Enhanced T cell responses against hepatitis C virus by ex vivo targeting of adenoviral particles to dendritic cells

Authors

  • Itziar Echeverria,

    1. Division of Hepatology and Gene Therapy, Center for Applied Medical Research, University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • Alexander Pereboev,

    1. Division of Human Gene Therapy, Department of Medicine, University of Alabama at Birmingham, Birmingham, AL
    Search for more papers by this author
  • Leyre Silva,

    1. Division of Hepatology and Gene Therapy, Center for Applied Medical Research, University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • Aintzane Zabaleta,

    1. Division of Hepatology and Gene Therapy, Center for Applied Medical Research, University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • José Ignacio Riezu-Boj,

    1. Division of Hepatology and Gene Therapy, Center for Applied Medical Research, University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • Marta Bes,

    1. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBERehd), Vall d'Hebron University Hospital, Barcelona, Spain
    Search for more papers by this author
  • María Cubero,

    1. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBERehd), Vall d'Hebron University Hospital, Barcelona, Spain
    Search for more papers by this author
  • Francisco Borras-Cuesta,

    1. Division of Hepatology and Gene Therapy, Center for Applied Medical Research, University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • Juan José Lasarte,

    1. Division of Hepatology and Gene Therapy, Center for Applied Medical Research, University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • Juan Ignacio Esteban,

    1. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBERehd), Vall d'Hebron University Hospital, Barcelona, Spain
    Search for more papers by this author
  • Jesús Prieto,

    1. Division of Hepatology and Gene Therapy, Center for Applied Medical Research, University of Navarra, Pamplona, Spain
    2. CIBERehd, University Clinic, University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • Pablo Sarobe

    Corresponding author
    1. Division of Hepatology and Gene Therapy, Center for Applied Medical Research, University of Navarra, Pamplona, Spain
    • University of Navarra, Center for Applied Medical Research, Division of Hepatology and Gene Therapy, Pío XII 55, 31008 Pamplona, Spain
    Search for more papers by this author
    • fax: (34)-948-194717


  • Potential conflict of interest: Alexander Pereboev received funding from Digna Biotech, which holds a patent on adapter molecules.

Abstract

Injection of dendritic cells (DCs) presenting viral proteins constitutes a promising approach to stimulate T cell immunity against hepatitis C virus (HCV). Here we describe a strategy to enhance antigen loading and immunostimulatory functions of DCs useful in the preparation of therapeutic vaccines. Incubation of murine DCs with CFm40L, an adapter molecule containing the coxsackie-adenovirus receptor fused to the ecto-domain of murine CD40L-induced DC maturation, produced high amounts of interleukin-12 and up-regulation of molecules associated with T helper 1 responses. Accordingly, targeting of an adenovirus encoding HCV NS3 protein (AdNS3) to DCs with CFm40L strongly enhanced NS3 presentation in vitro, activating interferon-γ–producing T cells. Moreover, immunization of mice with these DCs promoted strong CD4 and CD8 T cell responses against HCV NS3. CFh40L, a similar adapter molecule containing human CD40L, enhanced transduction and maturation of human monocyte-derived DCs. Comparison of DCs transduced with AdNS3 and CFh40L from patients with chronic HCV infection and healthy donors revealed similar maturation levels. More importantly, DCs from the patients induced NS3-specific responses when transduced with AdNS3 and CFh40L but not with AdNS3 alone. Conclusion: DCs transduced with AdNS3 and the adapter molecule CFm/h40L exhibit enhanced immunostimulatory functions, induce robust anti-HCV NS3 immunity in animals, and can induce antiviral immune responses in subjects with chronic HCV infection. This strategy may serve as therapeutic vaccination for patients with chronic hepatitis C. (HEPATOLOGY 2011;)

