Hepatitis C virus (HCV), a member of the Flaviviridae

Transcription

1 GASTROENTEROLOGY 2008;135: Neutralizing Host Responses in Hepatitis C Virus Infection Target Viral Entry at Postbinding Steps and Membrane Fusion ANITA HABERSTROH,*, EVA K. SCHNOBER,*,, MIRJAM B. ZEISEL,, PATRIC CAROLLA,*,,, HEIDI BARTH,* HUBERT E. BLUM,* FRANÇOIS LOIC COSSET,,#, ** GEORGE KOUTSOUDAKIS, RALF BARTENSCHLAGER, ANN UNION, ERIK DEPLA, ANIA OWSIANKA, ARVIND H. PATEL, CATHERINE SCHUSTER,, FRANÇOISE STOLL KELLER,, MICHEL DOFFOËL, MARLÈNE DREUX,,#, ** and THOMAS F. BAUMERT,, *Department of Medicine II, University of Freiburg, Freiburg, Germany; Inserm, U748, Strasbourg, France; Université Louis Pasteur, Strasbourg, France; Faculty of Biology, University of Freiburg, Freiburg, Germany; Inserm, U758, Lyon, France; # Université de Lyon, (UCB-Lyon1), IFR128, Lyon, France; **Ecole Normale Supérieure de Lyon, Lyon, France; Department of Molecular Virology, University of Heidelberg, Heidelberg, Germany; GENimmune NV, Ghent, Belgium; MRC Virology Unit, Institute of Virology, University of Glasgow, Glasgow, United Kingdom; and Service d Hépatogastroénterologie, Hôpitaux Universitaires de Strasbourg, Strasbourg, France Background & Aims: Hepatitis C virus (HCV) is a leading cause of chronic hepatitis worldwide. Viral attachment and entry, representing the first steps of virus-host cell interactions, are major targets of adaptive host cell defenses. The mechanisms of antibody-mediated neutralization by host neutralizing responses in HCV infection are only poorly understood. Retroviral HCV pseudotypes (HCVpp) and recombinant cell culture-derived HCV (HCVcc) have been successfully used to study viral entry and antibody-mediated neutralization. Methods: In this study, we used these model systems to investigate the mechanism of antibody-mediated neutralization by monoclonal antienvelope antibodies and polyclonal anti-hcv immunoglobulins purified from HCV-infected patients. Results: Using a panel of monoclonal antienvelope antibodies, we identified an epitope within the E1 glycoprotein targeted by human neutralizing antibodies during postbinding events. Interestingly, we observed that host neutralizing responses in the majority of HCVinfected individuals include antibodies targeting HCV entry after binding of the virus to the target cell membrane. Using a kinetic assay based on HCVpp and HCVcc entry, we demonstrate that purified antiviral immunoglobulins derived from individual HCV-infected patients appear to inhibit HCV infection at an entry step closely linked to CD81 and scavenger receptor BI (SR-BI). Conclusions: Our results indicate that host neutralizing responses in HCV-infected patients target viral entry after HCV binding most likely related to HCV-CD81, and HCV-SR-BI interactions, as well as membrane fusion. These findings have implications not only for the understanding of the pathogenesis of HCV infection but also for the design of novel immunotherapeutic and preventive strategies. Hepatitis C virus (HCV), a member of the Flaviviridae family, is a major cause of chronic liver disease worldwide. The majority of infected individuals develops chronic hepatitis, which can progress to cirrhosis and hepatocellular carcinoma. Both viral and host factors appear to play an important role for resolution of acute infection. 1 A large body of evidence suggests that a strong, multispecific, and long-lasting cellular immune response appears to be important for control of acute hepatitis C. 1 By comparison, the impact of neutralizing immune responses for viral clearance has been less welldefined. 2,3 Viral attachment and entry representing the first interaction of the virus with the host cell are major targets of adaptive humoral responses. 2 Viral proteins are recognized as nonself by the host s immune system and induce the production of antibodies. A small proportion of these antibodies exhibits antiviral activity in vitro and is defined as virus-neutralizing antibodies. Neutralizing antibody responses often provide a first-line adaptive defense against infection by limiting virus spread. 3 Antibodies with HCV-neutralizing properties have been first described in experimental infection of chimpanzees. 4 The recent development of model systems to study early steps of viral infection including retroviral HCV pseudotypes (HCVpp), HCV-like particles, and recombinant cell culture-derived infectious virions (HCVcc) allowed to detect and characterize neutralizing antibodies in HCV infection 2 and resulted in a better understanding of the impact of host neutralizing responses for resolution of HCV infection. 3,5 7 Although viral and host factors for the initi- Abbreviations used in this paper: E1, E2, envelope glycoproteins E1 and E2; HCVcc, recombinant cell culture-derived HCV; HCV-LP, HCV-like particle; HVCpp, retroviral HCV pseudotypes; SR-BI, scavenger receptor BI by the AGA Institute /08/$34.00 doi: /j.gastro

