Circulating CD4⁺ CD25⁺ regulatory T cells correlate with chronic hepatitis B infection

Abstract

Circulating CD4⁺ CD25⁺ regulatory T cells (Tregs) have been demonstrated to maintain immunotolerance and suppress the antigen-specific or antigen-non-specific T-cell responses, but their role in chronic hepatitis B (CHB) infection in humans has not been well characterized. In this study, we analysed the frequency and phenotypic characteristics of CD4⁺ CD25⁺ Tregs in patients of different hepatitis B virus (HBV) infection status, and investigated the effect of Tregs on antiviral immune responses in CHB patients, and the mechanism of this effect. A total of 137 subjects, including 79 CHB patients, 26 asymptomatic HBV carriers (ASCs), 12 acute hepatitis B (AHB) patients and 20 healthy controls, were enrolled in the study. We found that the frequency of CD4⁺ CD25^high Tregs in AHB patients was comparable to that in healthy controls, while it was significantly increased in CHB patients. CD4⁺ CD25⁺ Tregs produced interleukin (IL)-10 but little or no interferon (IFN)-γ under anti-CD3 stimulation. In CHB patients, the frequency of CD4⁺ CD25^high Tregs positively correlated with serum viral load, and the Tregs were capable of suppressing the proliferation and IFN-γ production of autologous peripheral blood mononuclear cells (PBMC) mediated by HBV antigen stimulation in vitro. However, combined administration of anti-programmed death-1 (PD-1) and anti-cytotoxic lymphocyte antigen-4 (CTLA-4) monoclonal antibody slightly enhanced the cellular proliferation and significantly increased the IFN-γ production of PBMC cocultured with Tregs at a ratio of 2 : 1. Thus, the frequency of circulating CD4⁺ CD25⁺ Tregs is increased in patients with CHB, and this may play an important role in viral persistence by modulating virus-specific immune responses.

Introduction

Chronic hepatitis B (CHB) is a common and serious infectious disease of the liver caused by the hepatitis B virus (HBV). About 350 million patients world-wide are chronically infected and become HBV carriers. Current antiviral therapy options include interferon (IFN)-γ and nucleotide analogues, but the problems of drug resistance and severe side effects have not yet been resolved. It is therefore of great interest to identify and establish alternative approaches. During HBV infection, the host immune responses, particularly the cellular immune response, mediate clearance of HBV infection,¹ although the exact mechanisms remain unclear. Unfortunately, the patient often exhibits impairment of HBV-specific T-cell activity during chronic HBV infection.² A goal in the immunotherapy of CHB is to find a way of breaking T-cell tolerance and rebuilding the activity of HBV-specific T cells.

In recent years, regulatory T cells (Tregs) have become a popular subject of immunological research. Sakaguchi et al. were the first to identify a population of CD4⁺ T cells that showed high levels of expression of interleukin (IL)-2Rα (CD25) and that prevented autoimmunity in a murine model.³ Numerous reports in the years that followed shed light on the major aspects of Treg biology in humans, with the characterization of different T-cell subpopulations, including naturally occurring CD4⁺ CD25⁺ Tregs, induced Tregs [IL-10 producing CD4⁺ type I regulatory T cells (Tr1) and T helper type 3 (Th3) cells], and CD4⁺ CD25⁺ T cells that develop in the periphery by conversion of CD4⁺ CD25^– T cells. All these different T-cell populations with regulatory function coexist and contribute to immune suppression.^4–6 There is now considerable evidence that CD4⁺ CD25⁺ Tregs represent a stable population of peripheral lymphocytes. These CD4⁺ CD25⁺ Tregs represent about 5 to 10% of human CD4⁺ T cells, and are commonly identified by the constitutive expression of CD25, as well as the transcription factor scurfin, encoded by the forkhead family transcription factor 3 (Foxp3) gene,⁷ and are also characterized by their very low levels of proliferation on T-cell receptor (TCR) stimulation in vitro.⁸ The CD4⁺ CD25^high subset (which has high levels of CD25 expression) in humans comprises about 1–3% of circulating CD4⁺ T cells. Unlike the total population of CD4⁺ CD25⁺ T cells, these CD4⁺ CD25^high cells can significantly inhibit the proliferation and cytokine secretion induced by TCR cross-linking of CD4⁺ CD25^– responder T cells, CD8⁺ T cells, dendritic cells (DCs), natural killer (NK) cells, and B cells.^9–12