The antiviral T cell immunity found in patients with chronic hepatitis C virus (HCV) infection is characterized by its low magnitude and the narrow repertoire of epitopes recognized by T cells. By contrast, individuals who clear the virus have strong and multispecific T cell responses against viral antigens.1 These results suggest that T cell immunity may be responsible for viral clearance and support the notion that induction of these responses may be advantageous for the treatment of chronic hepatitis C. Strategies combining HCV antigens in different formats (e.g., peptides, proteins, DNA, bacterial or viral vectors) with adjuvants have been used to induce T cell responses against HCV.2 Some of these strategies have reached clinical trials, and although they induce anti-HCV T cell responses, they have a low impact on viral load.3-5 Thus, more potent strategies are required to induce efficient anti-HCV responses. During recent years, there has been growing interest in using dendritic cells (DCs) as immunogens.6 DCs are a heterogeneous population of professional antigen-presenting cells, which capture antigens that are processed and presented to T lymphocytes, linking innate and adaptive immunity.7 To accomplish this function, DCs undergo an activation process induced either by pathogen-associated molecular patterns8 or by molecules expressed or released during inflammation (e.g., CD40L, cytokines).9 These steps involve up-regulation of MHC, adhesion, and costimulatory molecules, as well as production of chemokines and cytokines, which attract and activate T cells. Ex vivo–prepared DCs have been used to induce immune responses against viral or tumor antigens in clinical trials.10-12 Recently, a phase I clinical trial in patients with chronic HCV infection using autologous DCs loaded with lipopeptides has been reported.13 Treatment was safe and induced anti-HCV CD8 T cell responses, although it did not modify viral load. Thus, more potent strategies to load DCs with antigens and induce their maturation need to be developed. In a previous study, we demonstrated that immunization with murine DCs transduced with a recombinant adenovirus encoding HCV NS3 protein (AdNS3) induced CD4 and CD8 T cell responses against NS3.14 Although these DCs showed some degree of maturation due to adenoviral infection, they did not reach the maturation level attained by typical stimuli commonly used, such as Toll-like receptor ligands, cytokine mixtures, or CD40L. Moreover, DCs, compared with other cell types, are not easily transduced by adenoviruses, requiring high multiplicities of infection to obtain expression of antigen in most cells.15 Several strategies have been developed to target adenoviruses to DCs and improve their transduction efficiency.16-18 We have reported that CFm40L, an adapter molecule combining the coxsackie-adenovirus receptor fused to the ecto-domain of CD40L by way of a trimerization motif, was able to efficiently target adenoviruses to DCs.19 Moreover, direct immunization with adenoviral particles coated with this adapter molecule was able to induce stronger immune responses than uncoated adenoviral particles.19, 20

Because we14 and others have demonstrated that DC-based immunization strategies are more potent than direct adenoviral immunization, we studied the ability of CFm40L to increase the immunogenicity of DCs transduced ex vivo with AdNS3. We show that the use of the adapter promoted a T helper 1 (Th1)-inducing activation status in DCs, enhanced their antigen-presenting capabilities in vitro, and strengthened their ability to elicit T cell responses against HCV NS3 in vivo. Moreover, we show that human monocyte-derived DCs (MoDCs) from HCV patients are efficiently transduced using this strategy and acquire robust immunostimulatory properties enabling them to trigger anti-HCV NS3 immunity.

Abbreviations

AdGFP, adenovirus encoding enhanced green fluorescent protein; AdNS3, adenovirus encoding NS3; CTL, cytotoxic T lymphocyte; DC, dendritic cells; ELISA, enzyme-linked immunosorbent assay; ELISPOT, enzyme-linked immunosorbent spot; FITC, fluorescein isothiocyanate; HCV, hepatitis C virus; IFN, interferon; IL, interleukin; MLR, mixed lymphocyte reaction; MoDC, monocyte-derived dendritic cells; PCR, polymerase chain reaction; rNS3, recombinant NS3; RPMI-1640, Roswell Park Memorial Institute 1640 medium; RT-PCR, real-time PCR; Th1, T helper 1.

Materials and Methods

Reagents, Antigens, and Recombinant Adenovirus.

