Escape of Hepatitis C Virus from Epitope I Neutralization Increases Sensitivity of Other Neutralization Epitopes

Abstract

The hepatitis C virus (HCV) E2 glycoprotein is a major target of the neutralizing antibody (nAb) response, with multiple type-specific and broadly neutralizing antibody (bnAb) epitopes identified. The 412-to-423 region can generate bnAbs that block interaction with the cell surface receptor CD81, with activity toward multiple HCV genotypes. In this study, we reveal the structure of rodent monoclonal antibody 24 (MAb24) with an extensive contact area toward a peptide spanning the 412-to-423 region. The crystal structure of the MAb24-peptide 412-to-423 complex reveals the paratope bound to a peptide hairpin highly similar to that observed with human MAb HCV1 and rodent MAb AP33, but with a different angle of approach. In viral outgrowth experiments, we demonstrated three distinct genotype 2a viral populations that acquired resistance to MAb24 via N415D, N417S, and N415D/H386R mutations. Importantly, the MAb24-resistant viruses exhibited significant increases in sensitivity to the majority of bnAbs directed to epitopes within the 412-to-423 region and in additional antigenic determinants located within E2 and the E1E2 complex. This study suggests that modification of N415 causes a global change in glycoprotein structure that increases its vulnerability to neutralization by other antibodies. This finding suggests that in the context of an antibody response to viral infection, acquisition of escape mutations in the 412-to-423 region renders the virus more susceptible to neutralization by other specificities of nAbs, effectively reducing the immunological fitness of the virus. A vaccine for HCV that generates polyspecific humoral immunity with specificity for the 412-to-423 region and at least one other region of E2 is desirable.IMPORTANCE Understanding how antibodies neutralize hepatitis C virus (HCV) is essential for vaccine development. This study reveals for the first time that when HCV develops resistance to a major class of bnAbs targeting the 412-to-423 region of E2, this results in a concomitant increase in sensitivity to neutralization by a majority of other bnAb specificities. Vaccines for the prevention of HCV infection should therefore generate bnAbs directed toward the 412-to-423 region of E2 and additional bnAb epitopes within the viral glycoproteins.

Keywords: glycoproteins; hepatitis C virus; neutralizing antibodies; vaccines.

FIG 1

Generation of genotype 2a HCVcc MAb24 neutralization resistance variants. (A) At day 1, HCVcc, produced by transfection of Huh7.5 cells, was incubated in the presence of MAb24. At day 3, cell-free virus, together with infected cells from the passage, was added to naive Huh7.5 cell monolayers in the presence of MAb24 and repeated every 3 days for a total of nine passages. The amount of MAb24 was progressively increased, and the proportion of infected cells was gradually decreased at the passages indicated, as shown at the top. Luciferase activity in the supernatant fluid (in relative light units [RLU]) was measured on the days indicated. (B) At day 1, HCVcc, produced by transfection of Huh7.5 cells, was incubated in the presence of MAb24, and at day 3, tissue culture supernatant fluid containing cell-free virus was applied to naive Huh7.5 monolayers in the presence of MAb24; this was repeated for a total of seven passages, and the RLU in the supernatant fluid were measured on the days indicated. (C) Virus harvested at passage 9 (P9) from panel A cultured at a starting MAb24 concentration of 9.8 μg/ml or cultured in the absence of MAb24 was incubated with serial dilutions of MAb24 and added to naive Huh7.5 cells. Infectivity was measured 72 h later. GND is a mutation in NS5B that renders HCVcc replication incompetent.

FIG 2

Sanger sequencing of cDNA clones obtained following the generation of MAb24-resistant virus. cDNA was prepared from viral RNA present in the supernatant fluid of infected cells at passage 9 and sequenced with BigDye Terminator. The entire E1E2 region was sequenced, and the region where mutations were detected is shown (residues 384 to 443).

FIG 3

Viruses with N415D, N417S, and N415D/H386R mutations are replication competent, infectious, and resistant to MAb24 neutralization. (A) RNA was prepared from reverse-engineered HCVcc containing mutations Asp415, Ser417, and Arg386/Asp415 and transfected into Huh7.5 cells. Luciferase activity was measured in the supernatant fluid every 24 h. RLU, relative light units. (B) Supernatant fluid was collected 72 h after RNA transfection into Huh7.5 cells, normalized for infectivity, and used to infect naive Huh7.5 cell monolayers. Luciferase activity was measured in the supernatant fluid every 24 h. (C) Normalized amounts of HCVcc were added to naive Huh7.5 cells. After 72 h, monolayers were fixed and stained with antibody to NS5A and counterstained with propidium iodide and the number of nuclei per focus was determined. Data are representative of two independent experiments performed in duplicate. (D) Reverse-engineered viruses with N415D and N417S were incubated with serial dilutions of MAb24 and applied to naive Huh7.5 cell monolayers. Luciferase activity in the supernatant fluid was measured at 72 h postinfection. Graphs were drawn in Prism v7 by using nonlinear regression.