An abundance of experimental data has confirmed that CD4⁺ CD25⁺ Tregs can suppress effective antiviral immune responses. In particular, in the chronic infections caused by the human immunodeficiency virus (HIV) and hepatitis C virus (HCV),^13,14 the frequency and functional properties of Tregs are important because increased numbers of Tregs might favour chronic virus development and influence the course of the disease. Chronic HBV infection is associated with impairment of the proliferative, cytokine production, and cytotoxic effector functions of HBV-specific T cells which probably contributes significantly to viral persistence, and CD4⁺ CD25⁺ Tregs have been found to inhibit effective virus-specific immune reactions,^15,16 although the exact mechanism has not yet been well defined. Moreover, because the findings of different studies have been contradictory,^15–18 controversy remains as to whether the frequency and function of circulating CD4⁺ CD25⁺ Tregs change in CHB patients and whether these factors are correlated with HBV persistence. In the light of these considerations, the aim of the present study was to investigate the number, phenotypes, and cytokine production of CD4⁺ CD25⁺ Tregs in HBV-infected patients of different infection status, and to evaluate the suppressive capacity of Tregs as well as their relationship with HBV DNA replication in CHB patients.

Materials and methods

Subjects

Seventy-nine HLA-A2-positive CHB patients, including 44 HBV envelope antigen (HBeAg)-positive and 35 HBeAg-negative individuals, 26 asymptomatic HBV carriers (ASCs) and 12 patients in the early phase of acute hepatitis B (AHB) were enrolled in the study. The standards for diagnoses of AHB and CHB have been described in detail previously,^18,19 and all patients were hospitalized or followed up in our unit. The diagnosis of ASC was made according to the diagnostic standard of the National Program for Prevention and Treatment of Viral Hepatitis,²⁰ and the ASC subjects were selected from patients undergoing routine conventional examination. No one received anti-HBV agents or steroids 6 months before sampling, and as control, 20 healthy blood donors were also selected. Our study was approved by the local ethics committee, and all patients provided written informed consent.

Virological assessments

HBV serum markers were determined using commercial enzyme immunoassay kits (AXSYM System, Abbott, Wiesbaden, Germany). HBV DNA was extracted from serum samples and quantified using a commercial polymerase chain reaction (PCR) diagnostic kit with a detect limit of 200 copies/ml (PG Biotech, Shenzhen, China).

Isolation of PBMC and CD4⁺ CD25⁺ Tregs

PBMC were isolated from fresh heparinized blood by Ficoll-Hypaque density gradient centrifugation (Biochrom, Berlin, Germany) and suspended at the indicated concentrations in RPMI-1640 supplemented with 10% fetal calf serum. For isolation of CD4⁺ CD25⁺ and CD4⁺ CD25^– T cells, PBMC were further separated using a regulatory T-cell isolation kit (Miltenyi Biotech, Bergisch Gladbach, Germany) and the AutoMACS separation unit (Miltenyi Biotech) according to the manufacturer's instructions. The purity of enriched cells for CD4⁺ CD25⁺ (> 90%) and CD4⁺ CD25^– T (> 86%) was determined by flow cytometry with antibodies against CD3, CD4 and CD25.

Flow cytometry analysis

To determine the frequencies and phenotypes of Tregs in PBMC, multicolour flow cytometry was performed using the following antibodies: anti-CD4, anti-CD25, anti-CD45RA, anti-CD45RO, anti-cytotoxic lymphocyte antigen-4 (CTLA-4) (Coulter, Fullerton, CA), anti-forkhead family transcription factor 3 (Foxp3) and anti-programmed death-1 (PD-1) (eBioscience, San Diego, CA). For surface staining, 10⁶ cells were incubated with specific antibodies at 4° for 30 min. Cells were then washed three times with phosphate-buffered saline (PBS)/2% fetal calf serum (FCS), and 10⁵ cells were usually analysed. For intracellular staining, cells were treated with Fix & Perm Reagent (Coulter) after labelling with surface antibodies, and then incubated with anti-CTLA-4 or anti-Foxp3, washed twice, and submitted to FACScan flow cytometry (EPICS-XL; Coulter). Isotype-matched antibodies were used as controls, and the data were analysed using cellquest software (Coulter).

