Unmasking Stem-Specific Neutralizing Epitopes by Abolishing N-Linked Glycosylation Sites of Influenza Virus Hemagglutinin Proteins for Vaccine Design

Influenza virus hemagglutinin (HA) protein consists of two components, i.e., a globular head region and a stem region that are folded within six disulfide bonds, plus several N-linked glycans that produce a homotrimeric complex structure. While N-linked glycosylation sites on the globular head are variable among different strains and different subtypes, N-linked glycosylation sites in the stem region are mostly well conserved among various influenza virus strains. Targeting highly conserved HA stem regions has been proposed as a useful strategy for designing universal influenza vaccines. Since the HA stem region is constituted by an HA1 N-terminal part and a full HA2 part, we expressed a series of recombinant HA mutant proteins with deleted N-linked glycosylation sites in the HA1 stem and HA2 stem regions of H5N1 and pH1N1 viruses. Unmasking N-glycans in the HA2 stem region (H5 N484A and H1 N503A) was found to elicit more potent neutralizing antibody titers against homologous, heterologous, and heterosubtypic viruses. Unmasking the HA2 stem N-glycans of H5HA but not H1HA resulted in more CR6261-like and FI6v3-like antibodies and also correlated with the increase of cell fusion inhibition activity in antisera. Only H5 N484A HA2 stem mutant protein immunization increased the numbers of antibody-secreting cells, germinal center B cells, and memory B cells targeting the stem helix A epitopes in splenocytes. Unmasking the HA2 stem N-glycans of H5HA mutant proteins showed a significantly improvement in the protection against homologous virus challenges but did so to a less degree for the protection against heterosubtypic pH1N1 virus challenges. These results may provide useful information for designing more effective influenza vaccines. IMPORTANCE N-linked glycosylation sites in the stem regions of influenza virus hemagglutinin (HA) proteins are mostly well conserved among various influenza virus strains. Targeting highly conserved HA stem regions has been proposed as a useful strategy for designing universal influenza vaccines. Our studies indicate that unmasking the HA2 stem N-glycans of recombinant HA proteins from H5N1 and pH1N1 viruses induced more potent neutralizing antibody titers against homologous and heterosubtypic viruses. However, only immunization with the H5N1 HA2 stem mutant protein can refocus B antibody responses to the helix A epitope for inducing more CR6261-like/FI6v3-like and fusion inhibition antibodies in antisera, resulting in a significant improvement for the protection against lethal H5N1 virus challenges. These results may provide useful information for designing more effective influenza vaccines.