Therapeutic recovery of hepatitis B virus (HBV)-induced hepatocyte-intrinsic immune defect reverses systemic adaptive immune tolerance

Authors

  • Peixiang Lan,

    1. Institute of Immunopharmacology & Immunotherapy, School of Pharmaceutical Sciences, Shandong University, Shandong, China
    Search for more papers by this author
  • Cai Zhang,

    Corresponding author
    1. Institute of Immunopharmacology & Immunotherapy, School of Pharmaceutical Sciences, Shandong University, Shandong, China
    • Institute of Immunopharmacology & Immunotherapy, School of Pharmaceutical Sciences, Shandong University, 44 Wenhua West Road, Jinan 250012, China; fax: 86-531-8838-3782===

    Search for more papers by this author
    • fax: 86-531-8838-3782

  • Qiuju Han,

    1. Institute of Immunopharmacology & Immunotherapy, School of Pharmaceutical Sciences, Shandong University, Shandong, China
    Search for more papers by this author
  • Jian Zhang,

    1. Institute of Immunopharmacology & Immunotherapy, School of Pharmaceutical Sciences, Shandong University, Shandong, China
    Search for more papers by this author
  • Zhigang Tian

    Corresponding author
    1. Institute of Immunopharmacology & Immunotherapy, School of Pharmaceutical Sciences, Shandong University, Shandong, China
    2. Department of Microbiology and Immunology, School of Life Sciences, University of Science and Technology of China, Anhui, China
    • Institute of Immunopharmacology & Immunotherapy, School of Pharmaceutical Sciences, Shandong University, 44 Wenhua West Road, Jinan 250012, China; fax: 86-531-8838-3782===

    Search for more papers by this author
    • fax: 86-531-8838-3782


  • Potential conflict of interest: Nothing to report.

  • Supported by grants from the National 973 Basic Research Program of China (#2013CB944901), the Natural Science Foundation of China (#81273220, #31200651), and the National 115 Key Project for HBV Research (#2008ZX10002-008).

Abstract

Hepatitis B virus (HBV) persistence aggravates hepatic immunotolerance, leading to the failure of cell-intrinsic type I interferon and antiviral response, but whether and how HBV-induced hepatocyte-intrinsic tolerance influences systemic adaptive immunity has never been reported, which is becoming the major obstacle for chronic HBV therapy. In this study, an HBV-persistent mouse, established by hydrodynamic injection of an HBV-genome-containing plasmid, exhibited not only hepatocyte-intrinsic but also systemic immunotolerance to HBV rechallenge. HBV-specific CD8+ T-cell and anti-HBs antibody generation were systemically impaired by HBV persistence in hepatocytes. Interestingly, HBV-induced hepatocyte-intrinsic immune tolerance was reversed when a dually functional vector containing both an immunostimulating single-stranded RNA (ssRNA) and an HBx-silencing short hairpin RNA (shRNA) was administered, and the systemic anti-HBV adaptive immune responses, including CD8+ T-cell and anti-HBs antibody responses, were efficiently recovered. During this process, CD8+ T cells and interferon-gamma (IFN-γ) secreted play a critical role in clearance of HBV. However, when IFN-α/β receptor was blocked or the Toll-like receptor (TLR)7 signaling pathway was inhibited, the activation of CD8+ T cells and clearance of HBV was significantly impaired. Conclusion: These results suggest that recovery of HBV-impaired hepatocyte-intrinsic innate immunity by the dually functional vector might overcome systemic adaptive immunotolerance in an IFN-α- and TLR7-dependent manner. The strategy holds promise for therapeutic intervention of chronic persistent virus infection and associated cancers. (Hepatology 2013;)

Hepatitis B virus (HBV) infection, with 400 million carriers worldwide, is a major risk factor for hepatocellular carcinoma (HCC),1 particularly in Asia and Africa. Both innate and adaptive immunity are capable of controlling HBV replication.2, 3 However, recent studies report that the innate response is weak and poorly able to sense HBV during infection, partly due to active suppression strategies by persistent HBV in liver. For example, HBV suppresses type I interferon (IFN) production by disrupting the virus-induced signaling adapter (VISA)-associated complex4 and inhibits innate immune Toll-like receptor (TLR)-mediated antiviral signaling in hepatocytes,5 leading to cell-intrinsic immunotolerance. CD8+ T cells also play a critical role in HBV clearance, especially intrahepatic HBV-specific CD8+ T cells.3, 6 Although HBV-specific CD8+ T-cell numbers remain low during infection, their cytokines, including IFN-γ and tumor necrosis factor alpha (TNF-α), are essential for suppressing HBV gene expression and replication.3 Unfortunately, in chronic HBV (CHB)-infected patients, CD8+ T cells lose their ability to proliferate and mediate antiviral function; this dysfunctional state is characterized by coinhibitory molecule overexpression (e.g., PD-1, Tim-3, CTLA-4), low cytokine production, and T-cell exhaustion.7 In addition to exhibiting impaired HBcAg-specific CD8+ T-cell responses, studies on HBV-carrier mice revealed that anti-HBs antibody (Ab) production is also suppressed.8 Lower HBV-specific Abs are also reflected in CHB patients, indicating that HBV persistence impairs both CD8+ T-cell and humoral arms of adaptive immunity. To achieve effective HBV therapy, there is a pressing need to develop strategies to break cell-intrinsic tolerance and reconstitute adaptive immunity against HBV.

One promising strategy to treat CHB infection is simultaneous use of immune stimulation and HBV gene-expression silencing to reduce antigen load; recently, bifunctional 5′-triphosphate-small interfering RNAs (siRNAs) (3p-siRNAs) silenced HBV expression and simultaneously activated the host retinoic acid inducible gene I (RIG-I) signaling pathway to successfully reverse hepatocyte-intrinsic immunotolerance.9, 10 However, whether reversing cell-intrinsic tolerance promotes recovery of adaptive immunity in vivo is unknown.

