Toll-like receptor-mediated control of HBV replication by nonparenchymal liver cells in mice†
Article first published online: 10 OCT 2007
Copyright © 2007 American Association for the Study of Liver Diseases
Volume 46, Issue 6, pages 1769–1778, December 2007
How to Cite
Wu, J., Lu, M., Meng, Z., Trippler, M., Broering, R., Szczeponek, A., Krux, F., Dittmer, U., Roggendorf, M., Gerken, G. and Schlaak, J. F. (2007), Toll-like receptor-mediated control of HBV replication by nonparenchymal liver cells in mice. Hepatology, 46: 1769–1778. doi: 10.1002/hep.21897
Potential conflict of interest: Nothing to report.
- Issue published online: 28 NOV 2007
- Article first published online: 10 OCT 2007
- Manuscript Accepted: 25 JUN 2007
- Manuscript Received: 19 MAR 2007
- Deutsche Forschungsgemeinschaft. Grant Numbers: SCHL 377/2-2, Lu 669/2-1, GRK1045/1, Lu 669/5-1
Hepatitis B virus (HBV) infection is one of the most frequent causes of chronic liver disease worldwide. Because recent studies have suggested that Toll-like receptor (TLR)-based therapies may be a promising approach in the treatment of HBV infection, we studied the role of the local innate immune system of the liver as a possible mediator of this effect. Murine nonparenchymal cells, including Kupffer cells (KCs) and sinusoidal endothelial cells (LSECs), were isolated from C57/BL6 wild-type or MyD88−/− mice and stimulated by agonists of TLR1 to TLR9. Supernatants were harvested and assayed for their antiviral activity against HBV in HBV-Met cells. No direct antiviral effect of TLR agonists could be observed. In controls (myeloid dendritic cells), TLR1, TLR3, TLR4, TLR7, and TLR9 activation lead to production of antiviral cytokines. By contrast, only supernatants from TLR3-stimulated and TLR4-stimulated KCs and TLR3-stimulated LSECs from wild-type mice were able to potently suppress HBV replication as assessed via Southern blotting. Similar results were found with cells from MyD88−/− mice, indicating that the effect was independent of this signaling pathway. Cellular HBV RNA and hepatitis B surface antigen or hepatitis B e antigen levels in supernatants remained unchanged. Using neutralizing antibodies, we demonstrated that the TLR3-mediated effect but not the TLR4-mediated effect is mediated exclusively through interferon-β. Conclusion: Our data indicate that the innate immune system of the liver can control HBV replication after activation by TLR agonists. This has implications for the development of TLR-based therapeutic approaches against HBV. (HEPATOLOGY 2007.)
Hepatitis B virus (HBV) infection is among the most common causes of acute and chronic hepatitis worldwide and can lead to liver cirrhosis, hepatic decompensation, and hepatocellular carcinoma.1 Currently, only interferon (IFN)-α and nucleoside or nucleotide analogues have been shown to be effective in suppressing HBV replication and in inducing clinical remission of liver disease.2 However, increasing evidence suggests that immune mediators such as Toll-like receptor (TLR) ligands could be used successfully as therapeutic agents in HBV3 or other viral infections.4–6