Real-time PCR analysis of Foxp3 expression

Total RNA was isolated from freshly purified CD4⁺ CD25⁺ T cells if a sufficient PBMC sample could be obtained, according to the protocol for Trizol Reagent (Invitrogen, Carlsbad, CA). cDNA was generated by Moloney murine leukaemia virus reverse transcriptase (BBI, Markham, ON, Canada) using an oligo(dT)17 primer (Bioengineer, Shanghai, China) according to the manufacturer's instructions. Relative quantitative real-time PCR was performed using SYBR-green I Premix Ex Taq on the ABI Prism 7500 (Applied Biosystems, Foster, CA) following the manufacturer's instructions. The primers for β-actin were available commercially, and the primers for Foxp3 were synthesized by Invitrogen (sense: 5′-GGCACTCCTCCAGGACAG-3′; antisense: 5′-GCTGATCATGGCTGGGCTCT-3′). Thermal cycle parameters were 30 seconds at 95°, and 40 cycles of denaturation at 95° for 10 seconds followed by annealing at 60° for 15 seconds and extension at 72° for 40 seconds. All PCR assays were performed in duplicate, and data were analysed with the ABI Prism Detection system using the comparative threshold cycle method as previously described.²¹ Standard curves were generated and indicated excellent amplification efficiency (90–100%).

Cytokine profile of CD4⁺ CD25⁺ Tregs

Freshly purified CD4⁺ CD25⁺ T cells (1 × 10⁵) and PBMC (2 × 10⁵) were stimulated either with 1 µg/ml plate-bound anti-CD3 (BD Pharmingen, San Diego, CA) or 10 µg/ml purified homogenous HBV surface antigen (purity > 99%; National Vaccine & Serum Institute, Beijing, China) in 96-well plates for 5 days. Each well was supplemented with medium to a volume of 200 µl. The IFN-γ and IL-10 concentrations of the culture supernatant were then determined using an enzyme-linked immunosorbent assay (ELISA) kit (BioSource International, Camarillo, CA) as previously described.¹⁷ Purified HCV envelope antigen (10 µg/ml; Invitrogen) was used as a control.

Suppression assay for CD4⁺ CD25⁺ Tregs

Purified CD4⁺ CD25⁺ T cells were cultured alone or together with autologous PBMC at different ratios, and stimulated with 10 µg/ml HBV surface antigen or 1 µg/ml anti-CD3, in the presence of 20 U/ml recombinant human IL-2 (rhIL-2; Sigma, St Louis, MO) in 96-well plates for 6 days. Purified HCV envelope antigen (10 µg/ml) was also used as a control. In alternative experiments, before mixing with PMBC, anti-human PD-1 monoclonal antibody (mAb) (10 µg/ml; BD Pharmingen) and anti-human CTLA-4 mAb (10 µg/ml; BD Pharmingen) were added alone or together to the culture medium of CD4⁺ CD25⁺ T cells, and the culture was maintained for 24 hr and then washed twice with PBS. Cellular proliferation was measured by the addition of 0·5 µCi/well ³H-thymidine for the last 18 hr, and the amount of incorporated ³H-thymidine was determined by liquid scintillation spectroscopy as previously described.²² The supernatants from the proliferation assay were removed before the addition of ³H-thymidine, and the IFN-γ concentration was measured by ELISA as described above.

Statistical interpretation

Data were analysed with spss 12·0 for Windows (SPSS, Chicago, IL). Statistical analysis was performed using the Kruskal–Wallis H-test to assess differences between groups. Spearman correlation analysis was performed to compare the frequency of CD4⁺ CD25^high Tregs and HBV viral titres. P-values less than 0·05 were considered statistically significant.

