Cross-reactive antibodies binding to H4 hemagglutinin protect against a lethal H4N6 influenza virus challenge in the mouse model

Abstract

Influenza viruses of the H4 subtype are widespread in wild birds, circulate in domestic poultry, readily infect mammals, and tolerate the insertion of a polybasic cleavage site. In addition, serological evidence suggests that humans working with poultry are exposed to these viruses. While H4 viruses are not of immediate pandemic concern, there is a lack of knowledge regarding their antigenicity. In order to study viruses of the H4 subtype, we generated and characterized a panel of antibodies that bind a wide variety of H4 hemagglutinins from avian and swine isolates of both the Eurasian and North American lineage. We further characterized these antibodies using novel recombinant H4N6 viruses that were found to be lethal in DBA/2J mice. Non-neutralizing antibodies, which had activity in an antibody dependent cell-mediated cytotoxicity reporter assay in vitro, protected mice against challenge in vivo, highlighting the importance of effector functions. Our data suggest a high degree of antigenic conservation of the H4 hemagglutinin.

Keywords: H4N6; Influenza virus; avian influenza; hemagglutinin; monoclonal antibodies.

Figure 1.

Cross-reactive antibodies bind divergent HAs from avian and mammalian isolates of both the Eurasian and North American H4 lineage. (A) Phylogenetic analysis of various selected H4 HAs. The HAs of H4 viruses cluster into two groups, Eurasian or North American lineage. The scale bar represents a 1% difference in amino acid sequence. (B–G) Binding of mAbs to recombinant H4 proteins of avian or mammalian origin. Standard ELISA assays were performed to test the binding of the eight mAbs to (B) A/duck/Czechoslovakia/1956 H4, (C) A/swine/HuBei/06/2009 H4, (D) A/duck/Zhejiang/D9/2013 H4, (E) A/blue-winged teal/Illinois/10OS1563/2010 H4, (F) A/shorebird/Delaware Bay/718/1988 H4, and (G) A/swine/Missouri/A01727926/2015 H4. (H–I) Binding of mAbs to recombinant chimeric HA proteins, cH4/3 and cH4/1, consisting of an H4 head and H3 or H1 stalks respectively. These ELISAs were performed on Ni-NTA plates onto which the chimeric proteins were added. (J) Binding of mAbs to a Cal09 H1 recombinant protein. The positive control used was mAb CR9114, which is a broadly reactive human antibody binding all HA subtypes. A mAb binding to the Lassa virus glycoprotein was used as negative control. (K) IVPM. An influenza virus protein microarray was utilized to further test the breadth of the mAbs to HAs from all influenza A virus HA subtypes. Microarray slides were printed with all the respective proteins and three different dilutions per antibody were tested. Area under the curve (AUC) values from each antibody were calculated and the data is presented as a heat map. Shown are group 1 HA proteins followed by group 2 HA proteins including various H4 proteins, and a neuraminidase (N2) protein as a control. dCZ 56 corresponds to A/duck/Czechoslovakia/1956 H4, dZJ13 corresponds to A/duck/Zhejiang/D9/2013 H4, sHB 09 corresponds to A/swine/HuBei/06/2009 H4, bwtIL 10 corresponds to A/blue-winged teal/Illinois/10OS1563/2010 H4, sMO 15 corresponds to A/swine/Missouri/A01727926/2015 H4, and sDE 88 corresponds to A/shorebird/Delaware Bay/718/1988 H4. One influenza virus neuraminidase, N2, was added as an irrelevant protein and the negative control used was an anti-Lassa glycoprotein antibody.

Figure 2.

