GASTROENTEROLOGY 2007;132:667– 678 Hepatitis C Virus Continuously Escapes From Neutralizing Antibody and T-Cell Responses During Chronic Infection In Vivo THOMAS VON HAHN,* JOO CHUN YOON,‡ HARVEY ALTER,§ CHARLES M. RICE,* BARBARA REHERMANN,‡ PETER BALFE,储,¶ and JANE A. MCKEATING*,¶ *Center for the Study of Hepatitis C, The Rockefeller University, New York, New York; ‡Immunology Section, Liver Diseases Branch, and the §Department of Transfusion Medicine, Warren Grant Magnuson Clinical Center, National Institutes of Health, Bethesda, Maryland; 储Division of Infectious Diseases, Columbia University, New York, New York; and the ¶Division of Immunity and Infection, Institute of Biomedical Research, Medical School, University of Birmingham, Birmingham, United Kingdom Background & Aims: Broadly reactive neutralizing antibodies (nAbs) and multispecific T-cell responses are generated during chronic hepatitis C virus (HCV) infection and yet fail to clear the virus. This study investigated the development of autologous nAb and HCV-glycoprotein–specific T-cell responses and their effects on viral sequence evolution during chronic infection in order to understand the reasons for their lack of effectiveness. Methods: Numerous E1E2 sequences were amplified and sequenced from serum samples collected over a 26-year period from patient H, a uniquely well-characterized, chronically infected individual. HCV pseudoparticles (HCVpp) expressing the patient-derived glycoproteins were generated and tested for their sensitivity to neutralization by autologous and heterologous serum antibodies. Results: A strain-specific nAb response developed early in infection (8 weeks postinfection), whereas cross-reactive antibodies able to neutralize HCVpp-bearing heterologous glycoproteins developed late in infection (>33 wk postinfection). The humoral response continuously failed to neutralize viruses bearing autologous glycoprotein sequences that were present in the serum at a given time. The amplified glycoprotein sequences displayed high variability, particularly in regions corresponding to defined linear B-cell epitopes. Mutations in defined neutralizing epitopes were associated with a loss of recognition by monoclonal antibodies against these epitopes and with decreased neutralization of corresponding HCVpp. Viral escape from CD4 and CD8 T-cell responses also was shown for several novel epitopes throughout the glycoprotein region. Conclusions: During chronic infection HCV is subjected to selection pressures from both humoral and cellular immunity, resulting in the continuous generation of escape variants. H epatitis C virus (HCV) is an important human pathogen infecting about 170 million people worldwide. In the United States, it is the single most common cause of chronic liver disease requiring liver transplantation.1,2 Cellular and humoral responses are generated during acute infection, however, they are insufficient to achieve viral clearance in the majority of individuals, with approximately 60%– 80% of new infections becoming persistent. In vivo, HCV replicates to high levels using an errorprone viral RNA polymerase, which leads to a spectrum of related but distinct sequences within infected individuals, often referred to as a quasispecies.3 The immune system is thought to exert unequal selective pressure on variants within the circulation, favoring the rapid emergence of escape mutants.4 Escape from CD8 T-cell responses by mutation is well documented and an important predictor of progression to chronic HCV infection.5– 8 In keeping with this, a vigorous and broad CD8 T-cell response during the acute phase of infection is associated with viral clearance.9 –11 Studies of acute HCV infection in chimpanzees previously exposed to the virus provide compelling evidence that protective CD8 T-cell–mediated immunity exists.12 CD4 T cells also are required for control of HCV on re-exposure, but their role is less well defined.13 Even less is known about the impact of the humoral immune response on HCV pathobiology. Without the ability to culture HCV, there was, until recently, no simple in vitro method to evaluate viral escape from the antibody-mediated immune response. The development of HCV glycoprotein-bearing retroviral pseudoparticles (HCVpp) has made it possible to assess antibody-dependent neutralization of HCV entry.14 –18 Neutralizing anAbbreviations used in this paper: EIA, enzyme-linked immunosorbent assay; gp, glycoprotein; HCVpp, HCV pseudoparticle; HVR, hypervariable region; MAb, monoclonal antibody; nAb, neutralizing antibody; PBMC, peripheral blood mononuclear cell; PCR, polymerase chain reaction; sE2, soluble HCV E2. © 2007 by the AGA Institute 0016-5085/07/$32.00 doi:10.1053/j.gastro.2006.12.008 BASIC–LIVER, PANCREAS, AND BILIARY TRACT See editorial on page 801. 668 VON HAHN ET AL BASIC–LIVER, PANCREAS, AND BILIARY TRACT tibody (nAb) responses often provide the first-line adaptive defense against infection by limiting virus spread. Serum antibodies from chronically HCV-infected individuals show broadly reactive neutralizing properties and yet fail to clear viral infection.15,17,19 The reasons for their lack of effect are understood poorly, however, one possible explanation is that Ab response(s) are less able to neutralize autologous glycoprotein species circulating within an individual at the time of sampling. In setting out to address these questions we were fortunate to have access to sequential serum samples from patient H, an individual who was infected with HCV in 1977 and has been meticulously followed-up since then. Moreover, patient H was the source of the prototype HCV strains H and H77, and thus a wide range of tailor-made reagents are available for virologic and immunologic analyses.20 –22 For these reasons this patient is a unique subject for studies into the immunologic history of chronic HCV infection. To assess the impact of the HCV quasispecies on the nAb response we generated HCVpp-bearing glycoprotein variants cloned from sequential samples from chronically infected patient H. By using a series of samples obtained between 3 weeks postinfection and throughout the 26 years thereafter, we sought to investigate the process of antigenic escape in the viral glycoproteins from humoral and cellular immune surveillance. We found compelling evidence for repeated mutational change resulting in loss of recognition of the HCV glycoprotein by the cognate antibody response and escape from antibody-mediated neutralization. Similarly, mapping of T-cell responses to the E1E2 region identified 4 novel T-cell epitopes in which mutations occurred, leading to escape from CD4 and CD8 T-cell recognition. Materials and Methods Antibodies, Cells, and Reagents Both 293T and Hep3B cells were propagated in Dulbecco’s modified Eagle medium with 10% fetal bovine serum. Antibodies against HCV E2 have been described previously.14,23 To generate soluble CD81 an expression plasmid encoding the CD81 large extracellular loop fused to Glutathione S-transferase (GST) was transformed into Rosetta-gami Escherichia coli (Novagen, La Jolla, CA). Fusion proteins were prepared by lysis with an Avestin (Avestin, Ottawa, Canada) air emulsifier (3 passages at 15,000 psi) and subsequent centrifugation (25,000 ⫻ g for 30 min at 4°C). Cleared lysates were purified over a GSTrap FF affinity column according to the manufacturer’s instructions (Amersham Biosciences, Piscataway, NJ). HCV Antigens For