A multifunctional human monoclonal neutralizing antibody that targets a unique conserved epitope on influenza HA

Abstract

The high rate of antigenic drift in seasonal influenza viruses necessitates frequent changes in vaccine composition. Recent seasonal H3 vaccines do not protect against swine-origin H3N2 variant (H3N2v) strains that recently have caused severe human infections. Here, we report a human V_H1-69 gene-encoded monoclonal antibody (mAb) designated H3v-47 that exhibits potent cross-reactive neutralization activity against human and swine H3N2 viruses that circulated since 1989. The crystal structure and electron microscopy reconstruction of H3v-47 Fab with the H3N2v hemagglutinin (HA) identify a unique epitope spanning the vestigial esterase and receptor-binding subdomains that is distinct from that of any known neutralizing antibody for influenza A H3 viruses. MAb H3v-47 functions largely by blocking viral egress from infected cells. Interestingly, H3v-47 also engages Fcγ receptor and mediates antibody dependent cellular cytotoxicity (ADCC). This newly identified conserved epitope can be used in design of novel immunogens for development of broadly protective H3 vaccines.

Fig. 1

Prophylactic and therapeutic efficacy of mAb H3v-47 in mice. Groups of DBA/2J mice were inoculated i.p. with 10 (high dose) or 1 mg/kg (low dose) of mAb H3v-47 on the day before (day −1, prophylaxis), or on day one (day +1, treatment) after lethal respiratory (i.n. route) challenge with ∼ 6 × 10⁴ focus forming units (FFU) of A/Minnesota/11/2010 × −203 virus and monitored for 10 days. The control groups included mice treated with PBS (mock control) or the previously described mAb CR8020. The protective efficacy of mAbs was assessed by weight change kinetics (a, b), clinical score (c, d), and virus titers in lungs on 4 dpi (e, f). The dotted line (a, b) indicates the IACUC stipulated endpoint for humane euthanasia. Data in a, b shows body weight only for the animals that survived based on actual death between watches or endpoint for euthanasia, and represents the mean value ± SEM, using 10 mice per group. Each group was compared to the PBS-treated group using two-way ANOVA with Dunnett‘s post-hoc test for a, b, and one-way ANOVA with Dunnett‘s post-hoc test for e, f. Dots in e, f indicate individual mice (n = 5 mice per group). The median titer is shown with the line, and the dotted line indicates the limit of detection (LOD). *p < 0.05; **p < 0.01; ***p < 0.001 were considered significant

Fig. 2

Competition binding of H3v-47 with other influenza head- or stem-binding antibodies. Competition-binding assays were performed using biolayer interferometry. The His-tagged A/Minnesota/11/2010 H3N2v HA was loaded onto Ni-NTA tips, and binding of two successively applied antibodies (IgG) was tested. MAb H3v-47 was competed against mAb C05, a receptor binding site mAb or each of four stem-binding antibodies: CR8020, CR9114, 39.29, or FI6v3. Numbers indicate normalized percent level of association to HA, compared to uncompeted control (100%). The colored boxes indicate each competition-binding group. There was partial overlap between group 1 and 2, shown in cyan and blue, respectively. Competition was not detected between H3v-47 and the stem antibodies indicated in the red competition-binding group box. The experiment was conducted twice independently

Fig. 3

Crystal structure of antibody H3v-47 Fab in complex with H3N2v HA (A/Minnesota/11/2010) and identification of the epitope. a Overall structure of the H3v-47 Fab- H3N2v HA complex. One HA/Fab protomer of the trimeric complex is colored with HA1 in yellow, HA2 in orange, Fab heavy chain in green, and Fab light chain in cyan. N-linked glycans are depicted as colored balls representing their atom types (carbon in pink, oxygen in red and nitrogen in blue). The other two protomers are in gray, but the third Fab molecule is hidden behind the HA trimer. b Zoomed-in view of the interaction between H3N2v HA and H3v-47 Fab with color coding as in a. H3v-47 Fab binds to the region spanning the receptor-binding and vestigial esterase sub-domains (shown as a solid surface in yellow and wheat, respectively) mainly using CDRs H2, H3, L1, and L3. The contact regions ascribed to antigenic sites A and E are highlighted in orange or red, respectively. c H3v-47 epitope mapped onto the H3N2v HA surface. The footprint of antibody H3v-47 on the HA shows the central role of antigenic E site residues for H3v-47 binding. The interacting surface contributed by residues from antigenic sites A or E is colored in red or orange, respectively, whereas residues that contribute to the epitope outside of these two antigenic sites, are colored in yellow. d Antibody contact residues (sticks) that interact with the HA with CDRs H2, H3, L1, L3, and FR3 in the light chain. e Specific interactions (H-bonds, salt bridges, hydrophobic interactions, and VDW contacts) between H3v-47 and its footprint residues shown as sticks. The H-bonds and salt bridges are labeled using dash lines