Abbreviations

APC, antigen-presenting cell; CHB, chronically HBV infected patients; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; HBx, hepatitis B virus X gene; IFN, interferon; ISG, interferon-stimulated gene; MxA, myxovirus resistance protein A; NF-κB, nuclear factor-κB; PRRs, pathogen recognition receptors; TGF, transforming growth factor; TLR, toll-like receptor; TNF, tumor necrosis factor; PD-1, Programmed Death-1.

The nucleotide-sensing pattern recognition receptors (PRRs), TLR7 and TLR8, are located on endosomes and recognize specific viral single-stranded RNA (ssRNA) sequences, such as GUGUU,11 U-rich sequences, and a GU-rich 4-mer.11, 12 Receptor activation stimulates IFN-regulatory factor 7 (IRF7), nuclear factor-κB (NF-κB), and other downstream signal pathways to induce type I IFN and inflammatory cytokine production.11 TLR7/8 also recognize synthetic imidazoquinoline derivatives and siRNAs with U-rich RNA sequences,12 and these agonists show potential to enhance innate and adaptive immunity in immunotherapy against cancer and infection.13, 14

In this study we show that HBV persistence induced hepatocyte-intrinsic immunotolerance that suppressed the generation of both HBV-specific CD8+ T-cell and anti-HBs Ab responses, leading to systemic tolerance in vivo. Treating HBV-carrier mice with a dual-function short hairpin RNA (shRNA) vector, exerting both immunostimulatory and HBx-silencing effects in vivo, efficiently inhibited HBV and increased type I IFN production. Most important, this therapy reversed HBV-induced hepatocyte-intrinsic immunotolerance and recovered systemic anti-HBV adaptive immunity by restoring hepatic CD8+ T-cell activation and proliferation as well as HBV-specific Ab responses.

Materials and Methods

Cell Lines and Reagents.

HepG2 cell lines were maintained in our laboratory and cultured in RPMI-1640 medium (GIBCO/BRL, Gaithersburg, MD) containing 10% fetal bovine serum (FBS). HepG2.2.15 cells (derived from HepG2 cells transfected with a plasmid carrying two head-to-tail copies of HBV genome DNA serotype ayw) were maintained in complete Dulbecco's modified Eagle's medium (DMEM) (GIBCO/BRL) supplemented with 10% FBS. All cultures were incubated at 37°C and 5% CO2 in a humidified atmosphere. The TLR7 inhibitor was endotoxin-free oligodeoxyribonucleotide IRS661 (5′-TGCTTGCAAGCTTGCAAGCA-3′)15 (Takara, Japan). Neutralizing α-IFNR I Ab was from Millipore (Bedford, MA).

HBV Carrier Mouse Model.

HBV-carrier mice were established by hydrodynamic injection of pAAV/HBV1.2 plasmid (kindly provided by Pei-Jer Chen, National Taiwan University College) by way of the tail vein into wild-type (WT) C57BL/6, IFN-γ−/− and Rag-1−/− mice. Four weeks later, hepatitis B surface antigen (HBsAg) was highly expressed in liver tissue, and HBV-carrier mice (HBV+) were defined as harboring serum HBsAg levels >500 ng/mL. For HBV vaccination, HBV vaccine (rHBs/CFA) was injected subcutaneously. All animal experiments and protocols were approved by the Committee on the Ethics of Animal Experiments of the Shandong University.

HBV DNA Analysis.

Viral particles in supernatants and in mice sera were quantified by real-time polymerase chain reaction (PCR) according to the kit's instructions (Da-An, Guangzhou, China). Primers detecting the HBV S region were 5′-ATCCTGCTGCTATGCCTCATCTT-3′ and 5′-ACAGGGGGAAAGCCCTACGAA-3′ as well as the 5′-FAM-TGGCTAGTTTACAGTGCCATTTG-TAMRA fluorescent probe. Quantitative PCR (qPCR) was performed in the iCycleriQ for 42 cycles.

Flow Cytometry.

Multiparameter flow cytometry was performed according to a standard protocol. Surface or intracellular staining was performed using the following antimouse monoclonal Abs (mAbs) or Ab controls: FITC-conjugated immunoglobulin G (IgG) isotype, α-NK1.1, α-PD-1, α-PD-L1, α-CD8, and α-CD4; PE-conjugated IgG isotype, α-CD69, α-CTLA-4, α-IFN-γ, and α-perforin; PE-Cy5.5-conjugated IgG isotype, α-CD3, α-CD8, and α-CD25; allophycocyanin (APC)-conjugated IgG isotype, α-CD28, and α-CD107a. All Abs were purchased from eBioscience (San Diego, CA). Dimeric H-2Kb:Ig fusion protein (BD Biosciences, San Jose, CA) was complexed with HBc 93-100 peptide (AnaSpec, Fremont, CA). Lymphocytic choreomeningitis virus (LCMV) gp33-41 peptide was purchased from AnaSpec.

Mouse liver lymphocytes were separated and washed with phosphate-buffered saline (PBS) and stained with fluorochrome-conjugated Abs using control IgG isotype. Intracellular staining was performed using fixation and permeabilization buffers (eBioscience) according to the manufacturer's instructions. Flow cytometry was performed using FACSCalibur and data were analyzed with CellQuest software (BD Biosciences).

mAb and Cell Depletion.

Cell depletion mAbs were purified from 2.43 (α-CD8β), GK1.5 (α-CD4), and PK136 (α-NK1.1) hybridoma cell lines. To deplete cells, mice were injected intraperitoneally with 1 mg of mAb. Three days later the dual vector was administered intravenously.

Cell Sorting and Adoptive Transfer.

Splenic lymphocytes were separated from WT C57BL/6 or IFN-γ−/− mice. Lymphocytes were incubated with microbeads directly conjugated to antimouse CD8 Ab (10 μL /107 cells) at 4°C for 20 minutes. Labeled cells were removed on MACS columns in a magnetic field. After washing twice with PBE solution, the column was removed from the magnet and flushed with PBE. After washing with PBS, CD8+ T cells were sorted and purity was analyzed by fluorescent-activated cell sorter FACS (>90%).

