Nanoparticles encapsulating hepatitis B virus cytosine-phosphate-guanosine induce therapeutic immunity against HBV infection

Authors

  • Shujuan Lv,

    1. Institute of Immunology, Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei, China
    2. Department of Microbiology, Anhui Medical University, Hefei, China
    Search for more papers by this author
  • Jun Wang,

    Corresponding author
    1. School of Life Sciences and Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei, China
    • Address reprint requests to: Haiming Wei, M.D., Zhigang Tian, Ph.D., or Jun Wang, Ph.D., School of Life Sciences, University of Science and Technology of China, 443 Huangshan Road, Hefei City, 230027, Anhui, China. E-mail: ustcwhm@ustc.edu.cn, tzg@ustc.edu.cn, or jwang699@ustc.edu.cn; Fax: +86-551-6360-6783.

    Search for more papers by this author
  • Shuang Dou,

    1. School of Life Sciences and Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei, China
    Search for more papers by this author
  • Xianzhu Yang,

    1. School of Life Sciences and Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei, China
    Search for more papers by this author
  • Xiang Ni,

    1. Institute of Immunology, Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei, China
    Search for more papers by this author
  • Rui Sun,

    1. Institute of Immunology, Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei, China
    Search for more papers by this author
  • Zhigang Tian,

    Corresponding author
    1. Institute of Immunology, Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei, China
    • Address reprint requests to: Haiming Wei, M.D., Zhigang Tian, Ph.D., or Jun Wang, Ph.D., School of Life Sciences, University of Science and Technology of China, 443 Huangshan Road, Hefei City, 230027, Anhui, China. E-mail: ustcwhm@ustc.edu.cn, tzg@ustc.edu.cn, or jwang699@ustc.edu.cn; Fax: +86-551-6360-6783.

    Search for more papers by this author
  • Haiming Wei

    Corresponding author
    1. Institute of Immunology, Hefei National Laboratory for Physical Sciences at Microscale and School of Life Sciences, University of Science and Technology of China, Hefei, China
    • Address reprint requests to: Haiming Wei, M.D., Zhigang Tian, Ph.D., or Jun Wang, Ph.D., School of Life Sciences, University of Science and Technology of China, 443 Huangshan Road, Hefei City, 230027, Anhui, China. E-mail: ustcwhm@ustc.edu.cn, tzg@ustc.edu.cn, or jwang699@ustc.edu.cn; Fax: +86-551-6360-6783.

    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

  • Supported by grants from Ministry of Science & Technology of China (973 Basic Science Project 2009CB522403, 2012CB519004) and the Natural Science Foundation of China (#81330071, #30730084).

Abstract

Infection with hepatitis B virus (HBV) is the most common cause of liver disease worldwide. However, because the current interferon (IFN)-based treatments have toxic side effects and marginal efficacy, improved antivirals are essential. Here we report that unmethylated cytosine-phosphate-guanosine oligodeoxynucleotides (CpG ODNs) from the HBV genome (HBV-CpG) induced robust expression of IFN-α by plasmacytoid dendritic cells (pDCs) in a Toll-like receptor 9 (TLR9)-dependent manner. We also identified inhibitory guanosine-rich ODNs in the HBV genome (HBV-ODN) that are capable of inhibiting HBV-CpG-induced IFN-α production. Furthermore, nanoparticles containing HBV-CpG, termed NP(HBV-CpG), reversed the HBV-ODN-mediated suppression of IFN-α production and also exerted a strong immunostimulatory effect on lymphocytes. Our results suggest that NP(HBV-CpG) can enhance the immune response to hepatitis B surface antigen (HBsAg) and skew this response toward the Th1 pathway in mice immunized with rHBsAg and NP(HBV-CpG). Moreover, NP(HBV-CpG)-based therapy led to the efficient clearance of HBV and induced an anti-HBsAg response in HBV carrier mice. Conclusion: Endogenous HBV-CpG ODNs from the HBV genome induce IFN-α production so that nanoparticle-encapsulated HBV-CpG may act as an HBsAg vaccine adjuvant and may also represent a potent therapeutic agent for the treatment of chronic HBV infection. (Hepatology 2014;59:385–394)