the initial screening of T-cell responses against the HCV E1 and E2 proteins, 15-mer peptides (total, 111) (Mimotopes, Clayton, Australia), overlapping GASTROENTEROLOGY Vol. 132, No. 2 by 10 amino acids and covering the complete HCV H77 (genotype 1a) E1E2 sequence, were resuspended at 20 mg/mL in dimethyl sulfoxide and further diluted with phosphate-buffered saline (PBS) solution to obtain a final concentration of 1 ␮g/mL. For direct comparison of responses against the mutant epitopes, the corresponding mutant peptides were synthesized at Rockefeller University (New York, NY) to more than 95% purity. Polymerase Chain Reaction Amplification, Cloning, and Sequence Analysis of HCV E1E2 The plasmid encoding H77 E1E2 has been described previously.14,21,24 Total RNA was prepared from HCV-infected plasma using commercial kits (Qiagen, Valencia, CA).25 Briefly, complementary DNA (cDNA) was synthesized in a reaction volume of 20 ␮L, containing 2–5 ␮L of template RNA, 2.5 U of Multiscribe MMuLV reverse transcriptase with 400 ␮mol/L each of the 4 deoxynucleoside triphosphates and 200 nmol/L of the antisense primer p7-2710 (AGC AGG AGG AGN GGC CAY ATC CCR TAG A, Y ⫽ C/T mixture, R ⫽ A/G, N ⫽ A/G/C/T) in the manufacturer’s recommended buffer (N808-0234; ABI, Foster City, CA) for 2 hours at 42°C. This cDNA was used as the template for polymerase chain reaction (PCR) amplification of the E1E2 region as previously described.24 Briefly, a 50-␮L PCR was set up containing 2.5 ␮L of cDNA, 2.5 U of the proofreading Expand polymerase mixture (1 681 834; Roche, Mannheim, Germany) in 1⫻ Expand buffer 3 (2.25 mmol/L Mg2⫹), with 400 ␮mol/L each of the 4 deoxynucleoside triphosphates and 200 nmol/L each of the primers core⫹813 (GAG GAC GGY RTR AAY TAY GCA ACA GG; sense) and p7-2710. The PCR consisted of 30 cycles at 92°C for 45 seconds, 45°C for 45 seconds, and 68°C for 300 seconds, and was performed in an Eppendorf (Westbury, NY) thermal cycler. A total of 2 ␮L of the completed reaction was used as template for a second amplification, containing the same reaction components as described previously with 200 nmol/L of the primers core⫹843 (CACC ATG GGT TGC TCT TTC TCT ATC TT; sense) and E2-2580H (CTA CTA CGC CTC CGC TTG GGA TAT GAG TAA CAT CAT CCA, antisense). This second round of PCR comprised 25 cycles at 92°C for 35 seconds, 55°C for 35 seconds, and 68°C for 150 seconds. In those cases in which the input RNA was more than 2000 viral copies, full-length E1E2 was amplified readily. PCR products were cloned into pcDNA3.1D-TOPO (K490001; Invitrogen, Carlsbad, CA) and the sense and antisense strands were sequenced (Big Dye 3.1 Terminator Kit; ABI). All sequences were deposited with Genbank and have the accession numbers DQ897773–DQ897818. Several sequences for the E1E2 region of HCV within this patient have been deposited in Genbank previously. For comparison with the new sequences described here we included the following sequences from 1977 in our analyses: H77C (AF01175120), HPCST77 (M6238122), H77IMC (AF00960621), H21 (AF01175320), H (M6746326), and H11 (AF01175220). An additional clone, H90, obtained in 1990, was available and was included in the analysis (M6238222). In addition to the sequences recorded in Genbank, several sequences for the E2 hypervariable region (HVR) were reported by Ogata et al,22 and several matched the HVR sequences of clones obtained in this study (data not shown). Phylogenetic Analyses The nucleotide sequences were aligned and translated using the SeAL2.0 program (A. Rambaut, Oxford University, available at: http://evolve.zoo.ox.ac.uk). Synonymous and nonsynonymous distances were estimated using the Nei and Gojobori27 method implemented in the PAML3.14 program.28 Phylogenetic analyses were performed using the PAUP 4.0 software package29 using a modified HKY85 evolutionary model for the data, selected by hierarchic likelihood testing (program Modeltest 3.630), the transition/transversion ratio (7.04), proportion of invariable sites (.46), and gamma distribution shape parameter (␣) for variable sites (.86) were estimated by maximum likelihood methods; distance estimates were averaged within and between groups using Excel 2001 (Microsoft, Redmond, WA). Pseudoparticle Production, Infection, and Neutralization Assays Pseudoparticles were generated by transfection of 293T cells with pNL4-3.Luc.R⫺E⫺ plasmid containing the env-defective proviral genome and an expression plasmid encoding the HCV glycoproteins (gps), as previously described.14 The virus containing extracellular media was collected 48 and 72 hours after transfection. Heat-inactivated sera, monoclonal antibodies (MAbs), or soluble CD81 were incubated with virus at their appropriate dilution in 3% fetal bovine serum/Dulbecco’s modified Eagle medium plus 4 ␮g/mL polybrene at 37°C for 1 hour. The virus-Ab mixture was transferred to Hep3B cells seeded in 96-well plates (8 ⫻ 103 cells/well) and infections were centrifuged at 400g for 1 hour, incubated at 37°C for 6 hours, and unbound virus was removed and incubated for a total of 72 hours. Cells were lysed with cell lysis buffer (Promega, Madison, WI) and tested for luciferase activity as previously described.14 The percentage neutralization was determined by comparing pseudoparticle infectivity (luciferase relative light units) in the presence of a test serum or MAb with infection in the presence of a control HCV-negative serum or an irrelevant isotype-matched immunoglobulin (Ig)G at the same dilution. Quantitative Enzyme-Linked Immunosorbent Assay GNA lectin (Sigma, St. Louis, MO) was used to coat Immulon II enzyme-linked immunosorbent assays HCV ESCAPES NEUTRALIZING ANTIBODIES AND CTL 669 (EIA) plates (Nunc, Rochester, NY) at 1 ␮g/mL for 4 hours at 37°C. After washing with PBS, the plates were blocked with 5% bovine serum albumin/PBS and lysates of cells expressing HCV E1E2 or pelleted virus allowed to bind overnight at 4°C. A preparation of truncated E2661 was used as an internal calibrator in all EIAs and allowed comparison of data between different assays. Bound antigen was visualized with MAbs specific for E2 or pooled HCV-positive human sera, an antispecies IgG– horseradish-peroxidase conjugate (Jackson, West Grove, PA) and tetramethylbenzidene (BioFX Laboratories, Owings Mills, MD). Absorbance values were measured at 450 nm (fusion plate reader; Perkin Elmer, Boston, MA). Isolation of CD4 and CD8 T Cells Peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation as previously described.31 For selected experiments, CD8 T cells were isolated from PBMCs using CD8 microbeads and the autoMACS separator (Miltenyi, Bergisch-Gladbach, Germany). Subsequently, CD4 T cells were isolated from the CD8-negative population using CD4 microbeads. The purity of the CD8 T-cell population was 95% and the purity of the CD4 T-cell population was 92%, as analyzed by flow cytometry. The negatively selected CD4⫺CD8⫺ cells were irradiated (3000 rad) and 105 cells per well were used as feeder cells in enzyme-linked immunospot assays. Enzyme-Linked Immunospot Assay Interferon-␥ enzyme-linked immunospot assays were performed as described32 using duplicate cultures of either the indicated number of CD4 and CD8 T cells and 105 irradiated (3,000 rad) CD4⫺CD8⫺ cells, or 3 ⫻ 105 CD25-depleted PBMCs. Cells were stimulated with either 1 ␮g/mL of the individual HCV E1E2 peptides, 10 ␮g/mL of either wild-type or mutant epitope peptides, 1 ␮g/mL phytohemagglutinin (Murex Biotech