Results

Frequency analysis of circulating CD4⁺ CD25⁺ Tregs

We analysed peripheral blood from 79 CHB patients, 12 early-stage AHB patients, 26 ASC patients and 20 healthy donors (the clinical parameters are listed in Table 1) to determine the percentage of CD25⁺ T cells in the total CD4⁺ T-cell population. In CHB patients, the CD4⁺ CD25^high (fluorescence intensity of CD25 > 50) population represented 2·62 to 4·53% of CD4⁺ T cells; however, the total CD4⁺ CD25⁺ T-cell population comprised 9·80 to 14·90% of CD4⁺ T cells (Figs 1a and b). As a result, the frequency of CD4⁺ CD25^high in HBeAg-positive CHB patients [median 3·28%; mean ± standard deviation (SD) 3·42 ± 0·53%] was significantly higher than in healthy controls (median 2·45%; mean ± SD 2·72 ± 0·46%; P < 0·01) and AHB patients (median 2·05%; mean ± SD 2·25 ± 0·37%; P < 0·01), but not HBeAg-negative CHB or ASC patients (Fig. 1c). However, there was no significant difference in the frequency of the total CD4⁺ CD25⁺ T-cell population between groups (Fig. 1d; all P > 0·05; Kruskal–WallisH-test).

Table 1.

Clinical parameters for patients and healthy blood donors

	Sex					HBV DNA (copies/ml)
Subjects	Male	Female	Age (years)	ALT (U/l)	AST (U/l)	ND	< 103	103−107	> 107
Normal1	12	8	29·1 ± 6·8	–	–	20	0	0	0
AHB2	7	5	33·1 ± 5·8	729·1 ± 122·7	878·1 ± 94·8	3	4	5	0
CHB, E+3	26	18	37·7 ± 9·3	188·4 ± 52·6	144·7 ± 59·6	4	9	20	11
CHB, E–4	24	11	40·4 ± 8·6	124·5 ± 40·9	133·2 ± 48·6	5	7	18	5
ASC5	17	9	27·4 ± 4·9	< 40	< 50	7	9	10	0

¹Healthy blood donor.

²Acute hepatitis B (early phase) patient.

³Hepatitis B virus envelope antigen (HBeAg)-positive chronic hepatitis B (CHB) patient.

⁴HBeAg-negative CHB patient.

⁵Asymptomatic hepatitis B virus (HBV) carrier.

ALT, alanine aminotransferase; AST, aspartate aminotransferase; ND, not detected.

Figure 1.

Frequency of circulating CD4⁺ CD25⁺ regulatory T cells (Tregs) in various subjects. (a) CD4⁺ T cells were separated into CD25^high, CD25^low and CD25^– T-cell subsets, defined by the fluorescence intensity of CD25 obtained using isotypic control antibody. (b) The frequency of CD4⁺ CD25^high and total CD4⁺ CD25⁺ T cells from chronic hepatitis B (CHB) patients. Data are expressed as box plots, in which the horizontal lines represent the 25th, 50th and 75th percentiles of the measured frequencies of Tregs. (c) The percentages of CD4⁺ CD25^high Tregs in various subjects. The horizontal bars indicate the median percentage of Tregs. The individual frequency for each subject included in the analysis is shown. The significance of differences was calculated using the Kruskal–Wallis H test. (d) The percentage of total CD4⁺ CD25⁺ T cells from different groups. The bars represent the mean (± standard deviation) CD4⁺ CD25⁺ T-cell frequency. AHB, acute hepatitis B (early phase); ASC, asymptomatic hepatitis B virus (HBV) carrier; E⁺, HBV envelope antigen (HBeAg) positive; E^–, (HBeAg) negative; FITC, fluorescein isothiocyanate; PE, phycoerythrin.

Phenotypic analysis of circulating CD4⁺ CD25⁺ Tregs

Freshly isolated PBMC from HBV-infected patients and healthy controls were labelled with CD4, CD25 and a series of specific markers to further characterize these Tregs. CD4⁺ CD25^high, CD4⁺ CD25^low and CD4⁺ CD25^– were analysed for expression of cell surface markers for comparison with the characteristics of Tregs, as previously described.^19,23 For CHB patients, representative surface expression of CD45RO, CD45RA, PD-1 and intracellular CTLA-4 in these three CD4⁺ T-cell subsets is shown in Fig. 2(a). Compared with CD25^– T cells, CD25^high cells had significantly elevated expression of CD45RO and CTLA-4 (CD152) but not CD45RA and PD-1. Surface expression of these specific molecules did not differ between CD25^high and CD25^low T cells (Fig. 2b). On the whole, CD25^high and CD25^low T cells were phenotypically similar in healthy controls and in patients of different HBV infection status (data not shown).

