Hepatitis B virus (HBV) antigen-pulsed monocyte-derived dendritic cells from HBV-associated hepatocellular carcinoma patients significantly enhance specific T cell responses in vitro

Authors


Dr Fu-Sheng Wang, Research Centre of Biological Therapy, Beijing Institute of Infectious Diseases, Beijing 302 Hospital, 100 Xi Si Huan Middle Road, Beijing 100039, China.
E-mail: fswang@public.bta.net.cn

Summary

To investigate whether hepatitis B virus (HBV) antigen-pulsed monocyte-derived dendritic cells (MoDC) could mount a T cell response in hepatocellular carcinoma (HCC) patients associated with chronic HBV infection, peripheral blood mononuclear cells (PBMCs) from 36 HBV-associated HCC patients were induced into MoDC and pulsed with hepatitis B core antigen (HBcAg) and hepatitis B surface antigen (HBsAg), alone and in combination. Co-stimulatory molecules CD80, CD86 and CD40, as well as human leucocyte antigens D-related (HLA-DR) were found to express at the highest level on MoDC pulsed with HBcAg or HBsAg + HBcAg, at a median level on MoDC pulsed with HBcAg or HBsAg alone, and at the lowest level on non-antigen-pulsed MoDC. Interleukin (IL)-10 and IL-12 cytokines were released by antigen-pulsed MoDC at increased levels in the order: no-antigen <  HBsAg < HBcAg < HBcAg + HBsAg. MoDC pulsed with HBcAg or HBsAg + HBcAg also had the strongest ability to stimulate autologous T cell proliferation and intracellular interferon (IFN)-γ production. HBcAg- or HBsAg + HBcAg-pulsed MoDC could also induce HBV core peptide-specific CD8+ T cell proliferation determined by tetramer staining. In addition, the antigen-pulsed MoDC were found to have a stronger capacity to produce IL-12 and induce T cell response in vitro for patients with higher alanine transaminase (ALT) levels than those with lower ALT levels, indicating that antigen pulse could substantially reverse the impaired function of MoDC in primary HCC patients with active chronic hepatitis B. In conclusion, HBV antigen-pulsed MoDC from HCC patients with chronic hepatitis B could induce HBV-specific T cell response in vitro.

Introduction

Hepatitis B virus (HBV) infection is one of the major causes of chronic hepatitis worldwide. It is estimated that more than 350 million individuals are currently developing lifelong viral persistence and have a significant risk of progressive liver cirrhosis and primary hepatocellular carcinoma (HCC) with high morbidity and mortality [1]. In China, approximately 95% of HCC patients have a history of chronic HBV infection. Increasing evidence suggests that host immune responses are an important factor in determining the outcome of HBV infection. Acutely infected individuals typically develop a strong, polyclonal, multi-specific cytotoxic T lymphocyte (CTL) response to the virus while this response is attenuated markedly in chronically infected patients [2]. Based on these observations, the magnitude and diversity of the HBV antigen-specific T cell responses are thought to be responsible for viral clearance. An inadequate T cell response to the virus probably favours viral persistence in HBV-infected patients who usually have a defective function of antigen-presenting cells (APCs). Therefore, a compelling challenge is not only to gain more insight into the immune pathogenesis of chronic hepatitis B, but also to develop powerful novel approaches to boost the host immune response to control chronic HBV infection and prevent the development of HCC in these patients.