TLRs play a crucial role in early host defense by recognizing so-called pathogen-associated molecular patterns that are essential for the survival of the microorganism but are not present in eukaryotes.7 To date, 13 mammalian TLRs have been identified (10 in humans and 13 in mice), each containing a unique extracellular domain and a conserved cytoplasmic Toll/interleukin (IL)-1 receptor domain.8 After ligand binding, the cytoplasmic Toll/IL-1 receptor domain of TLRs associates with intracellular adapters and activates downstream signaling molecules, including the transcription factors nuclear factor κB, interferon regulatory factor-1, interferon regulatory factor-7, and mitogen-activated protein kinases, which lead to the activation of type I IFNs, proinflammatory cytokines, or costimulatory molecules.9 Five adapter proteins have been discovered, including MyD88 (myeloid differentiation primary-response protein 88), TIRAP (Toll/IL-1 receptor domain–containing adapter protein), TRIF (Toll/IL-1 receptor domain–containing adapter protein inducing IFN-β), TRAM (TRIF-related adapter molecule), and SARM (sterile alpha and HEAT-Armadillo motifs).9, 10 MyD88 is common to all TLRs except for TLR3,11 whereas TRIF is specifically involved in TLR3-mediated and TLR4-mediated signaling pathways that lead to the activation of IFN-β and subsequent expression of IFN-inducible genes in a MyD88-independent manner.12
Nonparenchymal cells (NPCs), including Kupffer cells (KCs) and sinusoidal endothelial cells (LSECs), play an indispensable role in pathogen clearance and innate and early adaptive immune responses.13, 14 It is widely accepted that the clearance of HBV virions in the incubation phase prior to the onset of clinical symptoms is induced by antiviral cytokines that are produced by cells of the innate and early adaptive immune system.15
Previously, it was shown that HBV replication could be inhibited in HBV transgenic mice by administration of a single intravenous injection of ligands specific for either TLR3, TLR4, TLR5, TLR7, or TLR9 within 24 hours in a type I IFN–dependent manner.3 The same article proposed that in vivo TLR ligands activate an antiviral program in dendritic cells but not in hepatocytes. We investigated whether TLR agonists can directly or indirectly (via NPCs) suppress HBV replication using the HBV-Met cell system. Our data indicate that NPCs can be activated by TLR3 and TLR4 to produce IFN-β and other mediators that can potently suppress HBV replication.
Materials and Methods
MyD88−/− mice were generously provided by Dr. S. Akira (Osaka University, Osaka, Japan).16 A homogenous population was established by back-crossing MyD88−/− mice to C57BL/6 mice for more than 6 generations. Inbred C57BL/6 wild-type and MyD88−/− mice (8-10 weeks old) were kept in the animal facilities at the University Hospital of Essen. All animals received humane care according to the criteria outlined in the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and published by the National Institutes of Health.
Reagents and Antibodies.
Agonists for TLR1/2 (palmitoyl-3-cysteine-serinelysine-4), TLR2 (heat-killed Listeria monocytogenes), TLR4 [lipopolysaccharide (LPS) from Escherichia coli 011:B4 strain], TLR5 (Salmonella typhimurium flagellin), TLR6/2 [S-(2,3-bispalmitoyloxypropyl)-Cys-Gly-Asp-Pro-Lys-His-Pro-Lys-Ser-Phe], TLR7 (single-stranded RNA 40), TLR8 (Gardiquimod), and TLR9 (CpG oligonucleotides) were purchased from InvivoGen (San Diego, CA). The TLR3 agonist polyinosine-polycytidylic acid (poly I:C) was obtained from Amersham Biosciences (Buckinghamshire, England). Neutralizing anti–IFN-α and anti–IFN-β rabbit antibodies were purchased from Calbiochem (Darmstadt, Germany). Neutralizing anti–IFN-γ rat antibodies and anti–tumor necrosis factor-α (TNF-α) rat antibodies were obtained from eBioscience (San Diego, CA). Liberase Blendzymes (mixtures of highly purified collagenase and neutral protease enzymes) were obtained from Roche (Mannheim, Germany). Cycloheximide (CHX) was purchased from Fluka (Seelze, Germany).
Isolation and Culture of Nonparenchymal Cells.