Limited, Dartford, England), or dimethyl sulfoxide control, respectively. The number of spot-forming cells was determined with a KS enzyme-linked immunospot Reader (Zeiss, Thornwood, NY). Numbers of antigen-specific spot-forming cells (in the presence of antigen minus spot-forming cells in dimethyl sulfoxide controls) are shown. Results Clinical Information Patient H, the subject of this study, was infected with HCV through a blood transfusion in 1977 while undergoing cardiac surgery. After an initial high alanine aminotransferase peak (2112 IU/L), indicative of severe acute hepatitis, he went on to develop mild chronic hepatitis with persistently low alanine aminotransferase levels and HCV RNA levels in the 104 and 106 genome copies/mL range over the following 26 years (Figure 1A). BASIC–LIVER, PANCREAS, AND BILIARY TRACT February 2007 670 VON HAHN ET AL GASTROENTEROLOGY Vol. 132, No. 2 Development of Antiglycoprotein Responses The appearance of the early strain-specific nAb response was associated with the detection of antibodies specific for H77 E1E2 and the HVR in an EIA assay (Figure 1B). However, high levels of antibodies capable of binding H77-soluble E2 (sE2) or J6 E1E2 were first detected at 111 weeks, coincident with the appearance of cross-reactive nAbs and an increase in the titer of H77specific nAbs (Figure 1A and B). In contrast to the other steadily increasing anti-gp responses, the Ig response specific for the H77 HVR peaked at 9 weeks after infection and decreased to undetectable levels by 33 weeks (Figure 1B), suggesting a dynamic and rapidly evolving immune response to the viral gps in the early phase of infection. As previously reported by others,34 the anti-gp response was restricted to the IgG1 isotype (data not shown). The Early Neutralizing Response Targets the HVR BASIC–LIVER, PANCREAS, AND BILIARY TRACT Figure 1. Evolving anti-HCV E1E2 response in patient H over time. (A) Lines show serum ALT (- - -) and viral load (〫 and grey line) over 26 years of chronic infection. Bars represent the serum dilution capable of neutralizing HCVpp-H77 (□) or HCVpp -J6 () infectivity by 90% (ID90). Arrows indicate samples from which E1E2 sequences were amplified and included in the phylogenetic analysis. (B) The reactivity of sequential serum samples with antigens representing HCV E1E2 sequences by EIA. Antigens tested were as follows: 〫, H77 HVR peptide; ⽧, H77 E1E2; ⌬, J6 E1E2; Œ, H77 sE2. (C) HCVpp H77 infection of Hep3B cells in the presence of neutralizing E2 antibodies 9/27 and 11/20 (at 1 ␮g/mL) or H serum (week 9 at 1:300; week 244 at 1:6000, and week 759 at 1:10,000 dilution) with or without H77 sE2 (5 ␮mol/L) or HVR peptide (5 ␮mol/L; mean of n ⫽ 3 ⫾ SD). MOCK, mock infected cells; NO ENV, cells infected with an envelope-deficient pseudoparticle; RLU, relative light units. Patient H underwent liver biopsy procedures on 4 occasions, each showing minimal inflammation with no fibrosis and no discernable progression over time. As expected, liver function was well preserved. Thus, patient H is representative of many chronically HCV-infected patients with mild, nonprogressive disease.33 Development of Strain-Specific and Cross-Reactive nAb Responses Serial serum samples from patient H covering 26 years of chronic HCV infection were used for this study. All samples were characterized for their ability to neutralize HCVpp-bearing autologous H77 gps (a sequence cloned at 3 weeks after infection) and heterologous gps of the closely related genotype 1b (Con1, OH8) and the distant 2a (JFH, J6). In keeping with our previous report,17 HCVpp-H77–specific nAbs were observed at seroconversion (8 weeks after infection) (Figure 1A and data not shown). Neutralization of HCVppbearing heterologous gps was first detected at 111 weeks after infection, at a time when chronic infection had been established. To characterize epitopes targeted by the polyclonal nAb response, sE2 and peptide pools covering the entire H77 E1E2 region were screened for their ability to compete with acute-phase serum neutralization of HCVpp-H77 (data not shown). The only peptide able to inhibit serum neutralization represented the HVR. As controls, increasing concentrations of H77 sE2 and HVR peptide did not affect HCVpp-H77 infectivity but specifically reduced the neutralizing activity of MAbs targeting the HVR (MAb 9/27) or a non-HVR E2 epitope (MAb 11/20) (Figure 1C). sE2 and HVR peptide prevented the neutralization of HCVpp-H77 by serum antibodies collected after 9 weeks, but not after 244 or 759 weeks, consistent with the early strain-specific neutralizing response targeting predominantly the HVR and later responses evolving to a breadth and/or titer that could not be blocked by H77 HVR or sE2. E1E2 Diversity and Generation of Autologous HCVpp Patient H developed an nAb response to HCVppH77 that increased in breadth and titer over time, acquiring the ability to neutralize distantly related HCV genotypes. Yet, paradoxically, this failed to control HCV replication, as evidenced by fluctuating yet robust HCV RNA levels throughout the course of infection. One possible explanation is that Ab neutralization of heterologous isolates does not reflect the response against the virus population present within the serum at the same time point. To address this issue we cloned the E1E2 region from plasma samples collected over a 26-year period. Although no reverse-transcription PCR products were obtained from serum samples between 1977 and 1991, we were able to generate and sequence multiple clones originating from 1977, 1991, 1992, 1993, 1994, 1995, 2002, and 2003. These were used to determine the consensus February 2007 HCV ESCAPES NEUTRALIZING ANTIBODIES AND CTL 671 sequence at the time of sampling and the quasispecies diversity. Four E1E2 clones included stop codons and were not studied further. Clones subsequently are referred to as Haa.bb, where aa indicates the year of the serum sample and bb indicates the clone number. Phylogenetic analysis of the sequences identified 3 major groups (Figure 2A). The diver- sity within a single sample from 1977 was 1.1% based on the entire 1734-bp sequence. This was comparable with that seen between 1991 and 1995 (1.3%) and 2002 and 2003 (1.5%). When only phylogenetically informative sites (ie, those with changes occurring in at least 2 independent clones) were analyzed, the intrasample diversity was 12.5%, 10.4%, and 13.3% for 1977, 1991–1995, and 2002–2003, respectively. To assess the autologous nAb response in patient H, all E1E2 clones that were free of stop codons were coexpressed individually with pNL4-3.Luc.R⫺E⫺ in 293T cells to generate HCVpp. All clones expressed E1E2 at comparable levels (data not shown), however, only 9 of 42 produced infectious HCVpp (Figure 2B and data not shown). Comparison of sequences generating infectious HCVpp with their nearest neighbors revealed that a single amino acid change often could distinguish between viable and nonviable sequences (eg, S234L in H77 vs H77.16). Often, the nonviable clones possessed unique, noninformative changes that were atypical of the group. Figure 2. Generation of HCVpp-bearing glycoprotein clones from H serum samples. (A) The phylogenetic tree displays the relationships between E1E2 sequences amplified from H serum samples. The 2 digits after H define the year of the serum sample from which the clone was obtained. *Clones that gave rise to infectious HCVpp. (B) Infectivity of