Tumor necrosis factor alpha inhibits the suppressive effect of regulatory T cells on the hepatitis B virus–specific immune response

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

Chronicity of hepatitis B virus (HBV) infection is characterized by a weak immune response to the virus. CD4+CD25+ regulatory T cells (Treg) are present in increased numbers in the peripheral blood of chronic HBV patients, and these Treg are capable of suppressing the HBV-specific immune response. The aim of this study was to abrogate Treg-mediated suppression of the HBV-specific immune response. Therefore, Treg and a Treg-depleted cell fraction were isolated from peripheral blood of chronic HBV patients. Subsequently, the suppressive effect of Treg on the response to HBV core antigen (HBcAg) and tetanus toxin was compared, and the effect of exogenous tumor necrosis factor alpha (TNF-α), interleukin-1-beta (IL-1β), or neutralizing antibodies against interleukin-10 (IL-10) or transforming growth factor beta (TGF-β) on Treg-mediated suppression was determined. The results show that Treg of chronic HBV patients had a more potent suppressive effect on the response to HBcAg compared with the response to tetanus toxin. Neutralization of IL-10 and TGF-β or exogenous IL-1β had no effect on Treg-mediated suppression of the anti-HBcAg response, whereas exogenous TNF-α partially abrogated Treg-mediated suppression. Preincubation of Treg with TNF-α demonstrated that TNF-α had a direct effect on the Treg. No difference was observed in the type II TNF receptor expression by Treg from chronic HBV patients and healthy controls. Conclusion: Treg-mediated suppression of the anti-HBV response can be reduced by exogenous TNF-α. Because chronic HBV patients are known to produce less TNF-α, these data implicate an important role for TNF-α in the impaired antiviral response in chronic HBV. (HEPATOLOGY 2007.)

Worldwide 400 million people suffer from a chronic hepatitis B virus (HBV) infection and approximately 1 million people die annually of HBV-related disease. Infection with HBV in adults results frequently in a self-limiting, acute hepatitis, which confers protective immunity and causes no further disease. In 10% of infected adults HBV leads to a chronic infection.1–3 In patients with an acute self-limiting HBV infection, a multispecific CD4+ and CD8+ T cell response with a type 1 cytokine profile is important for control of the infection.4, 5 Patients with a chronic HBV infection lack such a vigorous multispecific T cell response.6, 7

Regulatory T cells (Treg) play an important role in maintaining peripheral immune tolerance.8–10 Treg can be distinguished from CD4+ effector T cells by their continuous expression of the activation marker CD25 and the expression of the Treg-specific transcription factor FoxP3.11, 12 Besides regulating peripheral tolerance, Treg also can affect T cell responses to foreign antigens.13 Patients with a chronic HBV infection have increased percentages of Treg in their peripheral blood compared with healthy controls and individuals who have resolved their HBV infection.14 Recently, Xu et al15 showed that patients with a high viral load have an increased proportion of Treg compared with patients with a low viral load. Treg from patients with a chronic HBV infection are capable of inhibiting the HBV-specific CD4+ and CD8+ T cell response in a dose-dependent manner.14–16

The mechanism of the suppressive activity exerted by Treg on the T cell response remains unclear. There is inconsistency in the literature about the possible involvement of the immunoregulatory cytokines interleukin-10 (IL-10) and transforming growth factor beta (TGF-β).17, 18 Recently, Treg were shown to express the type II tumor necrosis factor (TNF) receptor.19 Tumor necrosis factor alpha (TNF-α) plays an important role in the control of viral infections. It is involved in the recruitment and activation of macrophages, it can polarize the T cell response toward the development of antiviral effector functions, and it has a direct antiviral effect20, 21 TNF-α is essential for the proliferation of HBV-specific cytotoxic T cells.22

Given the important role of Treg in the impaired antiviral immune response during a chronic HBV infection, this study is focused on which factors can abrogate Treg-mediated suppression and restore the anti-HBV immune response.