FIG 4

Neutralization of HCVcc with N415D, N417S, and H386R/N415D mutations by human MAbs. Serial dilutions of each MAb were incubated with parental HCVcc (WT) or N415D, N417S, or H386R/N415D mutant virus. After 72 h, the luciferase activity was measured. Data shown are the mean of at least three independent experiments performed in triplicate, with the exception of AR5A, which was performed once in triplicate. Neutralization curves were drawn by nonlinear regression analysis (Prism v7) and used to obtain IC₅₀s (Table 2).

FIG 5

Epitope 1, represented as a green ribbon, adopts a β-hairpin conformation when bound to MAb24 and HCV1 (PDB code 4DGY), in contrast to the extended conformations found in structures of HC33.1 (PDB code 4XVJ) and 3/11 (PDB code 4WHT). Note the difference in the angles of approach between MAb24 and HCV1 evident from the shift of the heavy (brown) and light (blue) chains in this orientation.

FIG 6

Recognition of E2 epitope I by MAb24. (A) Crystal structure of epitope I peptide E2_412-423 (green) bound to the Fab of MAb24. The heavy and light chains are brown and light blue, respectively. Side chains of MAb24 within 4 Å of epitope I are shown as sticks. (B) Diagram of the interactions between epitope I (stick-and-ball representation) and MAb24. Hydrogen bonds are shown as green dotted lines, and Van der Waals contacts are shown as red arcs. Residues from epitope I and the light and the heavy chains are labeled P, L, and H, respectively. Intramolecular contacts are omitted. Residues contacting W420 of epitope I are shaded in blue. (C) Wall-eyed stereo view of epitope I peptides from MAb24 (white) and AP33 (orange; PDB code 4GAG) structures. (D) Wall-eyed stereo comparison of the orientation of MAb24 (blue/brown) and AP33 (magenta) with regard to epitope I. The respective epitope I chains were used for alignment of the complexes. (E, F) Orthogonal views of the paratopes of MAb24 (E) and AP33 (F). The representation scheme is the same as in panel A omitting the molecular surface. The CDR3s of the heavy chains bearing most of the differences between the antibodies are labeled CDR3_H. (G) Comparison of contact residues between heavy- and light-chain CDRs with peptide between MAb24 and AP33. All contacts within 4 Å are shown. Differences between AP33 and MAb24 are red. The image shown was adapted from reference .

FIG 7

Structural impact of mutations at positions 415 and 417 on the recognition of epitope I. (A, B) Electrostatic surface of MAb24. The N415D mutation is modeled and shown in cyan. The peptide is shown as a green ribbon over the molecular surface of MAb24 colored according to its electrostatic potential. The red-white-blue ramp shows a scale of −2 to +2 kT. (C, D) Electrostatic surface of AP33 (C) and HCV1 (D) showing that N415 (cyan) faces W94 of the light chain, which is highlighted in magenta. (E) Epitope I of the MAb24-sensitive strain shown as a ribbon with a modeled basic glycan at position 417 (cyan) represented as spheres. (F) Epitope I of a MAb24-resistant strain bearing the N417S mutation modeled with a basic glycan at position 415 (cyan). (G, H) Orthogonal views of the N417S mutant in complex with HC33.1. A glycan is modeled at position N415 and is shown as cyan spheres.

FIG 8

Models of the proposed interaction of epitope I, HVR1, and other antigenic regions. (A, B) Mutations in epitope I increase the sensitivity to MAbs targeting epitope I itself (blue; MAb HC33.1) and remote antigenic regions such as domain C (cyan; CBH-7), E2 components of AR5 (red; AR5A), domain D (yellow; HC84.1 and HC84.27), and the CD81 binding loop comprising part of domain B (green; CBH-5). The gray ellipse and arrows represent HVR1 and epitope I, respectively, in the MAb24-sensitive (left) and MAb-resistant (right) viral strains.