The sorted splenic CD8+ T cells from WT C57BL/6, IFN-γ−/− or IFNAR−/− mouse, or CD8-depleted splenic lymphocytes, were transferred intravenously into recipient Rag 1−/− HBV carrier mice (5 × 106 cells/recipient).

Plasmid construction, lentiviral packaging, reverse transcription, and real-time PCR analysis, western blot, enzyme-linked immunosorbent assay (ELISA) assay, and immunohistochemistry are included in the Supporting Information.

Data Analysis.

Statistical analysis was performed using a paired Student's t test. P < 0.05 was considered statistically significant.

Results

HBV Induces Cell-Intrinsic and Systemic Immunotolerance.

HBV inhibits TLRs or other PRR-mediated innate immune responses4, 5 by suppressing the host antiviral type I IFN signal pathway, leading to cell-intrinsic immunotolerance. To explore this cell-intrinsic immunotolerance, we first evaluated expression of type I IFNs, IFN-inducible genes (ISG15 and MxA), and immunosuppressive cytokines (TGF-β and interleukin [IL]-10) in HBV-persistent HepG2.2.15 cells. We found that IFN-α, IFN-β, ISG15, and MxA expression was significantly lower, while TGF-β and IL-10 was higher, in HepG2.2.15 cells than in control HepG2 cells (Supporting Fig. 1A). We also evaluated gene expression in human primary HCC cells harvested from HBV+ CHB patients and found similar results (Supporting Fig. 1B). To further confirm this in vivo, we established HBV-persistent mice by hydrodynamic injection of pAAV/HBV1.2 plasmid into C57BL/6 mice (Supporting Fig. 2A-D). Four weeks later, a time when HBV-carrier established, liver tissue exhibited high HBsAg expression without liver injury (Supporting Fig. 2B). The higher levels of HBsAg and HBV-DNA in serum can persist for at least 6 months (Supporting Fig. 2C,D) with no nonspecific inflammatory and liver injury, suggesting that the HBV-persistent mice had been successfully established. Mice with serum HBsAg levels >500 ng/mL were defined as HBV-persistent mice (HBV+), and were inoculated with HBV vaccine (rHBs/CFA). HBV+ mice did not produce anti-HBs Ab like vaccinated control HBV mice (Fig. 1A), suggesting that HBV+ mice were systemically immunotolerant to HBV. Similar to infected human hepatocytes and liver tissues, IFN-α/β mRNA levels were lower in HBV+ than in HBV hepatocytes (Fig. 1B), while immunosuppressive cytokines significantly increased (Fig. 1C). These results collectively indicate that HBV infection induces hepatocyte-intrinsic innate immunotolerance.

Figure 1.

HBV-induced cell-intrinsic and systemic immunotolerance. (A) Plasmid pAAV/HBV1.2 (8 μg) was hydrodynamically injected intravenously into 4-5-week-old C57BL/6 mice to generate an HBV-carrier mouse model. Four weeks later, HBV vaccine (rHBs/CFA) was injected subcutaneously into the tail vein of HBV+ (HBsAg >500 ng/mL) and HBV (HBsAg <100 ng/mL) mice. Two weeks later, serum anti-HBs Ab level was determined by ELISA. (B,C) HBV-carrier mice were generated as in (A), and mouse primary hepatocytes were isolated from HBV or HBV+ mice by the liver perfusion method. Type I IFN mRNA was analyzed by qRT-PCR (B), and serum TGF-β1 and IL-10 were detected by ELISA (C). (D,E) HBV-persistent mice were generated as in (A), the percentage and absolute number of hepatic CD8+ T cells (D), and the percentage of PD-1+CD8+ T cells (E) from HBV+ or HBV mice was assayed by FACS. (F,G) Plasmids pAAV or pAAV/HBV1.2 (8 μg) were reinjected into HBV+ and HBV mice by way of the tail vein 4 weeks after first administration. Two weeks later, the percentages and absolute number of HBc93-100-specific CD8+ T cell (F), and the percentage of IFN-γ+ (G) within hepatic CD8+ T cells were detected by FACS. (H) LCMV gp 33-41 was injected into HBV+ and HBV mice by way of the tail vein and 2 weeks later LCMV-specific CD8+ T cells were detected by FACS. Data are expressed as percentage mean ± standard deviation (SD) from at least three separate experiments (A-C) or representative of three independent experiments with six mice per group (D-H). *P < 0.05, **P < 0.01, versus HBV mice.

Evaluating adaptive immunity generated in HBV+ mice, we found that the percentage and absolute number of hepatic CD8+ T cell (Fig. 1D) was reduced, and moreover, inhibitory PD-1 expression on hepatic CD8+ T cells was almost 3-fold higher than in HBV mice (Fig. 1E). To observe recall responses and to determine if HBV persistence was established in HBV+ mice, pAAV/HBV1.2 plasmid was readministered. Two weeks later, HBV mice eliminated HBV, but HBV+ mice remained HBV persistent (data not shown). Importantly, the percentage and absolute number of hepatic HBc-specific CD8+ T cells (detected by HBcAg93-100 pentamer staining) (Fig. 1F) as well as the percentage of hepatic IFN-γ+ CD8+ T cells (Fig. 1G) decreased significantly in HBV+ mice, indicating that HBV persistence impaired CD8+ T-cell responses. We also detected the specific response to LCMV infection by LCMV gp33 administration. Our data showed that the percentages of LCMV gp33+ CD8+ T cells were increased in both HBV and HBV+ mice with no significant differences (Fig. 1H). These results suggest that HBV-induced systemic immunotolerance is HBV-specific. All the results raised the possibility that impairing HBV-induced hepatocyte-intrinsic immune responses leads to systemic adaptive immunotolerance.

Dual-Function Immunostimulatory HBx-shRNA Vector Reverses HBV-Induced Intrinsic Immunotolerance.