Abbreviations

DC, dendritic cell; HBcAg, Hepatitis B virus core antigen; HBeAg, hepatitis B virus e antigen; HBV, Hepatitis B virus; IL-1β, interleukin 1 beta; IL-2, interleukin 2; IL-10, interleukin 10; PBMC, peripheral blood mononuclear cell; SEM, standard error of the mean; SI, stimulation index; TGF-β, transforming growth factor beta; TNF-α, tumor necrosis factor alpha; Treg, regulatory T cell.

Patients and Methods

Patients.

For functional experiments, peripheral blood samples were obtained from chronic HBV patients. All patients had detectable HBV deoxyribonucleic acid levels in serum. For the functional assays, peripheral blood mononuclear cells (PBMC) were used from 29 patients with a chronic HBV infection. Eleven patients were HBV e antigen (HBeAg) positive, the median viral load was 2.5 × 106 geq/ml (1.0 × 103 − 1.6 × 1010), and the median alanine aminotransferase level was 32 units/l (16–321). For TNF receptor staining, peripheral blood samples from 13 age-matched chronic HBV patients and healthy controls were used. Six patients were HBeAg positive, the median viral load was 3.5 × 105 geq/ml (1.0 × 103 − 1.6 × 1010), and the median alanine aminotransferase level was 32 units/l (11-137). Patients co-infected with either human immunodeficiency virus, hepatitis A virus, hepatitis C virus, or hepatitis D virus and patients with a resolved viral hepatitis were excluded from this study. Also patients and controls who were immunocompromised or pregnant and patients who received antiviral or immunomodulatory HBV treatment during the last 6 months before blood sampling were excluded from this study. The institutional review board of the Erasmus MC-University Medical Center Rotterdam approved this study, and informed consent was obtained from all patients before their inclusion in this study.

CD4+CD25+ T Cell Isolation.

PBMC from chronic HBV patients were obtained by Ficoll separation (Ficoll-Paque plus, Amersham Biosciences, Buckinghamshire, UK). CD4+ T cells were isolated from PBMC by negative selection using the untouched CD4+ T cell isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany). CD4+CD25+ T cells were isolated from CD4+ T cells using anti-CD25-microbeads (Miltenyi Biotec). The isolations were performed according to the manufacturer's instructions. The CD4 and the CD4+CD25 fraction were pooled and were used as CD25 (Treg)-depleted responder cells. The CD4+CD25+ Treg purification method resulted in a Treg fraction containing more than 90% pure CD4+CD25+ Treg and a Treg-depleted cell fraction containing all other cell types present in PBMC. These isolated cell fractions were used in proliferation assays. Depending on the yield of isolated CD4+CD25+ cells, the cells were used for the different functional experiments.

Flow Cytometry.

The purity of the isolated cell fractions was determined by flow cytometry with antibodies against CD4 and CD25. The following antibodies were used: anti-CD4-PerCP-Cy5.5 (SK3) (Becton Dickinson, San Jose, CA) and anti-CD25-PE (M-A251) (BD Pharmingen, San Jose, CA). An isotype-matched control antibody was used to determine the level of background staining for the CD25-PE antibody. The viability of Treg, preincubated with either IL-2 alone or IL-2 and TNF-α, was determined using 7-amino-actinomycin D (BD Pharmingen) and annexin V-APC (BD Pharmingen).

The FoxP3 antibody staining was performed according to the manufacturer's instructions. Briefly, cell surface markers were stained with anti-CD25-PE and anti-CD4-PerCP-Cy5.5. The cells were fixed and permeabilized with Fix/perm buffer, and permeabilization buffer (eBiosciences, San Diego, CA) anti-FoxP3-APC (PCH101) (eBiosciences) was added during permeabilization. For the CD120b (TNF receptor type II) staining, biotin-conjugated anti-CD120b (hTNFR-M1) (BD Pharmingen) and streptavidin-PerCP (Becton Dickinson), anti-CD25-PE, anti-FoxP3-APC and anti CD4-FITC (13B8.2; Immunotech, Marseille, France) were used. For anti-CD25-PE, anti-FoxP3-APC and biotin-conjugated anti-CD120b isotype-matched control antibodies were used. After staining, the cells were analyzed using a 4-color cytometer (FACScalibur, CELLQuest Pro software, Beckton Dickinson).