Isolation of murine NPCs was performed as described previously.17 Briefly, livers were perfused through the portal vein with 0.07 mg/mL Liberase Blendzymes in a calcium-free phosphate buffer. Livers were mechanically disrupted and incubated with 0.07 mg/mL Liberase Blendzymes in Grey's balanced saline solution for 20 minutes at 37°C with constant rotation (200 rpm). The fraction of NPCs was recovered via density centrifugation on a 30% metrizamide gradient (Nycomed, Oslo, Norway) at 400g (1410 rpm). LSECs were isolated using Anti-LSEC MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany), and separation of KCs was achieved via counterflow centrifugal elutriation. A J2-MC centrifuge (Beckman, Munich, Germany) with a JE-6B rotor and a standard elutriation chamber was employed. The rotor speed was kept constant at 2500 rpm (602g), and KCs were collected at a flow rate of 55 mL/minute. LSECs and KCs were seeded onto collagen I–coated 6-well plates (BD Biosciences. Bedford, MA) and cultured in RPMI 1640 medium (Gibco-BRL, Neu-Isenburg, Germany) supplemented with 10% fetal bovine serum (Gibco) and antibiotics (100 U/mL penicillin, and 100 μg/mL streptomycin, Gibco) at 37°C in the presence of 5% CO2. To induce cytokine production, LSECs and KCs were cultured for 20 hours at 1 × 106/2 mL in 6-well culture plates in the presence of 1 μg/mL palmitoyl-3-cysteine-serinelysine-4, 108cells/mL heat-killed Listeria monocytogenes, 100 μg/mL poly I:C, 10 μg/mL LPS, 1 μg/mL S. typhimurium flagellin, 1 μg/mL S-(2,3-bispalmitoyloxypropyl)-Cys-Gly-Asp-Pro-Lys-His-Pro-Lys-Ser-Phe, 10 μg/mL single-stranded RNA 40, 10 μg/mL Gardiquimod, 2 μM ODN1826. The cell supernatants were collected and kept at −80°C.
Primary Hepatocyte Culture.
Livers from 8-week-old C57BL/6 mice were perfused and collagenase digested (Liberase Blendzymes, Roche) as described previously.18 The cells were plated on collagen I–coated 6-well plates (BD Biosciences) in RPMI 1640 medium (Gibco) supplemented with 55 ng/mL epidermal growth factor (BD Biosciences), 16 ng/mL insulin growth factor II (Sigma, Munich, Germany), 10 μg/mL insulin (Sigma), 10% fetal bovine serum (Gibco), and antibiotics [100 U/mL penicillin and 100 μg/mL streptomycin (Gibco)] at 37°C in the presence of 5% CO2.
Isolation and Culture of Myeloid Dendritic Cells.
Myeloid dendritic cells (MDCs) were generated from single-cell bone marrow suspensions prepared from mouse femurs as described previously.19 Bone marrow cells were seeded at a density of 2 × 106 cells/10 mL in RPMI 1640 medium supplemented with 10% fetal bovine serum, streptomycin, penicillin, and 2-mercaptoethanol (Sigma) in the presence of 5 ng/mL recombinant granulocyte–monocyte colony-stimulating factor and 10 ng/mL IL-4. At day 8, MDCs were purified with anti-CD11c magnetic microbeads (Miltenyi Biotec).
HBV-Met Cell Culture.
The HBV-Met cell line (kindly provided by S. Wieland and F. Chisari, La Jolla, CA) is an immortalized hepatocyte cell line established from HBV-Met double-transgenic mice that is produced by mating homozygous 1.3.46 mice with homozygous cyto-Met transgenic mice. The HBV-transgenic mouse lineage 1.3.46 contains a terminally redundant, 1.33× genome length transgene (ayw subtype) that starts just upstream of the X promoter and enhancer I and ends just downstream of the unique HBV polyadenylation site.20 Cyto-Met transgenic mice express the constitutively active cytoplasmic domain of the human hepatocyte growth factor receptor (c-Met) in their livers. HBV-Met cells were cultivated as described previously.21
Analysis of HBV Replicative Intermediates and HBV RNA.
Analysis of HBV replicative intermediates and HBV RNA was performed as described previously.22
Hepatitis B Surface Antigen and Hepatitis B e Antigen Chemiluminescent Microparticle Immunoassay.
Levels of hepatitis B surface antigen (HBsAg) and hepatitis B e antigen (HBeAg) in cell supernatants were determined using the Architect system/Abbott HBsAg and HBeAg chemiluminescent microparticle immunoassay kit (Abbott, Wiesbaden, Germany) according to the manufacturer's instructions.
RNA Isolation and Real-Time PCR.