2 1720 HABERSTROH ET AL GASTROENTEROLOGY Vol. 135, No. 5 ation of HCV infection have been identified, 2 the mechanisms of virus neutralization still remain elusive. Materials and Methods Antibodies and Cells Purified human monoclonal anti-e1 antibodies produced from HCV-infected patients (IGH520, IGH526, IGH534), 8 chimpanzee monoclonal antibody (mab) anti- E1 (IGH481), purified mouse anti-e1 mab (IGH433), and chimpanzee anti-e2 mab (IGH461, IGH466) were provided by GENimmune NV (Ghent, Belgium). Mouse mabs anti-e2 AP33, E2G, 917, and 3E5; anti-e1 11B7; and polyclonal antibody (pab) anti-scavenger receptor BI (SR-BI) have been described Preimmune and nonimmune rat serum were used as controls for experiments performed with rat anti-sr-bi. Purified anti-human CD81 mab and purified mouse and human isotype monoclonal control IgGs were obtained from Biosciences Pharmingen. Heparin and chondroitin sulfate were obtained from Merck Biosciences. Production and purification of recombinant envelope glycoprotein E1 (amino acids ) and HCV-like particles have been described. 13 Huh7 and Huh7.5.1 cells were maintained in Dulbecco s modified Eagle medium (PAA Laboratories, Pasching, Austria) with 10% fetal bovine serum (PAN Biotech, Aidenbach, Germany). 11,14 Huh7.5.1 cells 14 were obtained from F. V. Chisari, Scripps Research Institute, La Jolla, CA. The origin and maintenance of Huh7 and HEK-293T cells have been described. 11,14 Patients Anti-HCV-positive serum samples were obtained from 22 patients with chronic HCV infection (9 patients were followed at Department of Medicine II, University Hospital, Freiburg; 13 patients were followed at the Gastroenterology-Hepatology Service, University Hospital, Strasbourg). IgG was purified from sera by protein G antibody affinity chromatography (MabTrap; Amersham Biosciences) as described. 11 Viral load was quantified using Bayer Versant HCV RNA 3.0 (Bayer Diagnostics) or Amplicor (Roche Diagnostics). Detection of anti-hcv antibodies was performed using ARCHITECT Anti-HCV EIA (Abbott Laboratories). Purified control IgG was derived from anti-hcv-negative individuals. Approval was obtained from the institutional review boards of the University Hospital, Freiburg, and Strasbourg University hospitals. Antibody-Mediated Neutralization of HCVpp Infection Murine leukemia virus based retroviral HCVpp expressing envelope proteins of strain H77C (1a), HCV-J (1b), and HCV-J6 (2a) were generated as described. 5,15 In protocol I, Huh7 cells ( cells/well) were spinoculated with a mix of HCVpp and anti-e1/e2 mab (100 g/ml) or control IgG (100 g/ml), anti-cd81 mab or control IgG (20 g/ml), anti-sr-bi polyclonal serum or preimmune control serum (dilution, 1:50), anti-hcv IgG derived from HCV-infected patients, or control individuals (100 g/ml) for 1 hour (2000 rpm, 4 C). Following removal of nonbound viral proteins by washing with ice-cold phophate-buffered saline (PBS), cells were supplemented with fresh medium containing the antibodies and temperature was shifted to 37 C. After 4 hours, inhibitors were removed, and, 72 hours later, GFP reporter gene expression was detected as described. 15 To study whether mabs or purified human anti-hcv IgG interfered with early steps of HCVpp entry postbinding, experiments were performed according to protocol II: Huh7 cells were spinoculated with HCVpp for 1 hour, washed with prechilled PBS to remove unbound HCVpp, and then incubated with antibodies for 4 hours at 37 C. To allow fine mapping of the kinetics and contribution of antibodies during HCV entry, antibodies were added every 20 minutes for up to 120 minutes after viral binding. Data of antibody-mediated neutralization are presented as mean percent cells positive for GFP relative to neutralization of HCVpp without antibodies. Antibody-Mediated Neutralization of HCVcc Infection Production and titration of recombinant HCVcc (Luc-Jc1; containing the HCV structural proteins of strain J6 and a luciferase reporter element) were carried out as described. 16,17 Inhibition of HCVcc entry into Huh7.5.1 cells by human anti-hcv IgG or human control IgG (100 g/ml) and anti-cd81 mab or isotype control IgG (5 g/ml) during binding and postbinding steps (protocol I and II) and kinetic assays was performed as described. 10,17 Interference of Monoclonal Antibodies With Cellular Binding of E1, HCV-LP, and HCVpp Recombinant envelope glycoprotein E1 (1b; 10 g/ml), HCV-like particles (HCV-LPs) (1a; 2.5 g/ml), or HCVpp (1b; purified by sucrose-gradient ultracentrifugation 13 ) were incubated with human or chimpanzee anti-e1 mab (100 g/ml), isotype control IgG, or heparin (100 g/ml) for 1 hour at 37 C prior to the addition to Huh7 cells as described. 13 Following binding of antibody-ligand complexes for 1 hour at 4 C to cells/100 L PBS, nonbound viral proteins were removed by washing with PBS, and binding of ligands was analyzed by flow cytometry using mouse anti-e1 (11B7) mab and phycoerythrin (PE)-conjugated anti-mouse IgG. 13 Fusion Assay HCVpp/liposome lipid mixing assays with rhodamine-labelled liposomes were performed as previously described Purified HCVpp were preincubated with anti-hcv IgG derived from HCV-infected patients or HCV-negative control individuals (100 g/ml) or apo-

3 November 2008 NEUTRALIZING HOST RESPONSES IN HCV 1721 lipoprotein C-I acting as a positive control. Fusion kinetics were recorded on an SLM Aminco 8000 spectrofluorimeter over a 30-minute time period, with exc at 560 nm and em at 590 nm as described The initial rates of fusion were taken as the value of the slope of the tangent, determined at the initial part of the kinetics at time 0. Results Patient-Derived Human anti-hcv IgGs Neutralize HCVpp Entry at Postbinding Steps To map the entry steps targeted by mabs or polyclonal antisera, we established an HCVpp-based neutralization assay allowing to distinguish between antibody-mediated interference during binding and postbinding events by adapting a previously published method described for cell culture grown HCV (HCVcc). 17 HCVpp binding to Huh7 cells was performed for 1 hour at 4 C in the presence or absence of antibodies. Under these conditions, HCVpp bind to the cells but do not efficiently enter. Synchronous entry and infection occurs when the inoculum is removed and the cells are shifted to 37 C. In protocol I, inhibitory antibodies were added before binding of HCVpp to the target cell. In protocol II, antibodies were added after HCVpp binding to target cells, thus allowing to study the effect of antibodies on viral entry after HCVpp binding (Figure 1A). In this assay, a molecule inhibiting HCVpp infection during postbinding steps is able to inhibit HCVpp entry when added following binding of virus (protocol II). To validate this system, we analyzed HCVpp entry into Huh7 cells in the presence of anti-sr-bi and anti-cd81 antibodies recently shown to inhibit HCV entry at a step postbinding of virus. 