Dendritic cells (DC) orchestrate the immune response against invading pathogens including bacteria, parasites and viruses by acting as professional APCs. APCs take up various antigens and initiate a primary immune response, driving naive T lymphocytes to develop into effector cells and activating epitope-specific memory CD8+ T cells [3]. Impaired DC function has been found in chronic diseases caused by hepatitis C virus (HCV) [4], human immunodeficiency virus (HIV) [5] and HBV [6–15], usually in an immunocompromised state such as immunological tolerance or ignorance, and peripheral anergy. In particular, viral infected DC could down-regulate the expression of surface co-stimulatory molecules, decrease cytokine production and reduce their capability to function as APCs, which generally fail to trigger efficient cellular immune responses against HBV, and may represent a common mechanism for persistent virus infection [16,17]. Thus, therapeutic induction and/or activation of the T cell response by DC pulsed with HBV antigens have a potential to boost anti-HBV adaptive immunity, including a specific CTL response for the efficient control of HBV infection. Recent reports have indicated that hepatitis B surface antigen (HBsAg)-pulsed DC can effectively break the cytotoxic T lymphocyte tolerance in HBV transgenic mice [18–21] and induce an immune response in healthy volunteers or in immunosuppressed HBV carriers. A clinical trial has shown that autologous HBsAg-loaded DC vaccine could suppress HBV replication in chronic hepatitis B patients [22]. Another report suggested that therapeutic strategies should be designed to simultaneously target multiple viral antigens, one of which is the preparation of a DC vaccine pulsed by one or multiple HBV antigens [23]. More importantly, DC immunization has been demonstrated to be able to reverse immune tolerance and induce specific anti-tumour immune responses in cancer patients [24–26].

However, although HBsAg-pulsed DC have been administered in patients with chronic hepatitis B and shown promise for treatment of patients with chronic hepatitis B, until now few studies have investigated whether HBV antigen-pulsed DC have the capability to stimulate lymphocytes in HBV-associated HCC patients. To address these issues, we prepared HBV antigen-pulsed monocyte-derived dendritic cells (MoDC) from primary HCC patients associated with chronic hepatitis B, and then evaluated their function and phenotypic properties in vitro. We found that MoDC pulsed with HBV antigens could significantly enhance autologous HBV-specific T cell responses in these patients. Although few specific HCC antigens have been identified, some evidence has shown that HBV antigen might serve as one of the tumour-associated antigens of HCC [27–29]. In this study, our findings indicate that HBV antigen-pulsed DC from HCC patients with chronic hepatitis B could induce HBV antigen-specific T cell responses in vitro. This investigation favours the development of HBV antigen-pulsed MoDC against HCC as well as HBV in clinical trials.

Materials and methods

Study patients

Blood samples were obtained from 36 primary HCC patients (30 males and six females with average ages of 48·5 ± 10·6 years) who had had at least a 20-year history of chronic HBV infection and were diagnosed simultaneously with liver cirrhosis. All these patients were found routinely to have space-occupying anomalies examined by ultrasonography, computed tomography or magnetic resonance imaging and the abnormally elevated levels of serum alpha fetoprotein (499·5 ± 468·8 ng/ml) [30]. They did not receive any anti-viral treatment, surgical tumour removal and chemical therapy before the blood sampling. The study protocol was approved by the ethics committee of our unit, and written informed consent was obtained from all participants. Individuals with concurrent HCV, hepatitis G virus (HGV) or HIV infections and autoimmune liver disease were excluded. The baseline clinical data are shown in Table 1.

Table 1.  Clinical profile of chronic hepatitis B patients.
ParameterNumberMean ± s.d. or range
  • a

    Values are expressed as mean ± s.d. (range).

  • b

    b Values are expressed with range. ALT: alanine transaminase; HCC: hepatocellular carcinoma; HBV: hepatitis B virus; HBcAg: hepatitis B core antigen; HBsAg: hepatitis B surface antigen.

Gender (M/F)30/6 
Age (years)a 48·5 ± 10·6 (29–76)
Case history (years)a 24·3 ± 8·8 (20–31)
Serum HBV DNAb
 < 105 copies/ml160–9·35 × 104
 > 105 copies/ml201·69 × 105−1·72 × 108
ALT(U/l)a
 ≤ 80 U/l1936·6 ± 16·1 (23–73)
 > 80 U/l17197·7 ± 130·4 (121–499)
HBV molecular marker
 HBsAg(+), anti-HBc(+)13 
 HBsAg(+), HBeAg(+), anti-HBc(+)8 
 HBsAg(+), anti-HBe(+), anti-HBc(+)15 
Liver cirrhosis35 
HCC stage
 I and II3 
 III and IV33 

