Potential conflict of interest: Nothing to report.
Supported by Hepato-Regio-Net (Europäischer Fonds für regionale Entwicklung des Oberrheins) and Deutsche Forschungsgemeinschaft (Fi 801/2-1), Germany.
CD40, a member of the tumor necrosis factor receptor family, and its ligand, CD40L (CD154), are important regulators of the antiviral immune response. CD40L is up-regulated on lymphocytes and CD40 on hepatocytes during infection with hepatitis C virus (HCV); we investigated the role of CD40 signaling during HCV replication in hepatocytes. Viral replication was studied in primary human hepatocytes (PHH) and Huh7.5 cells using the infectious HCV Japanese fulminate hepatitis 1 isolate (JFH1) culture system, and in coculture with HCV antigen-specific CD8+ T cells. CD40L rapidly and transiently inhibits expression of the HCV nonstructural proteins NS3 and NS5A as well as HCV structural proteins core and E2 in Huh7.5 cells. Similarly, CD40L prevented replication of HCV in PHH, in synergy with interferon (IFN)-alpha. In Huh7.5 cells with replicating HCV, CD40L prevented production of infectious viral particles. When HCV antigen-specific CD8+ T cells were cocultured with HLA-A2-expressing Huh7 cells that had replicating virus, the T cells became activated, up-regulated CD40L, and inhibited HCV replication. Inhibition of CD40L partially prevented the antiviral activity of the CD8+ T cells. The antiviral effect of CD40L required activation of c-Jun N terminal kinases (JNK)1/2, but not induction of apoptosis or the JAK/STAT pathway that is necessary for the antiviral effects of IFNs. Conclusion: CD40 inhibits HCV replication by a novel, innate immune mechanism. This pathway might mediate viral clearance, and disruptions might be involved in the pathogenesis of HCV infection. (HEPATOLOGY 2013;)
Hepatitis C virus (HCV) infection persists in up to 80% of patients and is a leading cause of liver cirrhosis and hepatocellular carcinoma (HCC). Worldwide, about 170 million individuals are infected with HCV. Mechanisms of endogenous viral clearance are still under investigation, and both, the acquired and the innate immune system, seem to be essential.1 Several studies point to a major impact of non-cytolytic mechanisms to control HCV infection.2, 3
The receptor CD40 is a 48 kDa type I transmembrane protein and a member of the tumor necrosis factor (TNF) receptor family that is expressed mainly on the surface of immune cells and by non-immune cells in inflammation. The ligand of CD40, CD40L (CD154), is primarily expressed by activated CD4+ T-cells. In inflammation conditions it is also induced on natural killer cells, B cells and monocytes.4 The wide expression of this costimulatory receptor-ligand pair indicates the pivotal roles they play in different cellular immune processes. A soluble form of CD40L has been reported with activities on CD40-expressing cells similar to the transmembrane form.5 After interaction with CD40L, CD40 recruits adapter proteins known as TNFR-associated factors (TRAFs) to its cytoplasmic domain that are consecutively degraded.6 TRAF proteins activate different signaling pathways including the mitogen-activated protein kinases (MAPKs), c-Jun activating kinases (JNK), phosphoinositide 3-kinase (PI3K), and others.4 In lymphocytes, TRAF2-dependent activation of JNK has been demonstrated to depend on the activation of mitogen-activated protein kinase kinase kinase 1 (MEKK1),7 sometimes involving mitogen-activated protein kinase kinase 7 (MKK7) or MKK4.8-10 The membrane-proximal region of the intracellular CD40 tail contains a binding domain for Janus kinase 3 (Jak3), allowing signaling independently of TRAF proteins involving the Jak3 and signal transducer and activator of transcription 5 (STAT5).4, 11 Interferon (IFN)-dependent inhibition of HCV replication depends on the Jak/STAT pathway.12 Nevertheless, IFN-independent inhibition of HCV replication has been described.13, 14 CD40-CD40L interaction also affects key processes in immune cell activation, differentiation, proliferation and apoptosis. CD40 ligation increases antiapoptotic Bcl-xL and c-FLIP in follicular lymphomas, thereby protecting them against apoptosis induction.15 Conversely, CD40 activation is able to directly induce apoptosis in other cells.4 Clinical trials in cancer patients using agonistic CD40 antibodies and recombinant CD40L have been performed.16-18 In contrast to immune cells, the role of CD40 on hepatocytes is largely unknown, and its potential function in the innate immune response has not been evaluated so far. Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) has recently been described to induce apoptosis in HCV-infected but not in normal hepatocytes, thereby promoting viral clearance.19, 20 The role of CD40 in hepatocyte apoptosis is controversial, with only few data available to date.21-24
To determine the impact of CD40L-induced CD40 activation on HCV replication, we studied mechanisms of CD40-dependent antiviral activity in the infectious HCV Japanese fulminate hepatitis 1 isolate (JFH1) tissue culture model system25 in primary human hepatocytes (PHH) and Huh7.5 hepatoma cells and in a coculture model of HCV antigen-specific CD8+ T-cells with HLA-A2 expressing, HCV SGR-JFH1-replicating, Huh7 cells.2
HCV, hepatitis C virus; JFH1, Japanese fulminate hepatitis 1 isolate; JNK, c-Jun N terminal kinases; MEKK1, mitogen-activated protein kinase (MAPK) kinase kinase 1; MKK7, mitogen-activated protein kinase kinase 7; PARP, poly (ADP-ribose) polymerase; PHH, primary human hepatocytes; TRAF, tumor necrosis factor receptor-associated factor; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand.
Materials and Methods
For details please see Supporting Materials and Methods.
Primary human hepatocytes (PHH), Huh7.5 hepatoma cells, Huh7-Lunet transduced with JFH1-based selectable subgenomic luciferase replicon and with HLA-A2 gene with mutated NS5B2594-2602 epitope and HCV NS5B2594-2602-specific CD8+ T-cell clone were isolated and cultured as described in the Supporting Information.
Transfection of HCV RNA, HCVcc Production, Infection of Huh7.5 cells, and PHH.
Transfection of Huh7.5 cells by way of electroporation of RNA derived from HCV genotype 2a plasmids pJFH1, pJFH1/GND, pSGR-JFH1,25 pFK I389Luc/EMCV/N3-3′/JFH, and pFK I389Luc/EMCV/JFH/deltaGDD.26-29 Cell culture supernatants from JFH1 HCV RNA transfected cells were used to infect PHH and naïve Huh7.5. Replication-deficient JFH1/GND HCV RNA served as control.
Incubation With Cytokines, Ligand, Inhibitors, and HCV-Peptide.
Incubation with CD40 ligand, IFN-α, IFN-γ (Cell Signaling) or TRAIL (Axxora), and pan-caspase-inhibitor (Z-VAD-FMK; R&D Systems), Jak-inhibitor I (Pyridone 6, Merck) JNK-inhibitor II (SP600125; Merck), MEKK1-inhibitor (LL Z1640-2; SantaCruz Biologicals), neutralizing of CD40-mediated bioactivity with antagonistic CD40 monoclonal antibody (mAb) (R&D Systems) or CD40L mAb (Ancell) are described in the Supporting Information. HCV-specific T-cell clone was stimulated for 6 hours with HCV NS5B2594-2602 or phytohemagglutinin (PHA, 2 μg/mL) as described in Jo et al.2
Immunoblot Analysis, Immunofluorescence Staining of Viral and Cellular Proteins, and Enzyme-Linked Immunosorbent Assay (ELISA) of Cytokeratin 18 Fragment.
Immunoblot analysis, immunofluorescence staining, used antibodies, and ELISA Kits are described in detail in the Supporting Information.
Reverse-Transcription Polymerase Chain Reaction (RT-PCR) of HCV RNA.
Infection of PHH with HCV JFH1 was performed by incubation of cells for 3 hours with culture supernatants of JFH1 transfected Huh7.5 cells as described. Seventy-two hours postinfection, the medium was changed and CD40L was added for 24 hours. Viral particles were enriched from the supernatant using the viral nucleic acid purification kit from Roche. The purification was performed according to the instructions of the manufacturer. Secreted viral genomes were quantified by real-time PCR using COBAS Ampli PREP/COBAS TaqMan HCV test (Roche Diagnostics) according to the manufacturer's instructions.