Total RNA was isolated from 1-10 × 106 cells using Trizol reagent (Gibco) following the manufacturer's protocol. One-step reverse-transcription PCR with real-time detection was performed with the QuantiTect SYBR Green RT-PCR Kit (Qiagen, Hilden, Germany) on the Rotor-Gene 2000 real-time amplification system (Corbett Research, Mortlake, Australia) according to the manufacturer's instructions. TLR1-TLR9 expression was detected via commercial Quantitec Primer Assays (Qiagen; primer sequences not available).
TLR-Mediated Elimination of HBV Replicative Intermediates from HBV-Met Cells by NPCs, Hepatocytes, and Myeloid Dendritic Cells.
HBV-Met cells were grown to confluence and then kept in complete medium supplemented with 1% dimethyl sulfoxide (DMSO) for 10 days, allowing HBV replication to reach steady-state levels. To determine whether TLR ligands can induce suppression of HBV replication, murine NPCs, hepatocytes, and MDCs were isolated from C57/BL6 wild-type mice and stimulated by TLR1-TLR9 agonists for 20 hours. The culture supernatants from stimulated and unstimulated NPC hepatocytes and MDCs were harvested after 2 days and assayed for their antiviral activity against HBV in HBV-Met cells via Southern blot analysis.
As shown in Fig. 1A, HBV replication is not reduced by treatment with supernatants from unstimulated NPCs, hepatocytes, or MDCs (no stimulation control) or TLR1-TLR9 ligands alone (no cell control). Therefore, we can rule out the possibility that the effects described below result from TLR1-TLR9 ligands remaining in the supernatant or from antiviral mediators secreted by unstimulated NPCs, hepatocytes, or MDCs. By contrast, HBV replicative intermediates are almost completely abolished by treatment with supernatants from TLR3-stimulated and TLR4-stimulated KCs (Fig. 1B) and TLR3-stimulated LSECs (Fig. 1C). Supernatants from TLR3-stimulated and TLR4-stimulated hepatocytes could significantly suppress HBV replication, whereas moderate suppression was seen with TLR1 and TLR6 (Fig. 1D). Similar results were obtained when the supernatants from TLR-stimulated MDCs were used. HBV replication was significantly suppressed by TLR3, TLR4, and TLR7, whereas moderate suppression was seen with TLR1 and TLR9 (Fig. 1E). Interestingly, as observed after IFN induction in HBV-transgenic mice23 or IFN treatment of HBV-Met cells,21 single-stranded DNA was almost completely abolished, whereas relaxed circular double stranded DNA was still detectable.
In preparation for these studies, a virus protection assay (encephalomyocarditis virus) was used to test the supernatants from experiments using different doses of stimuli (at least 4 concentrations of each TLR agonist; data not shown) and duration of stimulation (20 hours and 6 hours; data not shown). The experimental conditions were chosen based on these data. As shown in Fig. 1F-H, when the stimulation time was shortened from 20 hours to 6 hours in NPCs, similar results were obtained (Fig. 1F-G), with the exception that supernatants from TLR3-stimulated LSECs produce less antiviral activity (Fig. 1H).
To determine whether the lack of direct effects of TLR agonists on HBV replication may be due to the absence of TLR receptor molecules in HBV-Met cells, we assessed the expression of TLR1-TLR9 on the transcriptional level using quantitative reverse-transcription PCR. Our data indicate that messenger RNAs for all TLRs are detectable in HBV-Met cells with comparatively high (TLR2, TLR4, TLR5), medium (TLR1, TLR3, TLR6, TLR7) or very low (TLR8, TLR9) levels (Table 1). In addition, the data suggest that messenger RNAs for TLR1-TLR9 are present in murine KCs and LSECs (Table 1). To determine whether HBV-Met cells are totally unresponsive to TLR stimulation or simply lack the production of antiviral mediators against HBV (as shown in Fig. 1A), we assessed the production of IFN-α, IFN-β, TNF-α, and IL-6 on the transcriptional level after 6 hours of stimulation. Consistent with our functional data, TLR1-TLR9 stimulation does not induce the expression of IFN-α and IFN-β (data not shown); however, TLR1, TLR3, TLR4, TLR5, and TLR6 stimulation can upregulate TNF-α and IL-6 expression, whereas TLR7, TLR8, and TLR9 stimulation lead to modest suppression of these mediators (Fig. 1I). These findings imply that HBV-Met cells are not unresponsive to TLR stimulation but appear to lack the ability to produce type I IFNs in response to these stimuli. Whether this is an effect mediated by HBV or inherent to the cell line and not related to HBV is currently under investigation.