To test whether intrinsic innate immunotolerance can be reversed in vivo, we constructed a dually functional vector containing an immunostimulatory ssRNA and an HBx-gene-silencing shRNA. We designed four different sequences encoding ssRNAs and HBx-shRNA, and inserted them into the shRNA pSIREN expression vector. Transfection with ssRNA1- and ssRNA4-containing vectors significantly enhanced IFN-α production in supernatants, while all four shRNA vectors effectively silenced HBx expression at both the messenger RNA (mRNA) and protein levels (Supporting Fig. 3A,B). We selected ssRNA4 and HBx-shRNA3 to construct the dual-function vector (Supporting Fig. 3C).

The dual-function (dual), single immunostimulatory RNA (ssRNA), single HBx-shRNA (shRNA), or pSIREN (empty control) vectors were separately transfected into HBV-persistent HepG2.2.15 cells. Although shRNA and dual vectors significantly reduced HBx expression at both the mRNA and protein levels, the dual vector more effectively reduced HBV DNA replication and HBsAg/HBeAg production (Supporting Fig. 4A). Furthermore, the dual vector induced higher IFN-α, IFN-β, ISG15, and MxA production (Supporting Fig. 4B-D) as well as lower TGF-β and IL-10 (Supporting Fig. 4B). Compared to the ssRNA vector, the dual vector induced higher IFN-α and IFN-β, and we speculated that this was due to shRNA-mediated interruption of HBV-induced immune inhibition. Further evidence from human primary HBV+ HCC patient cells supported this idea, as dual vector also more efficiently inhibited HBV replication than shRNA, and recovered hepatocyte-intrinsic innate response by inducing more IFN-α and IFN-β secretion while less IL-10 and TGF-β expression than ssRNA (Supporting Fig. 5).

To determine whether dual vector also mediated a therapeutic effect in vivo, the dual, shRNA, ssRNA, and pSIREN vectors were hydrodynamically injected into HBV+ mice (Fig. 2A). Although both dual and shRNA vectors significantly inhibited HBV replication, dual vector exerted stronger inhibitory effects on HBx mRNA, HBx protein, HBV DNA, HBsAg, HBeAg, and HBcAg in hepatocytes, serum, or liver tissue (Fig. 2B-E). These results suggest that the added immunostimulation function aids the dual vector in more strongly inhibiting HBV replication and transcription than HBx silencing alone. Moreover, the HBV suppressive effect of dual functional vector lasted for at least 6 months after treatment without inducing liver injury (Fig. 2F). Meanwhile, although both dual and ssRNA vectors induced IFN-α and -β mRNA expression in hepatocytes from HBV+ mice, the dual vector induced significantly more than ssRNA (Fig. 3A). Similarly, systemic serum IFN-α and -β protein levels in the dual-treated group were higher than in the ssRNA-treated group (Fig. 3B). Dual vector also more effectively reduced inhibitory mediators in hepatocytes, such as immunosuppressive cytokines (Fig. 3C) and surface PD-L1 expression (Fig. 3D), compared to ssRNA or shRNA vector alone. These results strongly suggest that the ssRNA-HBx-shRNA dual vector powerfully inhibits HBV replication and successfully reverses HBV-induced hepatocyte-intrinsic immunotolerance.

Figure 2.

In vivo inhibition of HBV replication by a therapeutic immunostimulatory HBx-shRNA vector. (A) HBV-persistent mice were generated. Four weeks later, pSIREN, ssRNA, shRNA, and dual (6 μg) vectors were injected intravenously into HBV+ mice. (B,C) Two weeks later, mouse primary hepatocytes were isolated by the liver perfusion method. HBx mRNA (B) and protein (C) levels in hepatocytes were detected by qRT-PCR and western blot. (D) Serum HBV DNA, HBsAg, and HBeAg were detected by qPCR and ELISA, respectively. (E) HBsAg and HBcAg expression in liver tissue was detected by immunohistochemical staining. Original magnification: 100×. (F) Six months after vector therapy, serum levels of HBsAg and ALT were detected by ELISA. Data are representative of three independent experiments with six mice per group. *P < 0.05, versus shRNA transfection group.

Figure 3.

In vivo reversal of liver cell-intrinsic immunotolerance by a therapeutic dual-function immunostimulatory HBx-shRNA vector. HBV-persistent mice were generated. Four weeks later, pSIREN, ssRNA, shRNA, and dual (6 μg) vectors were injected intravenously into HBV+ mice. Two weeks later, (A,B) IFN-α and -β mRNA in hepatocytes were detected by qRT-PCR (A), and serum IFN-α and -β proteins were detected by ELISA (B). (C) IL-10 and TGF-β mRNA in hepatocytes were detected by qRT-PCR. (D) The primary hepatocytes were separated by the liver perfusion method and then the percentage of PD-L1+ hepatocytes was detected by FACS. Data are representative of three independent experiments with six mice per group. *P < 0.05, **P < 0.01, versus pSIREN transfection group; #P < 0.05, ##P < 0.01, versus ssRNA or shRNA transfection group.

Anti-HBV Recall Immunity Is Recovered After Reversing Hepatocyte-Intrinsic Tolerance.

As HBV persistence can paralyze systemic immune responses, we explored whether reversing hepatocyte-intrinsic immunotolerance recovers efficient adaptive immunity. As shown in Fig. 1A, HBV+ mice lose the ability to produce anti-HBs Ab in response to subcutaneous HBV vaccine. Upon dual or single vector treatment in HBV+ mice after subcutaneous vaccination, shRNA and dual vectors significantly inhibited serum HBsAg levels, indicating HBV replication was inhibited (Fig. 4A). More important, the treated mice regained anti-HBs Ab production, especially after dual-vector treatment (Fig. 4A), showing that dual vector enhanced anti-HBs Ab production after vaccination almost up to Ab levels in HBV mice (Fig. 1). Furthermore, to observe whether dual-vector treatment recovered recall responses in HBV-persistent mice, pAAV/HBV1.2 plasmid was hydrodynamically injected into shRNA or dual-vector-treated HBV+ mice. Interestingly, only dual vector treatment generated the highest serum anti-HBs Ab and lowest serum HBsAg (Fig. 4B), indicating that reversing hepatocyte-intrinsic immunotolerance may recover anti-HBV adaptive immunity and reverse the HBV viral persistence.