Data are expressed as means ± standard error of the mean (SEM). In general, at least three independent experiments were performed. Results were compared to untreated control, otherwise identified with clamps, using the two-tailed Student's t test: *P < 0.05 was considered statistically significant.
CD40L Down-regulates HCV Protein Expression in HCV Replicating Huh7.5 Cells and PHH Through Activation of CD40.
Although an important role of CD40L/CD40 in the immune regulation of HCV infection is well established, the exact role of CD40L on HCV replication remains elusive. Therefore, we studied the effect of recombinant CD40L on HCV replicating Huh7.5 cells and PHH. CD40L induced a time- and dose-dependent down-regulation of HCV nonstructural protein NS3 and HCV structural protein E2 expression in Huh7.5 cells (Fig. 1A,B). Inhibition of CD40 or CD40L using antagonistic antibodies reversed CD40L-dependent down-regulation of HCV proteins (Fig. 1C), suggesting that CD40L-induced antiviral effects depend on the activation of CD40 receptor. CD40L inhibited HCV nonstructural and HCV structural protein expression in HCV JFH1 infected PHH (Fig. 1D). CD40L-dependent inhibition of HCV protein expression was similar in Huh7.5 cells containing the full-length HCV JFH1, able to produce infectious virions, as compared to Huh7.5 cells containing the HCV SGR-JFH1 mutant, able to replicate, but unable to produce virions25 (Fig. 1E).
CD40L Inhibits HCV RNA Replication but Not Translation.
In order to distinguish HCV translation from HCV RNA replication, we transfected Huh7.5 cells with the bicistronic replicon I389Luc/EMCV/N3-3′/JFH carrying luciferase under control of the HCV IRES, and its mutant I389Luc/EMCV/JFH/deltaGDD unable to replicate. Two to 24 hours after transfection luciferase activity was observed (Fig. 2A). Eight hours posttransfection CD40L and IFN-α start to reduce luciferase activity in I389Luc/EMCV/N3-3′/JFH transfected cells, in contrast to replication-deficient I389Luc/EMCV/JFH/deltaGDD. Hence, CD40 primarily inhibits RNA replication.30
CD40L and IFN-α Synergistically Inhibit HCV Replication in PHH.
The current antiviral treatment of HCV-infected patients is based on IFN-α. In order to determine a potential additional suppression of HCV replication, HCV-infected PHH were incubated with IFN-α and CD40L. CD40L furthermore reduced IFN-α-dependent down-regulation of HCV RNA by nearly 2-logs of HCV RNA copies/mL, CD40L and INF-α act synergistically and this supernatant was no longer infectious (Fig. 2B).
CD40L Prevents Production of Infectious HCV Virions and Reduces Secretion of HCV RNA.
Down-regulation of HCV protein may not have a significant impact on HCV infection. In our HCV JFH1 cell culture system, the supernatant of HCV JFH1 replicating Huh7.5 cells contains infectious virions.19, 25 In order to address the impact of CD40 on the release of infectious virions, HCV JFH1 replicating Huh7.5 cells were incubated with CD40L or control medium. Although HCV JFH1 supernatant infected naïve Huh7.5 cells, supernatant of CD40L-incubated HCV JFH1 infected less naïve Huh7.5 cells (Fig. 3A-C). Secreted HCV RNA in the cell culture supernatant was reduced from 3.7 × 106 to 2.5 × 103 copies/mL in HCV-replicating PHH and from 1.8 × 105 to 7.3 × 103 in Huh7.5 cells copies/mL (Fig. 3D). Addition of neutralizing CD40L antibody to the CD40L-containing supernatant after collection and prior to incubation with naïve Huh7.5 cells did not alter infectivity of the supernatant (Fig. 3C). These data demonstrate that CD40L primarily inhibits HCV replication, and not viral entry.