Generation of HBV antigen-pulsed MoDC

Peripheral blood mononuclear cells (PBMCs) were obtained from each individual using a blood cell separator (Spectra version 6·1 Cobe; Denver, CO, USA) as described previously [31]. Mature MoDC were generated as described previously with some modifications [13]. Briefly, PBMCswere suspended at 5 × 106 cells/ml in six-well plates in serum-free AIM-V medium (Gibco/Invitrogen, Carlsbad, CA, USA). After incubation for 1·5 h at 37°C, non-adherent cells were removed gently. The remaining adherent cells were incubated in serum-free AIM-V medium in the presence of granulocyte–macrophage colony-stimulating factor (GM-CSF) (800 U/ml) and interleukin (IL)-4 (1000 U/ml) (PeproTech, Rocky Hill, NJ, USA) at 37°C. Cells were fed every 2 days with fresh medium containing IL-4 and GM-CSF and harvested at 5 days as immature DC. To load HBV antigens, immature DC were pulsed with various doses of HBV antigens that had been verified as lipopolysaccharide-free alone and in combination at day 6. After 24 h, DC were cultured for another 2 days in AIM-V medium supplemented with the cytokine cocktail, including IL-1α, tumour necrosis factor (TNF)-α, IL-6 (all at 1000 U/ml; PeproTech) and prostaglandin E2 (PGE2) (1 mg/ml, PeproTech) together with CD40 ligand [32] to mimic DC maturation. Recombinant HBsAg (HBV genotype C) was kindly supplied by the Kang-Tai Biological Company (Shen Zhen City, Guangdong, China). Purified recombinant hepatitis B core antigen (HBcAg) was kindly supplied by Ke-Wei Reagent Company (Bejing, China).

Phenotypical analysis of MoDC

Both immature and mature MoDC were analysed phenotypically using a fluorescence-activated cell sorter (FACS) (FACSCalibur; Becton Dickinson, San Diego, CA, USA). Isotype-matched monoclonal antibodies (MoAbs) were used as negative controls. Cells were stained with conjugated monoclonal mouse anti-human antibodies: anti-CD80-fluorescein isothiocyanate (FITC), anti-CD40-FITC, anti-CD86-phycoerythrin (PE), anti-CD83-PE and anti-human leucocyte antigen D-related (HLA-DR)-PE (Becton Dickinson), and analysed using colour flow cytometry and CellQuest software (Becton Dickinson).

Cytokine release of MoDC

Supernatants of MoDC stimulated with HBcAg (20 µg/ml) and HBsAg (20 µg/ml) alone and in combination were harvested at day 3. The IL-10 and IL-12 levels were measured using commercially available enzyme-linked immunosorbent assay (ELISA) kits according to the manufacturer's instructions (Genzyme, Cambridge, MA, USA).

Proliferation assay of autologous T cells

The T cell stimulatory capacity of MoDC was analysed by CellTiter 96®AQueous non-radioactive cell proliferation assay kit (Promega, Madison, WI, USA) according to the manufacturer's instructions. Briefly, irradiated (30 Gy) MoDC (washed twice before use) were suspended in serum-free AIM-V medium and placed in 96-well round-bottomed culture plates (Costar, NY, USA) at densities of 1·0 × 104 cells/well. Autologous responder T cells were added to each well at a density of 1·2 × 105, 2·5 × 105, 5·0 × 105 and 10·0 × 105 cells/well (rate of DC/T cell as 1/12, 1/25, 1/50 and 1/100), respectively, and the final volume of each well was 100 µl. Twenty µl CellTiter 96®AQueous non-radioactive reagent was added to each well, and the plate was incubated for 4 h at 37°C in a humidified 5% CO2 atmosphere. After incubation, the absorbance at 490 nm was recorded using an ELISA plate reader.

Analyses of HLA genotyping and HBV epitope-specific Tetramer positive CD8+ T cells

Screening for the HLA-A2 haplotype was performed by staining PBMCs with an anti-HLA-A2 antibody (BD/PharMingen, San Jose, CA, USA) followed by a FITC-conjugated goat anti-mouse IgG secondary antibody and flow cytometric analysis. Patients were further confirmed to have the HLA-A2·01 allele by polymerase chain reaction (PCR) DNA typing according to our previous protocols [33].

