An estimated 180 million people worldwide are infected with Hepatitis C virus (HCV). Only about 20% of the infected individuals are able to spontaneously clear the virus during acute infection leading to chronic infection in 80% of the cases. Chronic HCV infection is a major cause of liver cirrhosis and liver cancer and therefore became the leading indication for liver transplantation 
, but the rapid re-infection of the engrafted liver leads to poor survival rates of transplanted patients 
. One of the major challenges in HCV therapy is the great genetic diversity of the virus resulting from the rapid and error-prone activity of the RNA polymerase NS5B. Consequently, the six major genotypes differ by up to 30% at the nucleotide level 
and within the major glycoprotein E2 by up to 34% at the amino acid level. The rapid replication results in generation of up to 1012
virus particles per day in an infected individual, representing a population of circulating variants that can quickly react to selective pressures such as the adaptive host immune response or antiviral therapies. This requires special considerations for the design of vaccines and therapeutics.
The current HCV therapy includes pegylated alpha interferon (IFN-α), ribavirin and one of the recently approved HCV NS3 protease inhibitors Boceprevir and Telaprevir for genotype I infections 
, and IFN-α and ribavirin for infections with other genotypes. However, the limitations of these regimens are the associated severe side effects 
and sustained virological response (SVR) rates that vary considerably with the viral genotype. The natural emergence of viruses resistant to both of the available direct-acting antivirals 
suggests that HCV will remain a major global health burden despite the approval of the recently developed antiviral strategies, illustrating the urgent need for development of a safe and efficient HCV vaccine.
The role of neutralizing antibodies in the course of HCV infection in vivo
has been analyzed by a number of studies. A protective effect for anti-HCV antibodies was suggested by screening of HCV-infected patients receiving Hepatitis B polyclonal immunoglobulins containing anti-HCV antibodies 
. Also, antibodies directed against the major envelope glycoprotein E2 were shown to prevent non-homologous virus infection after vaccination in chimpanzees 
. Broadly neutralizing human polyclonal and monoclonal antibodies (mAbs) protected in a passive transfer experiment against heterologous virus challenge in human liver–chimeric Alb-uPA/SCID mice 
. Various studies provided evidence that the presence of high titers of neutralizing antibodies are associated with viral clearance during acute HCV infection 
, and that these antibodies are directed to specific epitopes 
. More recently, a broadly neutralizing human mAb was reported to prevent and treat HCV infection in chimpanzees 
. Lastly, immunization of immunocompetent humanized mice with vaccinia virus expressing HCV structural proteins resulted in a robust antibody response that protected from challenge with heterologous HCV in some of the animals, and correlated with the serum level of antibodies to E2 
A key challenge for the design of a safe and efficient B-cell vaccine is to know whether the elicited antibodies permit the virus to escape from neutralization by mutations in the viral glycoproteins. This has been suggested by several studies analyzing the mechanism of escape from neutralizing antibodies 
, identifying three patterns of virus escape for JFH-1 HCVcc propagated under selective immune pressure by increasing concentrations of a neutralizing antibody 
. These results underscore the ability of the virus to react to selective pressure exerted by neutralizing antibodies and emphasize the need to find highly conserved epitopes that are not associated with virus escape.
Recently, we identified a group of broadly neutralizing, human monoclonal antibodies termed HC84-1–HC84-27 
from a random paired scFv-expressing yeast display library. This group of antibodies recognizes a cluster of conformational epitopes that lie within a continuous region in HCV E2 encompassing residues 434 to 446 (according to H77 polyprotein numbering) termed “epitope II” 
. In addition, most of these mAbs appear to have an additional contact at tryptophan 616. The fact that no virus escape was observed upon passaging of a genotype 2a isolate in the presence of any of those antibodies indicated that this cluster of epitopes is resistant to neutralization escape 
. Epitope II is involved in binding the cellular receptor CD81 
, which is in line with the fact that the HC84 antibodies inhibit binding of E2 to CD81 
. Based on the assumption that HCV E2 adopts a class II fusion protein fold we have recently reported a model of the domain organization of HCV E2 suggesting that the glycoprotein is composed of three domains: DI, DII and DIII, as in the “class II” fusion proteins from flaviviruses and alphaviruses 
. The recently reported structure of the major glycoprotein E2 of the closely related pestiviruses showed, however, that this assumption does not hold for the pestivirus E2, which is an elongated molecule consisting of four β-sandwich domains arranged linearly from N to C terminus 
and does not have a class II fold. These results therefore cast doubt on a class II based model for HCV E2.
Epitope II elicits neutralizing antibodies 
as well as non-neutralizing antibodies that interfere with neutralizing antibodies directed against a conserved linear epitope located within residues 412 to 423 
. Notably, conflicting findings on the relationship of antibodies to epitope II and aa412–423 have been reported recently, showing mainly additive neutralizing activities when antibodies against both epitopes were combined 
. The crystal structure of a synthetic peptide mimicking the epitope aa412–423 has been reported recently in complex with Fab fragments derived from broadly neutralizing antibodies. It shows a β-hairpin conformation that exists as an exposed flap-like structure with an N-linked glycan at one side and the antibody binding to the other side 
In the present study we report the crystal structures of Fab/peptide complexes derived from two human mAbs of the HC84 group (HC84-1 and HC84-27, respectively). The peptide is derived from the epitope II (434-NTGWLAGLFYQHK-446; residues in bold are highly conserved across genotypes). These structures reveal the determinants of the mAb interaction with this cluster of epitopes, allowing for improved design of immunogens properly presenting one of these epitopes.