Proliferation Assay.

The proliferation of CD4+ T cells was determined as described earlier.14 Briefly, the pooled CD25-depleted cells and the pooled CD25-depleted cells supplemented with 20% CD4+CD25+ Treg were cultured in triplicate. CD25-depleted cells were cultured in a concentration of 1 × 105cells per well in 100 μl RPMI-1640 (Bio Whittaker, Verviers, Belgium) containing 5% pooled human serum (Department of Immunohematology and Bloodbank, Leiden University Medical Center, Leiden, The Netherlands) and penicillin/streptomycin (Gibco, Paisley, UK). For the culture condition with 20% Treg, 25,000 of the isolated CD4+CD25+ cells were added to the Treg-depleted cells. The cells were cultured for 6 days either in the presence of 1 μg/ml HBV core antigen (HBcAg) (gift of M. van Roosmalen, Biomerieux, Boxtel, The Netherlands), 15 limits of flocculation/ml purified tetanus toxin (SVM, Bilthoven, The Netherlands), or without extra stimuli. For IL-10 neutralization, 5 μg/ml rat anti-human and viral IL-10 (JES-9D7) (BD Pharmingen) was used, and rat immunoglobulin G1 (R3-34) (BD Pharmingen) was used as an isotype-matched control antibody. For TGF-β neutralization, 0.5 μg/ml goat anti-human TGF-β1 [anti-human latency associated peptide (LAP) (TGF-β1); R&D systems, Oxon, UK] was used, and purified normal goat immunoglobulin G (R&D systems) was used as an isotype-matched control antibody. To determine the effect on TNF-α and interleukin-1β (IL-1β) on Treg-mediated suppression, 25 ng/ml TNF-α (Strathman Biotech) or 50 ng/ml IL-1β (Strathman Biotech) was added to the cell culture medium. For the Treg preincubation experiment, the CD4+CD25+ were cultured overnight with 25 ng/ml TNF-α and 100 u/ml IL-2 (Chiron Corp.) or with IL-2 alone. After this overnight incubation, the Treg were washed twice with RPMI-1640 medium containing 5% human serum and added to HBcAg-stimulated CD25-depleted cells.

After 5 days of incubation, the cells were pulsed with 0.25 μCi/well of [3H]thymidine (Radiochemical Centre, Amersham, Little Chalfont, UK). The cells were harvested 16 hours later. Proliferation was determined by liquid scintillation counting of the harvested cells and expressed as stimulation index (SI).

Mathematics and Statistics.

All results are given as mean ± standard error of the mean (SEM). The SI was calculated from the counts per minute with antigen divided by the counts per minute without antigen. The percentage of suppression by Treg was determined using the following formula: [1 − SI (Treg-depleted cells + 20% Treg) / SI (Treg-depleted cells)] × 100%. The data were analyzed with SPSS 11.5 for Windows (SPSS, Chicago, IL). The Wilcoxon paired signed rank sum test was used to analyze the data from the experiments in which the percentage of suppression between 2 different culture conditions was compared.

Results

Treg from HBV Patients Have a Stronger Suppressive Effect on the HBV-Specific Response Compared with the Response Against Tetanus Toxin.

Treg were isolated using a CD4+ and CD25+ isolation kit; subsequently, the FoxP3 expression of the isolated CD4+CD25+ cells (Treg) and the CD25 (Treg)–depleted cell fraction was determined. To determine whether the suppression of the response against HBV by Treg was antigen specific, the suppressive effect of Treg on the response against tetanus toxin and HBcAg was compared. Data from 1 representative patient are shown in Fig. 1A. Treg showed a stronger inhibitory effect on the response against HBcAg compared with the response against tetanus toxin (59.7% × 4.3% vs 40.8% ± 6.7%, P = 0.02; Fig. 1B). No correlation was observed between the suppressive capacity of Treg and viral load or alanine aminotransferase levels of the patients. Also, the suppressive capacity between Treg isolated from HBeAg-positive or HBeAg-negative patients was not different (data not shown).