The amino acid sequences of the epitope peptides (core amino acids 18–27 FLPSDFFPSI, envelope amino acids 183–191 FLLTRILTI and polymerase amino acids 575–583 FLLSLGIHL) used in this study were consistent with the indicated regions of genotype C of an HBV strain that is highly prevalent in northern China. Peptides were synthesized by Time Biomedecs Corporation (LID; Guanzhou, Guangdong, China).

Three tetrameric peptide–major histocompatibility complexes (MHC), respective of the HLA-A0201 allele, were purchased commercially (Proimmune, Oxford, UK) and used to directly quantify the CD8+ T cells specific for the HBV core (amino acids 18–27), envelope (amino acids 183–191) and pol (amino acids 575–583) epitopes. MoDC were matured by HBcAg and HBsAg, either alone or in combination, and co-incubated with autologous T cells at a 1 : 20 ratio for 72 h. T cells were incubated with 10 µg/ml of the corresponding peptides (core 18–27 and envelope 183–191) alone and in combination for another 72 h. After incubation, cells were stained with 1 µg of PE-labelled tetrameric peptide–MHC and saturating amounts of FITC-anti-CD8 and peridinin chlorophyll protein (PerCP)-anti-CD3 (both Sigma, St Louis, MO, USA) in phosphate-buffered saline (PBS) at 4°C for 30 min. Analysis was performed on a FACSCalibur (Becton Dickinson) using CellQuest software. T cells stimulated with no-antigen-pulsed DC were used as a negative control.

Analysis of IFN-γ-producing T cells by enzyme-linked immunospot (ELISPOT) assay

Frequencies of HBV core and envelope antigen-specific CTL precursors were analysed using an ELISPOT assay (U-Cytech, Utrecht, the Netherlands), as describedpreviously [34] with modifications. Polyvinylidene difluoride (PVDF) membrane-bottomed 96-well ELISPOT plates (Millipore, Bedford, MA, USA) were coated with 10 µg/ml of anti-IFN-γ MoAb (U-Cytech). Antigen-pulsed MoDC were seeded in triplicate (1 × 104 cells/well) together with 1·25 × 105 cells/well of autologous T cells. After incubation for 48 h, cells were discarded and plates were washed in PBS-0·05% Tween and incubated with biotinylated anti-IFN-γ MoAb (U-Cytech). After washings, plates were incubated with gamma aminobutyric acid (GABA), washed again and incubated with Activator I/II solution. The staining was stopped by rinsing with demineralized water and red spots were counted as single spot-forming cells (SFC). All results were expressed as means of triplicates.

Virological assessment

The protocols for monitoring the serum HBV antigen/antibody markers and HBV DNA levels are the same as described recently [35].

Statistical analysis

Data were expressed as mean ± standard deviation (s.d.). All data were analysed using spss software (SPSS Inc., Chicago, IL, USA). Statistically significant differences between the two groups were determined by applying an unpaired Student's t-test, and a P-value less than 0·05 was considered statistically significant.

Results

Phenotypic and functional restoration of antigen-pulsed MoDC

The densities of CD80 and CD83 expressions were substantially lower on immature MoDC than mature MoDC without antigen stimulation (9·3 ± 3·3% versus 34·0 ±  20·7% for CD80; 4·5 ± 2·7% versus 13·2 ± 6·5% for CD83; both P < 0·05). There were no differences in the expression of surface markers such as CD86, CD40 and HLA-DR molecules between immature and mature MoDC (46·0 ± 22·1% versus 45·2 ± 22·0%; 49·0 ± 45·1% versus 54·2 ± 23·2%; and 50·8 ± 20·3% versus 54·6 ± 20·7% for CD86, CD40 and HLA-DR, respectively) (all P > 0·05) (data not shown).