Abbreviations
cccDNA

covalently closed circular DNA

CpG ODN

cytosine-phosphate-guanosine oligodeoxynucleotide

HBcAg

hepatitis B c antigen

HBsAg

hepatitis B surface antigen

HBV

hepatitis B virus

IFN

interferon

PBMC

peripheral blood mononuclear cell

pDC

plasmacytoid dendritic cell

NP

nanoparticle

PCR

polymerase chain reaction

PEG-PLA

polyethylene glycol-polylactic acid

TLR

Toll-like receptor

Persistent infection with the hepatitis B virus (HBV) affects more than 360 million people worldwide and has become a severe public health problem owing to the increased risk of liver cirrhosis and hepatocellular carcinoma in infected individuals. The current recombinant hepatitis B surface antigen (rHBsAg) vaccine provides protection against HBV infection but fails to protect ∼10% of those who are vaccinated and is also ineffective for individuals who are already infected with HBV. Conventional antiviral drugs used for the treatment of HBV, including lamivudine and interferon-alpha (IFN-α), suppress viral replication and reduce hepatic symptoms.[1] However, the persistence of HBV covalently closed circular DNA (cccDNA) and defective immune responses lead to treatment failure and progression to liver disease.[2] Therefore, more efficient therapeutic strategies are needed to eradicate HBV infection.

HBV seems to avoid inducing strong innate immune responses including the type I IFN response.[2] Therefore, methods of inducing vigorous immune responses against HBV may play a critical role in the clearance of HBV infection. The unmethylated cytosine-phosphate-guanosine (CpG) motifs presented in bacterial DNA can stimulate the immune system by interacting with the pattern-recognition receptor Toll-like receptor 9 (TLR9).[3] TLR9 is predominantly expressed in plasmacytoid dendritic cells (pDCs) and B cells in humans, and pDCs are known as interferon-producing cells (IPCs).[4, 5] Unmethylated CpG motifs can trigger an immune cascade that improves antigen presentation and the secretion of cytokines, especially IFN-α, which is induced at high levels.[6] Ongoing studies indicate that CpG oligodeoxynucleotides (ODNs) can provide a basis for improved vaccines and immunotherapy for cancer, allergies, and infectious diseases.[7] Several clinical trials have demonstrated that CpG ODNs can enhance vaccine responses and contribute to the prevention or treatment of HBV and HCV infections.[8-10] However, the responses to CpG ODN-based therapies were generally not sustained and was accompanied by deleterious autoimmune reactions or toxic shock.[11-13] Therefore, therapeutic application of CpG ODNs to HBV has not been achieved. Importantly, the genome of HBV is partially double-stranded DNA which contains some CpG islands. Whether the HBV-derived CpG ODNs can induce type I IFN is still unknown.

In this study we identified CpG ODNs from the HBV genome (HBV-CpG) that are capable of inducing IFN-α production. We also identified inhibitory guanosine-rich ODNs from HBV DNA (HBV-ODNs) capable of inhibiting HBV-CpG-mediated IFN-α production. Nanoparticle-encapsulated HBV-CpG, termed NP(HBV-CpG), reversed the HBV-ODN-mediated suppression of IFN-α production and activated the innate immune system. Our results demonstrate that NP(HBV-CpG) not only acts as an adjuvant for the development of vaccines against hepatitis B but also efficiently eradicates HBV infection.

Materials and Methods

HBV-CpG, HBV-ODN, and NP(HBV-CpG)

Partially or completely phosphorothioate-modified HBV-CpG, 488-labeled HBV-CpG and inhibitory HBV-ODNs were synthesized by Sengong (Shanghai, China) and purified by high-performance liquid chromatography (HPLC). The final products contained more than 95% full-length sequences and undetectable levels of endotoxin. The HBV-CpG molecules and inhibitory HBV-ODNs used in this study are listed in Supporting Tables 1, 2. Nanoparticles loaded with HBV-CpG or CpG-2216 were prepared by a double emulsion-solvent evaporation technique. For example, an aqueous solution of HBV-CpG (200 μg) in 25 μL of phosphate-buffered saline (PBS) was emulsified by a 30-second sonication in 0.5 mL of chloroform containing 1.0 mg of BHEM-Chol and 10 mg of mPEG5K-PLA10K over an ice bath. The mixture was then added to 5 mL of H2O and further emulsified by sonication for 30 seconds over an ice bath; then it was stirred for 20 minutes to allow the chloroform to evaporate. These HBV-CpG-loaded or CpG-2216-loaded nanoparticle formulations are henceforth referred to as NP(HBV-CpG) and NP(2216), respectively.