10,17 As expected for antibodies Figure 1. Neutralizing host responses in HCV infection target HCVpp entry at postbinding steps. (A) Experimental setup. HCVpp binding to Huh7 cells was performed for 1 hour at 4 C in the presence or absence of antibodies. Under these conditions, HCVpp bind to the cells but do not efficiently enter. Synchronous entry and infection occurs when the inoculum is removed and the cells are shifted to 37 C. In protocol I, inhibitory antibodies were added before binding of HCVpp to the target cell. In protocol II, antibodies were added after HCVpp binding to target cells, thus allowing to study the effect of antibodies on viral entry after HCVpp binding. Dashed lines indicate the time intervals at which antibodies were present. Seventy-two hours later, HCVpp infection was assessed by GFP reporter gene expression determination, expressed relative to control infections without addition of inhibitors. (B) Inhibition of HCVpp (1a; H77c strain) entry into Huh7 cells by anti-sr-bi, anti-cd81, antibodies, heparin, or chondroitin sulfate (CSA) in protocols I and II. Mean SD of 6 independent experiments is shown. (C) Inhibition of HCVpp (1a; H77c strain) entry into Huh7 cells by purified human anti-hcv IgG (100 g/ml) from patients with chronic HCV infection or anti-hcv-negative control individuals (CTRL) in protocols I and II. Mean SD of at least 4 independent experiments is shown. (D) Dose-dependent inhibition of HCVpp (1a; H77c strain) entry by anti-hcv IgG derived from patient 7 (solid diamonds) or control IgG (open triangles) in experimental protocol II. Mean values SD of a representative experiment performed in triplicate are shown. Statistically significant differences (P.01; t test) in inhibition of HCVpp entry between protocol I and II in panels B and C are indicated (*).

4 1722 HABERSTROH ET AL GASTROENTEROLOGY Vol. 135, No. 5 Table 1. Inhibition of HCVpp Entry Pre- and Postbinding by Patient-Derived anti-hcv IgG Neutralization of HCVpp (H77C-1a) entry, % Patient No. Age Genotype Viral load (IU/mL) I II b a b b b a d b b b b b b a a b a No infection No infection a a a NOTE. IgG was purified from 22 patients with chronic HCV infection and 2 anti-hcv-negative control individuals using protein G affinity chromatography. HCVpp entry was assessed as described (Figure 1A). Patient age, genotype, load, and percent neutralization of HCVpp (1a; H77C strain) entry by purified patient IgG (100 g/ml) in protocols I and II are shown (mean SD of at least 4 independent experiments). targeting entry factors mediating postbinding events, anti-sr-bi and anti-cd81 markedly inhibited HCVpp entry when added postbinding (protocol II). Heparin a well-characterized molecule shown to interfere with HCV infection predominantly during viral attachment 10,13,17 markedly inhibited HCVpp entry when added before binding of HCVpp to target cells (protocol I) but inhibited HCVpp entry less efficiently when added postbinding (protocol II). Because the effect of antireceptor antibodies and heparin on HCVpp entry in protocols I and II was similar to the effect of these reagents on HCVcc infection (Figure 5 and previous studies 10,17 ), we used the HCVpp system as a screening assay to identify the entry steps targeted by a large panel of poly- or monoclonal anti-hcv antibodies. Purified IgG from anti-hcv-positive serum samples was obtained from 22 individuals with chronic HCV infection infected with HCV genotypes 1 4 (Table 1). Purified IgG from 2 anti-hcv-negative individuals served as controls (Table 1). As shown in Figure 1C and D and Table 1, purified anti-hcv IgG from 18 out of 22 patients inhibited HCVpp infection by more than 25%. Interestingly, IgG from almost all patients with detectable neutralizing antibodies showed a marked neutralization of HCVpp entry in protocol II (Table 1). These data indicate that the majority of host neutralizing responses in HCV-infected individuals include antibodies that inhibit HCVpp entry at a step following HCV binding. Monoclonal anti-e2 and anti-e1 Antibodies Target HCVpp Entry at Postbinding Steps To identify HCV envelope epitopes involved in postbinding steps during HCV entry, we screened a panel of monoclonal antienvelope antibodies using the HCVpp entry assay (Figure 1). As shown in Figure 2 and Supplementary Table 1 (see Supplementary material online at several monoclonal anti-e1 and -E2 antibodies markedly inhibited HCVpp entry when added postbinding in protocol II. These findings suggest that both envelope glycoproteins contribute to virus-host entry factor interactions during postbinding events. To confirm further the impact of E1 for postbinding steps, we investigated the interference of anti-e1 antibodies with cellular binding of the viral envelope glycoprotein E1. In the absence of a robust system to study cellular binding of serumderived or recombinant HCV, we investigated whether the respective anti-e1 antibodies inhibited binding of recombinant E1, HCV-LPs, or HCVpp to Huh7 cells (Figure 2E and F). Using previously established assays based on detection of ligand binding using flow cy-

