Boosting neuraminidase immunity in the presence of hemagglutinin with the next generation of influenza vaccines

Abstract

Neuraminidase (NA), the second most abundant surface glycoprotein on the influenza virus, plays a key role in viral replication and propagation. Despite growing evidence showing that NA-specific antibodies correlate with resistance to disease in humans, current licensed vaccines focus almost entirely on the hemagglutinin (HA) antigen. Here, we demonstrate that recombinant NA (rNA) protein is highly immunogenic in both naïve mice and ferrets, as well as in pre-immune ferrets, irrespective of the level of match with preexisting immunity. Ferrets vaccinated with rNA developed mild influenza disease symptoms upon challenge with human H3N2 influenza virus, and anti-NA antibody responses appeared correlated with reduction in disease severity. The addition of rNA to a quadrivalent HA-based vaccine induced robust NA-specific humoral immunity in ferrets, while retaining the ability to induce HA-specific immunity. These results demonstrate that the addition of rNA is a viable option to increase immunogenicity and potentially efficacy versus currently licensed influenza vaccines by means of boosting NA immunity.

Fig. 1. Structural analysis of rTET-NA constructs.

A Diagram of recombinant tetrabrachion neuraminidase (rTET-NA) compared to the full-length NA. B Molecular weight determined by SEC-MALS and enzymatic activity of four different influenza virus strain rTET-NAs (dots represent independent experiments and solid lines represent the average measurement). C Oseltamivir-binding kinetics analysis for four different influenza virus strain rTET-NAs.

Fig. 2. Immunogenicity of A/Singapore/INFIMH-16-0019/2016 N2 and A/Michigan/45/2015 N1 rTET-NAs, LVNA and IIV preparations in naïve mice.

Female BALB/c mice (n = 8 per group) were immunized twice (intramuscularly) and terminally bled 2 weeks after second immunization (A). Sera were collected two weeks after booster vaccination; sera pools from two animals were created and in turn tested via ELLA to assess NAI antibody activity using reassortant H6N2 A/Singapore/INFIMH-16-0019/2016 (B) or H6N1 A/Michigan/45/2015 (C) viruses as sources of sialidase, respectively. º symbol represents IC₅₀ NAI titers without AF03 addition and ■ represents groups with AF03 addition. The solid lines represent the average measurement, and the dashed lines indicate the starting serum dilution used for testing. *p < 0.05; **p < 0.01; ***p < 0.001 (the red text/symbols are comparisons vs adjuvanted PBS/control). p < 0.001 for all comparisons between adjuvant and non-adjuvanted vaccines at same dose level.

Fig. 3. Immunogenicity of A/Singapore/INFIMH-16-0019/2016 N2 and A/Michigan/45/2015 N1 rTET-NAs in naïve and pre-infected ferrets.

Sera were collected three weeks after initial dose or prime and three weeks after booster vaccination and tested via ELLA to assess NAI antibody activity using reassortant H6N2 A/Singapore/INFIMH-16-0019/2016 (A–C) or H6N1 A/Michigan/45/2015 (D–F) viruses as sources of sialidase, respectively. In A, B and D, E, º symbol represents IC₅₀ NAI titers three weeks after initial prime (D21), ■ represents IC₅₀ NAI titers three weeks after booster (D42), while dashed lines represent starting sera dilution or lower limit of detection for ELLA assay. The solid lines represent the average measurement. In C and F, all symbols represent boost/prime ratio for individual animals in the indicated vaccine groups, while the bar represents average ratio per vaccine group. *p < 0.05; **p < 0.01; ***p < 0.001 (the red text/symbols are comparisons vs PBS/control).

Fig. 4. rTET-NA is highly immunogenic irrespective of the level of pre-immune mismatch, with adjuvant contributing to breadth of NAI responses.

Ferrets were primed intranasally with A/Perth/16/2009 (PE09) or A/Kansas/14/2017 (KS17) H1N2 reassortant viruses, prior to receiving booster vaccination with PE09 rTET-NA or KS17 rTET-NA three weeks later. Sera were collected three weeks after booster vaccination and tested with ELLA to assess NAI antibody activity using A/Perth/16/2009 (A) or A/Kansas/14/2017 (B) H6N2 reassortant viruses as sources of sialidase, respectively. In (A, B), circles represent individual IC₅₀ NAI titers three weeks after booster vaccination and solid lines represent average NAI titers per vaccine group, while dashed lines represent starting serum dilution or lower limit of detection for ELLA assay. *p < 0.05; **p < 0.01; ***p < 0.001 (the red text/symbols are comparisons vs PBS/control).

Fig. 5. rTET-NA protects against challenge with H3N2 influenza virus.

Naïve ferrets were vaccinated twice with A/Perth/16/2009 (PE09) rTET-NA (0.2–45 μg, +/− AF03), challenged intranasally with PE09 H3N2 wild-type influenza virus three weeks after last immunization and monitored for disease symptoms and viral shedding for additional two weeks (A). Pre-challenge PE09 anti-N2 NAI titers were measured in sera collected three weeks after second vaccination (B). Longitudinal analysis of mean body weight loss up to 14 days post-challenge (C). Cumulative body weight change (AUC), peak temperature rise and cumulative viral shedding were also determined for each vaccination group (D). Data shown for individual animals, with whiskers representing the upper and lower ranges, and the upper and lower box edges representing the upper and lower quartiles, respectively, and the median value shown by the lines splitting the box. Only significant differences shown: *p < 0.05; **p < 0.01; ***p < 0.001 (the red text/symbols are comparisons vs PBS/control).

Fig. 6. Post-vaccination NAI titers are a correlate of protection in the ferret model.

A Ferrets with severe symptoms showed significantly lower NAI titers than those classified as non-severe. B NAI titers can correctly predict disease severity in ferrets since the area under the curve (AUC) of the receiver operating characteristics (ROC) curve for NAI titers was significantly higher than an uninformative biomarker. C Log2 NAI titer of 8.9 (NAI titer of 477) was the optimum NAI titer cutoff at D42 that best distinguishes severe from non-severe cases as determined by the Youden index.

Fig. 7. rTET-NA retains its immunogenicity following octavalent HA and NA vaccination in ferrets.

Naïve ferrets were vaccinated twice (21 days apart; (A)) with quadrivalent rTET-NA plus quadrivalent baculovirus-HA (B) from strains included in the quadrivalent 2018–19 Northern hemisphere seasonal influenza vaccine. NAI titers against vaccinal N2, N1, B/Victoria/2/87-like and B/Yamagata/16/88-like strains were determined in sera from vaccinated ferrets three weeks after each immunization (C). *p < 0.05; **p < 0.01; ***p < 0.001 (the red text/symbols are comparisons between 4xHA and 4x NA or 4xHA + 4xNA at matched dose and time). For all comparisons between 1 dose and 2 doses, all groups that received NA-containing formulations had significantly higher NAI titers after 2 doses (p < 0.001), but there were no significant differences for groups that received formulations containing only HA (4xHA). The solid lines represent the average measurement, and the dashed lines indicate the lower limit of detection (LLOD).

Fig. 8. rTET-NA supplementation does not interfere with the magnitude and breadth of HA responses elicited by octavalent recombinant HA and NA vaccine.

HAI titers against vaccinal H1 and H3 strains (A, B) as well as total anti-HA antibody titers against panel of seasonal H3 (C, D) and H1 (E, F) strains were determined in sera from vaccinated ferrets three weeks after the second immunization. The line and whiskers represent the mean and 95% CI. *p < 0.05; **p < 0.01; ***p < 0.001 (the red text/symbols are comparisons vs 4x NA group at matched dose and time).