In Vitro Cell Culture and Stimulation

Peripheral blood mononuclear cells (PBMCs) were isolated from the buffy coats of healthy donors, which were obtained from the Blood Bank of Anhui Province (Hefei, China). Fifteen peripheral blood samples of patients with chronic hepatitis B were obtained from the Department of Infectious Diseases of the First Affiliated Hospital of Anhui Medical University. Ethical approval to use the PBMCs was obtained from the Ethics Committee of the University of Science and Technology of China. The PBMCs were prepared by density-gradient centrifugation on Ficoll-Hypaque (Solarbio, Beijing) and suspended in RPMI 1640 culture medium supplemented with 10% fetal bovine serum (FBS). Single-cell suspensions of mouse splenocytes were prepared by passing spleen cells through a 200G stainless steel mesh. The erythrocytes were lysed with RBC lysis buffer (BioLegend) and the splenocytes were cultured as described above. PBMCs (6 × 106/mL/well) and splenocytes (1 × 107/mL/well) were cultured alone or with HBV-CpG, inhibitory HBV-ODNs, NP(HBV-CpG), nanoparticles (NPs), control CpG-2216, or chloroquine (Sigma, St. Louis, MO) at the concentrations indicated in the text in 24-well plates. The cytokine levels in the supernatants were measured using enzyme-linked immunosorbent assay (ELISA) kits for human IFN-α, murine IFN-α, and murine IFN-β (PBL Biomedical Laboratories, Piscataway, NJ). The cells were retained for phenotypic analysis.

Flow Cytometry

Human PBMCs were prepared and stained with the following monoclonal antibodies: anti-ICOS-L (eBioscience); anti-CD40, anti-CD69, anti-CD3, anti-CD56, anti-Lineage1, anti-CD123, and anti-HLA-DR (BD Bioscience). The following antibodies were used to stain the murine splenic lymphocytes: anti-CD11c, anti-CD11b, anti-B220, anti-CD3, anti-NK1.1, anti-CD69, anti-CD40, anti-CD80, anti-CD4, anti-CD8 (BD Bioscience); anti-ICOS (eBioscience). Normal mouse or rat serum was used to block nonspecific Fc-receptor binding and isotype controls were used as negative controls in all of the experiments. The stained samples were run on a FACSCalibur flow cytometer and the data were analyzed with WinMDI and FlowJo software.

Confocal Microscopy

Gen2.2[14] was incubated with 488-labeled HBV-CpG alone or with unlabeled HBV-ODN for 90 minutes at 37°C. Cells were washed, fixed, permeabilized, and stained with the primary antibody anti-TLR9 (Santa Cruz Biotechnology) and the secondary antibody donkey antimouse 546 (Invitrogen). Images were acquired on an LSM 710 Laser Scanning microscope (Zeiss, Germany).

Immunization of Mice

BALB/c or C57BL/6 mice (male, 6-8 weeks old, SLAC Laboratory Animal) were divided into four groups with six animals in each group that received the following treatments: (1) PBS, (2) 2 μg of the rHBsAg vaccine, (3) blank nanoparticles with 2 μg of the rHBsAg vaccine, or (4) 0.5 mg of nanoparticles loaded with 10 μg of HBV-CpG combined with 2 μg of the rHBsAg vaccine. Each animal was treated subcutaneously on days 0, 7, and 14. The mice were bled by way of the lateral tail vein at 2 and 4 weeks after the last immunization and the serum was isolated to analyze the antibody response.

Treatment of HBV Carrier Mice

The pAAV/HBV1.2 vector was delivered into C57BL/6 mice using the hydrodynamic tail vein injection method.[15] Four weeks later, the mice were divided into three groups, with nine animals in each group. The animals in group 1 remained untreated and served as the negative controls. Groups 2 and 3 were continuously administered NPs or NP(HBV-CpG) (20 μg of HBV-CpG: 1 mg of NPs) by way of the intraperitoneal route for 3 days and were subsequently treated subcutaneously with one injection of 2 μg of rHBsAg plus NPs or NP(HBV-CpG) (20 μg of HBV-CpG: 1 mg of NPs). The above treatments were repeated for 3 consecutive weeks. The mice were bled by way of the lateral tail vein at regular intervals, and the serum was isolated to analyze the HBsAg and anti-HBsAg levels. The mice were euthanized 1 week after the last treatment. The levels of HBV DNA and intrahepatic HBcAg were assayed by quantitative polymerase chain reaction (PCR) and immunohistochemical staining, respectively.

Details of the intracellular detection of IFN-α, TLR9 and TLR7, ELISAs, RIAs, immunohistochemical staining, and detection of serum HBV DNA are included in the Supporting Information.

Statistical Analysis

Data were analyzed using two-tailed Student t tests. All analyses were performed using Prism software (GraphPad Software). Differences were considered significant at a P < 0.05.