CD40L Contributes to CD8+ T-Cell-Mediated Inhibition of Viral Replication.
In order to examine the role of CD40 in a more physiological model, we used a coculture model with a subgenomic HCV JFH1 replicating and HLA-class allele A2 gene expressing Huh7-Lunet cell line (Huh7A2HCVEM) recognized by an HCV antigen-specific CD8+ T-cell clone that has been described recently.2 In line with the previous experiments, CD40L reduced HCV replication in Huh7A2HCVEM cells (Fig. 4A). Activation of HCV-antigen-specific CD8+ T-cells, but not of non-HCV-specific CD8+ T-cells, with HCV peptides induced strong CD40L expression (Fig. 4B). In coculture, HCV replication in Huh7A2HCVEM cells was significantly inhibited by the HCV-specific CD8+ T-cell clone mainly by way of IFN-γ.2 Inhibition of CD40L partly reversed T-cell-dependent inhibition of HCV replication, whereas addition of CD40L enhanced the inhibition of HCV replication (Fig. 4C).
Antiviral Effects of CD40 Do Not Depend on Apoptosis Induction.
Recently, inhibition of HCV replication through TRAIL-induced apoptosis has been demonstrated.19, 20 Because HCV JFH1 renders infected cells susceptible to death receptor ligand-induced apoptosis, these cells are preferentially killed.19, 31 Some groups demonstrated CD40L-dependent apoptosis of hepatocytes and hepatoma cells, mostly of murine origin.21, 24, 32 In HCV JFH1 replicating PHH and Huh7.5 cells CD40L neither reduced cell viability (Fig. 5A), nor induced poly (ADP-ribose) polymerase (PARP) cleavage (Fig. 5B; Supporting Fig. S1A), cytokeratin-18 fragmentation (Fig. S1A,B), or cytochrome C release (Fig. 5D), indicators of apoptosis induction.33 HCV JFH1 replicating Huh7.5 were incubated with CD40L for up to 72 hours without detectable morphological evidence of apoptosis or PARP/cytokeratin-18 cleavage (not shown), and CD40L-dependent down-regulation of HCV protein did not depend on caspase activation (Fig. 5C), similar to interferons2, 34 (Fig. S1C), but in contrast to TRAIL (Fig. S1D).19
CD40-Dependent Inhibition of HCV Replication Does Not Involve the Jak/STAT Signaling Pathway.
Antiviral effects of IFNs mainly depend on activation of the Jak/STAT signaling pathway,35 and CD40-dependent activation of Jak3/STAT3 has been demonstrated.11 Incubation of HCV replicating Huh7.5 cells with IFN-α rapidly induced phosphorylation of STAT1, STAT2, STAT3, and STAT5, whereas CD40L had no such effect (Fig. 6A; Fig. S2B). In line with this, inhibition of Jak prevented IFN-dependent phosphorylation of STAT1 and STAT2 and down-regulation of HCV protein expression (Fig. 6D,E), whereas CD40L- and TRAIL-induced inhibition of HCV protein expression and loss of infectivity remained unchanged (Fig. 6B,C; Fig. S2C).
CD40L-Dependent Antiviral Activity Involves TRAF2 and TRAF3 and Activation of Stress-Activated Protein Kinases JNK1/2, MEKK1, and MKK7.
