hour. Then1μg/mL Brefeldin A was added for an additional5hours. The cells were harvested, stained with fluorochrome-conjugated antibodies specific for the cell surface markers (CD3,CD8,CD4and so on), washed, fixed and permeabilized, washed again, and then staining with fluorochrome-conjugated antibodies specific for cytokine or isotype controls for30minutes before flow cytometric analysis.5. The influence of γδDNT cells on the function of CD8+T cells.5.1The influence of γδDNT cells on the HBV-specific CD8+T cells response. Isolation of γδDNT cells and γδT(-)PBMC was carried out using an indirect magnetic labeling method. The proportion of CD4+γδT cells in γδT cells were less than1%and the proportion of CD8+γδT cells in γδT cells were27%(10%-52%). So CD8+cells were deleted from PBMC before that γδT cells were isolated in order to separate out γδDNT cells. The γδT:γδT(-)PBMC co-cultivations were set up at ratio of1:2. Both group of "γδT:γδT(-)PBMC" and γδT(-)PBMC were cultured with Core (3μg/mL) and the production of INF-γ (PerCP-Cy5) were detected by ICS (anti-TCRγδ-FITC, anti-TCRαβ-PE-Cy7and anti-CD8-APC-Cy7were used for surface stain). The γδT(-)PBMC cultivated with or without Gag (3μg/mL) were repectively set for negative or blank control.5.2The influence of different frequencies of γδT cells on the HBV-specific CD8+T cells response. Four CHB patients with γδDNT/γδT>80%were recruited in this study. The γδT cells and TCRγδ(-)PBMC were isolated as above. The γδT(-)PBMC:γδT co-cultivations were set up at ratios of1:1,2:1,4:1,16:1,16:0with a total of5×106cells/tube. And the production of IFN-γ and TNF-α (PE) was measured by ICS (anti-TCRγδ-FITC, anti-TCRαβ-PE-Cy7and anti-CD8-APC-Cy7were used for surface stain)5.3The influence of different frequencies of γδT cells on the non-specific production of IFN-y in CD8+cells. Four else CHB patients with γδDNT/γδT>80%were recruited in this study. The γδT:γδT(-)PBMC co-cultivations were set up at ratios of0:16,1:16,1:4,1:2,1:1with a total of5x105cells/tube in the presence of anti-CD3(OKT,5μg/mL) and anti-CD28(1μg/mL) antibodies. And the production of IFN-γ (PerCP-Cy5) was measured by ICS (anti-TCRγδ-FITC, and anti-CD8-APC-Cy7and anti-CD4-APC were used for surface stain)5.4The influence of different frequencies of γδT cells on the non-specific CD8+cells. Four else CHB patients with γδDNT/γδT>80%were recruited in this study. γδT(-)PBMC were label with CFSE and the γδT:γδT(-)PBMC co-cultivations were set up at ratios of0:16,1:16,1:4,1:2,1:1with a total of5x105cells/tube. The cells were cultured for3days in R10supplemented with anti-CD3and anti-CD28antibodies, stained with anti-CD8-PE and anti-CD4-APC for cytometry analysis of the proliferation of CD8+cells6. Characterization of DNT subsets in CHB. The PBMC from cross-sectional study were thawed and stained for characterization of DNT as follow:6.1. Activation marker (CD69and HLA-DR):anti-CD3-APC-Cy7, anti-CD4-APC, anti-CD8-APC and anti-TCRγδ-FITC, anti-CD69-PE-Cy7, anti-HLA-DR-PE6.2. coinhibitory molecular(PD-1and PD-L1, CD160and BTLA4, NKG2A):6.2.1. anti-CD3-APC-Cy7, anti-CD4-APC, anti-CD8-APC, anti-TCRγδ-FITC, anti-PD-1-PE-Cy7and anti-PD-L1-PE6.2.2. anti-CD3-APC-Cy7, anti-CD4-FITC, anti-CD8-FITC, anti-TCRαβ-PE-Cy7, anti-BTLA-PE and anti-CD160-Alex fluor6476.2.3. anti-CD3-APC-Cy7, anti-TCRyS-FITC, anti-CD4-PE-Cy7, anti-CD8-PerCP, anti-CD56-APC and anti-NKG2A-PE6.3. Potential production of cytokines:6.3.1. Fresh whole blood was collected from CHB patients and directly cultured with brefeldin A for6hours. The cells were harvested and stained with anti-CD3-APC-Cy7,anti-CD4-APC,anti-CD8-APC,anti-TCRγδ-FITC, anti-TGF-β-PerCp Cy5.5and anti-IFN-y-PE by ICS.6.3.2. Fresh PBMC isolated from CHB patients whose resting PBMC were TGF-β negative were stimulated with PMA