Results

HBV-CpG Induced a Potent IFN-α Response by Human pDCs

Because the HBV genome contains CpG islands, we hypothesized that endogenous CpG ODNs from the HBV genome (HBV-CpG) could induce an immune response by interacting with TLR9. Therefore, an extensive screen was performed to identify the HBV-CpG islands in the HBV genome. When the immunostimulatory effects of the HBV-CpG candidates were examined by ELISA, two candidates, A1 and B3, were found to potently induce IFN-α release by PBMCs (Supporting Table 1; Fig. 1A). However, the nonresponse modified HBV-CpG, which contained different phosphorothioate modifications from the effective HBV-CpG candidates, and the sequence irrelevant HBV-CpG, which contained different CpG ODNs from the HBV genome, both failed to induce IFN-α release by PBMCs (Fig. 1B). Therefore, HBV-CpG requires specific CpG islands with specific phosphorothioate-modified guanines to induce high levels of IFN-α.

Figure 1.

HBV-CpG-induced a robust IFN-α response by human PBMCs. (A) Freshly isolated PBMCs (6 × 106/mL) were stimulated for 42 hours with 5 μg of each HBV-CpG candidate alone (A1, A2, A3, A4, A5, A6, A7, A8, B1, B2, or B3). The level of secreted IFN-α was measured by ELISA. (B-E) The IFN-α and pDC response was analyzed after PBMCs (6 × 106/mL) had been cultured with 5 μg of HBV-CpG (A1 and B3), nonresponse modified HBV-CpG, or sequence irrelevant HBV-CpG. (B) The IFN-α level in the supernatant was measured by ELISA after a 42-hour incubation. (C) Lineage1, CD123+, HLA-DR+ cells were gated and analyzed for intracellular IFN-α levels after 8 hours of stimulation. (D) Statistical analysis of the percentage of IFN-α producing pDCs. (E) The expression of TLR9 was measured by flow cytometry after 8 hours of stimulation. The data presented are representative of three independent experiments (mean ± SEM). **P < 0.01.

We next measured the intracellular IFN-α levels to determine which cell type within the PBMC populations produced IFN-α in response to HBV-CpG. After stimulation with HBV-CpG, IFN-α was exclusively produced by Lineage1, CD123+, HLA-DR+ cells (pDCs).[16, 17] Thus, pDCs are the only cells within the PBMC population that produced IFN-α in response to HBV-CpG (Fig. 1C,D). To determine whether TLR9 acted as the HBV-CpG receptor, PBMCs were stimulated with HBV-CpG for 8 hours and TLR9 expression was determined by flow cytometry. This experiment illustrated that TLR9 was significantly upregulated in cells treated with HBV-CpG compared to cells treated with interleukin (IL)-3 (Fig. 1E). We next made use of chloroquine, an endosomal acidification inhibitor that can block TLR9 and TLR7 signaling and was able to significantly inhibit HBV-CpG-induced IFN-α production (Supporting Fig. 1). pDCs are known to trigger type I IFN production by two intracellular TLR, TLR7 and TLR9, which recognizes single-stranded RNA and unmethylated CpG motifs, respectively.[4] We also found that there was little change in expression of TLR7 on pDCs when PBMCs were incubated with HBV-CpG (Supporting Fig. 2). These data demonstrate that HBV-CpG potently induces the production of IFN-α by human pDCs in a TLR9-dependent manner.

NP(HBV-CpG) Reversed the HBV-ODN-Mediated-Block of IFN-α Production

Previous studies have reported that guanosine-rich ODNs specifically inhibit TLR9 signaling.[18, 19] We found a high frequency of guanosine repetitive elements in the HBV genome. To determine whether these guanosine-rich ODNs encoded by the HBV genome suppress IFN-α expression, inhibitory ODNs derived from HBV genomic sequences (HBV-ODNs) with guanosine-rich motifs were screened for suppressive activity (Supporting Table 2). When PBMCs were incubated with HBV-CpG and different inhibitory HBV-ODNs, the HBV-CpG-induced IFN-α production was inhibited (Fig. 2A). However, nonguanosine-rich HBV ODNs (negative control) failed to inhibit IFN-α production. Additionally, HBV-ODNs had no effect on the IFN-α production induced by CpG-2216 (Fig. 2B). These data indicate that the endogenous HBV-CpG-mediated induction of IFN-α can be specifically blocked by inhibitory HBV-ODNs, consistent with the defective type I IFN production observed during HBV infection.[20]

Figure 2.