5 November 2008 NEUTRALIZING HOST RESPONSES IN HCV 1723 Figure 2. Targeting of HCVpp entry during postbinding steps by monoclonal antienvelope antibodies. (A and B) Inhibition of HCVpp entry (1a; H77c strain) into Huh7 cells by anti-e1 (A) and anti-e2 (B) monoclonal antibodies (mabs). Inhibition of entry was assessed as described in Figure 1A. (C) Inhibition of HCVpp (1b; HCV J strain) entry by anti-e1 mabs. (D) Dose-dependent inhibition of HCVpp (1a; H77c strain) entry by anti-e2 mab AP33 (solid circles), anti-e1 mab IGH520 (solid squares), or control mouse IgG in experimental protocol II (addition of IgG postbinding). (E) Inhibition of recombinant envelope glycoprotein E1 binding to Huh7 cells by purified anti-e1 antibodies or heparin. Experiments were performed as described in the Materials and Methods section. Cells stained with detection antibodies alone served as negative control ( NC, light shaded histograms). (F) Inhibition of E1 (subtype 1b), HCV-LP (1a), and HCVpp (1b) binding to Huh7 cells by anti-e1, control IgG, or heparin. Data are shown as percentage of inhibition of binding of ligands incubated with PBS (100%). For panels A C and F, mean values SD of at least 6 experiments are shown. For panels D and E, mean values of a representative experiment performed in triplicate are shown. Statistically significant differences (P.01; t test) in inhibition of HCVpp entry between protocol I and II (panels A C) are indicated (*). tometry, 13 we demonstrate that none of the human anti-e1 antibodies targeting E markedly inhibited binding of HCV ligands. In contrast, heparin a well-characterized molecule shown to interfere with HCV infection predominantly during viral attachment 10,13,17 markedly inhibited cellular binding of ligands (Figure 2E and F). These data suggest that epitope E is targeted by neutralizing human anti-e1 antibodies during postbinding steps. Mapping of Entry Steps Targeted by Virus Neutralizing Antibodies Using an HCVpp-Based Kinetic Assay To identify further the cellular targets of anti- HCV antibodies in the HCV entry process, we investigated the inhibitory capacity of anti-hcv IgG, anti- CD81, and anti-sr-bi antibodies and heparin in a kinetic study in which the inhibitory agents were added at vari-

6 1724 HABERSTROH ET AL GASTROENTEROLOGY Vol. 135, No. 5 added at 60 minutes compared with anti-cd81, suggesting that they also interfere at steps in entry earlier than the interaction of HCVpp with CD81. Interestingly, anti- HCV IgG from several patients was able to inhibit HCV infection when added later than 60 minutes postbinding (Figure 4C). It is conceivable that these immunoglobulin fractions contain antibodies that may interfere with HCV entry at later stages of HCV infection such as HCVclaudin interaction or membrane fusion within the host cell endosome. Host Neutralizing Antibodies Inhibit HCV Membrane Fusion The fusion between viral and cellular membranes involves a complex, multistep conformational change of the viral glycoproteins. 22 HCV entry is ph dependent, 19 20,23 25 suggesting that the low ph induces the refolding of the E1E2 glycoprotein complex. The critical domains and the molecular events that mediate HCV membrane fusion remain poorly defined. 20 Such events may provide valuable targets for neutralizing antibodies. 26 To address further whether some antibodies may inhibit cell entry at the step of membrane fusion, we used an in vitro liposome/hcvpp fusion assay that we recently developed This lipid mixing assay is based on the direct measurement of mixing between HCVpp and liposome lipids. In line with previous results, 18 preincubation of HCVpp with ApoC-1 increased lipid mixing with lipo- Figure 3. Kinetics of anti-hcv IgG-mediated inhibition of HCVpp entry. (A) Schematic drawing of the experimental setup. Inhibition of HCVpp entry into Huh7 cells by purified anti-hcv-igg derived from patient or control sera (100 g/ml) was performed as described in legend to Figure 1A, except that patient-derived anti-hcv IgG or control IgG was added every 20 minutes for up to 120 minutes after HCVpp (1a; strain H77c) binding. Dashed lines indicate the time intervals at which antibodies were present. (B) Kinetics of antibody-mediated inhibition of HCVpp (1a; strain H77c) entry by anti-cd81, anti-sr-bi, heparin, and control IgG. (C) Kinetics of antibody-mediated inhibition of HCVpp entry by patient-derived IgG. Purified anti-hcv-igg derived from patients 1, 7, 8, 9, 12, 17, and 19 (Table 1); IgG derived from an anti-hcv-negative control individual 20 (Table 1), and anti-cd81 mab were added to HCVpp bound to target cells as depicted in A. Mean values of a representative experiment performed in triplicate are shown. ous time points postbinding (Figure 3A). The similar kinetic of inhibition of HCV infection observed for anti- CD81, anti-sr-bi, and several patient-derived anti-hcv IgG (Figure 3B and C) suggest that the entry steps targeted by host neutralizing antibodies are closely linked to HCV-CD81 and HCV-SR-BI interaction. Other IgG preparations showed less potency in blocking entry when Figure 4. Inhibition of membrane fusion of HCVpp-lipsomes by patient-derived anti-hcv IgG. The fusion capacities of HCVpp (1a; H77c strain) were tested using lipid mixing assays. The results are expressed as percentages of maximal fluorescence, obtained by addition of Triton X-100 (final 0.1% vol:vol) to the pseudoparticle/liposome suspensions. The experiments were repeated 3 times, and the most representative profiles of the fusion kinetics are presented here. (A) HCVpp were preincubated in the presence or absence of ApoC-I (0.7 g/ml) or control IgG from control individual 20 (100 g/ml). (B D) HCVpp were preincubated in the presence of IgG (100 g/ml) from anti-hcv-negative control individual 20 or purified anti-hcv IgG (100 g/ml) from patients 1, 7, 8, 9, 12, 17, and 19.