To determine the most optimal dose of antigen to pulse immature MoDC, we tested different concentrations of HBsAg or HBcAg at 10, 20, 40 or 60 µg/ml. CD80, CD86, CD40 and HLA-DR surface molecules were expressed in mature MoDC stimulated with 20 µg/ml of HBcAg or 20 µg/ml HBsAg at significantly higher levels than in MoDC stimulated with 10 µg/ml of antigen or non-antigen-pulsed MoDC. Increasing the HBcAg or HBsAg dose to 40 or 60 µg/ml did not change the CD80 and CD86 levels significantly. These results indicate that increasing the dose of HBcAg or HBsAg over 20 µg/ml did not enhance expression of CD80 and CD86 on MoDC (Fig. 1a, b). Thus, we chose the optimal dose of HBcAg or HBsAg as 20 µg/ml for further experiments and used HBsAg + HBcAg (both at 20 µg/ml) to load the MoDC. CD83, CD40 and HLA-DR expression was significantly higher on MoDC pulsed with HBsAg and HBcAg in combination compared with HBsAg and HBcAg alone or no-antigen (Fig. 1c). The results indicated that HBsAg + HBcAg in combination could enhance the expression of co-stimulatory molecules on MoDC in chronic HBV-infected patients with primary HCC.

Figure 1.

Expression of surface molecules on monocyte-derived dendritic cells (MoDC) pulsed with various doses of hepatitis B virus (HBV) antigen as analysed by flow cytometry. (a) Levels of CD86 and/or human leucocyte antigen D-related (HLA-DR) expressions were higher at a concentration of 20 µg/ml hepatitis B surface antigen (HBsAg) than those without antigen or with 10 µg/ml antigen. Increasing the dose of HBsAg to above 20 µg/ml has no further enhanced expression of surface molecules (*P < 0·05 compared to the no-antigen group, ▴, P < 0·05 compared to 10 µg/ml group). (b) The levels of CD80, CD86, CD40 and HLA-DR expressions were higher in mature MoDC stimulated with 20 µg/ml HBcAg. Increasing the dose of HBcAg to above 20 µg/ml has no effect on the expression of surface molecules on MoDC (▪, P < 0·05 compared to 0 and 10 µg/ml groups). (c) The levels of CD83, CD40 and HLA-DR expression on the surface of DC were significantly higher when pulsed with HBsAg (20 µg/ml) and HBcAg (20 µg/ml) in combination than those pulsed with no-antigen, HBsAg and HBcAg alone. The levels of CD80 and CD40 expressions were also significantly higher in HBcAg-pulsed DC than in no-antigen or HBsAg-pulsed DC (*P < 0·05 compared to the no-antigen group; ●, P < 0·05 compared to HBsAg group; ▴, P < 0·05 compared to HBcAg group).

The levels of IL-10 produced by MoDC loaded with HBcAg alone and HBsAg + HBcAg were higher than those loaded with HBsAg alone or no-antigen (Fig. 2a). Significantly higher levels of IL-12 p70 were released by MoDC after loading with HBsAg or HBcAg alone or in combination compared to the no-antigen group (Fig. 2b). Furthermore, the levels of IL-12 p70 and IL-10 in the supernatant of MoDC loaded with HBcAg alone and HBsAg + HBcAg were significantly higher than those loaded with HBsAg alone. The results indicated that HBcAg alone and HBcAg + HBsAg were more efficient than HBsAg alone at inducing IL-12 and IL-10 release by MoDC of HCC patients with chronic hepatitis B in vitro.

Figure 2.

Secreting capacity of interleukin (IL)-10 and IL-12 by hepatitis B core antigen (HBcAg) and HBsAg alone or in combination with pulsed monocyte-derived dendritic cells (MoDC). (a) IL-10 production by HBcAg alone and combined with HBsAg pulsed DC was significantly higher than that produced by no-antigen-pulsed DC or HBsAg-pulsed DC. (b) IL-12 production by HBcAg alone or combined with HBsAg pulsed DC was significantly higher than that produced by no-antigen-pulsed DC or HBsAg-pulsed DC. IL-12 production by HBsAg-pulsed DC was also higher than that produced by the no-antigen-pulsed DC (*P < 0·05, **P < 0·01).