Fig. 4

Inhibition of egress of A/Minnesota/11/2010 H3N2v virus by IgG of mAbs H3v-47 or CR8020 or by the small molecule neuraminidase inhibitor zanamivir. Twenty-hours prior to the experiment, MDCK cells were seeded on 6-well plates in DMEM + 5% FBS. The cells were washed twice with PBS and inoculated with a previously optimized amount of A/Minnesota/11/2010 H3N2v virus that is required to achieve 90–100% infection. After 1 h incubation at 37 °C, the cells were washed twice with PBS and replenished with 1.5 mL plain Opti-MEM I medium. After 3 h, the cells were washed again and replenished with 1.5 mL of Opti-MEM I medium containing serial dilutions of the antibodies or zanamivir. The cells were incubated at 37 °C for 12 h, and the supernatants were collected. The supernatants were diluted serially 11 times and added to an equal volume of 0.5% turkey RBCs in v-bottom plates to determine the virus titer by hemagglutination assay. Dotted line represents virus titer in supernatant in the absence of antibody treatment. The experiment was conducted three times independently. The hemagglutination assay to determine virus titer was also conducted three times independently (n = 3). The significance in the reduction of HA titer between H3v-47 and CR8020 was calculated at each concentration using 2-way ANOVA and displayed on the graph as ***(P < 0.001). Error bars indicate standard error of the mean (SEM)

Fig. 5

H3v-47 localizes to interfaces between virus and cell surface or between viral particles. a TEM images of the surface of MDCK cells inoculated with A/Minnesota/11/2010 H3N2v virus and infected cells were incubated with mAb H3v-47, zanamivir, mAb CR8020, or plain Opti-MEM at 3 h postinoculation and fixed for imaging at 14 h. For each image, the virus particles in the red squares are shown at higher magnification, below. Representative images of two independent experiments are shown. The white scale in each image represents 500 nm. b TEM images are shown of the surface of MDCK cells that were inoculated with virus and subsequently incubated with H3v-47 mAb similar to a, with addition of anti-human IgG conjugated to 10 nm gold particles at 13 h post-inoculation. The black opaque dots indicated by the red triangles represent the gold particles. The white scale in each image represents 100 nm. The experiment was conducted twice independently

Fig. 6

H3v-47 mAb exhibits ADCC activity. a Cross-linking of FcγRIIIa. Binding curves were obtained by performing ELISA with serial dilutions of each antibody (H3v-47, H3v-12, or control mAb VRC01 [to HIV]) onto HA-coated plates and assessing the ability of HA-bound mAbs to engage both Fc-binding sites on the soluble FcγRIIIa dimer. b Primary NK cell activation. Antibody (H3v-47, H3v-12, or VRC01) at 0.1, 1, or 10 µg/mL each were added independently on 96-well plates coated with purified A/Sydney/5/1997 H3 HA and incubated for 2 h. The plates were washed, and 5 × 10⁵ purified NK cells were added to each well. The cells were stained with anti-human CD107a allophycocyanin-H7 Ab, anti-human CD3 PerCP, anti-human CD56 allophycocyanin, and anti-IFNγ AF700. The data for 20,000–50,000 events were acquired using an LSRFortessa flow cytometer. The percentage of NK cell activation was calculated as the percentage of NK cells that expressed CD107a and/or IFNγ. The dotted lines in both a and b indicate the limit of detection. The results are representative of three independent experiments. Error bars indicate standard error of the mean (SEM)

Fig. 7

H3v-47 binds a unique epitope in the HA head domain. a Comparison of the binding site of H3v-47 (in magenta) wih structurally characterized H3-binding antibodies binding to the HA head. The three protomers of the HA trimer are shown as surface in light, middle or dark gray, respectively. The HAs from all these HA-antibody complexes are aligned. These antibodies include HC45 (PDB code 1QFU, green) and F005-126 (PDB code 3WHE, purple) binding to the globular head below the receptor-binding site (RBS). For comparison, mAbs C05 (PDB code 4FP8), S139 (PDB code 4GMS), HC19 (PDB code 2VIR), HC63 (PDB code 1KEN), and F045-092 (PDB code 4O58) all bind to the RBS. Because the epitopes of the RBS-binding antibodies overlap extensively, their epitopes (in wheat) are depicted approximately around the RBS, to indicate their relative location compared to H3v-47. b–d Comparison of the detailed epitopes recognized by H3v-47 (b), HC45 (c), and F005-126 (d)

Fig. 8

Vestigial esterase subdomain-binding antibodies mediate neutralization by diverse mechanisms. a Comparison of H3v-47 (magenta) with structurally characterized esterase subdomain-binding antibodies, including H5M9 binding to H5 HA (PDB code 4MHH, blue) and CR8071 (PDB code 4FQJ, orange) binding to influenza B HA. The HAs from all these complexes are aligned, and only the three protomers of A/Minnesota/11/2010 H3N2v HA trimer are shown as surface in light, middle or dark gray, respectively. The HAs from all these HA-antibody complexes are aligned. The vestigial esterase subdomain is highlighted in red. These antibodies target the HA vestigial esterase subdomain using different angles of approach. b-c The epitopes of CR8071 (b) and H5M9 (c) are mapped onto the surfaces of the corresponding HAs