7 November 2008 NEUTRALIZING HOST RESPONSES IN HCV 1725 alternatively, by blocking some E1E2 epitopes unmasked by acidic ph during their structural modifications. The results presented in Figure 4 are representative of 3 experiments, on average, for each patient. Although the extent of membrane fusion shows interassay variation because of the relative quality of HCVpp and R18-labeled liposome batches, the inhibitory effect of antibodies on membrane fusion was highly reproducible in independent experiments. For example, the mean percentage of inhibition, calculated here using the initial rate of membrane fusion, was 38.3% 8.2% (n 3) for anti-hcv IgG derived from patient 19 and 34.5% 10.0% (n 3) for patient 12, respectively. Figure 5. Inhibition of HCVcc infection by human anti-hcv neutralizing antibodies. (A) Inhibition of Luc-Jc1 HCVcc entry into Huh7.5.1 cells by heparin, anti-cd81, control mouse IgG, anti-sr-bi serum, or control rat serum was studied as described in Figure 1A and the Materials and Methods section. (B) Inhibition of HCVcc entry into Huh7.5.1 cells by human anti-hcv IgG from patients 3, 7, 9, 12, 17, 19, 22, 23, and 24; control IgG from control individual 21 (100 g/ml); anti-cd81; or anti- SR-BI was performed in protocols I and II as described in Figure 1A. Results are expressed as percentage inhibition relative to control infections performed in the same way but without addition of inhibitors and represent mean SD from 4 independent experiments performed in duplicate. somes as shown by the increased rates of fusion kinetics (Figure 4A). Incubation of HCVpp with control IgG derived from healthy individuals (patient 20) had no significant effect on HCVpp fusion kinetics (Figure 4A). However, when HCVpp were incubated with the same concentration of anti-hcv IgG derived from patients 1, 8, 9, 12, or 19, the fusion rates were markedly reduced as compared with the fusion kinetics of HCVpp incubated with control IgG (Figure 4B and D). These results indicate that neutralizing antibodies from these patients were able to inhibit membrane fusion events induced by the HCV glycoproteins. Such a block may occur either by steric hindrance of antibody-bound HCV E1E2 hence impairing the conformational rearrangements of the glycoproteins necessary for allowing membrane fusion or, Confirmation of Antibody-Mediated Neutralization Occurring at Postbinding Steps Using HCVcc To confirm key results obtained with the HCVpp model system, anti-hcv IgG-mediated inhibition of HCVcc infection of Huh7.5.1 cells was studied using protocols I and II as described above. Because Luc-Jc1 (J6-JFH1) HCVcc bear the envelope glycoproteins of isolate J6 (genotype 2a), we determined the ability of anti- HCV-IgG isolated from patients with sufficient amounts of IgG for side-by-side analyses in neutralization of HCVpp (J6) entry and HCVcc (J6-JFH1) infection. As shown in Figure 5B and Supplementary Table 2 (see Supplementary material online at org), anti-hcv-igg from patients showed a marked inhibition of HCVcc infection in protocol II, similar as seen for HCVpp. Targeting of viral entry steps following postbinding was also confirmed in side-by-side kinetic experiments using HCVcc (Figure 6). As seen for HCVpp (Figure 3), a similar kinetic of inhibition of HCV infection was observed for anti-cd81, anti-sr-bi, and several patient-derived anti-hcv IgG (Figure 6). These results demonstrate (1) that HCV-infected individuals contain antibodies inhibiting HCVcc infection at steps following HCV binding and (2) confirm results obtained with the HCVpp model system (Figures 1 3). Discussion In this study, we observed that host neutralizing antibodies in the majority of HCV-infected individuals target HCV entry during entry steps after binding of virus to the target cell, corresponding to HCV-host cell entry factor interaction and membrane fusion. Furthermore, we identified viral epitopes in envelope glycoproteins representing potential targets for neutralizing antibodies in postbinding events and membrane fusion. Kinetic and fusion assays suggest that host neutralizing responses in HCV-infected patients target HCV entry at postbinding steps most likely related to HCV-CD81, and SR-BI interactions as well as membrane fusion. These findings may have important implications not only for the understanding of the pathogenesis of HCV infection but also