Figure 2.

Characteristics of circulating CD4⁺ CD25⁺ regulatory T cells (Tregs) from chronic hepatitis B (CHB) patients. (a) Representative phenotypic profile of CD45RO, CD45RA, cytotoxic lymphocyte antigen-4 (CTLA-4) and programmed death-1 (PD-1) on different CD4⁺ subsets from a CHB patient. (b) The mean expression levels of CD45RO, CD45RA, CTLA-4 and PD-1¹ were determined, with gating on the CD4⁺ CD25^high, CD4⁺ CD25^low and CD4⁺CD25^– T-cell subsets, respectively. **P < 0·01. The bars represent mean ± standard deviation. The significance of differences was calculated using the Kruskal–Wallis H test. CMV, cytomegalovirus; HBeAg, hepatitis B virus envelope antigen.

Expression of Foxp3 in CD4⁺ CD25⁺ Tregs

Foxp3 is an important transcription factor and the most specific molecular marker for CD4⁺ CD25⁺ Tregs.²⁴ We measured Foxp3 mRNA in CD4⁺ CD25⁺ T cells from various patients and healthy controls. Because the volume of blood samples was limited, Foxp3 in CD4⁺ CD25^– T cells was measured only for CHB patients (n = 7). As shown in Fig. 3(a), purified CD25⁺ T cells from CHB patients expressed high levels of Foxp3, whereas CD25^– T cells expressed low levels of Foxp3. The relative Foxp3 mRNA level in CD25⁺ T cells was 21·5-fold higher than in the CD25^– T-cell subset, and the mean (± SD) relative Foxp3 mRNA levels in CD25⁺ T cells of healthy controls (n = 8), AHB patients (n = 8), CHB patients (n = 14) and ASC patients (n = 8) were 1·0 ± 0·10, 1·033 ± 0·124, 1·219 ± 0·187 and 1·182 ± 0·179, respectively (Fig. 3b; all P > 0·05). Thus, Foxp3 was selectively expressed in CD4⁺ CD25⁺ Tregs and CHB patients had higher levels of Foxp3 mRNA, but there was no statistical difference between the tested groups. Furthermore, we analysed Foxp3 expression in Tregs by intracellular staining of the anti-Foxp3 antibody. Similarly, in CHB patients (Fig. 3c), Foxp3 expression was greatly enhanced in CD4⁺ CD25^high T cells (84·4 ± 16·5%) compared with subpopulations of CD25^low (14·8 ± 6·2%) or CD4⁺ CD25^– T cells (3·9 ± 2·5%), which was consistent with the current description of Tregs. There was also no significant difference in intracellular Foxp3 expression in CD25^low or CD25^high cells between patients and healthy controls (data not shown).

Figure 3.

Forkhead family transcription factor 3 (Foxp3) expression in circulating CD4⁺ CD25⁺ regulatory T cells (Tregs). (a) Foxp3 mRNA level in CD4⁺ CD25⁺ T-cell population from chronic hepatitis B (CHB) patients is presented as expression index calculated by taking the cycle threshold (Ct) value ratio of CD4⁺ CD5⁻ T-cell population as 1. (b) Foxp3 mRNA levels in CD4⁺ CD25⁺ T cells from various subjects are presented as the Ct value ratio of Foxp3:β-actin.²¹ The bars represent the mean (± standard deviation) Foxp3 relative expression level. (c) The mean percentage of intracellular Foxp3-positive T cells in various CD4⁺ T-cell subsets from CHB patients. The significance of differences was calculated using the Kruskal–Wallis H test. AHB, acute hepatitis B (early phase); ASC, asymptomatic hepatitis B virus (HBV) carrier.

