Potential conflict of interest: Nothing to report.
This work was funded by grants from the UK Medical Research Council, ANRS, and recurrent funding from Institut Pasteur, CNRS, and Merck-Serono (to F. A. R.).
Address reprint requests to: Thomas Krey, Institut Pasteur, Structural Virology Unit, 25-28 Rue du Docteur Roux, 75015 Paris, France. E-mail: email@example.com; fax: (33)-145688993; or
Jonathan Ball, Virus Research Group, A Floor, West Block QMC, Queen's Medical Centre, Nottingham NG7 2UH, UK. E-mail: firstname.lastname@example.org; fax: (44)-1158230759.
Severe liver disease caused by chronic hepatitis C virus is the major indication for liver transplantation. Despite recent advances in antiviral therapy, drug toxicity and unwanted side effects render effective treatment in liver-transplanted patients a challenging task. Virus-specific therapeutic antibodies are generally safe and well-tolerated, but their potential in preventing and treating hepatitis C virus (HCV) infection has not yet been realized due to a variety of issues, not least high production costs and virus variability. Heavy-chain antibodies or nanobodies, produced by camelids, represent an exciting antiviral approach; they can target novel highly conserved epitopes that are inaccessible to normal antibodies, and they are also easy to manipulate and produce. We isolated four distinct nanobodies from a phage-display library generated from an alpaca immunized with HCV E2 glycoprotein. One of them, nanobody D03, recognized a novel epitope overlapping with the epitopes of several broadly neutralizing human monoclonal antibodies. Its crystal structure revealed a long complementarity determining region (CD3) folding over part of the framework that, in conventional antibodies, forms the interface between heavy and light chain. D03 neutralized a panel of retroviral particles pseudotyped with HCV glycoproteins from six genotypes and authentic cell culture–derived particles by interfering with the E2-CD81 interaction. In contrast to some of the most broadly neutralizing human anti-E2 monoclonal antibodies, D03 efficiently inhibited HCV cell-to-cell transmission. Conclusion: This is the first description of a potent and broadly neutralizing HCV-specific nanobody representing a significant advance that will lead to future development of novel entry inhibitors for the treatment and prevention of HCV infection and help our understanding of HCV cell-to-cell transmission. (Hepatology 2013;53:932–939)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
An estimated 180 million people worldwide are infected with hepatitis C virus (HCV). Chronic infection is frequent and often leads to progressive liver disease, with chronic HCV infection being the leading indication for liver transplantation. HCV exhibits significant genetic diversity, and at least six major genotypes, which differ by up to 30% in their nucleotide sequence, have been identified. Within an infected individual, the virus exists as a population of genetically related variants or quasispecies. This contributes to viral persistence by facilitating escape from antiviral selection pressures, with significant implications for the design of antiviral therapeutics and vaccines.
The standard treatment for chronic HCV infection is a combination of pegylated interferon-α and ribavirin. The recently introduced protease inhibitors boceprevir and telaprevir show improved treatment outcomes for genotype 1 infections in combination with interferon-α and ribavirin. However, therapy is limited by severe side effects and, dependent on the viral genotype, variable efficacy and drug resistance.
Therapeutic administration of anti-HCV neutralizing antibodies may contribute to future combination therapies with protease and/or polymerase inhibitors. The HCV glycoproteins E1 and E2 are the major target for neutralizing antibodies, and immunization studies have generated broadly reactive antibody responses. Neutralizing human monoclonal antibodies specific for HCV E2 have been shown to protect against heterologous virus challenge in a human liver-chimeric mouse model and in chimpanzees. More recently, a neutralizing human monoclonal antibody specific for HCV E2 was reported to delay viral rebound in patients following liver transplantation. Viral clearance during acute infection is associated with the presence of high-titer neutralizing antibodies.[11-13] However, reports that HCV can evade neutralizing antibodies by transmitting via cell-to-cell contacts have raised concerns on the efficacy of antibodies targeting the viral glycoproteins to limit viral transmission.[14-16]
In members of the family Camelidae, a large proportion of the humoral immune response is comprised of homodimeric immunoglobulin G (IgG) devoid of light chains (∼80 kD). These heavy-chain antibodies also lack the CH1 region, and their variable region is referred to as VHH or nanobody. Recombinant nanobodies (∼14 kD) are intact antigen-binding domains and exhibit a broad antigen-binding repertoire. They have unique characteristics, including an extended complementarity determining region 3 (CDR3) loop that can adopt a protruding conformation allowing interaction with concave epitopes that are occluded for conventional antibodies. To stabilize the enlarged CDRs, nanobodies often possess an additional disulfide bond between CDR1 and CDR3 in dromedaries, and CDR2 and CDR3 in llamas. Nanobodies have been raised to numerous viruses (reviewed by Vanlandschoot et al.) and despite being monovalent, they frequently exhibit biological activities comparable to conventional bivalent antibody molecules. As such, nanobodies are a promising tool for the targeted immunotherapy of viral infections.