Antibody binding to H4 on infected cells and in Western blot. (A–F) Immunofluorescence analysis to assess binding of antibodies to HA on the surface of infected cells. MDCK cells were infected with (A) A/duck/Czechoslovakia/1956 (H4N6-PR8), (B) A/shorebird/Delaware Bay/718/1988 (H4N6), (C) A/blue-winged teal/Illinois/10OS1563/2010 H4N6, (D) A/duck/Zhejiang/D9/2013 (H4N6-PR8), (E) A/Caspian seal/Russia/T1/2012 (H4N6-PR8), and (F) A/swine/Missouri/A01727926/2015 (H4N6-PR8) followed by staining with 30 μg/ml of each antibody and secondary antibody treatment (anti-mouse Alexa Fluor 488). A mAb binding to the Lassa virus glycoprotein served as negative control while a pan HA mAb, CR9114, served as positive control. (G) Western blot analysis. Recombinant H4 (A/duck/Czechoslovakia/1956) and recombinant H11 HA were denatured and reduced, run on an SDS-PAGE and then blotted onto a nitrocellulose membrane. Membranes were probed with 30 μg/ml of each antibody and then treated with anti-mouse IgG alkaline phosphatase secondary antibody. Both recombinant proteins feature a hexahistidine tag, and an anti-hexahistidine antibody was used as a positive control.

Figure 3.

Only one mAb neutralizes in vitro but several antibodies show ADCC activity in vitro. (A) Microneutralization assay against A/duck/Czechoslovakia/1956 (H4N6-PR8). A microneutralization assay was performed using all the eight antibodies at a starting concentration of 100 μg/ml. (B–C) ADCC reporter assay to assess engagement of the murine FcγIV receptor by the antibodies bound to infected cells. MDCK cells were infected with A/duck/Czechoslovakia/1956 (H4N6-PR8) (B) or A/swine/Missouri/A01727926/2015 (H4N6-PR8), then incubated with various dilutions of each antibody and effector/reporter cells expressing luciferase upon activation. Luminescence was measured as a readout. The negative control antibody was an anti-Lassa virus glycoproteins antibody and the positive control antibody used was mAb CR9114.

Figure 4.

Infection of DBA/2J mice with recombinant H4N6 viruses in the PR8 background. (A) Weight loss of mice (n = 4) after infection with A/duck/Czechoslovakia/1956 (H4N6-PR8). DBA/2J mice were infected with various doses of A/duck/Czechoslovakia/1956 (H4N6-PR8) and monitored for 14 days after infection. (B) Survival of mice after infection with A/duck/Czechoslovakia/1956 (H4N6-PR8). DBA/2J mice were infected with various doses of A/duck/Czechoslovakia/1956 (H4N6-PR8) and weight loss was monitored for 14 days after infection. (C) Weight loss of mice after infection with A/swine/Missouri/A01727926/2015 (H4N6-PR8). DBA/2J mice were infected with various doses of A/swine/Missouri/A01727926/2015 (H4N6-PR8) and monitored for 14 days after infection. (D) Survival of mice after infection with A/swine/Missouri/A01727926/2015 (H4N6-PR8). DBA/2J mice were infected with various doses of A/swine/Missouri/A01727926/2015 (H4N6-PR8) and weight loss was monitored for 14 days after infection.

Figure 5.

Protective prophylactic efficacy of mAbs in DBA/2J mice against challenge with A/duck/Czechoslovakia/1956 (H4N6-PR8) or A/swine/Missouri/A01727926/2015 (H4N6-PR8). (A–B) Mice were administered 10 mg/kg of each mAb intraperitoneally two hours prior to intranasal challenge with 5 mLD₅₀ of A/duck/Czechoslovakia/1956 (H4N6-PR8) virus and mice were monitored for 14 days after infection. Survival (A) and weight loss (B) curves are shown. The negative control antibody was an anti-Lassa virus glycoprotein antibody. (C–D) Viral lung titres after infection of DBA/2J mice with A/duck/Czechoslovakia/1956 (H4N6-PR8). Ten mg/kg of each respective antibody was administered intraperitoneally two hours prior to intranasal infection with 1 mLD₅₀ of virus. Mice were sacrificed on day 3 (C) and day 6 (D) and lungs were harvested and viral titre was determined using a plaque assay. (E-F). Mice were administered 10 mg/kg of mAb intraperitoneally two hours prior to intranasal challenge with 5 mLD₅₀ of A/swine/Missouri/A01727926/2015 (H4N6-PR8) virus and mice were monitored for 14 days after infection. Survival (E) and weight loss (F) curves are shown. The negative control antibody was an anti-Lassa virus glycoprotein antibody.