Cytokine profile of CD4⁺ CD25⁺ Tregs

In order to examine the cytokine profile of CD4⁺ CD25⁺ Tregs under conditions of TCR linkage and HBV antigen stimulation, PBMC and purified CD25⁺ Tregs were stimulated with plate-bound anti-CD3 or HBV surface antigen (HBsAg). As shown in Figs 4(a) and (b), under stimulation with anti-CD3 or HBsAg, PBMC from CHB patients (n = 12) predominantly secreted a high concentration of IFN-γ but little IL-10. As regards Tregs, under stimulation with anti-CD3, Tregs from CHB patients (n = 9) produced a low level of IL-10 (which is known to inhibit T-cell proliferation) and little or no IFN-γ. In control experiments with HCV envelope antigen, only four of 12 CHB patients tested had elevated IFN-γ production by PBMC, which was similar to that found under HBsAg stimulation, whereas Tregs exhibited no production of IFN-γ or IL-10 (data not shown). Similarly, under conditions of anti-CD3 stimulation, Tregs from healthy controls (n = 5), ASC patients (n = 7) and AHB patients (n = 6) had the same cytokine profile of IL-10 and IFN-γ as CHB patients (data not shown).

Figure 4.

The cytokine profile of circulating CD4⁺ CD25⁺ regulatory T cells (Tregs). Isolated peripheral blood mononuclear cells (PBMC) and CD4⁺ CD25⁺ T cells from chronic hepatitis B (CHB) patients (n = 12) were cultured separately either with anti-CD3 or with hepatitis B virus surface antigen (HBsAg) as described in the ‘Materials and methods’. The interferon (IFN)-γ (a) and interleukin (IL)-10 (b) levels in culture supernatants were measured by enzyme-linked immunosorbent assay (ELISA). Data are represented as mean ± standard deviation.

Suppressive function of CD4⁺ CD25⁺ Tregs

Several studies have shown that suppressive activity of CD4⁺ CD25⁺ Tregs is mainly cell contact-dependent in vitro,^25,26 and CD4⁺ CD25⁺ Tregs produced little or no IL-10 upon HBsAg stimulation in our study. To determine the effect of CD4⁺ CD25⁺ Tregs on responder cells and to investigate the mechanism underlying this effect, PBMC (4 × 10⁵) from CHB patients were cocultured with purified CD4⁺ CD25⁺ Tregs under stimulation with HBsAg or anti-CD3. Both PBMCs and Tregs proliferated strongly in response to anti-CD3 but more weakly in response to HBsAg stimulation (Fig. 5a). In contrast to the case of non-specific antigen stimulation with anti-CD3, when stimulated with HBsAg the proliferation (Fig. 5b) and IFN-γ production (Fig. 5c) of PBMC were inhibited in the presence of Tregs in a dose-dependent manner (in 14 CHB patients). Meanwhile, control HCV antigen had no effect on stimulation of cellular proliferation and IFN-γ production (data not shown). However, at PBMC:Tregs ratios of 5 : 1 to 1 : 2, exogenous rhIL-2 (200 U/ml) completely abrogated this suppression (unpublished data), which was consistent with other reports.^16,18 Interestingly, when PBMC were cocultured, at a ratio of 2 : 1, with Tregs that had been pretreated with anti-PD-1 and anti-CTLA-4 mAb, cellular proliferation was slightly increased (Fig. 5d), while the IFN-γ production (Fig. 5e) of responder T cells was significantly increased. However, alone, anti-PD-1, anti-CTLA-4 or isotypic antibody had no significant effect.

Figure 5.

Inhibitory capability of circulating CD4⁺ CD25⁺ regulatory T cells (Tregs). (a) The cellular proliferation of isolated peripheral blood mononuclear cells (PBMC) and CD4⁺ CD25⁺ T cells from chronic hepatitis B (CHB) patients under anti-CD3 or hepatitis B virus surface antigen (HBsAg) stimulation. (b) The proliferation of PBMC cocultured with Tregs was determined using the ³H-thymidine incorporation assay. (c) The interferon (IFN)-γ level in the coculture supernatant was measured by enzyme-linked immunosorbent assay (ELISA). *P < 0·05, **P < 0·01, compared with the data for the PBMC group. (d, e) The effects of anti-programmed death-1 (PD-1) and anti-cytotoxic lymphocyte antigen-4 (CTLA-4) monoclonal antibodies used alone or together on cellular proliferation and IFN-γ production of PBMC cocultured with Tregs at a ratio of 2 : 1. All data are shown as the mean ± standard deviation for four to six subjects in each group. The significance of differences was calculated using the Kruskal–Wallis H test. P:T, PBMC:CD4⁺ CD25⁺ Tregs.