Here, we report the isolation and characterization of four anti-HCV alpaca nanobodies raised by immunizing an alpaca with recombinant hepatitis C virus E2 glycoprotein. One of these nanobodies neutralized HCV pseudoparticles (HCVpp) representing diverse genotypes, authentic HCV cell culture–grown infectious particles (HCVcc) virions, and uniquely inhibited HCV cell-to-cell transmission. This provides the first evidence of nanobodies as potential candidates for immunotherapeutic administration in chronic hepatitis C.
Materials and Methods
HCVpp and HCVcc Neutralization Assays
HCVpp and 100 focus-forming units of Japanese fulminant hepatitis isolate 1 (JFH-1), respectively, were mixed with serum, nanobody, or monoclonal antibody for 1 hour before being added to Huh7.5 cells. Further experimental details are provided in the Supporting Information.
Nanobody Crystallization, Data Collection and Structure Determination
Crystals of the nanobody D03 were grown at 293 K using the sitting-drop vapor-diffusion method in drops containing 1.2 μL protein (∼20 mg/mL in 10 mM Tris [pH 8.0], 150 mM NaCl) mixed with 1.2 μL reservoir solution containing 2,005 mM LiSO4. Diffraction quality rod-like crystals belonging to spacegroup P65 appeared after 4 weeks. Data collection, processing, and structure solution are described in detail in the Supporting Information. The atomic coordinates and structure factors have been deposited in the Protein Data Bank, www.pdb.org, under the accession number 4JVP.
Inhibition of HCV Cell-to-Cell Transmission
Cell-to cell transmission was analyzed as described. HCV-infected producer Huh-7.5 cells were labeled with 5-chloromethylfluorescein diacetate (CMFDA) and cocultured with naïve cells in a 1:1 ratio for 2 hours. Antibodies, antibody fragments, or nanobodies were added, extracellular media was collected after 24 hours, and infectious virus was quantified by infecting naïve Huh-7.5 cells. Cell cocultures were fixed, permeabilized, stained for HCV NS5A expression, and analyzed by flow cytometry to quantify the number of newly infected target cells (NS5A+/CMFDA−). The frequency of target cells infected via the cell-to-cell transmission route in the presence of antibody was compared with the polyclonal anti-HCV Ig control. Statistical significance of neutralization was determined via one-way analysis of variance (ANOVA) with Bonferroni correction.
Nanobodies Neutralize HCV Infectivity
We used a phage-display library isolated from an alpaca immunized with HCV E2 to identify four nanobodies specifically recognizing E2 (Supporting Fig. 1). The nanobodies were expressed in Escherichia coli and antigen specificity was demonstrated via pull-down and immunofluorescence assay (Supporting Fig. 2). All four nanobodies were assessed for their ability to inhibit HCVpp and HCVcc infection. Determination of autologous neutralization of HCVpp bearing glycoproteins of the immunogen HCV isolate UKN2B2.8 revealed that D03 neutralized virus infection in a dose-dependent manner (>95% at 20 μg mL), while C09 possessed some neutralizing activity, and B11 and D04 had no effect on HCVpp infectivity (Supporting Fig. 3D). Subsequent analysis using JFH-1 HCVcc revealed that D03 had the strongest neutralizing effect, whereas C09 had a minor inhibitory effect (Fig. 1A). B11 and D04 did not show any neutralizing activity. Taken together, these data demonstrate that D03 neutralizes the infectivity of HCVpp and HCVcc expressing glycoproteins of HCV genotype 2.
To assess the breadth of neutralizing activity, all four nanobodies were screened at a single concentration for their inhibitory effect on entry of pseudoparticles bearing a well-characterized and diverse panel of HCV glycoproteins that exhibited different sensitivities to serum neutralizing antibodies. Only D03 possessed significant cross-neutralizing activity; C09 only neutralized HCVpp pseudotyped with genotype 2 glycoproteins (Fig. 1A). A more detailed analysis of the cross-reactive neutralization profile of D03 using a panel of HCVpp representing all six major HCV genotypes revealed that D03 neutralized across all genotypes, exhibiting 50% inhibitory concentrations that ranged between 1 and 10 μg/mL for most isolates. Some isolates, such as UKN2A1.2 (genotype 2a) and UKN2B1.1 (genotype 2b), were more easily neutralized by D03 than by monoclonal antibody (mAb) 1:7 (used as positive control). However, other strains such as UKN3A13.6 (genotype 3a) and UKN5.15.7 (genotype 5) were more refractory to neutralization by D03 and required significantly more nanobody to achieve 50% inhibition. These results indicated that the epitope recognized by D03 is conserved across genetically diverse isolates, but presentation of the epitope at the virion surface may differ between strains.
