Targeting neuraminidase: the next frontier for broadly protective influenza vaccines

Abstract

Current seasonal influenza vaccines, which mainly target hemagglutinin (HA), require annual updates due to the continuous antigenic drift of the influenza virus. Developing an influenza vaccine with increased breadth of protection will have significant public health benefits. The recent discovery of broadly protective antibodies to neuraminidase (NA) has provided important insights into developing a universal influenza vaccine, either by improving seasonal influenza vaccines or designing novel immunogens. However, further in-depth molecular characterizations of NA antibody responses are warranted to fully leverage broadly protective NA antibodies for influenza vaccine designs. Overall, we posit that focusing on NA for influenza vaccine development is synergistic with existing efforts targeting HA, and may represent a cost-effective approach to generating a broadly protective influenza vaccine.

Keywords: antibody; immunogen design; influenza virus; neuraminidase; vaccine.

Figure 1.. Key events in the history of influenza vaccine development.

The years of influenza pandemics since 1918 and the re-emergence of H1N1 (red) are indicated. The years when different types of influenza vaccines were first licensed (blue) are indicated [12, 13, 17]. The years of major discoveries and technological advancements that have facilitated influenza vaccine development (orange) are indicated [9, 10, 16, 18, 19]. In 2018, NIAID outlined a strategic plan for developing a universal influenza vaccine [20].

Figure 2.. Model for Improving the protection breadth of seasonal influenza vaccines via hemagglutinin (HA) plus neuraminidase (NA) antibody responses.

The schematic shows the change in vaccine effectiveness when HA antigenic drift occurs before NA antigenic drift. Seasonal influenza vaccines that elicit an antibody response to only HA poorly protect against a strain bearing an antigenically drifted HA (top) [46]. In contrast, seasonal influenza vaccines that elicit an antibody response to both HA and NA should offer some protection against a strain with antigenically drifted HA as long as its NA is not antigenically drifted (bottom).

Figure 3.. Most of the NA head domain surface is immunogenic.

(A) Protein surface representation of NA with one protomer colored in white and the other three colored in dark gray (PDB 8DWB [68]). Indicated are the locations of the active site and the highly conserved calcium-binding site in one of the protomers. Calcium ion is represented by a red sphere. Sialic acid in the active site is represented by yellow sticks. (B) Shown are structures of representative NA antibodies that bind to distinct epitopes, namely NC41 (PDB 1NCA [94]), 1G01 (PDB 6Q23 [5]), 3A10 (PDB 8EZ3 [8]), 2H08 (PDB 8E6K [7]), NA-63 (PDB 6PZF [1]), and NA-22 (PDB 6PZW [1]). In addition to NC41, which is a mouse antibody, other represented antibodies shown are from human.

Figure 4.. Feasibility of different NA immunogen design strategies.

Previous studies on broadly protective HA-based immunogen design have involved the consensus sequence approach [–78], glycan masking [80, 81], mosaic nanoparticle display [95], epitope removal [54, 55], and generating domain chimera [73]. Our proposed ranking of their feasibility for NA-based immunogen design is indicated via the decreasing black triangle gradient. This ranking is arbitrarily based on the degree of technical challenges that must be overcome.