and brefeldin A. The cells were harvested for cytokine detection by ICS:anti-TGF-β-PerCP Cy5.5, anti-IL-10-PE, anti-TNF-a-PE, anti-IL-21-PE, anti-IL-17A-Alex Fluor488or anti-IFN-y-PE,7. The Influence of HBV Antigen on the Frequencies of DNT subsets:Fresh PBMC were obtained from15non-responders and11responders after52weeks of LDT therapy, labeled with CFSE and cultured in R10for7days with or without either rHBcAg or rHBeAg. The cells were abtained and stain with anti-CD3-APC-Cy7, anti-CD4-APC, anti-CD8-APC and anti-TCRαβ-PE-Cy7for cytometric analysis. The frequencies of αβDNT cells, γδDNT cells and CD4+/CD8+y8T cells in the proliferating population were respectively compared in different stimulation group.8. Statistical Analysis. The serum HBV-DNA levels were normalized by log transformation for statistical analysis. Data were show with median (range) and analyzed using the Fisher’s Exact Test, Kruskal-Wallis H test, Mann-Whitney U test, Friedman Test, Wilcoxon matched pairs test, Spearman rank order correlation coefficient, logistic regression and receiver-operating characteristic (ROC) curves as appropriate using either SPSS13.0, GraphPad Prism5or SAS. P values (two tailed)<0.05were significant.Results1. A Cross-sectional Comparison of DNT Cell subsets Frequencies in Patients with HBeAg-positive and HBeAg-negative Chronic HBV Infections. The circulating frequencies of aPDNT cells, γδDNT cells, and total y8T cells in CD3+cells from, unstimulated PBMC were compared IT, CHB, IC and HC subjects, using Mann-Whitney U test between every two groups in the case that it was significant over all group using Kruskal-Wallis H test (overall). The frequencies of a(3DNT cells (overall P=0.011) were higher in CHB (1.02%), but not significantly enough in IT (0.97%), than in HC (0.72%, P=0.007) or in IC (0.70%, P=0.006). The frequencies of γδDNT cells dramatically increased in CHB and IT compared with either HC or IC. The median of frequencies of y8DNT cells in CHB (8.3%) was higher than in HC (5.4%, P=0.002) or in IC (5.5%, P=0.004) while it was also higher in IT (6.8%) than in HC (P=0.027) or in IC (P=0.030). Though the comparison of γδT cells frequencies (overall P=0.004) in different groups was similar to γδDNT cells, the difference was not significant enough between HC and IT (P=0.103). The frequency of CD4+/CD8+γδT cells was alsp calculated in each subject as the frequency of γδT cells minus the frequency of γδDNT cells. The frequencies of CD4+/CD8+y8T cells were similar in the four groups (P=0.305). There were no meaningful difference neither between CHB and IT nor between HC and IC (P>0.05), and no significant correlations between the frequencies of any of the4measured T cell subsets and either serum HBV-DNA levels, ALT or AST levels, or age within the CHB group of patients (P>0.05).2.γδDNT Cells Are Associated with HBeAg Seroconversion during Antiviral Therapy.Though the frequencies of αβDNT cells tended to decrease with time without stactistic significance (overall using Friedman test, P=0.125), they significant dercreased at24W (P=0.009) or52W (p=0.010) compared with baseline (beween every two timepoints using Kruskal-Wallis H test). There were no changes in the frequencies of γδDNT cells or total γδT cells over52weeks in either group (P>0.05). The frequencies of αβDNT cells, γδDNT cells, total γδT cells and CD4+/CD8+γδT cells at baseline,12weeks,24weeks and52weeks were compared between the groups of patients who achieved (responders) and did not achieve (non-responders) HBeAg seroconversion using Mann-Whitney U test. Since there were no differences in the frequencies of either αβDNT cells or CD4+/CD8+γδT cells (P>0.05) between the groups, the significantly lower frequencies of total γδT cells at both baseline (P=0.026) and12weeks (P=0.046) in the