CD40-dependent signal transduction involves TRAF activation and subsequent degradation. Recently, TRAF2 has been shown to directly interact with NS5A, whereas TRAF2 knockdown inhibited HCV replication.36-37 Incubation of HCV SGR-JFH1 replicating Huh7.5 cells with CD40L induced degradation of TRAF2 and TRAF3 (Fig. 7A). TRAF2 staining of HCV replicating Huh7.5 cells incubated with CD40L demonstrated an intracellular relocation of TRAF2 from nearby nucleus with consecutive loss of cellular TRAF2 expression (Fig. S3). CD40-dependent recruitment of TRAF2 has been demonstrated to activate the stress-activated protein kinases JNK1/2.4 We therefore examined phosphorylation of JNK1/2 and its downstream target, c-Jun. Incubation of HCV JFH1 replicating Huh7.5 cells with CD40L induced a rapid and transient phosphorylation of JNK1/2 and c-Jun (Fig. 7B). In order to analyze the impact of JNK1/2 activation on CD40-dependent antiviral activity, we blocked JNK1/2 activation using the specific inhibitor SP600125 (Fig. 7D). JNK1/2 inhibition completely prevented CD40-induced down-regulation of HCV protein expression (Fig. 7C), in contrast to IFN-dependent down-regulation (Fig. S4). These results point to the key role of JNK1/2 in CD40-induced inhibition of HCV replication. TNF-α as a strong JNK-pathway activator did not reduce HCV protein expression (Fig. S5), suggesting a CD40-specific effect on HCV inhibition. CD40-induced JNK1/2 activation involves phosphorylation of MKK7, but not MKK4 (Fig. 7E), and depends on MEKK1 activation, because inhibition of MEKK1 prevented CD40L-induced phosphorylation of c-Jun (Fig. 7F). In conclusion, CD40L/CD40 interaction results in the recruitment of TRAF2 and TRAF3 that is followed by a MEKK1-dependent activation of JNK1/2 that in turn mediates the antiviral activity of CD40.
In this study we demonstrate that the CD40L/CD40 interaction inhibits production of infectious HCV virions in PHH and Huh7.5 cells using the JFH1 cell culture model. In a coculture model of HCV antigen-specific CD8+ T-cells with HCV replicating, HLA-A2 gene expressing Huh7 cells, we demonstrate a distinct role of CD40 in the T-cell-dependent inhibition of viral replication. Detailed analyses indicate that this inhibition involves activation of TRAF2/3 with consecutive initiation of MEKK1 and MKK7, resulting in the activation of stress-activated protein kinases JNK1/2. The latter is essential for the antiviral effect of CD40. CD40-induced inhibition of viral replication does not involve the apoptotic signaling cascade and differs from the antiviral IFN signaling pathway, but acts synergistically with IFN-α. CD40L-expressing immune cells may contribute to the suppression of HCV infection.
CD40L Inhibits HCV Replication and HCV Virion Production.
IFNs are among the most important inhibitors of HCV infection,3 but may not be the only host defense mechanisms.2 In the infectious HCV JFH1 cell culture system, we demonstrate that CD40L inhibits HCV RNA and protein expression in Huh7.5 cells and PHH (Figs. 1A-E, 2B, 3D) by activating the receptor CD40 (Fig. 1C). CD40-mediated inhibition of HCV protein expression is not restricted to a particular HCV protein (Fig. 1A,D), and CD40L-mediated down-regulation of HCV replication has a major impact on HCV infection in our JFH1 cell culture system because we could demonstrate that CD40L prevents the infection of Huh7.5 cells (Fig. 3A-C). In our cell culture model, CD40L inhibits HCV RNA replication, but not viral entry, infection, or translation (Figs. 2A, 3C). In vivo, spontaneous HCV clearance depends on specific MHC-I expression of the target-cell and HCV antigen recognition by virus-specific lymphocytes. Using a coculture model with a subgenomic HCV JFH1 replicating and MHC-I allele HLA-A2 gene expressing Huh7-Lunet cell line (Huh7A2HCVEM) recognized by an HCV antigen-specific CD8+ T-cell clone,2 we demonstrate that HCV-specific activation of the T-cells successfully inhibited HCV replication, and this HCV antigen-specific CD8+ T-cell clone up-regulated CD40L after activation with HCV peptides after unspecific activation with PHA (Fig. 4B). Inhibition of the CD40/CD40L interaction partly reversed the T-cell-induced inhibition of HCV replication (Fig. 4C). These results suggest a small, but definite, role of CD40L in the T-cell-mediated inhibition of HCV replication in this novel coculture system. In HCV JFH1 replicating PHH, CD40L reduced HCV RNA by 3-logs copies/mL, far more as compared to the hepatoma cell lines Huh7.5 (Fig. 3D) or Huh7A2HCVEM (Fig. 4A). PHH may be more representative of the in vivo situation as compared to hepatoma cell lines and marked differences of both cell types in response to ligands of the TNF-superfamily have been described.19, 38 The CD40L/CD40 system may contribute to the spontaneous or IFN-induced clearance of HCV infection, because several studies point to additional mechanisms apart from IFNs in viral clearance,2, 39 and CD40L was synergistic with IFN-α in the inhibition of HCV replication in PHH (Fig. 2B). It is also interesting to note that several studies have suggested a specific impairment of HCV specific T-cells and NK cells in chronic HCV infection, e.g., characterized by a reduced CD40L expression, furthermore preventing HCV clearance.40-42
Signal Transduction of CD40 Differs from IFNs and TRAIL.