Figure 1.

Treg-depleted cells and Treg-depleted cells reconstituted with 20% CD4+CD25+ Treg were stimulated with either HBcAg or with tetanus toxin. After 6 days, the proliferation was determined by [3H]-thymidine incorporation. The proliferation is depicted as stimulation index (SI; counts stimulated cells/counts unstimulated cells). (A) SI of a representative experiment with cells from an HBV patient. The black bars are Treg-depleted cells and the gray bars are Treg-depleted cells reconstituted with 20% Treg. (B) The mean percentage of suppression by 20% Treg on the response to HBcAg and tetanus toxin (n = 18). The percentage of suppression was calculated using the following formula: [1 − SI (Treg-depleted cells + 20% Treg)/SI (Treg-depleted cells)] × 100%. *P < 0.05 compared with the tetanus toxin response.

Neutralization of IL-10 or TGF-β Does Not Abrogate Treg-Mediated Suppression.

Activated Treg can produce the immunoregulatory cytokines IL-10 and TGF-β; however, whether these cells exhibit their suppression via these cytokines is unclear. Therefore, we determined the effect of neutralization of IL-10 and TGF-β on HBV-specific Treg-mediated suppression for 10 patients. Treg-depleted cells and Treg-depleted cells reconstituted with 20% Treg were stimulated with HBcAg in the presence of neutralizing antibodies against IL-10 or TGF-β or isotype-matched control antibodies. Neutralization of IL-10 did not affect Treg-mediated suppression (48.5% ± 11.1% vs 48.1% ± 11.7%). Also, neutralization of TGF-β did not result in a decrease of Treg-mediated suppression of response against HBcAg (32.6% ± 12.3% vs 29.8% ± 11.2%).

TNF-α Partially Abrogates Treg-Mediated Suppression of the Response Against HBcAg.

To determine whether pro-inflammatory cytokines have a modulatory effect on Treg-mediated suppression of the response against HBcAg, Treg-depleted cells and Treg-depleted cells reconstituted with 20% Treg were stimulated with HBcAg in the presence of exogenous TNF-α or IL-1β. Incubation with exogenous TNF-α resulted in a decrease in the Treg suppression of the proliferation [62.4% suppression ± 5.3% (no cytokine) vs 29.8% ± 7.8% (TNF-α), P = 0.02]. This effect was not observed when IL-1β was added to the proliferation assay (54.2% ± 8.7%) (Fig. 2A). Because the receptor for TNF is not only expressed by Treg but also by a variety of other cells, TNF-α theoretically could affect all cell types present in culture. Therefore, we determined whether the decreased suppressive capacity was caused by a direct effect of TNF-α on Treg. Isolated Treg were preincubated overnight with either IL-2 alone or with IL-2 and TNF-α. After overnight culture Treg were extensively washed and added to HBcAg-stimulated Treg-depleted cells. Figure 2B shows that preincubation of Treg with TNF-α and interleukin-2 (IL-2) resulted in a partial abrogation of Treg-mediated suppression as well [42.0% suppression ± 6.0% (IL-2 alone) versus 19.5% ± 7.4% (IL-2 and TNF-α), P = 0.016). In addition, TNF-α partially abrogated the suppressive effect of Treg from patients with a chronic HBV infection and healthy controls when tetanus toxin was used as a stimulus (data not shown).

Figure 2.

(A) Treg-depleted cells and Treg-depleted cells reconstituted with 20% CD4+CD25+Treg were stimulated with HBcAg in the presence of no additional cytokines, 25 ng/ml TNF-α, or 50 ng/ml IL-1β (n = 10). (B) Treg-depleted cells were stimulated with HBcAg, and Treg were incubated overnight with either IL-2 or IL-2 and TNF-α (n = 7). After overnight preincubation, Treg were added to HBcAg-stimulated Treg-depleted cells. After 6 days, the proliferation was determined by [3H]thymidine incorporation. The percentage of suppression was calculated using the following formula: [1 − SI (Treg-depleted cells + 20% Treg)/SI (Treg-depleted cells)]. Data are depicted as mean ± SEM. *P < 0.05 compared with control.