NP(HBV-CpG) reversed the HBV-ODN-mediated suppression of IFN-α production. (A) PBMCs were incubated overnight with 5 μg of HBV-CpG in the presence of 5 μg of different inhibitory HBV-ODNs (t1, t2, t3, t4, t5, t6, t7, t8, or the negative control), and IFN-α was measured in the supernatant. (B) PBMCs were stimulated with 5 μg of CpG 2216 in the presence of 5 μg of the different HBV-ODNs for 42 hours, and IFN-α was measured in the supernatants. (C) Inhibitory HBV-ODN inhibited the localization of HBV-CpG together with TLR9. Gen2.2 cells were cultured with either 5 μg of HBV-CpG-488 or 5 μg of HBV-CpG-488 in combination with 5 μg of inhibitory HBV-ODN for 90 minutes. Cells were fixed, stained intracellularly with TLR9 antibodies (red), and imaged by confocal microscopy. Original magnification: 400×. (D) PBMCs were incubated overnight with HBV-CpG or NP(HBV-CpG) and inhibitory HBV-ODNs at HBV-CpG or NP(HBV-CpG) to HBV-ODN ratios of 1:1 or 2:1. The IFN-α level in the supernatant was measured. (E,F) PBMCs from patients with chronic HBV infection were stimulated with NP(HBV-CpG), and IFN-α levels was measured by flow cytometry and ELISA. All of the data are represented as the mean ± SEM. *P < 0.05. n.d.: not detected.

To investigate whether inhibitory HBV-ODNs inhibit the production of IFN-α by blocking TLR-9 signaling, the human pDC cell line Gen2.2[14] was incubated with fluorescent HBV-CpG and HBV-ODNs and stained for TLR9 by immunofluorescence and confocal microscopy. Consistent with previous reports,[21] HBV-CpG colocalized with TLR9. However, inhibitory HBV-ODN suppressed uptake of HBV-CpG by cells and blocked the colocalization of HBV-CpG with TLR9 in endosomal vesicles (Fig. 2C).

To reverse the inhibition of IFN-α induction, we used a double-emulsion method to encapsulate HBV-CpG into nanoparticles that were fabricated with biodegradable poly(ethylene glycol)-block-poly(d,l-lactide) (mPEG-b-PLA) with the assistance of a cationic lipid BHEM-Chol.[22] The average diameter of the nanoparticles was 120 ± 3.4 nm, and the encapsulation efficiency of HBV-CpG was 95.7%. The incubation of PBMCs with nanoparticles containing HBV-CpG (NP(HBV-CpG)) or HBV-CpG and equal concentrations of inhibitory HBV-ODNs resulted in IFN-α production at barely detectable levels that reached the baseline level (Fig. 2D). However, the coadministration of NP(HBV-CpG) and HBV-ODNs at a 2:1 ratio increased IFN-α production compared to coadministration of HBV-CpG and HBV-ODNs at the same ratio (Fig. 2D). Meanwhile, NP(HBV-CpG) activated pDCs, NK, and T cells to a greater extent than HBV-CpG and nanoparticles (Supporting Fig. 3). Together, these results suggested that NP(HBV-CpG) not only suppressed the activity of inhibitory HBV-ODNs but also stimulated stronger immune responses from PBMCs. Furthermore, the effect of NP(HBV-CpG) on HBV patient-derived PBMCs was investigated, and we also found that NP(HBV-CpG) induced IFN-α production in patient-derived PBMCs (Fig. 2E,F).

NP(HBV-CpG) Exerted a Strong Immunostimulatory Effect on the Lymphocytes of Wild-Type Mice

Because we found that NP(HBV-CpG) could activate human PBMCs in vitro, we subsequently challenged mice with NP(HBV-CpG) in vivo. After treated intravenously with NP(HBV-CpG), C57BL/6 mice displayed higher levels of CD40 and CD80 on pDCs (CD11cint/B220+/CD11b) and cDCs (CD11c+) than mice treated with PBS. In addition, the expression of CD69 and ICOS was strongly upregulated on the NK and T cells (CD4+ and CD8+) treated with NP(HBV-CpG) compared to the PBS control cells (Fig. 3A). In agreement with the in vivo experiment, the coincubation of murine splenic lymphocytes with NP(HBV-CpG) increased the expression of CD40 and CD80 on both pDC and cDC populations (Fig. 3B). The viability of the pDCs and cDCs was much higher in the presence of NP(HBV-CpG) than PBS (Fig. 3C). Furthermore, the expression ratio of CD69 on NK and T cells was higher in splenocytes cultured with NP(HBV-CpG) than in those cultured with PBS (Fig. 3D). These results suggest that NP(HBV-CpG) can induce the activation of murine lymphocytes in vitro and in vivo.

Figure 3.

NP(HBV-CpG) activation of lymphocytes from wild-type mice in vivo and in vitro. C57BL/6 mice were treated with NP(HBV-CpG), NPs, HBV-CpG or PBS intravenously, and the animals were euthanized after 12 hours. (A) pDCs and cDCs were analyzed for CD40 and CD80 expression. The surface expression of CD69 and ICOS on CD4+ T cells, CD8+ T cells, and NK1.1+ cells was analyzed by flow cytometry. Splenic lymphocytes (1 × 107/mL) from C57BL/6 mice were incubated with NP(HBV-CpG), HBV-CpG, 2216, or PBS for 36 hours. (B) The expression of CD40 and CD80 on pDCs and cDCs was analyzed by flow cytometry. (C) The absolute number of pDCs and cDCs was counted by flow cytometry. (D) The expression of the costimulatory molecule CD69 was analyzed on NK and T cells. All of the data are presented as the mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001, relative to PBS control.