CD4⁺ CD25⁺ Treg frequency was associated with HBV viral titres in CHB patients

To investigate whether the increase in circulating CD4⁺ CD25⁺ Tregs was correlated with HBV replication level, we measured the viral titres of serum from all tested CHB patients. Spearman analysis showed that the frequency of CD4⁺ CD25^high Tregs (also that of CD4⁺ CD25⁺ T cells; data not shown) positively correlated with serum HBV DNA load (Fig. 6a), but there was no correlation between Tregs and serum alanine aminotransferase (ALT) level (data not shown). The results suggested that an increased frequency of CD4⁺ CD25⁺ Tregs may be associated with a negative immune response, leading to poor viral clearance in CHB patients. As the persistence of HBV infection is associated with a high frequency of circulating Tregs, to find out whether this subset would be down-regulated as HBV viral load decreased, we followed up nine HBeAg-positive CHB patients who received antiviral treatment. Before initiation of therapy, a high percentage of CD4⁺ CD25^high Tregs in PBMC was detected in all subjects, while antiviral therapy resulted in a dramatic decline in plasma viral load with a simultaneous decrease in the frequency of CD25^high Tregs (Figs 6b and c). However, as the sample size in our study was small, these results need to be confirmed in a large-cohort study.

Figure 6.

Association between regulatory T-cell (Treg) frequency and hepatitis B virus (HBV) DNA load in chronic hepatitis B (CHB) patients. (a) In 71 CHB patients tested, the correlation between circulating CD4⁺ CD25^high Treg frequency and serum HBV DNA load was analysed by Spearman correlation analysis. (b) CD4⁺ CD25^high Treg frequency and HBV DNA load in nine CHB patients naïve to antiviral therapy. (c) CD4⁺ CD25^high Treg frequency and HBV DNA load in these nine CHB patients after antiviral therapy. Each bar represents the percentage of CD4⁺ CD25^high Tregs in each patient. Each triangle in the same panel depicts the viraemia level detected in each patient at the time of Treg analysis.

Discussion

Tregs have attracted a great deal of attention over the past few years as a consequence of their ability to suppress CD4 and CD8 effector T-cell responses. The immune regulation of CD4⁺ CD25⁺ Tregs is complicated, and the mechanisms of suppression of antiviral immune responses are not yet very clear. It has been suggested that Tregs may have evolved to prevent immunopathological damage but also contribute to viral persistence.²⁷ We investigated the theory that the immunodysregulation associated with chronic HBV infection is mediated partially by changes in Tregs. Our study did not show a significant increase in total CD4⁺ CD25⁺ Tregs in the peripheral blood of CHB patients, but a smaller population of CD4⁺ CD25^high T cells was significantly up-regulated in CHB (especially HBeAg-positive) and ASC patients compared with AHB patients or healthy controls.

Furthermore, we used several cell surface markers to distinguish CD4⁺ CD25⁺ Tregs from activated CD4⁺ cells. It is well known that CD25, OX40 and glucocorticoid-induced tumour necrosis factor receptor (GITR) are increased, mainly on Tregs but also on conventional T cells, upon activation.^28,29 Therefore, we also used other markers, namely CD45RO, CD45RA, CTLA-4, PD-1 and Foxp3, to identify Tregs precisely. CTLA-4 and PD-1, two inhibitory receptors expressed by subsets of activated T cells, could deliver a negative signal that results in down-regulation of T-cell activation.^30,31 In our study, in contrast to the situation in CD4⁺ CD25^– T cells, CTLA-4 was constitutively expressed in CD4⁺ CD25⁺ Tregs, while PD-1 was not over-expressed in comparison with conventional CD25^– T cells, findings that are similar to those of an earlier study.²⁵ Moreover, for CD4⁺ T cells purified either from patients or from healthy controls, the concentration of Foxp3 was significantly higher in the CD4⁺ CD25⁺ subset, suggesting that Foxp3 is essential for the genesis and function of CD4⁺ CD25⁺ Tregs.