Supernatants from TLR-Stimulated KCs and LSECs Do Not Suppress HBV RNA Levels and HBsAg/HBeAg Secretion.
To investigate whether TLR ligands can also induce suppression of HBV gene expression and protein secretion, murine NPCs were isolated from C57/BL6 wild-type mice and stimulated by TLR1-TLR9 agonists for 20 hours. The culture supernatants from stimulated and unstimulated NPCs were harvested after 4 days and assayed for their antiviral activity against HBV in HBV-Met cells via northern blot analysis and HBsAg and HBeAg chemiluminescent microparticle immunoassay.
As shown in Fig. 2, the cellular 3.5-kb and 2.1-kb HBV transcript levels and secretion of HBsAg and HBeAg into the culture medium are not suppressed by treatment with supernatants from either TLR1-TLR9–stimulated NPCs, unstimulated NPCs (no stimulation control), or TLR 1-9 ligands alone (no cell control), respectively.
De Novo Protein Synthesis Is Required for TLR3-Mediated and TLR4-Mediated Suppression of HBV Replication by KCs and LSECs.
To test the role of protein synthesis in TLR-mediated elimination of HBV replicative intermediates, we treated NPCs with 3 μM CHX for 1 or 4 hours to prevent protein synthesis. We then stimulated the NPCs with 100 μg/mL poly I:C or 10 μg/mL LPS for 20 hours. The culture supernatants were added to differentiated HBV-Met cells for 2 days, and HBV gene replication in treated cells was analyzed via Southern blotting.
Our data show that the elimination of HBV replicative intermediates mediated by TLR3-stimulated and TLR4-stimulated KCs and LSECs is completely inhibited by pretreating cells with CHX for 4 hours (Fig. 3A, lanes 3 and 6; Fig. 3B, lane 4). Pretreatment for 1 hour had no effect (Fig. 3B, lane 3), indicating that CHX does not modulate HBV replication directly.
IFN-β Is the Main Cytokine Mediating the Antiviral Activity of TLR3-Stimulated KCs and LSECs Whereas Other Cytokines of Undefined Nature Are Involved in TLR4-Mediated Antiviral Effects of KCs.
It has been shown that type I and II IFNs and TNF-α can inhibit HBV replication in the livers of HBV-transgenic mice24 or chimpanzees acutely infected with HBV15 and can inhibit HBV replication directly in HBV-Met cells.21 To further identify the antiviral cytokines in the supernatants of TLR-stimulated NPCs, we preincubated the supernatants with neutralizing antibodies against murine IFN-α, murine IFN-β, murine IFN-γ, and murine TNF-α for 2 hours. These culture supernatants were then added to differentiated HBV-Met cells for 2 days, and HBV gene replication was analyzed via Southern blotting. None of the anti-IFN antibodies showed any cross-reactivity with other IFNs (data not shown) or had any effect on unstimulated cells (Fig. 4A).
Antibodies against IFN-β, in contrast to anti–IFN-α or IFN-γ antibodies, totally blocked the TLR3-induced antiviral effect of NPCs (Fig. 4B-D). None of the anti-IFN antibodies showed any cross-reactivity with other IFNs (data not shown) or had any effect on unstimulated cells (Fig. 4A). Interestingly, neutralizing antibodies against IFN-β or TNF-α could only slightly block the TLR4-mediated antiviral effect of KCs and did not show any significant synergistic blocking activity (Fig. 4D-E). Therefore, we conclude that the antiviral effect of TLR3-stimulated NPCs is mediated primarily by IFN-β, whereas the TLR4-induced antiviral effects of KCs seem to be mediated at least in part by other undefined mediators. These findings are in accordance with other data from our group that indicate that encephalomyocarditis virus (Wu et al., in preparation) and hepatitis C virus (HCV) (Broering et al., submitted) replication is almost exclusively inhibited by IFN-β but not IFN-α or IFN-γ from TLR-stimulated NPCs.