functional HCVpp generated using sequences amplified from H serum samples. This constitutes the set of HCVpp that was used for subsequent analyses (mean of 3 experiments ⫾ SD). Development of Autologous nAb Responses All gp sequences cloned from patient H in 1977, 1991, 1992, 1995, and 2002 that gave rise to functional HCVpp and the H77 sequence were tested for their sensitivity to neutralization by autologous serum antibodies. As previously noted, the serum samples collected before BASIC–LIVER, PANCREAS, AND BILIARY TRACT Figure 3. Loss of antigen recognition by monoclonal antibodies raised against the H77 sequence over time. Lysates from 293T-cells expressing the H-derived E1E2 clones were immobilized on GNA-coated EIA plates and the ability of 4 neutralizing anti-E2 MAbs raised against the H77 sequence (3/11, 7/16, 9/27, 11/20) to bind the respective glycoproteins was tested. To control for variation in antigen expression level, optical density readings were normalized relative to the signal obtained with a nonneutralizing antibody, 9/75, that recognizes an epitope that was fully conserved among all H isolates. The value obtained with H77 E1E2 as the capture antigen was normalized to 1. 672 VON HAHN ET AL GASTROENTEROLOGY Vol. 132, No. 2 suggest that the nAb response fails to efficiently neutralize gp sequences that are dominant in vivo at a given time point. Interestingly, all of the patient H-derived HCVpp were neutralized by serum antibodies from 2 unrelated, chronically HCV-infected individuals, confirming that autologous glycoprotein sequences are specifically resistant to antibodies generated by their own host, but not unrelated hosts. Sequence Polymorphism Associates With Sites of Immune Recognition Figure 4. Identification of T-cell epitopes in the E1E2. (A) Patient H’s PBMCs were stimulated with 15-mer peptides corresponding to the H77 E1E2 sequence. The number of peptide-specific interferon-␥–producing cells was determined by enzyme-linked immunospot analysis. (B) The 3 most vigorously recognized peptides indicated in A were retested on purified CD4 and CD8 cells isolated from patient H. BASIC–LIVER, PANCREAS, AND BILIARY TRACT seroconversion in 1977 failed to neutralize any of the viruses tested, whereas subsequent samples showed hightiter neutralization of HCVpp-bearing H77.19, H77.20, and H77 gps (Table 1). In contrast, serum antibodies in 1991 neutralized HCVpp bearing some autologous gps from that same sampling time, but with variable efficiency, with some clones being resistant to neutralization (H91.A11 and H91.B1). Similarly, serum antibodies in 1992, 1995, and 2002 showed reduced neutralization of HCVpp-bearing concurrent and later gps, although generally showing high-titer neutralization of HCVpp-bearing gps cloned from earlier time points in infection. The last clone, obtained in 2002 (H02.E10), was resistant to neutralization by all of the serum samples. These data An alignment of patient H-derived E1E2 sequences with a map of known epitopes taken from the HCV Immunology Database (hcv.lanl.gov and Yusim et al35) revealed that of 52 residues showing phylogenetically informative changes, 30 were located within defined Bcell epitopes (Table 2). To assess the impact of sequence polymorphism on antibody neutralization we studied the sensitivity of HCVpp bearing the set of patient H-derived gps to neutralization by MAbs that were raised against a recombinant form of H77 E2 and have been reported to neutralize HCVpp-H77.14,23 HCVpp-bearing gps cloned from 1977 and 1991 were neutralized equally by all of the MAbs. However, with the exception of MAb 11/20, viruses bearing gps from later time points displayed increasing resistance to neutralization by all MAbs (Table 3). Although all viruses were neutralized by a soluble form of the large extracellular loop of CD81, HCVpp-bearing gps cloned in 1992, 1995, and 2002 showed a 3- to 4-fold reduced sensitivity. Analysis of epitope diversity in the patient H-derived E1E2 sequences identified several amino acid changes in linear MAb epitopes that may partly explain the neutralization patterns observed. For example, mutation of amino acid 417 N/K within the 3/11 epitope in clones 92.C1 and 95.C8 may reduce their sensitivity to neutralization. However, clone 02.E10, which is resistant to 3/11 neutralization, shows no amino acid changes within the epitope, suggesting that changes outside the linear epitope modulate 3/11 activity. To analyze the effect these amino acid Table 1. Sensitivity of HCVpp to Autologous and Heterologous Serum Neutralization Over Time HCVpp 1977 1991 1992 1995 2002 HCV⫹ 1 HCV⫹ 2 cH77 H77.19 H77.20 H91.64 H91.A6 H91.A11 H91.B1 H92.C1 H95.C8 H02.E10 ⬍50 ⬍50 ⬍50 ⬍50 ⬍50 ⬍50 ⬍50 ⬍50 ⬍50 ⬍50 10,000 4,000 5,000 12,000 7,000 500 500 ⬍50 ⬍50 ⬍50 10,000 5,000 5,000 20,000 1,000 3,000 ⬍50 500 ⬍50 ⬍50 10,000 6,000 8,000 5,000 2,000 1,500 ⬍50 500 1,000 ⬍50 28,000 5,000 5,000 23,000 25,000 500 ⬍50 1,000 2,000 ⬍50 16,000 12,000 15,000 12,000 14,500 15,000 12,000 15,500 14,500 14,000 18,500 16,000 16,500 12,000 14,500 13,000 15,500 16,000 15,500 16,500 H serum samples from 1977, 1991, 1992, 1995, and 2002, as well as 2 sera from unrelated chronically HCV infected individuals (HCV⫹ 1, HCV⫹ 2), were tested for their ability to neutralize HCVpp bearing E1E2 clones that had been amplified from the same serum samples. Values represent ID90 titers, ie, the dilution of serum that gave a 90% inhibition of the respective HCVpp. February 2007 HCV ESCAPES NEUTRALIZING ANTIBODIES AND CTL 673 Region E1 HVR E2 Amino acid position1 mAb epitope2 T-cell epitope2 22222333 33455037 46436814 –––––––– || |–– —— –––— 333333334444444444 889999990000000001 782456780123456781 |||||||||||||||||| —————————————————— –––––––––––––––––– 44444444444455555556666677 13333445678802233380123501 62689574761322928913177494 ||||||| |–—— ||––—— ||–––––––––– ——————— –— |||||––––––––––––––––––– ––————— 1977 sequences H77C H77.16 H77.18 H77.19 H77.20 H77.21 H77.22 H77.23 H77.24 H77.26 H77.25 H77.27 H77.28 HPCST77 H77IMC pH21 H PCVH11 H90 GATKTSAV –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– –––––––– ST–T–––– HVNGRTTALVGLLTPAKN –––––––––––––––––– ––––––––––S––––––– ––––––––––S––––––– ––––––––––S––––––– ––––––––––S––––––– ––––––––––S––––––– –A––––––––S––––––– ––S–H–A–IA––F–L––– ––––––––––S––––––– ––––––––––S––––––– –––––––––––––––––– –––––––––––––––––– ––S––––––––––––––– ––S––––––––––––––– –––––––––––––––––– –––––––––––––––––– ––S–H––––––––––––– ––S––SVLIASF––RP–– NETWLQKPAALESRSAFVLWDIVDSA –––––––––––––––––––––––––– –D––VY–S––P––K–E––R––––––T ––––––––––P––––––––N–––––– –––––––––––––––––––––––––– –––R–––––––––––––––––––––– –G–––––––––––––––––––––––– –––––––––––G–––––––––––––– –D––VY–S––P––K–E––R––––––T –D––VY–S––P––K–E––R––––––– –––––––––––––––––––––––––– –––––––––––––––––––––––––– ––––––––––––G––––––––––––– –––––H–––––––––––––––––––– –––––––––––––––––––––––––– –––––––––––––––––––––––––– –––––––––––––––––––R–––––– –D–––R–––––––––––––––––––– KDA–IHG–D–P–––NELIR–H–I––T 1991-2003 sequences H90 H91.58 H91.60 H91.61 H91.64 H92.72 H92.73 H92.74 H92.76 H93.85 H93.86 H93.87 H93.88 H94.A2 H94.A4 H94.A5 H94.A7 H95.C2 H95.C3 H95.C4 H95.C7 H02.E1 H02.E4 H02.E7 H02.E3 H03.G1 H03.G5 H03.G10 H03.G11 STTKTTAV ––––––T– –––––––– –––––––– –––––––– –––R–––– –––––––I G–––––TI –––––––– –––––––– –––––––– –––R–––– –––––––– –––R–––– ––AR–––– –––R–––– –––R–––– –––––––– –––R–––– G––R–––– –––R–––– –––RA––– –––RA––– –––R–––A –––RA––– ––ARA––– –––RA––– –––RA––– –––RA––– HVSGRSVLIASFLTRPKN ––––––––F––L–––––– –––A–––––––L–––––– –––––––FV––––––––– ––––––––F–––––K––K ––––––––F–––––Q––– ––––––––F––L–––––– ––––––––F––L–––––– –––––––––––IF––––– Y–––––––F–––––K––– Y–––––––F–––––K––– Y–––––––F–––––K––– Y–––––––F–––––K––– –––A––––L––––SQ––– –––A––––L––––SQ––– –––A––––L––––SQ––– –––––––FV–––––L––– –––––––––T–L–SL––– –––A––––L––––SQ––– R––A––––L––––SQ––– –––A––––L––––SQ––– Y––A––A–F–N–F–P–QR Y––A––A–F–N–F–P–QR Y––A––A–F–N–F–P–QR Y––A––A–F–N–F–P–QR YA–A––A–F–N–F–P–QK Y––A––A–F–N–F–P–QK Y––A––A–F–N–––P–QK Y––A––A–F–N–F–P–QK KDAWIHGPDAPESRNELIRWHIIDST N––––––––––––––––––––V–––– –––––––S––––––––P––––V–––– N–––L––––––––K–––––––––––– N–––L––––––––––––––––––––A –––––––––––––––––––––––GG– N–––––––––––––––––––––––G– N––––––T–––––––––––––––––– –––––––T––––––––––––R––––– N–––L––––––––––––––––V–N–– N–––L––––––––––––––––V–N–– N–––L––––––––––––––––V–N–– N–––L––––––––––––––––V–N–– –––––––––––––––––––––––––– –––––––––––––––––––––––––– –––––––––––––––––––––––––– N–––L––––––––––––––––––––– –––––––T–––––––––––––––––– –––––––S–––––––––T–––––––– –––––––S–––––––––––R–––––– –––––––S–––––––––––––––––– N–––LYR––––QN–––F––––V–––– N–––LYR––––Q–––––––––––N–– N–––LYR––V–Q––––F––––V–––– N–––LYR––––QN–––F––––V–––– N–––LYK––V–Q––––F––––––––– N––RLYK––V–Q––––F––––––––– N–––LYK––––QN–––F––––V–––– N––RLYK––V–Q––––F––––––––– All residues in the E1E2 sequence that showed phylogenetically informative aa changes are represented. Changes are indicated relative to the H77 or H90 (a clone that was previously obtained from patient H in 1990; accession M62382) sequence with “–” indicating conserved sequence. 1 - aa coordinates (shown in columns) are relative to the strain H sequence; 2 - “⫹” indicates that the respective site is within a known nAb epitope (information from own data as well as the HCV Immunology database at hcv.lanl.gov) or within a T-cell epitope identified in this study (Figures 4 and 5). BASIC–LIVER, PANCREAS, AND BILIARY TRACT Table 2. Amino Acid Changes in the H Derived E1E2 Sequence Set and Their Relation to Defined B- and T-cell Epitopes 674 VON HAHN ET AL GASTROENTEROLOGY Vol. 132, No. 2 Table 3. Sensitivity of HCVpp Bearing H E1E2 Glycoproteins to Neutralization by mAbs and Soluble CD81 Monoclonal antibodies Location Epitope1 cH77 77.19 77.20 91.64 91.A6 E1E2 clones 91.A11 91.B1 92.C1 95.C8 02.E10 9/27 396–407 TAGLVGLLTPG 3/11 412–423 NIQLINTNGSWHIN 2/69a 432–443 ESLNTGWLAGLP 1/39 432–443 ESLNTGWLAGLF 7/16b 436–447 TGWLAGLFYQHK 11/20 436–447 sCD81 TGWLAGLFYQHK undefined 0.8 ––––––––––– 0.7 ––––––––––– 0.7 ––––––––––– 0.7 VL–FASF––K– 1.1 VL–FASF––K– 1.2 VF–VASF––R– 1.4 VL–FASF––K– ⬎20 VL–IASIF–R– ⬎20 VL–ITSL–SL– ⬎20 AL–FANFF––– 1.1 –––––––––––––– 2.0 –––––––––––––– 1.3 –––––––––––––– 2.0 k––––––––––––– 1.1 k––––––––––––– 5.0 –––––––––––––– 2.0 k––––––––––––– ⬎20 –––––K–––––––– ⬎20 –––––K–––––––– ⬎20 k––––––––––––– 2.0 –––––––––––– 2.5 –––––––––––– 2.5 –––––––––––– 2.3 D–––A––––––– 2.3 D–––A––––––– 2.4 D–––A––––––– 2.5 D–––A––––––– ⬎20 D–––A––I–––– ⬎20 D–––A––I–––– ⬎20 D–––A––––––– 3.0 –––––––––––– 3.5 –––––––––––– 4.0 –––––––––––– 3.6 D–––A––––––– 4.0 D–––A––––––– 3.8 D–––A––––––– 3.7 D–––A––––––– ⬎20 D–––A––I–––– ⬎20 D–––A––I–––– ⬎20 D–––A––––––– 1.0 –––––––––––– 1.5 –––––––––––– 1.8 –––––––––––– 2.5 A––––––––H–G 1.1 A––––––––H–G 1.3 A––––––––H–G 1.6 A––––––––H–G ⬎20 A––I–––––H–G ⬎20 A––I–––––H–G ⬎20 A––––––––Y–R 0.4 –––––––––––– 0.2 –––––––––––– 0.4 –––––––––––– 0.3 A––––––––H–G 0.4 A––––––––H–G 0.7 A––––––––H–G 0.4 A––––––––H–G 1.5 A––I–––––H–G 1.5 A––I–––––H–G 2.0 A––––––––Y–R 0.4 0.2 0.4 0.3 0.4 0.7 0.4 1.5 1.5 2.5 Monoclonal antibodies against HCV E2 and soluble CD81 (sCD81) were tested for neutralization of HCVpp bearing sequential E1E2 glycoproteins cloned from patient H. Numbers indicate the concentration of inhibitor in ␮g/mL required to neutralize viral infectivity by 90% (ID90). Below each ID90 value changes in the linear epitope sequence compared to cH77 are indicated, with “-” denoting conserved residues. BASIC–LIVER, PANCREAS, AND BILIARY TRACT changes may have on antigen recognition we studied MAb binding to the diverse E1E2 gps by EIA. Four of the 6 neutralizing MAbs (3/11, 7/16, 9/27, 11/20) bound lysates from 293T expressing H77 E1E2 in a GNA lectin capture EIA. All of the MAbs showed reduced binding to H gps cloned late in infection (Figure 3). In contrast, 2 nonneutralizing MAbs (6/1a and 9/75) bound all H gps cloned over the course of the infection (data not shown). In the case of 3/11 and 9/27 the loss of reactivity in the EIA paralleled a decrease in neutralizing activity against viruses bearing gps cloned in 1992 and at later time points, whereas 11/20 showed reduced reactivity for 1992 and later E1E2 gps while maintaining its neutralizing ability, suggesting differences between MAbs in the amount of Ab required to bind and neutralize a virus particle and that required to saturate antigen in an EIA. Because we hypothesized that patient H serum antibodies selected for escape variants, we screened all serum samples for reactivity to a panel of overlapping peptides representing the H77 E1E2 sequence, some of which represent epitopes recognized by the neutralizing MAbs. Reactivity was observed only for the HVR peptide (data not shown), suggesting that the selecting Ab response was specific for conformational epitopes overlapping the defined MAb neutralization epitopes. HCV Escape From T-Cell Responses Given the association between sequence polymorphism in nAb epitopes and viral escape we were inter- ested to know whether any of the remaining polymorphic sites corresponded to sites recognized by T cells. To probe for escape mutations that may have arisen due to selection pressure exerted by HCV-specific T-cell responses, we screened PBMCs from patient H in 2003 for reactivity to a series of 15-mer overlapping peptides representing the H77 E1E2 sequence in an interferon-␥ enzyme-linked immunospot assay (Figure 4A). Three peptides, located at residues E1 226-240, E1 296-310, and E2 436-450, were recognized and represent new T-cell epitopes. Both E1 peptides were recognized by CD4 T cells, whereas the E2 peptide was recognized by both CD4 and CD8 T cells (Figure 4B). Interestingly, all 3 epitopes contained between 1 and 4 sites that showed changes within the H sequence set. The consensus sequence of the E1 226-240 peptide changed from CVPCVREGNASRCWV in 1977 to CVPCVRESNTSRCWV in 2003; similarly E1 296-310 changed from RRHWTTQDCNCSIYP to RRHWTTQDCNCTIYP and E2 436-450 from GWLAGLFYQHKFNSS to GWLAGLFYYHRFNSS. Furthermore, these mutations abrogated CD4 T-cell recognition of one of the E1 peptides and CD4 and CD8 recognition of the E2 peptide (Figure 5). These changes in the E1 T-cell epitopes had appeared between 1977 and 1990 (Table 2). The E2 T-cell epitope had accumulated 2 changes by 1990, these residues changed again between 1990 and 2002, with one of them showing a further change between 2002 and 2003 (Table 2 and Figure 5). Figure 5. Escape from CD4 and CD8 T-cell recognition owing to mutations in the H 2003 E1E2 consensus sequence. CD4 or CD8 T cells obtained in 2003 were stimulated with peptides representing the E1 226-240, E1 296-310, and E2 436-450 epitopes in their original 1977 (WT) form or with mutations that were present in the 2003 consensus (MUT). Interferon-␥ responses were assessed by enzyme-linked immunospot assay. Collectively, these data show that the HCV E1E2 sequence not only is targeted