NP (HBV-CpG) Synergistically Enhanced the Antibody Response to HBsAg

Cohorts of BALB/c and C57BL/6 mice were immunized with the rHBsAg vaccine alone, the rHBsAg vaccine plus NPs, the rHBsAg vaccine plus NP(HBV-CpG), or were left untreated. The level of anti-HBsAg antibodies in the sera of the mice was measured 2 and 4 weeks following immunization. Immunization with the rHBsAg vaccine plus nanoparticles containing 10 μg of HBV-CpG markedly increased the HBsAg-specific antibody levels in both BALB/c and C57BL/6 mice. Significantly, the most prominent effect of NP(HBV-CpG) administration was a large increase in the production of Th1-dependent IgG2a-antibodies (Fig. 4A,B). These results indicate that NP(HBV-CpG) can assist the rHBsAg vaccine in inducing a vigorous anti-HBsAg, Th1-biased antibody response.

Figure 4.

The antibody responses of BALB/c and C57BL/6 mice. (A) BALB/c mice were immunized with the rHBsAg vaccine and NP(HBV-CpG). The levels of total anti-HBsAg IgG, anti-HBs IgG1, and anti-HBs IgG2a were measured 2 and 4 weeks after the last immunization. (B) C57BL/6 mice were immunized as described above. The levels of anti-HBsAg antibodies were determined as described in (A). All of the data are presented as the mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001, ns, not significant, relative to rHBsAg control.

NP(HBV-CpG) Stimulated a Robust Innate Immune Response in HBV Carrier Mice

Given that NP(HBV-CpG) exerted a strong effect in wild-type mice, we evaluated the effect of NP(HBV-CpG) on the splenocytes of HBV carrier mice. C57BL/6 mice were hydrodynamically injected in the tail vein with the pAAV/HBV1.2 plasmid.[15] The HBV carrier mice were administered NP(HBV-CpG), HBV-CpG, NPs, or PBS. The NP(HBV-CpG) treatment resulted in a significantly greater up-regulation of CD40 and CD80 on cDCs, CD69, and ICOS on NK cells and CD69 on CD4+ T cells and CD8+ T cells than did the PBS treatment (Fig. 5A). Significantly higher levels of serum IFN-α, but not IFN-β, were also observed (Fig. 5B). These results suggest that NP(HBV-CpG) is capable of triggering an immune response in HBV carrier mice.

Figure 5.

NP(HBV-CpG) activated lymphocytes from HBV carrier mice. (A) HBV carrier mice were treated with NP(HBV-CpG) (20 μg of HBV-CpG: 1 mg of NPs), 20 μg of HBV-CpG, NPs, or PBS intravenously. Mice were bled for the detection of IFN after 6 hours and euthanized after 12 hours. The levels of the costimulatory molecules CD40 and CD80 (cDCs), CD69, and Icos (NK cells) and CD69 (CD4+ and CD8+ T cells) were determined by flow cytometry (mean ± SEM). *P < 0.05 and ***P < 0.001, relative to PBS control. (B) The serum IFN-α and IFN-β levels were measured by ELISA. n.d.: not detected. (mean ± SEM).

NP(HBV-CpG)-Based Therapy Effectively Cleared HBV in HBV Carrier Mice

Huang and colleagues have shown that HBsAg failed to induce an immune response in HBV carrier mice, indicating that these mice exhibit tolerance toward HBsAg.[15, 23] However, our aforementioned data indicated that NP(HBV-CpG) could trigger lymphocyte activation and IFN-α expression in HBV carrier mice (Fig. 5). Therefore, we sought to explore the role of NP(HBV-CpG) in the clearance of HBV in the HBV carrier mice model. We then found that NP(HBV-CpG) injected either intraperitoneally or intravenously could induce IFN-α expression (data not shown). Compared to the intravenous route, the intraperitoneal route is more convenient. Wild-type mice treated subcutaneously with the rHBsAg vaccine in combination with NP(HBV-CpG) showed enhanced antigen uptake and the induction of a protective immune response (Fig. 4). Therefore, we treated HBV carrier mice with NP(HBV-CpG) or NPs intraperitoneally three times and then treated the mice once with NP(HBV-CpG) or NPs combined with rHBsAg subcutaneously, and the treatment course lasted 3 weeks (Fig. 6A). The mice were bled regularly to monitor the serum levels of HBsAg. We observed that the HBsAg level in the serum of carrier mice treated with NP(HBV-CpG) combined with rHBsAg declined significantly. However, in accordance with the tolerance of HBsAg in the HBV carrier model, the HBsAg level remained high in the mice treated with NPs combined with rHBsAg (Fig. 6B). Likewise, this effect on elimination of HBV was not observed in HBV carrier mice treated with NP(2216) combined with rHBsAg or the NP(HBV-CpG) alone (Supporting Fig. 4). HBV surface antigenemia disappeared in ∼90% of the HBV carrier mice within 3 weeks of the NP(HBV-CpG)-based treatment (Fig. 6C). Moreover, the level of HBcAg in the liver tissues and the serum levels of HBV DNA declined significantly and were undetectable in some of the samples treated with NP(HBV-CpG) combined with rHBsAg; this effect was not observed in the untreated mice or the mice treated with the combination of NPs and rHBsAg (Fig. 6D,E). The rapid clearance of HBsAg, HBcAg, and HBV DNA in the NP(HBV-CpG)-immunized mice indicated a reversal of HBsAg tolerance. Furthermore, anti-HBsAg antibodies, protective neutralizing antibodies, began to appear after HBsAg clearance in the groups receiving the NP(HBV-CpG)-based treatment but remained undetectable in the other two groups (Fig. 6F). Moreover, the percentage of HBsAg-specific IFN-γ-producing CD8+ T cells were up-regulated by the NP(HBV-CpG) combined administration (Supporting Fig. 5), suggesting that the NP(HBV-CpG)-based treatment augmented cytotoxic T lymphocyte (CTL) function. Meanwhile, the livers of the mice receiving the combination treatment showed normal architecture with a mild inflammatory infiltrate. Normalization of serum aminotransferases and bilirubin suggested that HBV carrier mice exhibited good tolerance to the combination treatment (Supporting Fig. 6). These results suggest that NP(HBV-CpG)-based therapy can effectively clear HBV and aid in the short-term production of anti-HBsAg antibodies, thereby exerting strong anti-HBV activity.

Figure 6.

NP(HBV-CpG) combined with rHBsAg therapy cleared HBV in HBV carrier mice. (A) HBV carrier mice were administered by intraperitoneal injection of NP(HBV-CpG) or NPs on days 1, 2, 3, 8, 9, 10, 15, 16, and 17 and subcutaneously with NP(HBV-CpG) plus rHBsAg or NPs plus rHBsAg on days 4, 11, and 18. The HBV carrier mice were bled on days 7, 14, and 21. (B,C) The serum HBsAg levels were measured by RIA. The mice were euthanized on day 28, and (D) the HBcAg level in the liver was detected by immunohistochemical staining (scale bar = 50 μm). (E) The HBV DNA and (F) anti-HBsAg levels were measured by RT-PCR and RIA, respectively. n.d.: not detected. The data are expressed as the mean ± SEM. *P < 0.05, ns, not significant, relative to untreated control.

Discussion

CpG ODNs from the bacterial genome have been reported to activate pDCs through TLR9 and thereby induce IFN-α secretion. However, the type I interferon-inducing activity of CpG ODNs from the HBV genome remains to be defined. Our results are the first to demonstrate that HBV-derived CpG can induce potent IFN-α production by human pDCs (Fig. 1). Therefore, we propose that HBV-CpG sequences that mimic the motifs found in HBV DNA induce IFN-α by way of TLR9-dependent pathway. However, this hypothesis seems to be in disagreement with the finding that HBV does not induce any detectable changes in the expression of type I interferon in the early weeks of infection.[20] Previous studies have shown that HBsAg and HBcAg impair TLR9 and pDC responses, leading to impairments in IFN-α production,[24, 25] which partially explains this discrepancy. In this study, we observed that ∼14 guanosine-tetrad sequences exist in the HBV genome. Our findings indicate that these repetitive, guanosine-based inhibitory HBV-ODNs derived from the HBV genome blocked HBV-CpG-induced IFN-α production (Fig. 2A). A previous study found that G-tetrad-mediated disruption of the colocalization of CpG DNA with TLR9 down-regulated the innate immune response elicited by a TLR ligand.[19] We demonstrated that inhibitory HBV-ODNs interfered with the uptake of HBV-CpG and the colocalization of HBV-CpG with TLR9 in endosomal vesicles (Fig. 2C). It is interesting that the number of endogenous CpG-rich motifs in the HBV genome is lower than the number of inhibitory guanosine-rich motifs. This raises the intriguing possibility that guanosine-based DNA released by HBV could down-regulate immune responses driven by HBV-CpG and lead to the immune escape. These findings suggest that the relationship between HBV-CpG and HBV-ODN may be responsible for the lack of IFN-α in patients with chronic hepatitis B and the subsequent progression to persistent HBV infection.

Interestingly, higher concentrations of PEG-PLA nanoparticles containing HBV-CpG reversed the suppression of IFN-α production by inhibitory HBV-ODNs (Fig. 2C). Biodegradable polymers, such as PEG-PLA, are currently used to encapsulate vaccine antigens and adjuvants. Our previous studies have shown that PEG-PLA nanoparticles with small interfering RNA (siRNA) encapsulation are effectively internalized by cancer cells.[22] We propose that the nanoparticle-encapsulated HBV-CpG efficiently enters cells through an endocytic pathway and subsequently releases the HBV-CpG load rapidly, thereby reversing the HBV-ODN-mediated block of IFN-α production.

The current HBV vaccine does not stimulate the production of protective antibodies in ∼10% of healthy people, and the rates of seroconversion are also low among HIV-infected patients. Our data indicate that NP(HBV-CpG) induces a robust immune response when used as an rHBsAg vaccine adjuvant, not only by enhancing HBsAg-specific antibody levels but also by shifting a Th-2 biased response toward a Th-1 biased response (Fig. 4). Previous studies have demonstrated that CpG ODNs can improve vaccine immunogenicity.[8, 26] Our findings were confirmed in BALB/c mice (H-2d haplotype), which are strong immune responders, and in C57BL/6 mice (H-2b haplotype), which are intermediate immune responders.[15] Accordingly, it is reasonable to assume that NP(HBV-CpG) may benefit individuals who cannot respond to the rHBsAg vaccine alone.

We hypothesized that NP(HBV-CpG)-mediated activation of innate immune responses and the correction of imbalance of interferon by inhibitory HBV-ODNs may further enhance the function of pDCs, which could facilitate adaptive immune responses and, finally, help to break HBV-induced immune tolerance. Although pDCs obtained from patients chronically infected with HBV may have an altered function and may also down-regulate TLR9,[25, 27] in the present study we found that NP(HBV-CpG) was capable of inducing IFN-α production in PBMCs derived from HBV patients (Fig 2E,F). As expected, in terms of the therapeutic application for HBV infection, NP(HBV-CpG) combined with rHBsAg therapy drastically decreased the serum HBsAg and HBV DNA levels in HBV carrier mice in addition to causing their livers to become HBcAg-negative (Fig. 6). Previous studies have shown that CpG ODN treatments can enhance rHBsAg vaccine-induced immune responses and suppress HBV replication.[28-30] However, the goal of eradicating HBV without side effects has not yet been achieved.[31] Our data indicate that NP(HBV-CpG)-based therapy effectively aids in HBV clearance (Supporting Fig. 8), as demonstrated by HBsAg seroconversion and the inhibition of HBV DNA replication, which led to the development of protective anti-HBsAg antibodies. Additionally, CTL function (IFN-γ+ CD8+ T cells percentages increased) (Supporting Fig. 5) was also involved in the antiviral response.

The present study indicates that NP(HBV-CpG)-based therapy has a marked curative effect on HBV infection in HBV carrier mice compared to nonspecific CpG-2216-mediated therapy (Supporting Fig. 4). We speculate that NP(HBV-CpG)-based treatment may have a higher specificity against HBV than NP(2216)-based therapy because of the use of endogenous CpG derived from the HBV genome. On the other hand, TLR9 agonists can also trigger undesirable responses, including IL-10 production or the activity of T regulatory cells.[32] We also found that CpG-2216 triggered very large amounts of IL-10 compared to HBV-CpG (Supporting Fig. 7) and IL-10 could mediate liver tolerance. Therefore, these findings support the advantage of NP(HBV-CpG) in the treatment of hepatitis B infection.

Currently, there is no ideal HBV infection mouse model. In our present study, HBV carrier mice generated by hydrodynamic injection of the pAAV/HBV1.2 plasmid were employed to mimic human HBV persistent infection. Compared to other HBV-transgenic mouse models, this model is better for the observation of the immune response/tolerance.[15] However, this model cannot completely simulate HBV natural infection (e.g., infectious virus particles cannot be produced). Furthermore, whether NP(HBV-CpG)-based therapy can be used in the clinic requires further investigation.

In conclusion, NP(HBV-CpG) may provide a useful platform for the design of an HBV-targeted vaccine, and, more important, may represent a new strategy for the treatment of HBV infection in humans.

Acknowledgment

We thank Pei-Jer Chen (National Taiwan University) for the HBV plasmid pAAV/HBV1.2 and Zhenghong Yuan (Fudan University) for the Gen2.2 cell line.

Ancillary