Both Neutralizing and Non-Neutralizing Human H7N9 Influenza Vaccine-Induced Monoclonal Antibodies Confer Protection

Abstract

Pathogenic H7N9 avian influenza viruses continue to represent a public health concern, and several candidate vaccines are currently being developed. It is vital to assess if protective antibodies are induced following vaccination and to characterize the diversity of epitopes targeted. Here we characterized the binding and functional properties of twelve H7-reactive human antibodies induced by a candidate A/Anhui/1/2013 (H7N9) vaccine. Both neutralizing and non-neutralizing antibodies protected mice in vivo during passive transfer challenge experiments. Mapping the H7 hemagglutinin antigenic sites by generating escape mutant variants against the neutralizing antibodies identified unique epitopes on the head and stalk domains. Further, the broadly cross-reactive non-neutralizing antibodies generated in this study were protective through Fc-mediated effector cell recruitment. These findings reveal important properties of vaccine-induced antibodies and provide a better understanding of the human monoclonal antibody response to influenza in the context of vaccines.

Figure 1. Characterization of human H7-reactive antibodies

(A) Human monoclonal antibodies (mAbs) were cloned from plasmablasts isolated at day 7 after the inactivated H7N9 vaccine. (B) Twenty H7-reactive mAbs were screened for HAI, neutralization and cross-reactivity to diverse influenza A HAs (group 1 and 2). (C) Binding to recombinant HA proteins was tested by ELISA. Representative minimum positive concentrations (μg/ml) from 3 independent experiments are plotted as a heatmap. The different HAs were clustered by amino acid sequences phylogeny and multiple alignments were performed using the CLUSTALW algorithm. Phylogenetic rooted tree was constructed using the neighbor-joining method and was visualized using FigTree v1.4.0 software. mAbs are grouped by individuals (22, 07, 24 and 41) and HAI+ Neut+, HAI− Neut+ and HAI− Neut− categories are mentioned. Affinity measurements (K_D) determined by biolayer interferometry for A/Shanghai/1/2013 (H7N9) HA are displayed in purple. See also Table S1 and Figure S1.

Figure 2. Human H7-reactive antibodies exhibit various in vitro functional characteristics

(A–D) Minimum positive concentrations in μg/ml (mean± SEM) are reported for HAI; IC₅₀ in μg/ml (mean± SEM) are reported for neutralization. Experiments were done in duplicate 3 times. (A) HAI and neutralization assays were performed with A/Shanghai/1/2013 (H7N9) virus. (B) Neutralization with various group1 and 2 viruses based on previous ELISA results; n.d. for not determined. In color are reported previous neutralization results from (A) with H7N9. (C) HAI was performed with various H7 viruses from the Eurasian (H7N7) and North American (H7N3 and H7N1) lineages. Only the HAI+ Neut+ mAbs were tested. (D) HAI was performed with wild-type A/Anhui/1/2013 (H7N9) virus and with an escape mutant virus that displays mutations in the H7 HA antigenic site A (R149G). Only the HAI+ Neut+ mAbs were tested. No HAI activity was observed for 07-4D05 with the escape mutant (R149G). See also figure S1.

Figure 3. Human H7-reactive antibodies confer protection in vivo

6–8 week old female BALB/c mice were injected with 5, 1 or 0.3 mg/kg of each mAb and then infected with 7.5 LD₅₀ of A/Shanghai/1/2013 (H7N9) virus. Values represent mean ± SEM (n=5 mice per group). (A–J) Percent of initial weight and percent survival are plotted for each mAb: (A) 07-5D03, (B) 07-5F01, (C) 07-5G01, (D) 07-4B03, (E) 22-3E05, (F) 07-5B05, (G) 41-5E04, (H), 41-5D06, (I) 07-5E01 and (J) 24-4C01. Mice that received an H3-reactive but not H7-reactive mAb (011-2C01) died at day 6 after the challenge. The same control group of mice was used for all panels. See also Figure S2.

Figure 4. Human H7-neutralizing antibodies bind to various epitopes on the HA stalk and head domains

Escape mutant variants with A/Shanghai/1/2013 (H7N9) virus were generated for the neutralizing mAbs. Sequences for each mutant were compared to the wild-type virus and unique mutations were reported. (A) Critical residues for binding of Neut+ mAbs are reported. Nomenclature for substitutions follows the rule of original residue, position of the amino acid, new residue (H3 numbering). (B–C) Modeling of A/Shanghai/1/2013 H7 HA was done using PyMOL (PDB ID 4LN3). HAI+ Neut+ critical residues are shown in red and HAI− Neut+ residues are shown in yellow. The Arg65 residue, (HAI+ and HAI−) is shown in orange. See also Figure S3.

Figure 5. Neutralizing and non-neutralizing H7-reactive antibodies bind to different epitopes on hemagglutinin

(A) Biotinylated mAbs 41-5E04 (HAI− Neut+) and 41-5D06 (HAI− Neut−) were tested for binding to A/Shanghai/1/2013 HA by ELISA with or without the presence of a competitor mAb. The experiment was done in duplicate 3 times. The percentage of competition between the 41-5E04 and 41-5D06 (in red) and the non-neutralizing mAbs or CR9114 is shown. Absorbance value of each mAb against itself is scored at 100% inhibition and comparison of different mAbs was done as a percentage of this 100% inhibition. (B) Purified virus preparations were treated with various pH-buffered solutions (from pH 7.0 to 4.4) and binding (at pH 7.4) to HA was measured by ELISA. Antibody binding was tested in duplicate and shown here is one representative of two independent experiments. See also Figure S4.

Figure 6. Fc-FcγR interactions are important for in vivo protection by non- neutralizing antibodies

The HAI− Neut− mAbs and one HAI+ Neut+ antibody were cloned in an IgG2a mouse construct with or without the D265A mutation (DiLillo et al., 2014). (A) Binding of the mouse mAbs to A/Shanghai/1/2013 (H7N9) viruses by ELISA. Values represent mean ± SEM from triplicate wells. (B) In vitro MN with A/Shanghai/1/2013 (H7N9) viruses. Values represent mean ± SEM from triplicate wells. (C) 6–8 week old female BALB/c mice were injected with one dose of each mouse mAb and then infected with a sublethal dose of A/Shanghai/1/2013 (H7N9) virus. Values represent mean ± SEM (n=4 mice per group). Percent of initial weight are plotted for each mAb: 3 mg/kg for 07-5E01, 4 mg/kg for 41-5D06, 3 mg/kg for 24-4C01 and 4 mg/kg 07-5G01. Mice that received an irrelevant H6-reactive IgG2a mAb were used as a control. The same control group of mice was used for all panels. See also Figure S5 and Figure S6.