by antibodies, but by CD4 and CD8 T cells and that ongoing HCV sequence change mediates viral escape from these T-cell responses. Discussion This study is an in-depth analysis of the nAb response against autologous HCV E1E2 sequences and shows ongoing evasion of the virus-specific adaptive immune response during persistent infection. Patient H is a uniquely well-suited subject for this study because he is representative of many patients with mild, chronic hepatitis C and clinically is well characterized with serum samples spanning a long period of time. Furthermore, specific research tools such as sequences, infectious clones, established HCVpp, and well-characterized antibodies are available.11,22,36 Although several studies have investigated the CD8 T-cell responses during HCV infection and their impact on viral evolution and disease outcome, the role(s) of HCV ESCAPES NEUTRALIZING ANTIBODIES AND CTL 675 CD4 T cells and nAbs are much less well defined. Several earlier reports have suggested that nAbs develop late in primary HCV infection and have a minimal role in controlling acute HCV replication.15,17,37,38 However, these studies measured antibody neutralization of HCVppbearing gps of prototype laboratory strains and not the virus population within the patient. The generation of HCVpp-bearing gps cloned from serial serum samples from patient H (Figure 2) allowed us to study the dynamics of autologous and heterologous nAb responses over time (Figure 1A; Table 1). HCVpp provide an ideal tool for the study of nAb responses because they can be made to display a range of related glycoprotein sequences, allowing for direct comparison of their neutralization behavior. The percentage of patient H– derived E1E2 clones that produced infectious HCVpp, albeit sufficient for our purposes, was small (21.4%). Inspection of the 146 polymorphic sites in the sequence set showed a marked bias in the distribution of transitional changes (17 G to A vs 26 A to G and 11 C to T vs 32 T to C, P ⬍ .0001, Fisher exact test) that was indicative of reverse-transcriptase errors. However, it certainly is feasible that significant numbers of E1E2 sequences that are present in infected serum are either defective or, despite being functional in vivo, are not capable of generating infectious HCVpp in vitro. Our data show that nAbs against HCVpp-bearing gps representing the infecting strain are detected at seroconversion and that these early strain-specific nAbs appear to target the HVR (Figure 1C). However, the response against the H77 HVR, a sequence representative of the initially infecting virus, as detected by EIA, is short lived and undetectable by 28 weeks (Figure 1B), suggesting that additional strain-specific epitopes account for the low-titer anti-H77 nAb response observed at this point (Figure 1A). It is intriguing to speculate that the short half-life of the antibody response against this HVR sequence that was present and triggered a humoral response early in infection could be the result of a lack of CD4 T-cell help required for the establishment of memory B-cell responses. Cross-reactive nAb responses were first detected at 111 weeks after infection and increased in titer and breadth to recognize distant HCV genotypes. Despite this crossreactivity, the nAb response efficiently neutralized viruses bearing ancestral E1E2 sequences yet was less able to neutralize viruses expressing E1E2 sequences that were dominant in the serum at the time of sampling (Table 1). These data parallel what has been reported for human immunodeficiency virus39,40 and suggest that the nAb response lags behind the rapidly evolving gp sequences present within the quasispecies population. Sera from 2 unrelated chronically HCV-infected individuals could neutralize all H-derived E1E2 clones tested in this study, underscoring the conclusion that nAb responses elicited during chronic infection are broadly reactive to most BASIC–LIVER, PANCREAS, AND BILIARY TRACT February 2007 676 VON HAHN ET AL BASIC–LIVER, PANCREAS, AND BILIARY TRACT E1E2 sequences, including ones that the host immune system has not encountered. Thus, viruses bearing autologous E1E2 sequences are not insensitive to serum antibody–mediated neutralization per se. The observed resistance of viruses expressing gps cloned from samples collected in 1992 and later sequences to Abs present in earlier serum samples suggests that viral escape from the humoral immune response continues after decades of chronic infection. HCV continuously generates new sequences that are poorly neutralized by serum Abs. The dominance of neutralization-resistant sequences in the viral population at all time points (Table 1) suggests that viral evolution is driven by strong antibody-mediated selective pressures. Because serum antibodies from chronically HCV-infected individuals can neutralize diverse E1E2 sequences (Table 1 and Bartosch et al,15 Logvinoff et al,17 and Meunier et al19), it seems that E1E2 sequences that dominate during the course of chronic HCV infection must exploit specific gaps in the Ab repertoire of the individual host to evade the polyclonal humoral immune response. This strongly suggests the presence of a mechanism that removes nAb-sensitive sequences from the quasispecies pool. The high rate of amino acid substitutions in defined linear neutralization epitopes (Tables 2 and 3) suggests that these regions are under selective pressure(s). Rapid mutation in the HVR has long been suspected to be a consequence of antibody-mediated selection.41,42 Compelling evidence that antibodies exert selective pressure on the HVR in vivo stems from the observation that HVR evolution is reduced in hypogammaglobulinemic patients.43,44 Sequence analysis of envelope clones from patient H shows the expected high rate of amino acid substitutions in the HVR. HCVpp-bearing gps cloned from samples obtained in 1992, 1995, and 2002 were resistant to neutralization by an MAb targeting the HVR of the 1977 sequence (Table 3). Our data suggest that HCV can readily escape anti-HVR nAbs, consistent with a model that this region may act as an immune decoy.45,46 It should be noted, however, that changes in defined epitopes may only partly predict the ability of an antibody to neutralize a given E1E2 sequence. For example, clones 91.B1 and 02.E10 are identical in the 3/11 epitope, however, HCVpp-02.E10 escapes 3/11 neutralization, illustrating the complexity of the 3/11-gp interaction and its modulation by amino acid changes outside the minimally defined linear epitope. Several publications have reported that a broad and robust virus-specific CD4 and CD8 response is important for controlling the acute phase of HCV replication and achieving clearance of infection.9 –11 The emergence of viral variants bearing escape mutations in CD8 T-cell epitopes is associated with the development of chronic infection.7,47– 49 Moreover, with the development of chronic infection, T-cell responses seem to become narrower and less vigorous.50 –52 However, little is known GASTROENTEROLOGY Vol. 132, No. 2 about the impact of T-cell responses on the evolution of HCV E1E2. An analysis of T-cell reactivity against the consensus H77 sequence in PBMCs obtained from H in 2003 revealed strong reactivity against 4 epitopes, 2 of them located in a single peptide (Figures 4 and 5). All of these epitopes coincide with residues that show phylogenetically informative changes over time, changes that completely