Antigen-pulsed MoDC enhanced capacity of autologous T cell proliferation and IFN-γ production

The T cell proliferation-inducing capacities of MoDC were assessed in an autologous mixed lymphocyte reaction. MoDC were pulsed with no-antigen, or with HBcAg or HBsAg alone and in combination, then co-cultured with autologous T lymphocytes. The results showed that the T cell proliferation-inducing capacities of MoDC pulsed with either HBcAg and HBsAg alone or in combination were significantly higher than non-antigen-pulsed MoDC, and had a positive DC versus T cell ratio-dependent relationship (Fig. 3a).

Figure 3.

Autologous T cells proliferation and producing interferon (IFN)-γ stimulated by different antigen-pulsed monocyte-derived dendritic cells (MoDC). (a) The capacity of MoDC to induce T cell proliferation was assessed in an autologous mixed lymphocyte reaction. The T cell proliferation capacities induced by hepatitis B core antigen (HBcAg) and HBsAg alone or in combination pulsed MoDC were significantly higher than those pulsed with no-antigen (*P < 0·05 compared with HBsAg, HBcAg and HBsAg + HBcAg groups). (b) Analysis of hepatitis B virus (HBV) antigen-specific T cells was performed by IFN-γ enzyme-linked immunospot assay (ELISPOT) analysis. The numbers of spots produced by T cells co-cultured with HBcAg-pulsed DC and HBsAg + HBcAg-pulsed DC were significantly higher than by T cells co-cultured with no-antigen-pulsed DC and HBsAg-pulsed-DC. The number of spots was also higher for T cells co-cultured with HBsAg-pulsed DC than for no-antigen-pulsed DC. Boxes represent the 5th, 25th, 75th and 95th percentiles and the median value (solid line) of each group (*P < 0·05, **P < 0·01).

Mature MoDC were generated and co-cultured with autologous T lymphocytes for 48 h, and then T lymphocytes were tested by ELISPOT assay for IFN-γ production in vitro. The numbers of spots produced by T cells co-cultured with HBcAg-pulsed DC (41·0 ± 32·7 SFC/105 T cells), HBsAg-pulsed DC (18·0 ± 13·4 SFC/105 T cells) or DC pulsed with HBsAg + HBcAg (44·7 ± 38·0 SFC/105 T cells) were significantly higher compared to T cells co-cultured with no-antigen-pulsed DC (13·3 ± 9·1 SFC/105 T cells, all P < 0·05). The numbers of spots produced by MoDC loaded with HBcAg alone or HBsAg + HBcAg were also significantly higher than DC pulsed with HBsAg alone (Fig. 3b). The data indicated that MoDC of HCC patients with chronic hepatitis B that were pulsed with HBcAg and HBsAg alone and in combination could induce more IFN-γ-producing specific T lymphocytes, and HBcAg alone or combined with HBsAg was more efficient than HBsAg alone.

Increased production of HBV epitope-specific CD8+ T cells by antigen-pulsed MoDC

HLA-A2-positive peptide tetramers were used to directly quantify the CD8+ T cells specific for the HBV core (amino acids 18–27), env (amino acids 183–191) and pol (amino acids 575–583) epitopes. In general, HBV epitope-specific CD8 T cells in HCC patients with chronic hepatitis B (CHB) are very low, even at an undetectable level by using epitope–MHC tetramer staining. In our study, we found that the stimulation of HBV antigen-pulsed MoDC, together with the corresponding HBcAg and/or HBsAg epitopes alone and in combination, could efficiently induce proliferation of specific CD8 T cells of these patients. As shown in Fig. 4, the frequencies of HBcAg 18–27 epitope-specific tetramer-positive CD8+ T cells were higher in the HBcAg-pulsed and HBsAg + HBcAg-pulsed MoDC groups than those in the no-antigen-pulsed MoDC group. However, the frequencies of HBsAg 183–191 epitope-specific tetramer-positive T cells in HBsAg alone and the HBsAg + HBcAg-pulsed MoDC groups were the same as those in the no-antigen-pulsed MoDC group. These data indicate that MoDC that are pulsed with HBcAg and HBsAg alone or in combination could increase more significantly the frequencies of HBV epitope-specific CD8+ T cells.

Figure 4.