responder group were due to the frequencies of y8DNT cells. The median of y8DNT cells frequencies at baseline (5.1%, P=0.019) or12W (5.8%, P=0.041) were markedly lower in responders compared with non-responders (10.3%at baseline or10.4%at12W), though it didn’t achieve statistical difference at24W (P=0.071) or52W (P=0.070) due to a non-significant increased trend in responders.Univariate logistic regression showed that HBeAg seroconversion was associated with low baseline frequencies of both γδDNT cells (P=0.04) and HBV-DNA (P=0.005). On multivariate analysis, there was a significant effect of HBV-DNA (P=0.008) and a trend towards an independent effect of y8DNT cells (P=0.075). There was no significant interaction between these variables (P=0.77). Quantitative HBeAg data were available from the EFFORT trial only. HBeAg levels tended (P=0.08) to be lower in responders (n=12) than in non-responders (n=18), and there was no association (r=-0.02, P=0.93) between baseline HBeAg levels and baseline γδDNT cell frequency.Receiver operating characteristic curves (Fig.2E) were generated to analyze the relationship between the baseline frequency of γδDNT cells and the response to52weeks of LDT therapy. The baseline γδDNT cell frequency predicted the response to52weeks of LDT therapy (AUC=0.696, P=0.019). The optimal cut-off value for predicting a response to treatment was a γδDNT cell frequency of7.65%of total CD3-positive PBMC, which had a sensitivity of70.0%and a specificity of71.0%. The highest positive predictive value for HBeAg seroconversion was87.8%with a γδDNT cell frequency of less than4.8%. The highest negative predictive value for HBeAg seroconversion was71.4%with a γδDNT cell frequency of greater than15.2%. The12week frequency of γδDNT cells, and the baseline frequencies of total γδT cells gave similar results (data not shown). The baseline serum HBV-DNA level (AUC=0.76, P=0.002) also predicted the response to treatment.3. Intrahepatic DNT Cells. Liver infiltrating lymphocytes (LIL) were isolated from liver biopsy tissue taken from26CHB patients (9responders and17non-responders) after more than72weeks of LDT treatment. PBMC were obtained at the same time. There were significant correlations between the frequencies of both αβDNT cells (r=0.505, P<0.001) and γδDNT cells (r=0.231, P=0.020) in PBMC and LIL. The frequency of αβDNT cells was higher in LIL than PBMC (Wilcoxon matched pairs test, P<0.001). The frequency of γδDNT cells was similar in LIL and PBMC (P=0.107), though it was higher in LIL compared to PBMC in responders (P=0.021). There was a non-significant trend (P=0.067) for the frequency of γδDNT cells in LIL to be higher in non-responders (6.5%) than in responders(4.4%), while it was higher (P=0.019) in non-responders (8.4%) than in responders in PBMC(5.2%).4The influence of γδDNT cells on the function of CD8+T cells.4.1The influence of γδDNT cells on the HBV-specific CD8+T cells response. The proportion of CD4+y8T cells in y8T cells were less than1%and the proportion of CD8+γδT cells in γδT cells were27%(10%-52%). So CD8+cells were deleted and the purity γδDNT cells in γδT cells were88.9%(86.1%-96.1%). The production of IFN-γ in "γδT:γδT(-)PBMC" were higher than in γδT(-)PBMC(P=0.012), blank control control(P=0.012) or negative control(P=0.017).4.2The influence of different frequencies of γδT cells on the HBV-specific CD8+T cells response. In CHB patients (n=4) with γδDNT/γδT>80%, Along with the increasing of γδT cell frequncies, both production of IFN-γ (Friedman test, P=0.026) and TNF-α (Friedman test, P=0.044) in αβCD8+T cells sitimulated by Core decrease.4.3The influence of different frequencies of γδT cells on the non-specific production of IFN-γ in CD8+cells. In CHB patients (n=4) with γδDNT/γδT>80%, Along with the increasing of