Crystal Structure of D03
To gain insight into the conformation of its potential antigen-binding determinants, we crystallized D03 and determined its crystal structure to 1.8 Å resolution; details and statistics of the data collection, processing, and refinement are given in Supporting Table 1. As expected, the nanobody displayed an immunoglobulin fold (Fig. 2A). CDR1, CDR2, and CDR3 were 8, 8, and 20 residues long, respectively, according to the IMGT nomenclature (Supporting Fig. 1C), the CDR3 thus being longer than the average CDR3. It folded over part of the framework region, which—in conventional antibodies—forms the VH-VL interface. The flexibility of the extended CDR3 in D03 is restricted by a disulfide bridge between Cys50 directly upstream of the CDR2 and Cys103 in the CDR3 (Fig. 2A).
Antibody maturation in nanobodies frequently includes somatic mutations that improve shape or charge complementarity of the paratope with the antigen. These mutations occur mainly in residues that are not involved in antigen contacts, leading to reorganization of hydrogen bonding networks and electrostatic and van der Waals interactions, which often results in increased affinity for antigen binding. Amino acid alignment of D03 with its closest homologous germline gene IGHV1S1*01 and mapping of the somatic mutations on the molecular surface revealed somatic mutations in CDR1 (n = 1), CDR2 (n = 4), and CDR3 (n = 1) (Supporting Fig. 1C and Fig. 2B).
Epitope Mapping of Nanobody D03
The majority of reported anti-HCV E2 broadly neutralizing antibodies inhibit E2 binding to the receptor CD81, their epitopes overlapping the CD81 binding site (reviewed by Edwards et al.), which comprises mainly three discontinuous amino acid regions: 412-425, 428-446, and 523-540. Remarkably, all human conformation-sensitive antibodies recognizing the latter region bind to four main contact residues (G523, W529, G530, and D535), and different combinations of at least two of these have been reported for individual antibodies.
The antigenic region binding D03 was identified by competition analysis of D03 with binding of a well-characterized panel of mAbs to HCV E2 (Supporting Fig. 4B). D03 competed for binding of mAbs 1:7, AR3A and AR1A to HCV E2. These mAbs bind to epitopes localized in the CD81 binding region, suggesting that D03 also neutralizes HCV by interfering with E2-CD81 binding. This was further supported by the fact that no simultaneous binding of D03 and CD81-LEL to a soluble E2 ectodomain was detected, while the nonneutralizing nanobody B11 formed a ternary complex with E2 and CD81-LEL (Supporting Fig. 4C,D).
We defined D03 contact residues in E2 by binding analysis of D03 to a panel of HCV E2 mutants carrying individual alanine substitutions of conserved residues between amino acids 412 and 621 (Fig. 3A). D3 binding was reduced by more than 50% by substitutions at residues N415, G523, and T526, in line with an epitope overlapping with the CD81 binding site (Fig. 3).
D03 Inhibits HCV Cell-to-Cell Transmission
We and others have reported that cell-to-cell spread of HCV is resistant to several broadly neutralizing anti-E2 antibodies targeting the CD81 binding site, limiting their potential therapeutic capacity.[14-16] The mechanism used by HCV for cell-to-cell spread is unknown, and as such antibody resistance is not fully understood. We investigated the ability of D03 to limit cell-to-cell virus transmission using a recently described single-cycle infectious coculture assay. D03 neutralized more than 95% of the extracellular virus and inhibited cell-to-cell–transmitted events (Fig. 4).
HCV cell-to-cell transmission may be susceptible to the relatively small nanobody (∼15 kD), but not to full IgG molecules (∼150 kD). This would be in line with reports that the small molecules epigallocatechin-3-gallate and El-1 inhibit cell-to-cell transmission. Alternatively, neutralizing activity of D03 on cell-to-cell spread could be attributed to a specific binding mode of the nanobody to the novel epitope described above. We expressed recombinant antibody fragments of the broadly neutralizing human antibody A8 in Drosophila S2 cells.[31, 32] A8 binds to an epitope defined by contact residues G523, W529, G530, and D535 overlapping the novel epitope recognized by D03. Despite neutralizing >95% of extracellular JFH-1 virus, neither A8 IgG nor any of the recombinant fragments (Fab or scFv; ∼50 kD and ∼30 kD, respectively) (Fig. 4) reduced cell-to-cell transmission. Although these results suggest that epitope specificity rather than size per se underpins the inhibition of HCV cell-to-cell transmission, it is worth noting that the nanobody (∼15 kD) is still only half the size of the scFv.