Analysis of hepatitis B virus (HBV) epitope-specific CD8+ T cells using human leucocyte antigen (HLA)-A2-restricted epitope tetramer staining. The frequencies of HBV core (amino acids 18–27), env (amino acids 183–191) and pol (amino acids 575–583) epitope tetramer-positive cells were detected by flow cytometry in T lymphocytes stimulated by DC loaded with hepatitis B core antigen (HBcAg), HBsAg or HBsAg + HBcAg. The core 18–27 tetramer-positive T cells were detectable in HBcAg alone and combined with HBsAg-pulsed DC-stimulated T cells, and were not found in the no-antigen-pulsed or HBsAg-pulsed DC-stimulated T cells. These results indicate that HBcAg alone and combined with HBsAg-pulsed DC could induce core 18–27 epitope-specific CD8 T cells in hepatocellular carcinoma (HCC) patients associated with chronic HBV infection.

In contrast, MoDC pulsed with HBsAg alone or combined with HBcAg had no effect on the frequencies of HBV envelope-specific CD8+ T cells.

Correlation of MoDC function, alanine transaminase (ALT) level and virus load

We further conducted a retrospective analysis to determine whether the capability of HBV antigen-pulsed MoDC to stimulate T cell proliferation is influenced by the baseline status of these patients. The relation of MoDC function to serum ALT level and viral load is shown in Fig. 5. In patients with ≤ 105 copies/ml of HBV DNA there were no differences in IL-12 release and IFN-γ-producing T cell frequencies (data not shown). However, among the patients with > 105 copies/ml of HBV DNA, IL-12 release and IFN-γ-producing T cell frequencies induced by MoDC pulsed with HBsAg and HBcAg (alone and in combination) were higher in patients with ALT level > 80 U/l than with ALT level ≤ 80 U/l (P < 0·05). These results show that HCC patients with higher ALT levels have a stronger cellular response to HBV antigen-pulsed MoDC, which may imply that HBV antigen-pulsed MoDC vaccine treatment is probably more suitable for HCC + CHB patients who have a immune active status.

Figure 5.

The relationship between dendritic cell function and serum levels of alanine transaminase (ALT) or hepatitis B virus (HBV) DNA. Samples were divided into two groups as the level of ALT ≤ 80 U/l and > 80 U/l. In patients with HBV DNA load > 105 copies/ml, the levels of interleukin (IL)-12 production and the frequencies of interferon (IFN)-γ-producing T cells induced by hepatitis B surface antigen (HBsAg) and HBcAg alone or in combination with pulsed DC were higher in the group with ALT level > 80 U/l than in the group with ALT level ≤ 80 U/l. These results showed that hepatocellular carcinoma (HCC) patients with higher ALT levels had a stronger cellular response to HBV antigen-pulsed monocyte-derived dendritic cells (MoDC). Boxes represent the 5th, 25th, 75th and 95th percentiles and the median value (solid line) of each group. (○ No-Ag; ▵ HBsAg; □ HBcAg; ◊ HBsAg + HBcAg) (*P < 0·05).

Discussion

Chronic HBV or HCV hepatitis is the leading cause of hepatocellular carcinoma. Impaired allostimulatory function of DC has been described in patients with chronic hepatitis B and in patients with HCC accompanied by HBV or HCV infection. A decreased function of peripheral blood DC was described; also, DC from HCC patients were shown to be predominantly immature, with a reduced proliferation capacity and reduced IL-12 and IL-10 secretion [17,34]. Recent reports have indicated that DC-based immunotherapy was effective in patients with melanoma, prostate cancer and renal cell carcinoma [36–38]. In addition, HBsAg-pulsed DC could effectively induce an immune response in healthy volunteers or in immunosuppressed HBV carriers [11,20]. In this study, our findings show that MoDC pulsed with HBV antigens could restore their capacity to stimulate HBV epitope-specific T cell responses of HCC patients in vitro. Therefore, loading DC with HBV antigens may serve as a novel approach to improve defective cellular immune response in HBV-associated HCC patients [24–26]. In our study, the development of HCC has been associated with chronic HBV infection for more than 20 years. HBsAg often expresses on the HCC cell surface and may serve as a tumour-associated antigen. Adenoviral vector-transduced HBsAg DC have been shown to stimulate strong CTL responses to HBsAg-expressing HCC, and protect mice from lethal tumour challenge [27]. It has been shown recently that the adoptive transfer of immunity to HBsAg can suppress human HBsAg+ HCC xenografts effectively in athymic mice [28], and that immune modulation of anti-HBV immunity can induce an anti-tumour immune response in an animal model of HCC [29]. This finding suggests that HBV antigen-pulsed DC is a DC vaccine against HCC as well as HBV.