γδT cell frequncies, the production of IFN-γ in αβCD8+T cells sitimulated by anti-CD3and anti-CD28decrease (Friedman test, P=0.007).4.4The influence of different frequencies of γδT cells on the non-specific CD8+cells. In CHB patients (n=4) with γδDNT/γδT>80%, Along with the increasing of γδT cell frequncies, the proliferation of αβCD8+T cells sitimulated by anti-CD3and anti-CD28decrease(Friedman test, P=0.003).5. Characterization of DNT Cell Subsets in PBMC from CHB patients. More αβDNT cells and y8DNT cells expressed CD69in CHB than HC, and that more γδDNT expressed HLA-DR in CHB and IT than HC. The data suggest that the frequency of activated DNT cells was higher in CHB and IT than in HC. The frequencies of PD-1and PD-L1expression were higher in αβDNT cells than in γδDNT cells and CD8+T cells。The frequencies of BTLA and CD160expression, which are associated with suppressive activity in CD4+T cells, were higher in γδDNT cells than either αβDNT cells or CD8+T cells. The frequency of NKG2A expression, which is associated with suppressive activity in y8T cells, was higher in γδDNT cells than either NK cells, CD4+T cells or TCRγδ-DNT cells. The expression levels of LAG-3, FasL and CD107were low in all DNT cell subsets.Intracellular cytokine staining was used to measure the cytokine production profiles of the DNT subsets. αβDNT cells from unstimulated whole blood of5of10CHB patients produced TGF-β but not of IFN-y, IL-21, IL-10or IL-17. However TGF-β was produced by the αβDNT cells from the5negative patients after PMA-stimulation of PBMC. The αβDNT cells produced less IFN-y than CD4+/CD8+T cells. In contrast, γδDNT cells produced little TGF-β and high levels of IFN-y. Neither αβDNT cells nor y5DNT cells produced IL-10, IL-21or IL-17.6.Recombinant HBeAg Increases the Frequency of y8DNT Cells in Vitro. Recombinant HBeAg increased the frequency of y8DNT cells relative to both the unstimulated control (P=0.029) and rHBcAg (P=0.004) in the proliferating population, and rHBeAg did not change the frequency of either the αβDNT (P=0.262) or CD4+/CD8+γδT (F, P=0.175) cells. There was no significant influence of rHBeAg on γδDNT cell frequency when PBMC were taken from responders (n=11, data not shown).Conclusion1. Though αβDNT cells was not associated with HBeAg seroconversion during antiviral therapy, high frequencies of this T cell subsets correlated with the HBeAg positive phase of chronic HBV infection and the profile expression of CD69, HLA-DR, PD-1, PD-L1and TGF-β on it indicated the potential role of immune regulation of aPDNT on chronice HBV infection2. Both cross-sectional study and longitudinal study implied that high frequecies of y8DNT cells, but not CD4+/CD8+γδT, were not good for immune control of HBV. 3. Low baseline and12week frequencies of γδDNT cells in CD3+T cells can predict HBeAg seroconversion during LDT therapy for CHB.4. In vitro, γδT cells from CHB patients, mainly containing γδDNT cells, suppressed the HBV specific and non-specific CD8+T cell response. And the higher frequecies of γδDNT cells appeared more depression to function of CD8+T cells.5. The profile expression of co-inhibitor, CD160, BTLA, NKG2A, on γδDNT cells may be associated with the inhibition function of γδDNT6. There was a significant increase in the frequency of γδDNT cells when PBMC from non-responders but not responders were cultured with rHBeAg. This might mean that non-responders have a subset of HBeAg-sensitive γδDNT cells in PBMC that are not present in responders. However the failure to find an association between baseline HBeAg levels and baseline γδDNT cell frequency, and the failure to find a decrease in γδDNT cell frequency in the responder group during treatment suggests that viral factors not a major factor at a clinical level. |
PDF Full Text Request