Figure 4.

Recovery of anti-HBV adaptive immunity after reversing liver cell-intrinsic tolerance. Plasmid pAAV/HBV1.2 (8 μg) was hydrodynamically injected intravenously into 4-5-week-old C57BL/6 mice to generate an HBV-carrier mouse model. Four weeks later, pSIREN, ssRNA, shRNA, or dual (6 μg) vectors combined with HBV vaccine (rHBs/CFA) were injected subcutaneously into HBV+ mice. At 3 and 7 days, rHBs/CFA was repeatedly injected. Two weeks later, serum HBsAg and HBsAb was detected by ELISA. (B) Plasmid pAAV/HBV1.2 (8 μg) was hydrodynamically injected intravenously into 4-5-week-old C57BL/6 mice to generate an HBV-carrier mouse model. Four weeks later, pAAV/HBV1.2 plasmid combined with pSIREN, ssRNA, shRNA, or dual (6 μg) vectors was reinjected intravenously into HBV+ mice. Two weeks later, serum HBsAg and HBsAb levels were detected by ELISA. Data are representative of three independent experiments with six mice per group. *P < 0.05, **P < 0.01, versus pSIREN transfection group; #P < 0.05, ##P < 0.01, versus ssRNA transfection group.

We next explored whether dual-vector therapy could restore CTL responses by detecting the percentages and activation status of liver-resident NK, CD4+ T, and CD8+ T cells in HBV+ mice. While no significant changes were detected in hepatic CD4+ T, NK and NKT cells (Supporting Fig. 6), the percentage and absolute number of hepatic CD8+ T cells and activating CD8+ T cells were markedly up-regulated by dual-vector administration (Fig. 5A,B). Furthermore, dual-vector treatment markedly augmented CTL function (e.g., CD107a+ and IFN-γ+ CD8+ T-cell percentages increased) in liver compared to ssRNA vector (Fig. 5C), along with down-regulation of PD-1 and up-regulation of CD28 (Fig. 5D), suggesting that dual therapy reverses CD8+ T-cell anergy by regulating the PD-1/CD28 axis. Moreover, the percentages of HBV-specific hepatic CD8+ T cells as well as HBV-specific CD107a+ and IFN-γ+ CD8+ T cells were also significantly enhanced by dual-vector treatment (Fig. 5E). In addition, we also observed the T-cell responses in spleen and peripheral blood and found similar results in line with the microenvironment. As shown in Supporting Fig. 7, the percentages and absolute numbers of CD8+ T cells as well as activated CD8+ T cells in both spleen and peripheral blood increased after dual vector therapy, suggesting that the restoring of CD8+ T-cell response exerted by dual vector is systemic.

Figure 5.

Functional recovery of hepatic exhausted CD8+ T cells after reversing hepatocyte-intrinsic immune tolerance. (A,B) pSIREN, ssRNA, shRNA, and dual (6 μg) vectors were injected intravenously into HBV-persistent mice. Two weeks later, the absolute numbers and percentages of hepatic CD8+ T cells and CD69+ CD8+ T cells were detected by FACS. (C,D) The percentages of CD107a+, IFN-γ+ (C), PD-1+, and CD28+ T cells (D) in hepatic CD8+ T cells were detected by FACS. Data are representative of three independent experiments with 6 mice per group. (E) The percentages of HBV-specific, CD107a+ HBV-specific, and IFN-γ+ HBV-specific T cells within hepatic CD8+ T cells were detected by FACS. Data are statistics from three independent experiments with six mice per group. *P < 0.05, **P < 0.01, versus pSIREN transfection group; #P < 0.05, ##P < 0.01, versus ssRNA transfection group.

To further explore the role of CD8+ T cells, we depleted them from HBV+ mice with an anti-CD8β mAb. CD8+ T cells, but not CD4+ T, are critical for dual-vector-mediated inhibition of HBV replication, as HBV DNA copies, HBx expression, and HBsAg levels in serum became significantly higher, while serum IFN-γ levels declined, after CD8+ T-cell depletion (Fig. 6A,B). NK cell-depletion also partly attenuated dual-vector-mediated inhibition of HBV, suggesting that NK cells might also participate in dual-vector-mediated HBV inhibition. But the role of CD8+ T cells was the most prominent. To determine the role of CD8+ T cells, we adoptively transferred splenic CD8+ T cells or CD8+ T-cell-depleted splenic lymphocytes from wild-type (WT) mice into HBV+ Rag-1−/− mice. Adoptively transferring CD8+ T cells but not non-CD8+ T cells or IFN-γ−/− CD8+ T-cells (e.g., CD8+ T cells from IFN-γ−/− mice) inhibited HBV replication (serum HBsAg) in HBV-persistent Rag-1−/− mice when dual vector treatment was used (Fig. 6C). Although dual vector promoted CD8+ T cell activation, HBsAg inhibition was markedly impaired in HBV-persistent IFN-γ−/− mice (the inhibitory rate is 85.16 ± 4.75% and 55.24 ± 10.43% in WT HBV+ and IFN-γ−/− HBV+ mice, respectively) (Fig. 6D). These results suggest that recovering anti-HBV adaptive immunity by reversing hepatocyte-intrinsic tolerance is at least partially dependent on functional rescue by IFN-γ-producing CD8+ T cells.

Figure 6.