TRAIL has been demonstrated to exert antiviral effects by sensitizing HCV-replicating cells to apoptosis induction.19, 20 Both TRAIL-R1/-R2 and CD40 transduce intracellular signals by activating TRAFs.6, 43 In contrast to TRAIL, CD40 does not contain a death domain and apoptotic as well as survival signaling has been described.4 Similar to IFNs, CD40 induced apoptosis in neither Huh7.5 hepatoma cells nor in PHH (Fig. 5B,D; Fig. S1A,B), and CD40 down-regulated HCV proteins independently from caspase activation (Fig. 5C).19 Analysis of the receptor adapter proteins TRAF1-6, which are known to transduce CD40 signaling,6 demonstrate CD40-induced activation of TRAF2 and TRAF3 (Fig. 7A), but no IFNs-dependent TRAF degradation (not shown). Similarly, IFNs activated the STAT pathway, and their antiviral properties were found to depend on Jak activation, in contrast to CD40 (Fig. 6A-C; Fig. S2A+B). These results strongly suggest that CD40 activates signaling pathways that are distinct from IFNs and TRAIL and involve TRAF2 and TRAF3 activation.
CD40-Induced Inhibition of HCV Replication Depends on JNK1/2 Activation by Way of MEKK1.
CD40-mediated cellular effects by way of TRAF2/TRAF3 have been demonstrated to depend on JNK1/2 activation.7, 8 On the other hand, Jak/STAT-independent inhibition of HCV replication has been described.13, 14 CD40 induced strong and transient activation of JNK1/2 with consecutive c-Jun phosphorylation (Fig. 7B). Inhibition of JNK1/2 abolished CD40-dependent antiviral effects (Fig. 7C). The antiviral effect of CD40 is transient (Fig. 1A) and may reflect the transient activation of TRAF2/3 and of JNK1/2 (Fig. 7A,B). In lymphocytes, CD40-induced activation of JNK1/2 has been demonstrated to depend on the recruitment of TRAF2 with consecutive activation of MEKK1 and downstream MKK4/MKK7.7, 8 As depicted in Fig. 7E, CD40L induced phosphorylation of MKK7, but not MKK4. Inhibition of MEKK1 completely blocked CD40L-dependent activation of JNK1/2 (Fig. 7F). These results indicate CD40L-induced activation of MEKK1 and MKK7 that are necessary for activation of JNK1/2.
Taken together, CD40L/CD40 activation has a potent antiviral effect and may represent a target for urgently needed novel HCV treatment strategies. In clinical phase II studies, recombinant CD40L and agonistic CD40 antibodies have been delivered successfully and without major side effects to patients with solid tumors.16-18 Therefore, these agents may prove beneficial in combination with conventional therapeutic strategies aimed at the elimination of HCV infection in patients with chronic hepatitis C.
The authors thank Dr. C.M. Rice, New York, for the gift of Huh7.5 cells and Dr. V. Lohmann, Heidelberg, for pFK-I389Luc/EMCV/N3-3′/JFH and pFK-I389Luc/EMCV/JFH/deltaGDD and for helpful discussions. Author contributions: S.J. Rau: acquisition, analysis and interpretation of data, study design; E. Hildt and K. Himmelsbach: acquisition of PHH data; R. Thimme: support of coculture system; T. Wakita: material support; H.E. Blum: supervision; R. Fischer: supervision, analysis, and interpretation of data, study design, preparation of article.