To examine whether TNF-α affected the vitality of Treg, cells were preincubated with IL-2 alone or IL-2 and TNF-α. Their vitality after preincubation was determined by flow cytometry with 7-amino-actinomycin D and Annexin V. Preincubation of Treg with TNF-α did not result in decreased vitality; no increase in apoptosis was seen, and cell death was observed (Fig. 3A, B). During co-culture, no difference in cell death was observed between Treg preincubated with IL-2 and TNF-α or IL-2 alone as determined by flow cytometry with anti-FoxP3-APC and 7AAD (3.5% ± 0.8% vs. 3.7% ± 1.0%). Because FoxP3 expression may be related to the suppressive capacity of Treg, the effect of TNF-α on FoxP3 expression was determined. Overnight preincubation with TNF-α did not affect the FoxP3 expression of Treg (Fig. 3C, D). In addition, preincubation of Treg with TNF-α did result in a downregulation of CD120b on Treg (Fig. 3E, F, n = 5; P = 0.04).

Figure 3.

Treg were incubated overnight with either IL-2 or IL-2 and TNF-α (n = 4). Cell death and apoptosis was determined by staining with 7AAD and Annexin V. (A) Flow cytometry data from 1 representative 7AAD and Annexin V staining. (B) Percentages death and apoptotic cells. Data are expressed as mean ± SEM. (C) Flow cytometry data from a representative FoxP3 staining after incubation with either IL-2 and TNF-α or IL-2 alone. The dotted line depicts the FoxP3 expression of cells preincubated with IL-2, and the solid line depicts the FoxP3 expression of cells preincubated with IL-2 and TNF-α. The staining with the isotype-matched control antibody is depicted in gray. (D) The mean fluorescence intensity of FoxP3 of preincubated Treg. The data are expressed as mean ± SEM. (E) Representative staining for CD120b on Treg incubated with either IL-2 and TNF-α or IL-2 alone. The dotted line depicts the CD120b expression of cells preincubated with IL-2, and the solid line depicts the CD120b expression of cells preincubated with IL-2 and TNF-α. The staining with the isotype-matched control antibody is depicted in gray. (F) The mean fluorescence intesity of CD120b of preincubated Treg. Data are expressed as mean ± SEM. *P < 0.05 compared with control.

Similar Expression of TNF Receptor Type II Was Observed by Treg of Chronic HBV Patients and Healthy Controls.

TNF-α has a direct effect on Treg-mediated suppression; therefore, TNF receptor expression by Treg might be an indication of the Treg sensitivity for TNF-α of the Treg. The TNF type II receptor (CD120b) has been shown to be expressed by Treg.19 Therefore, freshly isolated PBMCs from patients with a chronic HBV infection and healthy controls were stained with antibodies against CD120b, CD4, and FoxP3. No difference was observed in the expression of CD120b by CD4+FoxP3+ cells between patients and healthy controls. CD120b is expressed by 26.6% ± 3.6% CD4+FoxP3+ cells in patients versus 26.8% ± 3.5% in healthy controls (Fig. 4).

Figure 4.

The proportion of TNF receptor type II (CD120b) positive Treg in patients and healthy controls determined by flow cytometry. Freshly isolated PBMCs stained with antibodies against CD120b, FoxP3, and CD4. Treg were defined as cells staining positive for CD4 and FoxP3. The bar represents the mean proportion of Treg expressing CD120b (n = 13).