In general, optimal antigen is a key factor for preparation of DC vaccine. Efficient DC vaccine used for induction of anti-tumour immunity has been limited by several factors, including the identification of appropriate tumour-associated antigens, delivery of antigens to DC and maintaining DC in a highly activated state. Our study found that MoDC stimulated by HBcAg alone and combined with HBsAg had higher levels of expression of co-stimulatory molecules and cytokine production, including IL-12 and IL-10, in comparison to no-antigen or HBsAg alone, and the optimal dose of HBcAg or HBsAg was 20 µg/ml for loading MoDC, indicating a certain potential recovery of HBV antigen-loaded DC function. It is well known that IL-12 and IL-10 secreted by DC are the key cytokines inducing Th1 or Th2 cell responses to various viral or bacterial products. IL-12 induces IFN-γ secretion by T cells driving Th1 differentiation, and IL-10 driving Th2 differentiation. The balance of IL-12 and IL-10 modulate anti-viral responses in HBV infection, favouring HBV suppression and protection of severe liver damage [39,40]. However, it seemed that matured MoDC had a much stronger capability to secrete IL-12 than IL-10 in this study, suggesting that HBV antigen-pulsed MoDC may induce Th1 immune responses preferentially, which has been demonstrated to be of importance for virus clearance in patients with HBV infection as well as in animal models [41,42].

More importantly, our data show that MoDC pulsed with HBV antigen could significantly increase the T cell stimulatory capacity and induce higher numbers of IFN-γ-producing T cells and virus epitope-specific CTL frequency compared to no-antigen-pulsed DC. The improvement in T cell response capacity may be due to the increased up-regulation of co-stimulatory molecules CD80 and CD86 after in vitro maturation of MoDC from chronic HBV-infected patients with primary HCC, and the sequential order for enhancement of costimulator molecules is as no-antigen, HBsAg, HBcAg and HBsAg combined with HBcAg. These results show that impaired function of DC in HBV infection may be reversed efficiently by pulsing with HBV antigen in vitro, in particular for HBcAg alone or combined with HBsAg, which would favour clinical trials of HBV antigen-pulsed MoDC.

In addition, we also found that MoDC from patients with higher levels of serum ALT had a stronger capacity of inducing HBV-specific IFN-γ-producing T cells than patients with lower ATL levels with more than 105 copies/ml high plasma HBV DNA, which was similar to one previous report [43] but different from another report [44]. The discrepancy might be due to differences in the pattern of stimulation or differences in the origin of the lymphocytes. In our study, we used HBcAg alone or in combination with HBsAg-pulsed DC to stimulate autologous T cells from PBMCs rather than using HBsAg antigen alone. Our study also revealed that the capacity of MoDC to produce IL-12 was correlated with peripheral blood ALT levels in these cohort patients. These results may provide guidance in selecting patients for autologous MoDC vaccines for clinical application.

HCC has a poor prognosis for the majority of patients. Increasing evidence shows that DC-based immunotherapy might be effective in patients with cancer [33–35]. Our findings show that loading with HBV antigen can not only restore functional impairment of MoDC, but also induce production of HBV epitope-specific T cells. Thus, our findings imply that it is a potential novel therapeutic strategy for HBV-associated HCC patients. Further studies will examine whether this is an efficient approach not only for the control of HBV replication, but also for treatment and prevention of the development of primary HCC in patients with chronic HBV infection.

Acknowledgements

This work was supported by grants from the National Key Basic Research Program of China (2006CB504305) and the National Natural Science Foundation of China (30228025, 30571738). The authors thank to Dr Dongping Xu in our institute for his valuable suggestions in the generation and functional analysis of MoDC pulsed with HBV antigens.

Ancillary