8 1726 HABERSTROH ET AL GASTROENTEROLOGY Vol. 135, No. 5 Figure 6. Kinetics of anti-hcv IgG-mediated inhibition of HCVcc entry. (A) Kinetics of inhibition of Luc-Jc1 HCVcc (genotype 2a) entry by anti- CD81 and anti-sr-bi were performed as described in Figure 3A. (B) Kinetics of antibody-mediated inhibition of HCVcc entry by anti-hcv IgG purified from patients with HCV 2a infection. Purified anti-hcv-igg derived from patients 22, 23, and 24; IgG derived from an anti-hcvnegative control individual 21; and anti-cd81 mab were added to HCVcc bound to target cells as depicted in Figure 3A. (C) Kinetics of antibody-mediated inhibition of HCVcc entry by patient IgG analyzed previously in the HCVpp system (Figure 4C). Purified anti-hcv-igg derived from patients 9, 7, 12, and 19 (Figure 3C) were added to HCVcc bound to target cells as described in B. Results are expressed as percentage HCVcc infection relative to infection performed in the absence of antibody and represent mean values from 1 representative experiment performed in duplicate. for the design of novel immunotherapeutic and preventive strategies. Targeting of Postbinding Entry Steps by Host Neutralizing Antibodies In analogy to other viral infections, antibodies neutralizing HCV may render virions noninfectious by interfering with different steps of the viral life cycle. 2 Binding of the antibodies to the virus may directly block binding of the virus to the host cell and thus inhibit dissemination of infection. Neutralizing antibodies may also interfere with postbinding steps as interaction of the virus with host entry factors such as SR-BI, CD81, and claudin-1. 2,3 If endocytosis is a mandatory step, internalization of the virus by endocytosis may also be a target for neutralization. Neutralization of viruses by antibodies may also take place during fusion at the cell surface or in endosomes: neutralizing antibodies may directly interfere with the fusogenic protein, hinder conformational changes necessary for the fusion process, or simply obstruct contact between cellular and viral membranes. In this study, we demonstrate that neutralizing antibodies in HCV-infected individuals interfere with HCVpp and HCVcc entry on postbinding steps, occurring at a similar time point as the HCV-CD81 and HCV-SR-BI interaction. Interference of anti-hcv IgG with HCV entry at later steps may comprise interference with HCV-claudin-1 interaction or membrane fusion as shown in Figure 4. Furthermore, our results do not exclude that binding events could also be affected. Antibody-mediated interference with postbinding steps and fusion have been observed for other members of the Flaviviridae family including West Nile virus 27 and Dengue virus. 28 Taken together, these results indicate that targeting of postbinding steps including membrane fusion is a key mechanism for efficient HCV neutralization. As shown in various model systems, envelope glycoproteins E1 and E2 are critical for host cell entry and thus represent important targets for virus neutralization. Antibodies targeting both linear and conformational epitopes of envelope glycoprotein E2 have been shown to inhibit HCV infection. 2,5,8,15,16,29 However, the precise entry step targeted by neutralizing antibodies has not yet been defined. In this study, we demonstrate that epitopes of envelope glycoprotein E1 and E2 are targeted by neutralizing monoclonal antibodies during postbinding events. The identification of epitope E as a viral factor involved in postbinding events (Figure 2) underlines the important functional impact of envelope glycoprotein E1 within the HCV entry process. Because the monoclonal anti-e1 antibodies directed against E have been generated from infected humans with HCV clearance, 8 it is conceivable that this epitope is targeted by host neutralizing responses in vivo. Confirmation of key findings in the HCVcc J6-JFH1 (Jc1) model system (Figures 5 and 6) suggests that mechanisms of neutralization in vitro appear to be similar between HCVpp and HCVcc infection of hepatoma cells. This is in line with recent studies demonstrating that in vitro neutralization in the HCVpp model system correlated with neutralization of HCVcc in vitro 30 and in the human liver upa-scid mouse in vivo. 30,31 Although differences in the export pathway of HCVpp and native HCV may result in consequences for the properties of

9 November 2008 NEUTRALIZING HOST RESPONSES IN HCV 1727 produced viral envelope in regard to the association with lipoproteins, 32,33 the HCVpp system has been used by many investigators because it is characterized by high robustness and the ability to perform high-throughput assays allowing the quantification of virus neutralization of a large number of patient IgG or mabs as required also for our study. Nevertheless, our results do not exclude that subtle differences between the systems exist. Impact of Interference of Antibodies With HCV Entry for Pathogenesis of HCV Infection Using the HCVpp model system, 2 studies have demonstrated that neutralizing antibodies are induced in the early phase of infection by patients who subsequently clear or control viral infection. 5,7 Patients who do not clear the virus develop high-titer and cross-neutralizing antibodies during the chronic phase of infection. 5,6,34 Paradoxically, these antibodies are not able to control HCV infection. 5,6,34 Studies in a chronically HCV-infected patient who has been followed for 30 years demonstrated that HCV continuously escapes the host neutralizing response by mutations, resulting in loss of recognition of HCV envelope glycoproteins by antibodies, 6 suggesting that the neutralizing antibody response of the host lags behind the rapidly evolving HCV envelope glycoprotein sequences of the quasispecies population. 