Adapter molecule CFm40L containing the mouse CD40L ecto-domain has been described.19 CFh40L, the adapter protein having the ecto-domain of human CD40L (amino acids 118-261 of GenBank #P29965) in place of mouse CD40L, was constructed similarly using plasmid pBS-hCD40L-6A9 (American Type Culture Collection # 9812) as a polymerase chain reaction (PCR) template. Both molecules were expressed and purified from supernatants of transfected HEK 293 cells as described.19 Synthetic peptides containing T cell epitopes from HCV NS3 protein were synthesized manually.14 Recombinant NS3 (rNS3) was purchased from Mikrogen (Martinsried, Germany). AdNS3 has been described.21 Adenovirus encoding enhanced green fluorescent protein (AdGFP) was a gift from R. Hernandez-Alcoceba (CIMA, Pamplona, Spain).

Mice.

HHD mice, transgenic for human HLA-A2.1 and beta-2 microglobulin molecules22 (a gift from F. Lemonnier, Institute Pasteur, Paris, France), and C57BL/6 and BALB/c mice (Harlan, Barcelona, Spain) were maintained in pathogen-free conditions and treated according to guidelines of our Institutional Review Board.

Patients.

Thirty-five patients with chronic HCV infection without cirrhosis (anti-HCV+, HCV RNA+ and with raised aminotransferase levels for more than 6 months) were included in this study (Supporting Table S1). Patients who underwent interferon (IFN)-α plus ribavirin therapy were studied at least 6 months after treatment. Twenty-nine healthy, sex- and age-matched individuals seronegative for HCV were also enrolled. Informed consent was obtained from individuals included in the study, which conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by our Institutional Review Board.

DC Generation and Transduction with Adenovirus.

Mouse bone marrow-derived DCs were grown using granulocyte-macrophage colony-stimulating factor and interleukin (IL)-4 as described.14 Human DCs were differentiated from monocytes obtained using anti-CD14–coated beads as described.23 In both cases, DCs were transduced with adenoviruses by incubating 107 cells/mL in Roswell Park Memorial Institute 1640 medium (RPMI-1640) with adenoviral particles at a multiplicity of infection of 30. When using adapter molecules, 6 μg of CFm40L or CFh40L were incubated at 37°C for 30 minutes with adenovirus in 50 μL of phosphate-buffered saline and then added to 106 DCs for further incubation. After 1 hour, complete medium (RPMI 1640 containing 2 mM L-glutamine, 0.05 mM 2-mercaptoethanol, 10% fetal bovine serum, and 1% antibiotics) containing cytokines was added. IFN-γ (1,000 U/mL; R&D Systems, Minneapolis, MN) was also added to culture medium of human DCs. Cells were harvested 24 hours later, extensively washed to discard any carry-over of adenoviral particles, and used for analysis by real-time PCR (RT-PCR), flow cytometry, in vitro stimulation experiments, or immunization.

Flow Cytometry.

Human DCs were stained with fluorescein isothiocyanate (FITC)-labeled anti-CD80, -CD86, and –HLA-DR antibodies and phycoerythrin-labeled anti-CD54 antibodies and their corresponding control isotypes (all from BD Biosciences, San Diego, CA). Murine T regulatory cells were enumerated using a kit (eBioscience; San Diego, CA) containing anti–CD4-FITC, anti–CD25-APC, and anti–FoxP3-phycoerythrin-labeled antibodies according to the manufacturer's instructions. In all cases, expression of the different molecules was analyzed using a FACScan flow cytometer (BD Biosciences).

Mixed Leukocyte Reaction.

Graded numbers of murine and human DCs were cultured in 96-well U bottomed plates with 105 nonadherent spleen cells obtained from BALB/c mice or with human allogeneic lymphocytes, respectively. Two or 4 days later, 0.5 μCi of tritiated thymidine were added to each well and cultured overnight to measure cell proliferation.

Immunization and Measurement of T Cell Responses.