The regulation of CD4⁺ CD25⁺ Tregs is mostly non-specific,³² while preferential inhibition of the HBV antigen-specific T-cell response has been reported in some cases.^16,18 In HCV- and HIV-infected subjects, Tregs may contribute to the persistence of infection by inhibiting HIV- or HCV-specific T-cell responses.^14,26,33 In our study, Tregs were capable of suppressing the proliferation and IFN-γ production of autologous PBMC mediated by HBsAg, indicating that circulating Tregs may suppress HBV-specific T-cell responses in CHB patients. Moreover, many studies have suggested that the suppressive effect of CD4⁺ CD25⁺ Tregs on responder cells is contact dependent.^34,35 In an additional study in which blocking with anti-IL-10 or anti-transforming growth factor (TGF)-β mAb was used, we found no effect on the suppression by CD4⁺ CD25⁺ Tregs of cocultured PBMC (unpublished data). Notably, however, when PBMC were cocultured with CD4⁺ CD25⁺ Tregs that had been pretreated with anti-CTLA-4 and anti-PD-1 mAbs, the cellular proliferation of PBMC was slightly increased and IFN-γ production was significantly increased. In light of previous findings,^36–38 we suggest that this may have been a result of the selective expression of high levels of PD-1 and CTLA-4 by CD4⁺ CD25⁺ Tregs after HBV infection, and the combined blocking of these two negative signalling pathways (PD-L1/PD-1 and CTLA-4/B7) by these mAbs. However, this effect was completely abrogated when Tregs and PBMC were cocultured at a high ratio (Treg:PBMC ratio) (data not shown), and Shevach et al. reported no inhibitory effect of blockade of CTLA-4 with anti-CTLA-4 mAb.³⁹ These findings suggest that the PD-L1/PD-1 or CTLA-4/B7 pathway might mediate only a small part of the Treg function that results in the inhibition of target CD4⁺ T cells. However, as pointed out in other studies,^40,41 it is still possible that IL-10 may play a role synergistically in vivo with a cell-contact mechanism to mediate suppression, so the mechanism of suppression of HBV-specific T-cell responses may not be unique. Thus, the precise mechanisms of regulation may depend on the tissue microenvironment and the nature of the inflammation.

Our study showed a positive correlation between CD4⁺ CD25⁺ Treg frequency and serum HBV DNA load, suggesting that the up-regulation of Tregs is associated with an increase in HBV replication. This result is identical to that reported recently by Yang et al.,⁴² but, in contrast to their findings, we did not find any association of HBeAg status with either CD4⁺ CD25^high or CD4⁺ CD25⁺ T cell frequency. Moreover, two earlier studies did not find a significant association between circulating Treg frequency and HBV DNA titre in CHB patients.^15,16 Differences in the methods, reagents and samples used in the study might account for the discrepancies. Our preliminary data also suggested that the elevation in the number of circulating Tregs in CHB patients decreased after the antiviral treatment, and the antigen-specific T-cell response to HBsAg was more significantly suppressed by Tregs. Taken together, these results support the hypothesis that chronic HBV infection leads to the induction of suppressive Tregs which inhibit antiviral immune responses.

In summary, our findings demonstrate that there is a marked increase in circulating CD4⁺ CD25⁺ Tregs in CHB patients. Tregs play a negative role not only in modulating the effectors of immune responses by inhibiting IFN-γ secretion and cellular proliferation upon HBV antigen stimulation, but also in influencing the viral load and disease persistence. These results indicate that modulation of CD4⁺ CD25⁺ Tregs might be one potential therapeutic strategy for the treatment of CHB. However, this study was limited by analysis of the circulating compartment only, and further detailed investigation of the frequency and function of intrahepatic CD4⁺ CD25⁺ Tregs needs to be carried out.

Abbreviations

AHB

acute hepatitis B

ALT

alanine aminotransferase

ASC

asymptomatic HBV carrier

CHB

chronic hepatitis B

CTLA-4

cytotoxic lymphocyte antigen-4

Foxp3

forkhead family transcription factor 3

GITR

glucocorticoid-induced tumour necrosis factor receptor family related receptor

PBMC

peripheral blood mononuclear cells

PD-1

programmed death-1

Treg

regulatory T cell