abrogated T-cell reactivity in 3 of 4 epitopes (Figure 5). Escape from cellular immune responses previously has been well documented only for CD8 T cells, yet our study uncovered 2 instances of viral escape in a CD4 epitope. These data strongly suggest that CD4 and CD8 T-cell– as well as B-cell–mediated selective pressures impact on HCV glycoprotein evolution. The quasispecies nature of HCV in vivo has long been viewed as important for the pathobiology of this disease. Several reports have shown that quasispecies complexity is associated with failure to clear acute infection7,36 and is associated inversely with the severity of HCV-related disease, in which patients with mild liver disease have a diverse repertoire of viral sequences.53,54 The latter observation has led several investigators to suggest that the HCV-specific immune responses that limit viral damage to the liver may lead to selective expansion of antigenically diverse viral variants.54 –58 This model of viral escape is analogous to that reported for human immunodeficiency virus.39,40,59 Here, we extend the earlier-described data by showing that amino acid changes in the gps directly reduce autologous serum antibody and MAb neutralization (Tables 1 and 3), and that the majority of sequence changes in E1E2 are located in regions identified to contain T-cell or antibody epitopes (Table 2 and Yusim et al35). These data provide a compelling explanation for the failure of the immune response in this patient to resolve HCV infection. Our study highlights the importance of studying autologous and heterologous nAb responses in individuals with different disease outcomes. References 1. Alter M. Hepatitis C virus infection in the United States. J Hepatol 1999;31(Suppl 1):88 –91. 2. Alter M, Margolis H, Krawczynski K, Judson F, Mares A, Alexander W, Hu P, Miller J, Gerber M, Sampliner R, et al. The natural history of community-acquired hepatitis C in the United States. The Sentinel Counties Chronic non-A, non-B Hepatitis Study Team. N Engl J Med 1992;327:1899 –1905. 3. Martell M, Esteban J, Quer J, Genesca J, Weiner A, Esteban R, Guardia J, Gomez J. Hepatitis C virus (HCV) circulates as a population of different but closely related genomes: quasispecies nature of HCV genome distribution. J Virol 1992;66:3225–3229. 4. Brown R, Juttla V, Tarr A, Finnis R, Irving W, Hemsley S, Flower D, Borrow P, Ball J. Evolutionary dynamics of hepatitis C virus envelope genes during chronic infection. J Gen Virol 2005;86: 1931–1942. 5. Weiner A, Erickson A, Kansopon J, Crawford K, Muchmore E, Hughes A, Houghton M, Walker C. Persistent hepatitis C virus infection in a chimpanzee is associated with emergence of a 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. cytotoxic T lymphocyte escape variant. Proc Natl Acad Sci U S A 1995;92:2755–2759. Chang K, Rehermann B, McHutchison J, Pasquinelli C, Southwood S, Sette A, Chisari F. Immunological significance of cytotoxic T lymphocyte epitope variants in patients chronically infected by the hepatitis C virus. J Clin Invest 1997;100:2376 – 2385. Erickson A, Kimura Y, Igarashi S, Eichelberger J, Houghton M, Sidney J, McKinney D, Sette A, Hughes A, Walker C. The outcome of hepatitis C virus infection is predicted by escape mutations in epitopes targeted by cytotoxic T lymphocytes. Immunity 2001;15: 883– 895. Seifert U, Liermann H, Racanelli V, Halenius A, Wiese M, Wedemeyer H, Ruppert T, Rispeter K, Henklein P, Sijts A, Hengel H, Kloetzel PM, Rehermann B. Hepatitis C virus mutation affects proteasomal epitope processing. J Clin Invest 2004;114:250 – 259. Cooper S, Erickson A, Adams E, Kansopon J, Weiner A, Chien D, Houghton M, Parham P, Walker C. Analysis of a successful immune response against hepatitis C virus. Immunity 1999;10: 439 – 449. Lechner F, Wong D, Dunbar P, Chapman R, Chung R, Dohrenwend P, Robbins G, Phillips R, Klenerman P, Walker B. Analysis of successful immune responses in persons infected with hepatitis C virus. J Exp Med 2000;191:1499 –1512. Thimme R, Oldach D, Chang K, Steiger C, Ray S, Chisari F. Determinants of viral clearance and persistence during acute hepatitis C virus infection. J Exp Med 2001;194:1395–1406. Shoukry NH, Grakoui A, Houghton M, Chien DY, Ghrayeb J, Reimann KA, Walker CM. Memory CD8⫹ T cells are required for protection from persistent hepatitis C virus infection. J Exp Med 2003;197:1645–1655. Grakoui A, Shoukry NH, Woollard DJ, Han JH, Hanson HL, Ghrayeb J, Murthy KK, Rice CM, Walker CM. HCV persistence and immune evasion in the absence of memory T cell help. Science 2003;302:659 – 662. Hsu M, Zhang J, Flint M, Logvinoff C, Cheng-Mayer C, Rice C, McKeating J. Hepatitis C virus glycoproteins mediate pH-dependent cell entry of pseudotyped retroviral particles. Proc Natl Acad Sci U S A 2003;100:7271–7276. Bartosch B, Bukh J, Meunier J, Granier C, Engle R, Blackwelder W, Emerson S, Cosset F, Purcell R. In vitro assay for neutralizing antibody to hepatitis C virus: evidence for broadly conserved neutralization epitopes. Proc Natl Acad Sci U S A 2003;100: 14199 –14204. Drummer HE, Maerz A, Poumbourios P. Cell surface expression of functional hepatitis C virus E1 and E2 glycoproteins. FEBS Lett 2003;546:385–390. Logvinoff C, Major M, Oldach D, Heyward S, Talal A, Balfe P, Feinstone S, Alter H, Rice C, McKeating J. Neutralizing antibody response during acute and chronic hepatitis C virus infection. Proc Natl Acad Sci U S A 2004;101:10149 –10154. McKeating J, Zhang L, Logvinoff C, Flint M, Zhang J, Yu J, Butera D, Ho D, Dustin L, Rice C, Balfe P. Diverse hepatitis C virus glycoproteins mediate viral infection in a CD81-dependent manner. J Virol 2004;78:8496 – 8505. Meunier JC, Engle RE, Faulk K, Zhao M, Bartosch B, Alter H, Emerson SU, Cosset FL, Purcell RH, Bukh J. Evidence for crossgenotype neutralization of hepatitis C virus pseudo-particles and enhancement of infectivity by apolipoprotein C1. Proc Natl Acad Sci U S A 2005;102:4560 – 4565. Yanagi M, Purcell R, Emerson S, Bukh J. Transcripts from a single full-length cDNA clone of hepatitis C virus are infectious when directly transfected into the liver of a chimpanzee. Proc Natl Acad Sci U S A 1997;94:8738 – 8743. HCV ESCAPES NEUTRALIZING ANTIBODIES AND CTL 677 21. Kolykhalov A, Agapov E, Blight K, Mihalik K, Feinstone S, Rice C. Transmission of hepatitis C by intrahepatic inoculation with transcribed RNA. Science 1997;277:570 –574. 22. Ogata N, Alter H, Miller R, Purcell R. Nucleotide sequence and mutation rate of the H strain of hepatitis C virus. Proc Natl Acad Sci U S A 1991;88:3392–3396. 23. Flint M, Maidens C, Loomis-Price LD, Shotton C, Dubuisson J, Monk P, Higginbottom A, Levy S, McKeating JA. Characterization of hepatitis C virus E2 glycoprotein interaction with a putative cellular receptor, CD81. J Virol 1999;73:6235– 6244. 24. Flint M, Logvinoff C, Rice C, McKeating J. Characterization of infectious retroviral pseudotype particles bearing hepatitis C virus glycoproteins. J Virol 2004;78:6875– 6882. 25. Hammond A, Lewis J, May J, Albert J, Balfe P, McKeating J. Antigenic variation within the CD4 binding site of human immunodeficiency virus type 1 gp120: effects on chemokine receptor utilization. J Virol 2001;75:5593–5603. 26. Inchauspe G, Zebedee S, Lee D, Sugitani M, Nasoff M, Prince A. Genomic structure of the human prototype strain H of hepatitis C virus: comparison with American and Japanese isolates. Proc Natl Acad Sci U S A 1991;88:10292–10296. 27. Nei M, Gojobori T. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol Biol Evol 1986;3:418 – 426. 28. Yang Z. Maximum likelihood estimation on large phylogenies and analysis of adaptive evolution in human influenza virus A. J Mol Evol 2000;51:423– 432. 29. Rogers J, Swofford D. A fast method for approximating maximum likelihoods of phylogenetic trees from nucleotide sequences. Syst Biol 1998;47:77– 89. 30. Posada D, Crandall K. MODELTEST: testing the model of DNA substitution. Bioinformatics 1998;14:817– 818. 31. Takaki A, Wiese M, Maertens G, Depla E, Seifert U, Liebetrau A, Miller J, Manns M, Rehermann B. Cellular immune responses persist and humoral responses decrease two decades after recovery from a single-source outbreak of hepatitis C. Nat Med 2000;6:578 –582. 32. Rahman F, Heller T, Sobao Y, Mizukoshi E, Nascimbeni M, Alter H, Herrine S, Hoofnagle J, Liang T, Rehermann B. Effects of antiviral therapy on the cellular immune response in acute hepatitis C. Hepatology 2004;40:87–97. 33. Afdhal NH. The natural history of hepatitis C. Semin Liver Dis 2004;24(Suppl 2):3– 8. 34. Chen M, Sallberg M, Sonnerborg A, Weiland O, Mattsson L, Jin L, Birkett A, Peterson D, Milich DR. Limited humoral immunity in hepatitis C virus infection. Gastroenterology 1999;116:135– 143. 35. Yusim K, Richardson R, Tao N, Dalwani A, Agrawal A, Szinger J, Funkhouser R, Korber B, Kuiken C. Los alamos hepatitis C immunology database. Appl Bioinformatics 2005;4:217–225. 36. Farci P, Shimoda A, Coiana A, Diaz G, Peddis G, Melpolder J, Strazzera A, Chien D, Munoz S, Balestrieri A, Purcell R, Alter H. The outcome of acute hepatitis C predicted by the evolution of the viral quasispecies. Science 2000;288:339 –344. 37. Lavillette D, Morice Y, Germanidis G, Donot P, Soulier A, Pagkalos E, Sakellariou G, Intrator L, Bartosch B, Pawlotsky J, Cosset F. 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;79:6023– 6034. 38. Netski D, Mosbruger T, Depla E, Maertens G, Ray S, Hamilton R, Roundtree S, Thomas D, McKeating J, Cox A. Humoral immune response in acute hepatitis C virus infection. Clin Infect Dis 2005;41:667– 675. 39. Frost SD, Wrin T, Smith DM, Kosakovsky Pond SL, Liu Y, Paxinos E, Chappey C, Galovich J, Beauchaine J, Petropoulos CJ, Little SJ, Richman DD. Neutralizing antibody responses drive the evolution BASIC–LIVER, PANCREAS, AND BILIARY TRACT February 2007 678 40. 41. 42. 43. 44. 45. 46. 47. BASIC–LIVER, PANCREAS, AND BILIARY TRACT 48. 49. 50. VON HAHN ET AL of human immunodeficiency virus type 1 envelope during recent HIV infection. Proc Natl Acad Sci U S A 2005;102:18514 – 18519. Richman DD, Wrin T, Little SJ, Petropoulos CJ. Rapid evolution of the neutralizing antibody response to HIV type 1 infection. Proc Natl Acad Sci U S A 2003;100:4144 – 4149. Kato N, Ootsuyama Y, Sekiya H, Ohkoshi S, Nakazawa T, Hijikata M, Shimotohno K. Genetic drift in hypervariable region 1 of the viral genome in persistent hepatitis C virus infection. J Virol 1994;68:4776 – 4784. Kato N, Sekiya H, Ootsuyama Y, Nakazawa T, Hijikata M, Ohkoshi S, Shimotohno K. Humoral immune response to hypervariable region 1 of the putative envelope glycoprotein (gp70) of hepatitis C virus. J Virol 1993;67:3923–3930. Kumar U, Monjardino J, Thomas HC. Hypervariable region of hepatitis C virus envelope glycoprotein (E2/NS1) in an agammaglobulinemic patient. Gastroenterology 1994;106:1072–1075. Booth JC, Kumar U, Webster D, Monjardino J, Thomas HC. Comparison of the rate of sequence variation in the hypervariable region of E2/NS1 region of hepatitis C virus in normal and hypogammaglobulinemic patients. Hepatology 1998;27:223– 227. Ray S, Wang Y, Laeyendecker O, Ticehurst J, Villano S, Thomas D. Acute hepatitis C virus structural gene sequences as predictors of persistent viremia: hypervariable region 1 as a decoy. J Virol 1999;73:2938 –2946. Mondelli M, Cerino A, Segagni L, Meola A, Cividini A, Silini E, Nicosia A. Hypervariable region 1 of hepatitis C virus: immunological decoy or biologically relevant domain? Antiviral Res 2001; 52:153–159. Cox A, Mosbruger T, Mao Q, Liu Z, Wang X, Yang H, Sidney J, Sette A, Pardoll D, Thomas D, Ray S. Cellular immune selection with hepatitis C virus persistence in humans. J Exp Med 2005; 201:1741–1752. Tester I, Smyk-Pearson S, Wang P, Wertheimer A, Yao E, Lewinsohn D, Tavis J, Rosen H. Immune evasion versus recovery after acute hepatitis C virus infection from a shared source. J Exp Med 2005;201:1725–1731. Timm J, Lauer G, Kavanagh D, Sheridan I, Kim A, Lucas M, Pillay T, Ouchi K, Reyor L, Schulze zur Wiesch J, Gandhi R, Chung R, Bhardwaj N, Klenerman P, Walker B, Allen T. CD8 epitope escape and reversion in acute HCV infection. J Exp Med 2004; 200:1593–1604. Cox AL, Mosbruger T, Lauer GM, Pardoll D, Thomas DL, Ray SC. Comprehensive analyses of CD8⫹ T cell responses during longitudinal study of acute human hepatitis C. Hepatology 2005;42: 104 –112. GASTROENTEROLOGY Vol. 132, No. 2 51. Ulsenheimer A, Gerlach JT, Gruener NH, Jung MC, Schirren CA, Schraut W, Zachoval R, Pape GR, Diepolder HM. Detection of functionally altered hepatitis C virus-specific CD4 T cells in acute and chronic hepatitis C. Hepatology 2003;37:1189 –1198. 52. Wedemeyer H, He XS, Nascimbeni M, Davis AR, Greenberg HB, Hoofnagle JH, Liang TJ, Alter H, Rehermann B. Impaired effector function of hepatitis C virus-specific CD8⫹ T cells in chronic hepatitis C virus infection. J Immunol 2002;169:3447–3458. 53. Qin H, Shire NJ, Keenan ED, Rouster SD, Eyster ME, Goedert JJ, Koziel MJ, Sherman KE. HCV quasispecies evolution: association with progression to end-stage liver disease in hemophiliacs infected with HCV or HCV/HIV. Blood 2005;105:533–541. 54. Sheridan I, Pybus O, Holmes E, Klenerman P. High-resolution phylogenetic analysis of hepatitis C virus adaptation and its relationship to disease progression. J Virol 2004;78:3447– 3454. 55. Chien D, Choo Q, Ralston R, Spaete R, Tong M, Houghton M, Kuo G. Persistence of HCV despite antibodies to both putative envelope glycoproteins. Lancet 1993;342:933. 56. Farci P, Alter H, Govindarajan S, Wong D, Engle R, Lesniewski R, Mushahwar I, Desai S, Miller R, Ogata N, et al. Lack of protective immunity against reinfection with hepatitis C virus. Science 1992;258:135–140. 57. Farci P, Alter H, Wong D, Miller R, Govindarajan S, Engle R, Shapiro M, Purcell R. Prevention of hepatitis C virus infection in chimpanzees after antibody-mediated in vitro neutralization. Proc Natl Acad Sci U S A 1994;91:7792–7796. 58. Farci P, Orgiana G, Purcell R. Immunity elicited by hepatitis C virus. Clin Exp Rheumatol 1995;13(Suppl 13):S9 –S12. 59. Ross H, Rodrigo A. Immune-mediated positive selection drives human immunodeficiency virus type 1 molecular variation and predicts disease duration. J Virol 2002;76:11715–11720. Received August 29, 2006. Accepted November 13, 2006. Address requests for reprints to: Thomas von Hahn, MD, Center for the Study of Hepatitis C, The Rockefeller University, 1230 York Avenue, New York, New York 10021. e-mail: vonhaht@rockefeller.edu; fax: (212) 327-7048. Supported by PHS grants CA57973, AI50798, U19 AI40034, Medical Research Council, UK (G0400802), the National Institute of Diabetes and Digestive and Kidney Diseases intramural research program, the Greenberg Medical Research Institute, and the Starr Foundation. T.v.H. is supported by a postdoctoral fellowship from the Deutsche Forschungsgemeinschaft. The authors thank Merna Torres and Ke Hu for expert technical assistance.