C57BL/6 or HHD mice (n = 3-5) were immunized subcutaneously with 2.5 × 105 DCs. One to 6 weeks later, animals were sacrificed and spleens were removed, homogenized, and pooled. Immune responses were measured in vitro using a murine IFN-γ enzyme-linked immunosorbent spot (ELISPOT) assay (BD Biosciences) as described.14 Briefly, splenocytes (4 × 105 cells/well) were cultured in triplicate in antibody-coated plates in the absence or presence of peptides (10 μM) or rNS3 (1 μg/mL). The next day, plates were incubated with biotinlylated anti–IFN-γ antibody and developed following the manufacturer's instructions. Spots were counted using an automated ELISPOT reader (CTL, Aalen, Germany). In vivo evaluation of immune responses was performed using in vivo killing assays. Briefly, 1 week after immunization, splenocytes from naïve HHD mice were pulsed with or without peptide 1073-1081 (10 μg/mL), labeled with different concentrations of carboxyfluoroscein succinimidyl ester (Molecular Probes), and injected intravenously into syngeneic immunized mice. Sixteen hours later, cell killing was measured by way of flow cytometry as described.24

Activation of T Cells by Adenovirus-Transduced DCs.

Murine splenocytes obtained from HHD mice immunized with NS325 were stimulated weekly with the HLA-A2–restricted epitope 1073-1081 and expanded with IL-2 to grow a T cell line. Expanded T cells (103 cells/well) were stimulated with 5 × 103 untreated DCs or DCs treated with different combinations of adenovirus and CFm40L, and T cell activation was measured by way of IFN-γ ELISPOT assay. For human cells, DCs (104 cells/well) were used to stimulate autologous CD14 cells (105 cells/well). The next day, IFN-γ–producing cells were detected and enumerated using a human IFN-γ ELISPOT kit (BD Biosciences) according to the manufacturer's instructions. In some cases, autologous CD14 cells (4 × 106 cells/well) were stimulated with DC-AdNS3+ CFh40L (2 × 105 cells/well) in 24-well plates. Three days later, IL-2 (20 U/mL) was added to expand cells until day 10. Next, T cells (2 × 104 cells/well) were stimulated with 5 × 103 autologous DCs with or without three peptide pools14 (10 μg/mL for each peptide [Supporting Table S3]) encompassing the whole NS3 sequence. T cell activation was measured as IFN-γ production by way of enzyme-linked immunosorbent assay (ELISA) or flow cytometry, analyzing CD69 up-regulation on peptide-stimulated CD4 or CD8 cells (all antibodies were from BD Biosciences).

Measurement of Cytokines from Culture Supernatants.

Murine IL-12, IL-10, IFN-γ, and IL-4 and human IL-12 and IFN-γ were measured in supernatants by way of ELISA (BD Biosciences) according to the manufacturer's instructions.

Statistical Analysis.

Results are presented as the mean ± SEM and were analyzed using a nonparametric Mann-Whitney U test. Paired samples were analyzed using a paired two-tailed Student t test. P < 0.05 was considered statistically significant.

Results

Transduction of DCs with AdNS3 Using the Adapter Molecule CFm40L Enhances Antigen Presentation and Activation of Th1 Responses.

We have shown that in vitro transduction of DCs with an adenovirus employing the adapter molecule CFm40L enhanced DC transduction, phenotypic maturation, and IL-12 production. We confirmed these results using AdNS3 (Supporting Fig. 1) and also found that these cells did not produce IL-10. Because our aim was to activate T cell responses, preferably those with a Th1 profile, we analyzed other molecules associated with T cell activation by way of RT-PCR. Costimulatory molecules belonging to the tumor necrosis factor family 4-1BBL, OX40L, and CD70 known to enhance T cell responses26 showed increased values in DC-AdNS3+CFm40L compared with DC-AdNS3, as occurred with the Th1-associated molecule Notch ligand Delta-like 4.27 By contrast, a weak expression of Th2-associated Notch ligands Jagged 1 and 2 was found in DC-AdNS3+CFm40L. An increase in CCR7 chemokine receptor was also observed in DC-AdNS3+CFm40L (Supporting Fig. 1). DC maturation as well as targeting of adenovirus by CFm40L was dependent on CD40–CD40L interactions (Supporting Fig. 2).

T cell stimulatory ability of DCs transduced with AdNS3 with or without CFm40L was initially assessed in vitro in mixed lymphocyte reactions (MLRs). DCs treated with CFm40L with or without AdNS3 induced higher proliferation (P = 0.002) (Fig. 1A) and IFN-γ production (P = 0.02) (Fig. 1B) by allogeneic lymphocytes than DC-AdNS3 or untreated DCs. However, they did not induce IL-4 (Fig. 1C), in accordance with a Th1 profile associated with high IL-12 production and Delta-like 4 up-regulation. Next, to analyze the effect of CFm40L on NS3-specific responses, DCs from HHD mice (transgenic for HLA-A2) were transduced with different combinations of AdNS3, control AdGFP, and CFm40L and were used to stimulate a murine CD8 T cell line recognizing the HLA-A2–restricted cytotoxic T lymphocyte (CTL) epitope 1073-1081 from NS3. We found that DC-AdNS3+CFm40L activated a higher number of IFN-γ–producing cells (P < 0.05) compared with control groups (Fig. 1D).

Figure 1.

Transduction of DCs with recombinant adenovirus AdNS3 using the adapter molecule CFm40L enhances antigen presentation and activation of Th1 responses in vitro. (A) Bone marrow-derived DCs from C57BL/6 mice were cultured in 96-well plates (n = 6) and treated with different combinations of AdNS3 and CFm40L. One day later, 3 × 103 DCs were used to stimulate 105 lymphocytes from BALB/c mice in MLR assays, and cell proliferation was measured after 3 days. (B) IFN-γ and (C) IL-4 content in 48-hour supernatants from MLR assays was measured by way of ELISA. (D) DCs from HHD mice treated with different combinations of AdNS3, AdGFP, and CFm40L (n = 4 wells/group) were used to stimulate a CD8 murine T cell line specific for the HLA-A2–restricted HCV NS3 peptide 1073-1081. T cell activation was measured using an IFN-γ ELISPOT assay. The results are representative of two to three different experiments.

DCs Transduced with AdNS3 Using CFm40L Induce Robust CD4 and CD8 T Cell Responses In Vivo.

In vivo experiments in C57BL/6 mice showed that, 1 week after immunization, a higher number of IFN-γ spot-forming cells were observed when using DC-AdNS3+CFm40L compared with DC-AdNS3 (Fig. 2A). This was manifest not only when evaluating CD8 T cell responses against peptides 1367, 1427, and 1447 (P = 0.02, P = 0.04, and P = 0.01, respectively), reported to be recognized by CD8 T cells,14 but also for CD4 T cell responses (P = 0.02), measured using rNS3 as an antigen. Equivalent results were obtained with HHD mice transgenic for human HLA-A2 molecules when responses were measured against HLA-A2–restricted CTL epitopes 1073-1081 and 1038-1047 (P = 0.005) (Fig. 2B). RT-PCR analysis of splenocytes showed that those from DC-AdNS3+CFm40L–immunized mice also produced enhanced amounts of IL-2 and tumor necrosis factor α, but not IL-4, confirming in vivo the Th1 profile (Supporting Fig. 3). Furthermore, in vivo killing assays of cells presenting CTL epitope 1073-1081 showed a higher killing activity (P = 0.01) in HHD mice immunized with DC-AdNS3+CFm40L, as opposed to those immunized with DC-AdNS3 (Fig. 2C). In long-term experiments, it was found that DC-AdNS3+CFm40L again induced stronger responses, which were still detectable at week 6 (Fig. 3A). This higher immunogenicity of DC-AdNS3+CFm40L was associated with their elevated migratory capacity, with a higher number of cells reaching draining lymph nodes 2 days after injection and a lower number of cells remaining at the injection site (Supporting Fig. 4). Finally, because stimulation through CD40 may induce tolerogenic DCs,28 IL-10 production by antigen-stimulated lymphocytes and T regulatory numbers were studied. No differences in short- and long-term experiments were observed in any of these parameters between mice immunized with DC-AdNS3 or DC-AdNS3+CFm40L (Fig. 3B,C). Analysis of innate immunity did not reveal significant differences between DC-AdNS3 or DC-AdNS3+CFm40L–immunized mice in the proportion of total and activated splenic natural killer cells and in splenic DCs (Supporting Fig. 5) at different time points. Thus, superior Th1 responses against peptides as well as against rNS3 were induced after immunization with DC-AdNS3+ CFm40L.

Figure 2.

DCs transduced with AdNS3 using CFm40L induce strong CD4 and CD8 T cell responses in vivo. Bone marrow-derived DCs from (A) C57BL/6 or (B) HHD mice were treated with AdNS3, with or without CFm40L, and 1 day later syngeneic mice (n = 3-5) received 2.5 × 105 DCs subcutaneously. After 1 week, splenocytes were stimulated in vitro with synthetic peptides encompassing CD8 epitopes or with rNS3 protein to measure T cell activation as IFN-γ production in ELISPOT assays. (C) HHD mice were immunized as described above, and 1 week later they were injected with syngeneic splenocytes pulsed with or without peptide 1073-1081 and labeled with two different carboxyfluorescein succinimidyl ester concentrations. Sixteen hours later, killing activity was evaluated by way of flow cytometry. The results are representative of two different experiments.

Figure 3.

DCs transduced with AdNS3+CFm40L induce long-term Th1 responses. C57BL/6 mice (n = 4) were immunized with DC-AdNS3 (white bars) or DC-AdNS3+CFm40L (black bars) and sacrificed at different time points. Splenocytes were then stimulated with peptides encompassing CD8 epitopes or with rNS3 protein to measure (A) IFN-γ production in ELISPOT assays or (B) IL-10 by way of ELISA. (C) The proportion of T regulatory cells in spleens of immunized mice was measured as the percentage of FoxP3+ cells in the CD4+CD25+ population. The results are representative of two independent experiments.

Activation of Human MoDCs by CFh40L.

An equivalent adapter molecule containing the human CD40L ecto-domain (CFh40L) was next developed and tested with human MoDCs. CFh40L improved MoDC transduction, as demonstrated using an AdGFP (Fig. 4A). Regarding maturation status, clear up-regulation of surface markers was observed after transduction of MoDCs with AdNS3 and CFh40L compared with cells transduced with AdNS3 alone (Fig. 4B). Analysis of cytokines showed that CFh40L increased IL-12 production by MoDCs treated with AdNS3, which was further enhanced by the addition of IFN-γ (Fig. 4C). Finally, the T cell stimulatory ability of MoDCs, as estimated in MLR assays, was higher for DC-AdNS3+CFh40L (P = 0.005) (Fig. 4D).

Figure 4.

CFh40L enhances adenoviral transduction of MoDCs and induces their maturation. (A) MoDCs from healthy individuals were treated with AdGFP with or without CFh40L. One day later, GFP expression was measured by way of flow cytometry as mean fluorescence index (MFI) or percentage of GFP+ cells. (B) MoDCs differentiated as in panel A were treated with AdNS3 or AdNS3+CFh40L or left untreated. One day later, they were stained with anti-CD80, -CD86, –HLA-DR and -CD54 antibodies (white histograms) or isotype controls (gray histograms) and analyzed by way of flow cytometry. Numbers indicate values of MFI. (C) IL-12 content in 24-hour culture supernatants was measured by way of ELISA. (D) Graded numbers of MoDCs treated as in panel A were used to stimulate allogeneic lymphocytes, and cell proliferation was measured 5 days later. The results are representative of three individuals.

MoDCs from Patients Chronically Infected with HCV Transduced with AdNS3 and CFh40L Are Functional and Stimulate Autologous T Cell Responses Against NS3.

In order to see the potential application of this strategy in chronic HCV infection, we next studied the effect of CFh40L in MoDCs from HCV patients and healthy seronegative controls. Baseline analyses performed in MoDCs without AdNS3 or CFh40L stimulation did not show any phenotypic difference between groups. However, MoDCs from HCV patients had a lower allostimulatory capacity (Supporting Fig. 6). After treatment with AdNS3+CFh40L, although there were heterogeneous responses in both groups, surface markers showed similar levels in patients with chronic HCV infection and in controls (CD54, P = 0.06; CD86, P = 0.85; HLA-DR, P = 0.95) (Fig. 5A). Moreover, the amount of IL-12 produced by MoDCs from both groups was equivalent (P = 0.16) (Fig. 5B). Importantly, the ability to stimulate allogeneic T cells was similar for MoDCs from patients and controls (P = 0.31) (Fig. 5C).

Figure 5.

MoDCs from patients chronically infected with HCV respond to CFh40L. MoDCs from healthy seronegative controls (SC) or from patients chronically infected with HCV (HCV) were treated with AdNS3, CFh40L, and IFN-γ. (A) Expression of maturation-associated surface markers was measured by way of flow cytometry. (B) IL-12 production in 1-day culture supernatants was analyzed by way of ELISA. (C) DCs (3 × 103/well) were used to stimulate 105 allogeneic lymphocytes in MLR assays, and T cell proliferation was measured 5 days later.

To analyze whether this strategy would induce T cell responses in HCV patients, T cells from patients and seronegative controls were stimulated for 24 hours with autologous DCs transduced with AdNS3 with or without CFh40L. As shown in an example of a patient (Fig. 6A), clear responses were induced by DC-AdNS3+CFh40L, as opposed to DC-AdNS3 or DC+CFh40L. This occurred not only in patients (P = 0.01) but also in healthy seronegative individuals (Fig. 6B), although in this case it did not reach statistical significance (P = 0.08). Experiments using the control adenovirus AdGFP showed similar enhancement in patients and controls (data not shown), suggesting that these responses probably recognized adenoviral antigens. To study specific anti-NS3 responses, T cells from an HCV patient were stimulated with DC-AdNS3 or DC-AdNS3+CFh40L and expanded with IL-2, and 10 days later, they were tested against three peptide pools encompassing the entire NS3 sequence. Specific production of IFN-γ was detected in cultures primed by DC-AdNS3+ CFh40L, but not in those primed by DC-AdNS3 (Fig. 6C). Using this stimulation protocol based on DC-AdNS3+CFh40L, T cell responses of different magnitude and specificity were detected against different peptide pools in two out of four additional patients (HCV#2 and HCV#3) (Fig. 6D). Interestingly, anti-NS3 immune responses were also observed in one out of three healthy controls (SC#1) similarly stimulated (Fig. 6E). No responses against NS3 peptide pools were observed after 1-day stimulation without previous priming with DC-AdNS3+ CFh40L in healthy controls, as opposed to responses against adenoviral antigens, observed both after short and long stimulation protocols (Supporting Fig. 7). Moreover, the lack of correlation in response against HLA-A2–restricted 1073 epitope (a peptide cross-reacting with influenza) and pool P1 (which contains a peptide encompassing 1073 epitope) in these individuals suggests in vitro priming of T cells with DC-AdNS3+CFh40L, and not cross-recognition by influenza-specific T cells. Finally, analysis of CD69 up-regulation revealed that CD4 and CD8 T cells were activated by peptides pools, both in HCV patients and seronegative controls (Supporting Fig. 8). These results suggest that DC-AdNS3+CFh40L have the ability to activate anti-HCV responses both in seronegative individuals and in HCV-infected patients.

Figure 6.

MoDCs from patients chronically infected with HCV transduced with AdNS3 and CFh40L activate autologous T cell responses against NS3. (A) MoDCs from a representative HCV patient treated with AdNS3, CFh40L, or AdNS3+CFh40L or left untreated were used to stimulate autologous T cells (n = 4 wells/group). One day later, IFN-γ spot-forming cells were quantified in ELISPOT assays. (B) Comparison of T cell responses induced by DCs treated with AdNS3 or AdNS3 + CFh40L in HCV patients (HCV) and seronegative controls (SC). (C-E) MoDCs treated with AdNS3 (C) or AdNS3 + CFh40L (C-E) were used to stimulate autologous CD14 cells from HCV patients (C,D) or seronegative controls (E). After 10 days of expansion with IL-2, NS3-specific responses were detected by measuring IFN-γ production after stimulation with three NS3 peptide pools (P1, P2, and P3).

Discussion

The low rate of response to current therapy in chronic hepatitis C,29 together with its adverse effects, long duration, and cost, makes necessary the development of novel therapies. Because strong anti-HCV T cell immune responses are associated with HCV clearance, only those methods capable of eliciting a robust antiviral T cell immunity could facilitate the resolution of chronic HCV infection. In the present study, we tested a new strategy to enhance immunogenicity of DCs, professional antigen-presenting cells. Due to the low ability that adenoviral vectors have to transduce DCs and induce their maturation,15 we used CFm40L (an adapter molecule containing the coxsackie-adenovirus receptor and CD40L previously designed to target DCs in vivo19) to improve the functional properties of DCs ex vivo and to increase their immunogenicity in DC-based immunization protocols using AdNS3. We found that DC-AdNS3+CFm40L are more mature and express cytokines and costimulatory molecules associated with Th1 responses. Moreover, these adapter molecules enhance transduction efficiency (as shown in experiments with AdGFP) and antigen-loading and presentation, as demonstrated using an NS3-specific T cell line, as well as their migratory capacity to lymphoid organs. All these changes ultimately result in the induction of robust Th1 responses. Most importantly, DC-AdNS3+CFm40L were more immunogenic in vivo and elicited stronger CD4 and CD8 T cell responses than DCs transduced with AdNS3 alone. Thus, as represented in the scheme in Supporting Fig. 9, the compound effect of antigen targeting of DCs together with the immunostimulatory properties of CFm40L increased the ability of DCs to act as efficient immunogens.

During recent years, considerable effort has been dedicated to developing new adapter molecules to target adenovirus to DCs directly in vivo,16-19 avoiding ex vivo loading of DCs, which is more laborious and time-consuming.30 However, previous work from our laboratory shows that DCs transduced in vitro with AdNS3 induced stronger responses compared with direct administration of adenovirus.14 On the other hand, the use of DCs loaded in vitro with antigen allows the control of the maturation status of the cells before their administration and ensures exclusive expression of the antigen by DCs. The alternative strategy relying on in vivo targeting of the adapter-coated adenovirus entails the disadvantage of possible binding to different cell types sharing target receptors with DCs. Indeed, CD40 is expressed not only by DCs, but also by monocytes, B lymphocytes, endothelial cells, and epithelial cells.31 Thus, it seems likely that the administration of ex vivo–manipulated DCs may stimulate T cell immune responses more efficiently than the direct injection of a targeted adenovirus.

Because immunotherapy of chronic hepatitis C was the final aim of this study, a human version of the adapter molecule, CFh40L, was also constructed. CFh40L also has a strong targeting ability and induces human MoDC maturation. Conflicting results regarding the functional status of MoDCs in chronic HCV infection have been reported.32-34 However, although baseline allostimulatory ability of MoDCs from HCV patients was lower than in controls, CFh40L, as we demonstrated with Toll-like receptor ligand polyinosinic:polycytidylic acid,23 induced similar functional responses in both groups. Finally, our study of antigen presentation by DCs to autologous T cells in patients with chronic HCV infection and healthy controls showed that, when using CFh40L, NS3-specific T cell responses could be detected in some HCV patients and controls. Thus, at least in a subset of patients, full DC maturation and antigen presentation induced by CFh40L is able to activate anti-NS3 T cell responses. Whether latent or newly induced T cells are detected in patients is not known, because they are also found in controls; however, our results suggest that these patients have functional T cells, which are only activated when stimulated with DCs treated with CFh40L. This subset of patients might be considered for a therapeutic vaccination trial using the strategy described in this article.

In conclusion, fusion molecules coxsackie-adenovirus receptor–fibritin–CD40L can target the antigen expressed by adenoviral vector to DCs, which become intensely activated and highly immunogenic. This strategy is capable of activating silent T cell responses in a subset of patients with chronic hepatitis C, suggesting that it can be a valuable tool in the preparation of therapeutic vaccines.

Acknowledgements

We thank R. Hernandez-Alcoceba, F. Lemonnier, and I. Melero for providing AdGFP, HHD mice, and EGFP mice, respectively.

Ancillary