TLR3-Mediated and TLR4-Mediated Suppression of HBV Replication by KCs and LSECs Is MyD88-Independent.
It is well established that the MyD88-independent signaling initiated by TLR3 or TLR4 activation can lead to the production of IFN-β. To test whether this also applies to NPCs, we isolated LSECs and KCs from MyD88−/− mice and C57BL/6 wild-type mice, respectively, and stimulated them with 100 μg/mL poly I:C or 10 μg/mL LPS for 20 hours. As shown in Fig. 5, HBV replicative intermediates are almost completely abolished by treatment with supernatants from TLR3-stimulated and TLR4-stimulated KCs and TLR3-stimulated LSECs in both MyD88−/− mice (Fig. 5A) and wild-type mice (Fig. 5B).
In this article, we established murine HBV-Met cells as a tool for studying direct and indirect antiviral effects of the innate immune system on HBV replication and expression. These cells can be induced to a highly differentiated hepatocyte phenotype and support HBV gene expression, replication, and virus secretion.21 Primary cells of the innate immune system can be cocultured with the HBV-Met cells, and direct antiviral effects can be determined. We studied the role of the TLR system for the control of HBV replication and expression. Whereas direct stimulation of HBV-Met cells with TLR1-TLR9 agonists had no effect, supernatants from TLR3-stimulated and TLR4-stimulated NPCs could potently suppress HBV replication but not HBV expression. The TLR3 effect was mediated by IFN-β, whereas TLR4 stimulation led to the production of undefined (non-IFN, non–TNF-α) antiviral factors.
The liver is exposed to pathogens from the gastrointestinal tract or from the systemic circulation.25 In this organ, LSECs and KCs function as major scavengers and effector cells of the innate immune system. To identify pathogens, the innate immune system detects pathogen-associated molecular patterns via pattern recognition receptors, including the TLR system. Therefore, we hypothesized that TLR-stimulated NPCs are able to suppress HBV replication, because some TLRs (TLR3, TLR7, TLR8, and TLR9) are known to detect viral components. Whereas TLR3 signaling is mediated by TRIF and results in interferon regulatory factor-3 activation and IFN-β induction, TLR4 leads to additional activation of nuclear factor-κB through MyD88.12 We demonstrate that TLR3-induced and TLR4-induced antiviral mediators of NPCs can control HBV replication in a MyD88-independent manner. Additional data suggest that stimulation of NPCs with other TLR agonists (e.g., TLR8 or TLR9) leads to up-regulation of cell surface markers such as CD40, CD80, CD86, and major histocompatibility complex class II without initiating an antiviral activity (Wu et al., in preparation). This is in contrast to MDCs, in which TLR9 stimulation induces both costimulatory molecules and IFNs. Taken together, these findings may indicate that TLR responses are cell type–specific.