6 In fact, neutralization of heterologous strains does not reflect neutralization of the viral variants present in the patient s serum at the time of sampling. 6 This observation may explain the fact that neutralizing antibodies identified in patients of this study were not able to control chronic HCV infection. Taken together, these findings provide strong evidence that both efficient antibody-mediated neutralization as well as viral escape from host neutralizing responses occurs during postbinding events of HCV entry. In conclusion, our results unravel the impact of HCV-host factor interactions during viral postbinding steps and fusion for control of viral infection and viral escape. The further identification and characterization of the mechanisms underlying antibody-mediated interference with viral entry will be important for the understanding of the pathogenesis of HCV infection as well as the development of antibody-based therapeutic or preventive strategies. Supplementary Data Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at and at doi: /j.gastro References 1. Bowen DG, Walker CM. Adaptive immune responses in acute and chronic hepatitis C virus infection. Nature 2005;436: Barth H, Liang TJ, Baumert TF. Hepatitis C virus entry: molecular biology and clinical implications. Hepatology 2006;44: Zeisel MB, Cosset FL, Baumert TF. Host neutralizing responses and pathogenesis of hepatitis C virus infection. Hepatology 2008;48: Farci P, Shimoda A, Wong D, et al. Prevention of hepatitis C virus infection in chimpanzees by hyperimmune serum against the hypervariable region 1 of the envelope 2 protein. Proc Natl Acad SciUSA1996;93: Pestka JM, Zeisel MB, Blaser E, et al. Rapid induction of virusneutralizing antibodies and viral clearance in a single-source outbreak of hepatitis C. Proc Natl Acad Sci U S A 2007; 104: von Hahn T, Yoon JC, Alter H, et al. Hepatitis C virus continuously escapes from neutralizing antibody and T-cell responses during chronic infection in vivo. Gastroenterology 2007;132: Lavillette D, Morice Y, Germanidis G, et al. Human serum facilitates hepatitis C virus infection, and neutralizing responses inversely correlate with viral replication kinetics at the acute phase of hepatitis C virus infection. J Virol 2005;79: Meunier JC, Russell RS, Goossens V, et al. Isolation and characterization of broadly neutralizing human monoclonal antibodies to the e1 glycoprotein of hepatitis C virus. J Virol 2008;82: Wellnitz S, Klumpp B, Barth H, et al. Binding of hepatitis C virus-like particles derived from infectious clone H77C to defined human cell lines. J Virol 2002;76: Zeisel MB, Koutsoudakis G, Schnober EK, et al. Scavenger receptor BI is a key host factor for hepatitis C virus infection required for an entry step closely linked to CD81. Hepatology 2007;46: Steinmann D, Barth H, Gissler B, et al. Inhibition of hepatitis C virus-like particle binding to target cells by antiviral antibodies in acute and chronic hepatitis C. J Virol 2004;78: Owsianka A, Tarr AW, Juttla VS, et al. Monoclonal antibody AP33 defines a broadly neutralizing epitope on the hepatitis C virus E2 envelope glycoprotein. J Virol 2005;79: Barth H, Schnober EK, Zhang F, et al. Viral and cellular determinants of hepatitis C virus envelope-heparan sulfate interaction. J Virol 2006;80: Zhong J, Gastaminza P, Cheng G, et al. Robust hepatitis C virus infection in vitro. Proc Natl Acad Sci U S A 2005;102: Bartosch B, Dubuisson J, Cosset FL. Infectious hepatitis C virus pseudo-particles containing functional E1-E2 envelope protein complexes. J Exp Med 2003;197: Wakita T, Pietschmann T, Kato T, et al. Production of infectious hepatitis C virus in tissue culture from a cloned viral genome. Nat Med 2005;11: Koutsoudakis G, Kaul A, Steinmann E, et al. Characterization of the early steps of hepatitis C virus infection by using luciferase reporter viruses. J Virol 2006;80: Dreux M, Boson B, Ricard-Blum S, et al. The exchangeable apolipoprotein ApoC-I promotes membrane fusion of hepatitis C virus. J Biol Chem 2007;282: Lavillette D, Bartosch B, Nourrisson D, et al. Hepatitis C virus glycoproteins mediate low ph-dependent membrane fusion with liposomes. J Biol Chem 2006;281: Lavillette DP, Pecheur EI, Donot P, et al. Characterization of fusion determinants points to the involvement of three discrete regions of both E1 and E2 glycoproteins in the membrane fusion process of hepatitis C virus. J Virol 2007;81: Dreux M, Pietschmann T, Granier C, et al. High-density lipoprotein inhibits hepatitis C virus-neutralizing antibodies by stimulating cell entry via activation of the scavenger receptor BI. J Biol Chem 2006;281: Kielian M, Rey FA. Virus membrane-fusion proteins: more than one way to make a hairpin. Nat Rev Microbiol 2006;4:67 76.