End-stage liver disease due to HCV infection is the leading indication of liver transplantation, and reinfection of the graft occurs universally. It is unlikely that new direct-acting antiviral agents such as telaprevir and boceprevir will prove effective in the liver transplant setting, because these drugs have been shown to interfere with commonly used immunosuppressive drugs. Moreover, combination therapy targeting multiple steps of the virus cycle is likely to be needed to limit the emergence of drug resistance. With their relatively safe profile, nanobodies have therapeutic potential in the particularly sensitive group of liver transplantation patients and as part of a combination therapy for the wider treatment of chronic HCV infection. We undertook the novel approach to generate HCV-specific nanobodies from the heavy-chain antibody repertoire of an alpaca. Nanobodies combine the advantages of high-affinity binding, protein stability, and an ability to bind epitopes that are not recognized by conventional IgG molecules. Characterization of the selected nanobodies revealed that D03 recognizes a novel conserved E2 epitope. This nanobody can inhibit cell-free virus and cell-to-cell transmission that has been reported to be resistant to broadly neutralizing antibodies and patient polyclonal Ig.
To maximize the likelihood of inducing cross-reactive nanobodies, we immunized with a truncated form of E2 lacking HVR1—the main target for strain-specific neutralizing epitopes (reviewed by Edwards et al.). We assessed neutralization breadth using HCVcc and also HCVpp supplemented with E1E2 that are resistant to antibody neutralization.
Epitope mapping revealed that possible contact residues recognized by D03 included N415, G523, and T526. N415 forms part of the epitope by broadly neutralizing mAbs AP33 and HCV1.[36, 37] G523 has been identified as a contact residue for conformation-sensitive antibodies AR3A, 1:7, A8, CBH-5 and to a lesser extent HC-1 and HC-11 (reviewed by Edwards et al.), all of which interact with at least one additional residue among W529, G530, and D535. Although D03 also interacts with G523, it does not depend on W529, G530, or D535 as contact residues. Instead, T526 acts as additional contact, a conserved residue that had not been implicated in a neutralizing epitope before. These data suggest that the epitope recognized by D03 overlaps with epitopes used by CD81 binding site-specific broadly neutralizing human antibodies but clearly possesses a novel pattern of contact residues.
Structural analysis of D03 revealed clustering of somatic mutations at the tip of D03, suggesting that the interaction between nanobody and HCV E2 occurs likely in close proximity to CDR1 and CDR2. This indicates that the additional disulfide bridge between the framework upstream of CDR2 and the CDR3 serves to restrict the latter in a conformation that allows maximal accessibility of the tip region of the nanobody. Both neutralizing nanobodies (D03 and C09) possessed this additional disulfide bridge (Supporting Fig. 1C), whereas both nonneutralizing nanobodies lacked this linkage. It is therefore tempting to speculate that the spatial restriction of the CDR3 contributes to a specific binding mode of the neutralizing nanobodies. Of note, it has been shown that binding of E2 to CD81 is modulated by N-linked sugars forming a glycan shield, which is likely in close proximity to the epitope recognized by D03. This glycan shield could interfere with binding of conventional antibodies but allow interaction with the tip of heavy-chain antibodies, the diameter of which is smaller due to the absence of a light chain. However, structural analysis of D03 in complex with its antigen is required to understand the particularities of this binding mode.
This study is the first description of a nanobody that neutralizes both cell-free virus and cell-to-cell transmission of HCV, which has been reported to be resistant to patient polyclonal immunoglobulin and broadly neutralizing antibodies targeting the CD81 binding site. As such, this entry inhibitor has a potentially unique application to limit HCV spread in the chronically infected liver and represents a significant advance toward therapeutic administration of antibodies to prevent neutralization-resistant infection of hepatocytes. The inhibitory effect of D03 on HCV cell-to-cell transmission demonstrates that the E2 glycoprotein at the surface of transmitted virions is exposed to extracellular antibodies. This suggests that the virus may be transmitting in protected pockets between cells, rather than directly forming synaptic or cellular contacts as observed for HIV-1, as the latter are likely inaccessible to all extracellular antibodies.
Although several nanobodies have entered clinical trials in the last few years, to our knowledge this study provides the first evidence that nanobodies can prevent both cell-free and direct cell-to-cell transmission of a virus, highlighting their potential to be clinically useful entry inhibitors.
We thank Ahmed Haouz and Patrick Weber for help with crystallization and the staff of the synchrotron beamline PX-I at the Swiss Light Source for assistance during data collection. We are grateful to Mats Persson, Steven Foung, Arvind Patel, Francois-Loic Cosset, Charles Rice, Takaji Wakita, and Mansun Law for the generous provision of reagents.