Recovery of hepatic exhausted CD8+ T cells is critical in reversing HBV-induced immunotolerance. (A,B) Depleting antibodies (α-CD8β, α-CD4, α-NK1.1) were injected intraperitoneally into HBV+ mice, respectively. Three days later, pSIREN, ssRNA, shRNA, and dual (6 μg) vectors were administered intravenously, and HBV DNA and HBx mRNA in hepatocytes were detected by qPCR 2 weeks later (A), and serum HBsAg (A) and IFN-γ were detected by ELISA (B). (C) Plasmid pAAV/HBV1.2 (8 μg) was injected intravenously into Rag-1−/− mice (4-5 weeks old) to generate an HBV-persistent mouse model. Splenic CD8+ T cells, or CD8-depleted splenic lymphocytes (4 × 106) isolated from WT C57BL/6 mice were transferred into HBV-persistent Rag-1−/− mice. Then pSIREN, ssRNA, shRNA, and dual (6 μg) vectors were injected intravenously. Three days later, serum HBsAg was detected. Data represent three independent experiments with 6 mice per group (A,C) or are expressed as mean ± SD from at least three separate experiments (B). *P < 0.05, **P < 0.01, versus dual transfection group; #P < 0.05, ##P < 0.01, versus CD8+ T+dual group. (D) Plasmid pAAV/HBV1.2 (2 μg) was injected intravenously into 4-5-week-old IFN-γ−/− mice to generate an HBV-carrier mouse model. Then pSIREN, ssRNA, shRNA, and dual (6 μg) vectors were injected intravenously. Serum HBsAg was detected. The inhibitory rate of HBsAg was calculated as: {(Level of HBsAgpSIREN − Level of HBsAgdual) / Level of HBsAgpSIREN}×100%. Data are representative of three independent experiments with six mice per group.

Type I IFN Is Important for Reversing CD8+ T-Cell Tolerance in a TLR7-Dependent Manner.

Our data showed that dual vector reversed HBV-inhibited type I IFN production and the subsequent function in HBV+ hepatocytes (Supporting Fig. S4B-D; Fig. 3A,B). Here, we further observed that blocking type I IFN signaling in vivo with a neutralizing antibody against the IFN-α/β receptor partially attenuated the dual-vector-mediated inhibition of HBV replication (Fig. 7A,B). Furthermore, when CD8+ T cells from type I IFN receptor (IFNAR−/−)-deficient mice were adoptively transferred into HBV-carrier Rag-1−/− mice, the HBV inhibition was attenuated in dual vector treatment (Fig. 7C). Type I IFN signal blockade also significantly reduced the recover of the exhausted CD8+ T cells by expression of CD69, CD28, and IFN-γ (Fig. 7D). Notably, the HBV-specific CD8+ T cells and anti-HBs responses also significantly decreased (Fig. 7E,F). These data suggest that type I IFN signaling is required for recovering CD8+ T-cell function and HBV clearance after dual-vector-reversed hepatocyte-intrinsic tolerance.

Figure 7.

Type I IFN is critical in recovering exhaustion of CD8+ T cells. (A,B) Anti-IFNR Ab blockade was administered intraperitoneally into HBV+ mice. One day later, pSIREN, ssRNA, shRNA, and dual (6 μg) vectors were administered intravenously. Two weeks later, HBx expression in hepatocytes (A) and serum HBsAg (B) were detected. (C) The pAAV/HBV1.2 plasmid (8 μg) was injected intravenously into Rag-1−/− mice (4-5 weeks old) to generate an HBV-persistent mouse model. Splenic CD8+ T cells (4 × 106) isolated from C57BL/6 mice or IFNAR−/− mice were transferred into Rag-1−/− HBV carrier mice. Then pSIREN, ssRNA, shRNA, and dual (6 μg) were injected intravenously. Two weeks later, serum HBsAg was detected. (D-F) Anti-IFNR Ab blockade was administered intraperitoneally into HBV+ mice. One day later, pSIREN, ssRNA, shRNA, and dual (6 μg) vectors were administered intravenously. Two weeks later, the percentages of CD69+, CD28+, and IFN-γ+ CD8+ T cells (D), as well as HBV-specific CD8+ T cells were detected by FACS (E); serum levels of anti-HBs Ab were detected by ELISA (F). Data are representative of three independent experiments with six mice per group. *P < 0.05, **P < 0.01, versus dual transfection group.

Since U-rich ssRNA sequences can function as TLR7/8 ligands, we further determined the mechanism underlying how innate ssRNA recognition leads to increased CD8+ T-cell activation during dual vector treatment. Both dual and ssRNA vectors promoted TLR7 mRNA and protein expression, while TLR3 expression was not affected in HepG2.2.15 cells (Fig. 8A,B). Similar up-regulation of TLR7 protein expression by dual and ssRNA vectors was also observed in murine primary hepatocytes (Fig. 8C). TLR7-siRNA knockdown attenuated dual-vector-mediated HBV inhibition and exhibited lower IFN-α production (Fig. 8D). This was further confirmed using the TLR7 inhibitor IRS661,15 showing that IRS661 significantly reduced serum IFN-α and -β production (Fig. 8E) and attenuated CD8+ T-cell activation (Fig. 8F). More important, the HBV-specific CD8+ T cells and anti-HBs responses significantly decreased (Fig. 7G), and HBV clearance was markedly impaired (Fig. 8H). These data suggest that TLR7 is required for type I IFN (and other inflammatory cytokine) production after dual-vector treatment, leading to recovery of CD8+ T-cell and humoral immunity by reversing HBV-induced hepatocyte-intrinsic immune tolerance.

Figure 8.

The TLR7 pathway is critical in type I IFN-dependent recovery of exhausted CD8+ T cells. (A,B) HepG2.2.15 cells were transfected with pSIREN, ssRNA, shRNA, or dual plasmid (1 μg/mL) for 24 hours. TLR3 and TLR7 mRNA (A) and intracellular protein (B) were detected. (C) HBV-persistent mice were generated. pSIREN, ssRNA, shRNA, and dual (6 μg) vectors were injected intravenously into HBV+ mice. Two weeks later, primary hepatocytes were isolated by the liver perfusion method. The expression level of TLR7 in hepatocytes was detected by FACS. (D) HepG2.2.15 cells were transfected by dual (1 μg/mL) or si-TLR7+dual for 24 hours, and HBx, TLR7, and IFN-α mRNAs were measured. (E-H) Dual vector or TLR7 inhibitor ISR661+dual was administered intravenously into HBV+ mice, and serum levels of IFN-α and IFN-β (E), the percentages of hepatic CD28+ or CD69+CD8+ T cell (F), HBV-specific CD8+ T cells (G), the serum levels of anti-HBs Ab (G) and HBsAg (H) were detected. Data are expressed as mean ± SD from at least three separate experiments (A,D,E) or representative of three independent experiments with six mice per group (B,C,F-H). *P < 0.05, **P < 0.01, versus dual transfection group.