10 1728 HABERSTROH ET AL GASTROENTEROLOGY Vol. 135, No Bartosch B, Vitelli A, Granier C, et al. Cell entry of hepatitis C virus requires a set of co-receptors that include the CD81 tetraspanin and the SR-B1 scavenger receptor. J Biol Chem 2003; 278: Hsu M, Zhang J, Flint M, et al. Hepatitis C virus glycoproteins mediate ph-dependent cell entry of pseudotyped retroviral particles. Proc Natl Acad Sci U S A 2003;100: Tscherne DM, Jones CT, Evans MJ, et al. Time- and temperaturedependent activation of hepatitis C virus for low-ph-triggered entry. J Virol 2006;80: Kwong PD, Doyle ML, Casper DJ, et al. HIV-1 evades antibodymediated neutralization through conformational masking of receptor-binding sites. Nature 2002;420: Nybakken GE, Oliphant T, Johnson S, et al. Structural basis of West Nile virus neutralization by a therapeutic antibody. Nature 2005;437: Se-Thoe SY, Ling AE, Ng MML. Alteration of virus entry mode: a neutralisation mechanism for dengue-2 virus. J Med Virol 2000; 62: Lindenbach BD, Evans MJ, Syder AJ, et al. Complete replication of hepatitis C virus in cell culture. Science 2005;309: Law M, Maruyama T, Lewis J, et al. Broadly neutralizing antibodies protect against hepatitis C virus quasispecies challenge. Nat Med 2008;14: Vanwollehgem T, Bukh J, Meuleman P, et al. Polyclonal immunoglobulins from a chronic HCV patient protect human liver-chimeric mice from infection with a homologous HCV strain. Hepatology 2008;48: Huang H, Sun F, Owen DM, et al. Hepatitis C virus production by human hepatocytes dependent on assembly and secretion of very low-density lipoproteins. Proc Natl Acad Sci U S A2007;104: Miyanari Y, Atsuzawa K, Usuda N, et al. The lipid droplet is an important organelle for hepatitis C virus production. Nat Cell Biol 2007;9: Logvinoff C, Major ME, Oldach D, et al. Neutralizing antibody response during acute and chronic hepatitis C virus infection. Proc Natl Acad Sci U S A 2004;101: Received October 11, Accepted July 17, Address requests for reprints to: Thomas F. Baumert, MD, Inserm Unit 748, Service d Hépatogastroentérologie, Université Louis Pasteur, 3 Rue Koeberlé, F Strasbourg, France. Thomas. Baumert@viro-ulp.u-strasbg.fr; fax: (33) The authors disclose the following: Supported by the EU (LSHM-CT to T.F.B. and R.B.; LSHB-CT to F-L.C.), the Deutsche Forschungsgemeinschaft (Ba1417/11-2 to T.F.B.), Inserm, ANR (ANR-05-CEXC-008 to T.F.B.), ANRS (No to T.F.B., No to F.S.-K.), the Ligue Nationale Contre le Cancer (to F.-L.C.), the Else Kröner-Fresenius Foundation (P17/07//A83/06 to T.F.B. and H.B.), and fellowships from Inserm Poste Vert and Région Rhône-Alpes (to M.B.Z. and M.D., respectively). The study was performed in the framework of Inserm European Associated Laboratory (EAL), University of Freiburg and Inserm U748 Strasbourg. The authors thank M. Houghton, H. B. Greenberg, F. V. Chisari for gift of reagents; B. Gissler for excellent technical assistance; C. Royer for IgG purification, M. Dimitrova for HCVcc production; and E. Schvoerer for genotyping. A.H. and E.K.S. contributed equally to this study. H.B. s current address is Liver Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland. E.D. s current address is Ablynx, Ghent, Belgium.

11 November 2008 NEUTRALIZING HOST RESPONSES IN HCV 1728.e1 Supplementary Table 1. Mapping of HCV Epitopes Involved in Postattachment Steps of Viral Entry Neutralization of HCVpp entry, % Genotype 1a Genotype 1b Antibody Source or reference Epitope (aa) I II I II Anti-E1 IGH433 GNI (1b) IGH520 GNI (1b) IGH526 GNI (1b) IGH481 GNI (1b) 0 0 ND ND IGH534 GNI (1b) B7 Wellnitz et al (1b) Anti-E2 AP33 Pestka et al 2 and Owsianka et al (1a) ND ND 3E5-1 Wellnitz et al (1a) E2G Wellnitz et al 1 ND (1a) Wellnitz et al (1b) ND ND IGH466 GNI (1b) IGH461 GNI (1b) NOTE. Antibodies were added to human hepatoma Huh7 cells pre- (protocol I) and postbinding (protocol II) of HCVpp. HCVpp entry was analyzed as described in Figure 1A. Viral epitopes targeted by the respective antibody, percent neutralization of HCVpp (1a; H77c strain and 1b; HCV-J strain) entry in protocols I and II, and source or reference of antibody are shown (mean SD from at least 6 independent experiments). aa, Amino acids; ND, not determined; GNI, GENimmune NV, Ghent, Belgium. Supplementary Table 2. Side-by-Side Analysis of Inhibition of HCVpp Entry and HCVcc Infection by Patient-Derived anti-hcv IgG Neutralization of HCVpp 2a entry, % Neutralization of HCVcc infection, % Patient No. Age, y Genotype Viral load (IU/ ml) I II I II b a a b b a a No infection a a a NOTE. IgG were purified from patients with chronic HCV infection. Patient-derived IgG were added to hepatoma cells pre- (protocol I) and postbinding (protocol II) of HCVpp or HCVcc. Neutralization of HCVpp (J6 strain; genotype 2a) entry and HCVcc (J6-JFH1/Jc1; genotype 2a) infection was assessed as shown in Figures 1 and 5. Results are expressed as percent inhibition relative to control infections performed in the same way but without addition of antibody (mean SD of at least 3 independent experiments for HCVpp and at least 4 independent experiments for HCVcc are shown). References 1. Wellnitz S, Klumpp B, Barth H, et al. Binding of hepatitis C viruslike particles derived from infectious clone H77C to defined human cell lines. J Virol 2002;76: Pestka JM, Zeisel MB, Blaser E, et al. Rapid induction of virusneutralizing antibodies and viral clearance in a single-source outbreak of hepatitis C. Proc Natl Acad Sci U S A 2007;104: Owsianka A, Tarr AW, Juttla VS, et al. Monoclonal antibody AP33 defines a broadly neutralizing epitope on the hepatitis C virus E2 envelope glycoprotein. J Virol 2005;79: