Influenza H3 hemagglutinin vaccine with scrambled immunodominant epitopes elicits antibodies directed toward immunosubdominant head epitopes

Abstract

Vaccination is the most effective countermeasure to reduce the severity of influenza. Current seasonal influenza vaccines mainly elicit humoral immunity targeting hemagglutinin (HA). In particular, the amino acid residues around the receptor-binding site in the HA head domain are predominantly targeted by humoral immunity as "immunodominant" epitopes. However, mutations readily accumulate in the head domain due to high plasticity, resulting in antigenic drift and vaccine mismatch, particularly with influenza A (H3N2) viruses. A vaccine strategy that targets more conserved immunosubdominant epitopes is required to attain a universal vaccine. Here, we designed an H3 HA vaccine antigen with various amino acids at immunodominant epitopes of the HA head domain, termed scrambled HA (scrHA). In ferrets, scrHA vaccination induced lower serum neutralizing antibody levels against homologous virus compared with wild-type (WT) HA vaccination; however, similar levels of moderately neutralizing titers against antigenically distinct H3N2 viruses were observed. Ferrets vaccinated with scrHA twice and then challenged with homologous or heterologous virus showed the same level of reduced virus shedding in nasal swabs as WT HA-vaccinated animals but reduced body temperature increase, whereas WT HA-vaccinated ferrets exhibited body temperature increases similar to those of mock-vaccinated animals. scrHA elicited antibodies against HA immunodominant and -subdominant epitopes at lower and higher levels, respectively, than WT HA vaccination, whereas antistalk antibodies were induced at the same level for both groups, suggesting scrHA-induced redirection from immunodominant to immunosubdominant head epitopes. scrHA vaccination thus induced broader coverage than WT HA vaccination by diluting out the immunodominancy of HA head epitopes. IMPORTANCE Current influenza vaccines mainly elicit antibodies that target the immunodominant head domain, where strain-specific mutations rapidly accumulate, resulting in frequent antigenic drift and vaccine mismatch. Targeting conserved immunosubdominant epitopes is essential to attain a universal vaccine. Our findings with the scrHA developed in this study suggest that designing vaccine antigens that "dilute out" the immunodominancy of the head epitopes may be an effective strategy to induce conserved immunosubdominant epitope-based immune responses.

Keywords: ferret; immunosubdominant epitopes; influenza A (H3N2) viruses; universal vaccines.

FIG 1

Preparation of Tokyo/14 17-AA scrambled HA vaccine. Side view (a) and top view (b) of the 17-AA mutagenesis positions. The 17 amino acid positions selected for mutagenesis are mapped in different colors on an HA monomer (shown in white; the other two monomers are shown in gray and black) of the A/Victoria/361/2011 (H3N2) influenza virus HA structure (PDBID: 4O5N). The amino acid residue at the receptor-binding pocket (Tyr98) is shown in red. (c) The table shows the amino acid residues at the 17 mutagenesis positions in wild-type A/Tokyo/UT-IMS2-1/2014 (Tokyo/14) HA and the amino acid substitutions in mutant HA #1–18. Dots indicate the identical amino acid residues to that found in wild-type HA. (d) Neutralizing antibody titers against Tokyo/14 WT and Mut #1 to #18 viruses were examined with A/Tokyo/UT-IMS2-1/2014 ferret antisera from four individual animals (ferret IDs 5724, 5725, 2940, and 2947).

FIG 2

Experimental design for the scrHA ferret vaccination study and pre-challenge serum neutralizing antibody titers. (a) Timeline and (b) vaccination regimens in the ferret virus challenge study. Groups of ferrets (N = 4/group) were intramuscularly (i.m.) vaccinated with Tokyo/14 recombinant HA (rHA) adjuvanted with Alhydrogel as part of a prime-only or prime-and-boost regimen. Pre-challenge sera were collected at 6 wk post-prime vaccination. Then, the animals were intranasally (i.n.) challenged with 10⁶ plaque-forming units (pfu) of homologous A/Tokyo/UT-IMS2-1/2014_PR8HY (Tokyo/14) virus or heterologous A/Kansas/14/2017_PR8HY virus (Kansas/17). Mut, mixture of mutants; –, not applicable. (c) All ferrets were bled at 6 wk (pre-challenge) and serum neutralizing antibody titers against Tokyo/14 virus and Kansas/17 were analyzed in micro-neutralization assays. Data points show the values of individual animals connected by lines for each animal. Data shown are the geometric means of duplicates. Dashed lines represent the detection limit of the assay. Prm, prime-only regimen; P & B, prime-and boost-regimen. Illustration created with BioRender.com.

FIG 3

Virus replication and clinical symptoms in ferrets after scrHA vaccination. Vaccinated or mock-vaccinated ferrets (N = 4/group) were intranasally infected with 10⁶ pfu of homologous A/Tokyo/UT-IMS2-1/2014_PR8HY (Tokyo/14) virus (a and b) or heterologous A/Kansas/14/2017_PR8HY (Kansas/17) virus (c and d). Virus titers in nasal swabs (a and c) and body temperature increase (b and d) as a clinical symptom were monitored for 7 days post-challenge. Data represent the means of each group. Statistical analyses (each vaccinated group was compared to mock-vaccinated group) were performed by using a one-way analysis of variance and corrected for multi-group comparison by using Dunnett’s test. (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001). Data of individual animals are shown in Supplementary Figures 1–4. Prm, prime-only regimen; P & B, prime-and-boost regimen.

FIG 4

Characterization of scrHA-elicited neutralizing antibodies. Neutralizing antibody titers of pre-challenge sera from the scrHA (Mut #1–18) prime-and-boost (P & B) group (N = 8), the WT HA P & B group (N = 8), and the mock-vaccinated group (N = 8) were analyzed against recent H3N2 viruses from different antigenic clades. Data points show the values of individual animals connected by lines for each animal. Data shown are the geometric mean of duplicates. P-values are indicated compared to the titers against Tokyo/14 virus (shown in Fig. 2c) analyzed by a one-way analysis of variance and corrected for multi-group comparison by using Dunnett’s test. n.s., not significant. Dashed lines represent the detection limit of the assay.

FIG 5

Characterization of scrHA-elicited HA-binding antibodies. (a–c) Binding antibody titers of pre-challenge sera from the scrHA (Mut #1–18) prime-and-boost (P & B) group (N = 8), the WT HA P & B group (N = 8), and the mock-vaccinated group (N = 8) were analyzed against Tokyo/14 wild-type HA (a), chimeric HA with the Vietnam/1203/2004 (VN1203; H5) head and Tokyo/14 stalk (b), and the Tokyo/14 17-AA mutant HAs (c; Mut #35, #39, and #53; the amino acid substitutions at 17 mutagenesis positions are shown in Supplementary Figure 6) in a cell-based ELISA by using full-length HA expressed on A549 cells. Data represent the means and SD of each group (N = 8). (d–f) AUCs (area under the curve) for individual animals (N = 8/group) in (a–c) were plotted. Bars show the median of the groups. Statistical analyses were performed by using a one-way analysis of variance and corrected for multi-group comparison by using Tukey’s test. n.s., not significant.