Discussion

Accumulating evidence suggests that HBV infection induces host immunotolerance.7, 8 Persistent HBV infection sustains suppression of antiviral immunity, and high HBV titers or particle load can inhibit innate or adaptive immune response activation, particularly innate PRRs (like TLR7) and their downstream signals in hepatocytes. For example, HBx, HBeAg, and even virion particles can directly suppress RIG-I-mediated innate immunity and inhibit antiviral protein expression (such as MxA) as well as type I IFN induction.4 HBV persistence also increases immunosuppressive cytokines like TGF-β and IL-10. Importantly, HBV impairs the antiviral function of hepatic lymphocytes, especially of CD8+ T cells in the adaptive immune response. HBV can up-regulate inhibitory PD-1 expression on CD8+ T cells, contributing to functional impairment and apoptosis that leads to CD8+ T-cell exhaustion and immune tolerance. In the present study, we demonstrated that the dual-function vector not only suppressed HBV replication by silencing HBx, which reduced HBV load, but also enhanced TLR7-mediated type I IFN and immunostimulatory cytokine production while inhibiting immunosuppressive cytokines, which helped to recover adaptive immunity and further promoted the inhibition of HBV replication.

By comparing the effects of ssRNA, shRNA, and ssRNA-shRNA dual functional vector on HBV replication or immune stimulation, we found that ssRNA-shRNA dual functional vector was the strongest inhibitor of HBV replication and the most efficient stimulator of innate immunity both in vitro and in vivo. The dual functional vector was more efficient to stimulate type I IFN response than ssRNA, possibly because the relief of HBV-induced immune tolerance after direct HBV load decreased. It exerted a stronger inhibitory effect on HBV replication than shRNA, possibly due to its arousing of IFN response by activation of the TLR7 pathway. We propose that both HBx silence and ssRNA-induced TLR7 activation contribute to the reversal of the HBV-intrinsic immune tolerance by a dually functional vector. Furthermore, dual-vector therapy reestablished HBV-specific CD8+ T-cell and anti-HBs Ab responses as well as effectively cleared HBV infection, demonstrating that reversing cell-intrinsic immunotolerance by this therapy further promotes the recovery of anti-HBV adaptive immunity. To our knowledge, this is the first report to propose that systemic immunotolerance can be overcome by reversing hepatocyte-intrinsic tolerance.

Up to now, there has been no ideal mouse model that can mimic HBV natural infection. Although several lines of transgenic mice expressing either HBs gene or full-length HBV genome have been established, the immune system in these HBV-transgenic mice is inherently tolerant to transgene products; HBV replication is generated from the integrated HBV sequence harbored in all hepatocytes, which is clearly different from the natural HBV persistence. So these HBV-Tg mice cannot be used to study peripheral tolerance. The mouse model established by a single hydrodynamic injection of pAAV/HBV1.2 DNA into the tail veins of C57BL/6 mice is the first good model to observe the immune response/tolerance in immunocompetent mice, in which HBV surface antigenemia persists for >6 months, and viral replication intermediates, transcripts, and proteins present in liver tissues for up to 1 year.8 Moreover, there was no neutralizing anti-HBs antibody production after HBsAg/CFA vaccination, suggesting the generation of tolerance. The characteristics of this mouse model for HBV persistence are thought comparable to those of human chronic HBV infections in the immune tolerant stage. A series of studies have used this model to research HBV persistence.8, 16-18 In the present study, we also displayed the chronic HBV infection and immune tolerant status in this model and further demonstrate that the expression of PD-1 was higher but CD69 and IFN-γ was lower in hepatic CD8+ T cells, which is in line with that in CHB patients. We think that this mouse model could at least partly mimic chronic HBV infection, supplying a tool to study the strategy of how to overcome immune tolerance in HBV chronic infection.

The liver is relatively immunotolerant, and previous evidence has established that this can lead to systemic immunotolerance.19 For example, allogeneic liver grafts are more easily accepted than other organ transplants with less host rejection. Interestingly, kidney transplant survival is enhanced when liver is also transplanted from the same donor.20 Ectopic expression of the neural autoantigen myelin basic protein (MBP) in the liver protects mice from experimental autoimmune encephalomyelitis (EAE) by inducing hepatic tolerance and generating MBP-specific T-regulatory cells (Tregs).21 One possible explanation is that the liver is a crossroads for systemic circulation, where the open architecture of the sinusoids allows direct and sufficient contact between circulating naïve T cells and diverse subsets of hepatic antigen-presenting cells (APCs), including hepatocytes. Based on our data, we are the first to propose that reversing hepatocyte-intrinsic immunotolerance by dually functional immunostimulatory HBx-shRNA therapy can induce the recovery of systemic immunotolerance. Notably, this HBV-induced cell-intrinsic and systemic immunotolerance is HBV-specific, for no immune tolerance was observed to non-HBV challenge, for example, LCMV (Fig. 1H).