The role of the adaptive immune system has been studied extensively. These studies have revealed that clearance of HBV infection is typically associated with vigorous, polyclonal, and multispecific CD4+ and CD8+ T cell responses to epitopes in all HBV proteins. However, the role of the innate immune system of the liver, and in particular the role of NPCs in controlling HBV infection, is not well understood. Using the duck hepatitis B virus (DHBV) model, it has been proposed that hepatotropic viruses including HBV are initially scavenged by LSECs and thereafter released to infect adjacent heaptocytes.26 Although LSECs are not productively infected by DHBV, internalization of DHBV into LSECs with the DHBV receptor, carboxypeptidase D, may help the virus escape from the immune system. To date, there is no direct in vitro or in vivo evidence that LSECs are involved in antiviral responses against HBV. In contrast, KCs activated by phagocytosis of merozoite-infected erythrocytes27 or LCMV infection28 contribute to early control of HBV replication in HBV transgenic mice via secretion of an array of cytokines (IFN-α/β, TNF-α, and IFN-γ) and recruitment of natural killer cells, natural killer T cells, and T cells. Using the DHBV model, it has been suggested that antiviral mediators released from hepatic NPCs upon LPS stimulation can control hepadnavirus replication at the posttranscriptional level.29 Experiments with liver biopsies of chronic HBV patients show that the absence of cells producing antiviral cytokines in the liver may be responsible for HBV persistence and that the inflammatory liver injury during chronic HBV infection is probably related to increased Fas-L expression by KCs.30
Wieland et al. found that the genes reflecting the innate immune response are not detectable during HBV entry and expansion.31 There are 2 possible explanations for the absence of innate immune responses in the initial phase of HBV infection. First, HBV may be a “stealth virus” that does not induce any genes during the early stages of infection. Second, HBV may have developed strategies to suppress the initial antiviral response of the host, including TLR pathways. Cheng et al. showed that recombinant HBsAg inhibits LPS-induced COX-2 expression, reduces IL-18 production by interfering with the nuclear factor-κB pathway, and down-regulates IFN-γ in the human monocytic cell line THP-1.32 Visvanathan et al. have shown that TLR2 expression is reduced in liver cells (hepatocytes and KCs) of HBeAg-positive patients and in HepG2 cells expressing HBeAg, indicating that HBeAg may be responsible for the lack of an antiviral effect of TLR2 in this model.33
Accumulating evidence suggests that the modulation of TLR signaling pathways can be used as a therapeutic approach in multiple diseases, including infectious diseases (HCV, sepsis, genital warts), immune-related disorders (allergy, actinic keratosis, ulcerative colitis, Crohn's disease), and cancer (basal cell skin cancer, melanoma, non-Hodgkin's lymphoma).34 Recently it has been shown that activation of TLR7/835 or TLR936 could elicit antiviral effects against HCV in an IFN-dependent manner or IFN-independent manner. Isogawa et al. demonstrated that administration of TLR3, TLR4, TLR5, TLR7, and TLR9 resulted in complete inhibition of HBV replication in a type I IFN-dependent manner in HBV transgenic mice.3 They suggested that NPCs but not hepatocytes are involved in activation of antiviral program by TLR ligands in their in vivo model. In the present study, we demonstrate that the antiviral activity of TLR3 and TLR4 agonists observed by Isogawa et al. may be largely mediated through activation of LSECs and KCs, whereas the antiviral activity induced by TLR5, TLR7, and TLR9 agonists are most likely mediated by other cells (e.g., DCs). Our data indicate that TLR3 and TLR4 agonists could be used as therapeutic agents against HBV. In fact, nature may already prove that TLR3 activation can suppress HBV as coinfection with HCV, which is known to activate TLR3 and induce endogenous type I IFNs, can efficiently down-regulate HBV replication.37 In addition, it has been shown that “eradication” of HCV (and thereby down-regulation of the endogenous IFN production in response to HCV) via IFN treatment may lead to reactivation of HBV replication.38
The mechanism by which TLR4-induced supernatants of KCs block HBV replication remains to be delineated. Previous studies suggested that activated macrophages can release a broad range of mediators: direct antiviral mediators (e.g., IFN-α/β, TNF-α, and nitric oxide), indirect immunoregulatory cytokines, chemokines, lipid mediators, low–molecular weight oxygen radicals, and peroxide.39 Nitric oxide has been shown to play a role in the defense against a variety of microbial pathogens, including bacteria, parasites, and viruses such as HBV40 and herpes simplex virus type 1.41 IL-1 has also been shown to have direct antiviral effects.42 Moreover, it is possible that other mediators produced by KCs have undefined antiviral functions on HBV replication.
In conclusion, our data show that TLR3-stimulated and TLR4-stimulated NPCs are potent suppressors of HBV replication through production of IFN-β (TLR3) and undefined factors (TLR4). These findings are of particular relevance for the local control of HBV replication by the innate immune system of the liver. In addition, these data may be used for the development of TLR-based antiviral strategies against HBV and shed new light on the viral cross-talk between HCV and HBV.