Discussion

Treg are involved in the impaired immune response during chronic HBV. They are capable of inhibiting HBV-specific proliferation and interferon gamma production of CD4+ and CD8+ T cells.14–16 Treg can have a direct inhibitory effect on T cells or can inhibit the T cell response in an indirect manner by inhibiting dendritic cell (DC) maturation.17, 23

Treg have to be activated through their T cell receptor to become suppressive. After activation, the suppression by Treg is non–antigen-specific.24 Treg suppress the immune response in a dose-dependent manner.25 This study shows that the suppressive effect of circulating Treg of a chronic HBV patient on the CD4+ T cell response to HBcAg is proportionally more potent compared with the suppressive effect on the response to tetanus toxin. This could explain the fact that chronic HBV patients have an impaired HBV-specific immune response, but still have an adequate immune response to tetanus toxin. In fact, the presence of HBV-specific Treg has been described in a recent publication.26

Treg can secrete IL-10 and TGF-β after activation8, 27, 28; however, abrogating Treg-mediated suppression of the HBV-specific response by means of IL-10 or TGF-β neutralization was not possible. This concurs with data from earlier in vitro studies in which neutralization of these 2 cytokines had no effect on the suppression of T cells stimulated with anti-CD3 and anti-CD28.25, 29

Our study shows that exogenous TNF-α can abrogate Treg-mediated suppression. Proinflammatory cytokines, such as IL-1β, IL-6, and TNF-α, are important mediators during the initiation of an adaptive immune response. IL-6 can influence suppression by Treg, however, not through a direct effect on Treg but by making effector cells anergic for Treg-mediated suppression.30 The current study shows that TNF-α was capable of directly affecting the suppressive capacity of Treg. IL-1β had no such effect. Treg can express the TNF receptor type II19; however, no difference was seen in the expression of this receptor between Treg of patients with a chronic HBV infection and Treg of healthy controls. Incubation of Treg with TNF-α resulted in a downregulation of the type II TNF receptor by Treg. Furthermore, the inhibitory effect of TNF-α on Treg-mediated suppression was not attributable to cell death of Treg, because preincubation of Treg with TNF-α did not result in increased cell death or apoptosis. Foxp3 expression has been suggested to correlate with the suppressive capacity of Treg. Valencia et al.19 showed that incubation of Treg with TNF decreased the Foxp3 expression of Treg and abrogated their suppressive capacity.19 In our study we showed that the TNF-induced abrogation of the Treg suppressive capacity was not mediated by downmodulation of Foxp3 expression. No downregulation of Foxp3 expression by Treg was observed after overnight incubation or incubation for 48 hours (data not shown) with TNF-α. The cytokine environment and the availability of cytokine receptors are factors that can influence the immune response in chronic HBV patients. Neutralization of the TNF receptor during therapy with infliximab for Crohn's disease resulted in (re)activation of HBV among chronic HBV carriers.31, 32 Infliximab has been shown to restore the suppressive capacity of Treg.19, 33 PBMCs from patients with an acute HBV infection produce more TNF-α compared with PBMCs from patients with a chronic HBV infection, and a previous study by Van der Molen et al. showed that the total amount of TNF-α produced by isolated myeloid-derived DC from patients with a chronic HBV infection was decreased compared with myeloid-derived DC from healthy controls.34, 35 The decreased TNF-α production observed in patients with chronic HBV infection could result in the presence of Treg that are more suppressive for the HBV-specific immune response.

TNF-α appears to be a key mediator in the chronicity of HBV; it can partially abrogate Treg-mediated suppression of HBV-specific T cells, it is essential for the proliferation of HBV-specific cytotoxic T lymphocytes,22 it has a direct anti-viral effect,20 and it can induce DC maturation.36 Because TNF-α is such an important cytokine during a chronic HBV infection, TNF-α might be a target for future immunotherapeutic strategies.

In conclusion, the suppression of the response against HBV by Treg is partially HBV specific and cannot be abrogated by neutralization of IL-10 and TGF-β or by exogenous IL-1β. TNF-α is capable of decreasing the suppressive capacity of Treg. Together with the decreased TNF-α production observed in patients with chronic HBV infection, this suggests that TNF-α is an important mediator of the impaired immune response observed in patients with chronic HBV infection.

Acknowledgements

We thank M. van Roosmalen for providing recombinant HBcAg and Dr. A. Boonstra for critical reading of the manuscript. Furthermore, we thank W.F. Leemans, M.J. ter Borg, E.H.C.J. Buster, A. Keizerwaard, L.A. van Santen-van der Stel and C. van de Ent-van Rij for their help with obtaining peripheral blood samples.

Ancillary

Advertisement