The characteristic liver immunotolerance derives from its unique immunosuppressive microenvironment, including the presence of TGF-β and IL-10 as well as a diverse repertoire of liver-resident APCs, such as DCs, Kupffer cells, liver sinusoidal endothelial cells (LSECs), stellate cells, and hepatocytes,22 that characteristically express low MHC class II and costimulatory molecules, high coinhibitory molecules (such as PD-L1), and secrete TGF-β and IL-10. T-cell priming by hepatic APCs typically leads to T-cell immunotolerance or apoptosis.22 Under steady-state conditions, hepatocytes mainly function as tolerogenic APCs; persistent HBV infection further enhances this effect. In the present study, we showed that HBV-mediated immune tolerance could be induced in both hepatocytes and HBV-carrier mice. Dual-function therapy abrogates this hepatocyte-intrinsic immune tolerance, possibly by switching hepatocyte function from tolerogenic to immunogenic for antigen presentation, thus leading to increased T-cell immunity and HBV clearance. The increased type I IFN and decreased TGF-β and IL-10 might also alter the inhibitory liver microenvironment. Moreover, the dual vector-induced type I IFN production by hepatocytes is essential for CD8+ T-cell activation, anti-HBs response, and HBV inhibition (Fig. 7). We propose that hepatocyte-produced type I IFNs not only directly promote CD8+ T-cell activation by way of triggering the type I IFN receptor on its surface, but also promotes maturation of other APCs as well as IL-12 production important for Th1 CD4+ T-cell priming and B cells involved in the antiviral response. The direct effect of type I IFN on CD8+ T cells also occurs in other viral infections.23 Type I IFNs also promote IL-15 production, which is a CD8+ T-cell survival factor24 and plays a critical role in stimulating and maintaining memory CD8+ T cells.23 Type I IFN and IL-12 are also involved in memory CD8+ T-cell development.25 Whether dually functional vector influences HBV-specific CD8+ memory T-cell generation needs further investigation.

CD8+ T cells play a critical role in HBV clearance.3 Intrahepatic HBV-specific CD8+ T cells are required for rapid viral clearance during acute HBV infection. Several HBV-specific CD8+ T-cell cytokines, such as IFN-γ and TNF-α, are essential for suppressing HBV gene expression and replication.3 However, in CHB patients the inhibitory PD-1 receptor is up-regulated on both peripheral blood mononuclear cells (PBMCs) and intrahepatic lymphocytes, particularly on HBV-specific CD8+ T cells, where PD-1 interaction with PD-L1 ligand on APCs results in functional suppression and apoptosis of CD8+ T cells.26 This loss of function is known as “T-cell exhaustion” that is common during persistent viral infection, including human immunodeficiency virus (HIV), HBV, and HCV infection. In the present study, we observed that immunotolerance in chronic HBV-carrier mice correlated with decreased hepatic CD8+ T-cell percentage and number (Fig. 1D), including IFN-γ+ and HBc-specific CD8+ T cells (Fig. 1F,G), as well as higher PD-1 expression (Fig. 1E) and augmented serum TGF-β and IL-10 (Fig. 1C). Treatment with ssRNA-shRNA dually functional vector reversed this immunotolerance. More important, CD8+ T cells were necessary for dual-vector-mediated HBV inhibition (Fig. 6A-C). In addition, NK cells may also be involved in dual-vector-mediated HBV inhibition, for NK cell-depletion partly attenuated dual-vector-mediated HBV suppression (Fig. 5A,B). Increasing data have shown that NK cells play an important role in clearance of HBV, especially in the early stage of infection. HBV persistence impairs NK cell function possibly by elevated TGF-β1 through down-regulation of NKG2D/DAP10 and 2B4/SAP expression.27 The dual-vector therapy might release the impairment of NK cell function by both reducing HBV load and arousing activation stimulated by increased type I IFN. The decreased TGF-β1 after dual-vector therapy may also contribute to restoring NK cell function.

PRRs play critical roles in defense against invading pathogens by way of recognition of pathogen-associated molecular patterns (PAMPs). Therapeutic strategies that incorporate both PRR activation and target-gene silencing have promising potential to treat cancer and viral infections. Poeck et al.28 first reported that bifunctional 5′-triphosphate-siRNA (3p-siRNA) targeting Bcl2 both silenced Bcl2 and stimulated innate immunity by way of RIG-I activation. This reagent displayed effective antitumor activity by promoting apoptosis of B16 melanoma, while simultaneously inhibiting lung metastasis. Later, this therapy was shown to exhibit beneficial therapy for chronic HBV infection.9, 10 TLR7 and TLR8 on endosomes recognize primarily U- or GU-rich ssRNA and transduce signals through MyD88, IRF7, NF-κB, mitogen-activated protein kinase (MAPK), and other signaling pathways that stimulate type I IFN and proinflammatory cytokine production.11, 12 TLR7 activation is important in generating anti-HBV responses and is impaired by persistent HBV infection.5 Similarly, TLR7 is essential to eliminate persistent LCMV infection in mice by generating antiviral adaptive immunity.29 Similar to the bifunctional 3p-siRNA, Gantier et al.30 designed immunostimulatory siRNAs by introducing a microRNA-like nonpairing uridine-bulge in the passenger strand that enhanced protection against Semliki Forest virus (SFV). Khairuddin et al.13 reported that siRNA-induced immunostimulation through TLR7 promoted antitumor activity against HPV-driven tumors in vivo, even independent of the gene-silencing effect. In the present study, we showed that both the immunostimulatory ssRNA and dually functional vectors significantly induced IFN-α and -β production (Fig. 3B). With this added immunostimulation function, the dual functional vector exerted more efficient HBV inhibition than shRNA vector alone (Fig. 2; Supporting Figs. 1-3). Moreover, the HBV suppressive effect of dual functional vector lasted for at least 6 months after treatment without inducing liver injury (Fig. 2F). And the dual functional vector treatment could prevent HBV-carrier mice from HBV reinfection (Fig. 4B). This suggests that the dually functional vector could efficiently clear HBV and reverse HBV viral persistence. To our knowledge, this is the first report showing that a bifunctional ssRNA-shRNA vector inhibits and clears HBV replication through both potent HBx-gene silencing and TLR7-dependent immunostimulation. This bifunctional therapeutic strategy that breaks adaptive immunotolerance by reversing cell-intrinsic immunotolerance to successfully clear HBV infection shows promise for treating other persistent viral infections (such as HCV and HPV) and associated cancers, including HCC.

Acknowledgements

We thank Pei-Jer Chen for kindly providing pAAV/HBV1.2 plasmid.

Ancillary