Vaccination with adjuvanted recombinant neuraminidase induces broad heterologous, but not heterosubtypic, cross-protection against influenza virus infection in mice

Abstract

In an attempt to assess the cross-protective potential of the influenza virus neuraminidase (NA) as a vaccine antigen, different subtypes of recombinant NA were expressed in a baculovirus system and used to vaccinate mice prior to lethal challenge with homologous, heterologous, or heterosubtypic viruses. Mice immunized with NA of subtype N2 were completely protected from morbidity and mortality in a homologous challenge and displayed significantly reduced viral lung titers. Heterologous challenge with a drifted strain resulted in morbidity but no mortality. Similar results were obtained for challenge experiments with N1 NA. Mice immunized with influenza B virus NA (from B/Yamagata/16/88) displayed no morbidity when sublethally infected with the homologous strain and, importantly, were completely protected from morbidity and mortality when lethally challenged with the prototype Victoria lineage strain or a more recent Victoria lineage isolate. Upon analyzing the NA content in 4 different inactivated-virus vaccine formulations from the 2013-2014 season via Western blot assay and enzyme-linked immunosorbent assay quantification, we found that the amount of NA does indeed vary across vaccine brands. We also measured hemagglutinin (HA) and NA endpoint titers in pre- and postvaccination human serum samples from individuals who received a trivalent inactivated seasonal influenza vaccine from the 2004-2005 season; the induction of NA titers was statistically less pronounced than the induction of HA titers. The demonstrated homologous and heterologous protective capacity of recombinant NA suggests that supplementing vaccine formulations with a standard amount of NA may offer increased protection against influenza virus infection.

Importance: Despite the existence of vaccine prophylaxis and antiviral therapeutics, the influenza virus continues to cause morbidity and mortality in the human population, emphasizing the continued need for research in the field. While the majority of influenza vaccine strategies target the viral hemagglutinin, the immunodominant antigen on the surface of the influenza virion, antibodies against the viral neuraminidase (NA) have been correlated with less severe disease and decreased viral shedding in humans. Nevertheless, the amount of NA is not standardized in current seasonal vaccines, and the exact breadth of NA-based protection is unknown. Greater insight into the cross-protective potential of influenza virus NA as a vaccine antigen may pave the way for the development of influenza vaccines of greater breadth and efficacy.

FIG 1

Vaccination with recombinant N1 protects mice from homologous and heterologous viral challenge. (A to C) Six- to 8-week-old naive BALB/c mice (n = 5 for PR8 N1 group and N2 control group; n = 10 for negative-control group and positive-control groups) were primed and boosted with 10 µg rNA from PR8 (5 µg delivered i.m. and 5 µg delivered i.n.) adjuvanted with poly(I ⋅ C). Negative-control mice were primed and boosted with 10 µg BSA (5 µg delivered i.m. and 5 µg delivered i.n.) adjuvanted with poly(I ⋅ C). Positive-control mice received a 1-µg i.m. prime and boost of a formalin-inactivated, unadjuvanted virus matching the challenge strain. Additionally, one experimental group was primed and boosted with rN2 in a fashion identical to the method used for the N1-vaccinated mice. Upon challenge, weight loss was monitored for 14 days postinfection as a measure of morbidity. Graphs plot the average amounts of weight loss as percentages of initial weight with standard deviation (SD). (D to F) Survival curves from the challenge experiments whose results are shown in panels A to C. (G to I) Pooled sera from individual mice (PR8 N1 vaccinated, rN2 vaccinated, or naive) in each experimental group were tested in triplicate for reactivity to purified virus via ELISA. (J to L) The same sera used in the experiment whose results are shown in panels G to I were tested in triplicate for NI activity against the respective challenge viruses. *, positive-control data shown in panels C and F were collected from the high-challenge-dose group (10 mLD₅₀). n = 5 mice per group unless otherwise stated.

FIG 2

Vaccination with recombinant N2 protects mice from homologous and heterologous viral challenge. The experimental design for these challenge studies was identical to that detailed in the legend to Fig. 1, except that mice (n = 5 per group) were primed and boosted with rNA from HK68/X-31 (H3N2) and challenged with homologous H3N2 reassortant strain HK68/X-31 or the heterologous H3N2 strain Phil82/X-79. Control mice were primed and boosted with rNA from PR8 or BSA. (A to D) Weight loss and survival of mice challenged with HK68/X-31 (A and C) or Phil82/X-79 (B and D). (E to G) Pooled sera from individual mice (HK68/X-31 N2 vaccinated, rN1 vaccinated, or naive) in each experimental group were tested in triplicate both for reactivity to purified virus via ELISA (E and F) and for NI activity against HK68/X-31 (G) and Phil82/X-79 (H).

FIG 3

Passive transfer of sera from vaccinated mice and i.m. versus i.n. vaccination. To demonstrate that humoral immunity against NA is sufficient for protection, passive transfer experiments were performed. Sera from animals vaccinated with HK68/X-31 N2, inactivated whole HK68/X-31 virus, or BSA were transferred into naive mice, which were subsequently challenged with HK68/X-31 virus. (A) Weight loss postchallenge. All mice that received HK68/X-31 N2 or the inactivated whole-virus vaccine survived the challenge. (B) Lung titers of virus in animals vaccinated with HK68/X-31 N2, BSA, or inactivated whole HK68/X-31 virus on day 3 and day 6 postchallenge with HK68/X-31. (C and D) To assess whether the route of vaccine administration had an impact on protection, a challenge experiment identical to the one whose results are shown in Fig. 2A was performed, except that the mice in one group (n = 10) were primed and boosted with 10 µg N2 [adjuvanted with poly(I ⋅ C)] exclusively intramuscularly (i.m.), while those in the other (n = 10) were primed and boosted exclusively intranasally (i.n.). Initially, there was a slight but not very distinguishable difference in weight loss (C); however, upon repeating the experiment with a higher challenge dose (25 LD₅₀), a clear difference in the percentages of weight lost was seen, with the i.n.-vaccinated mice displaying significantly less weight loss than the i.m.-vaccinated mice (D). Survival was 100% in both groups. (E) Reactivities to HK68/X-31 virus were similar for mice that received HK68/X-31 N2 via the i.m. route, the i.n. route, or both at the same time (i.m.+i.n.). n.s., not significant; P > 0.05; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001. n = 5 mice per group unless otherwise stated.

FIG 4

Vaccination with recombinant influenza B virus NA protects mice from homologous and heterologous viral challenge. The experimental design for these challenge studies was identical to those whose results are shown in Fig. 1 and 2, except that mice (n = 5 per group) were primed and boosted with rNA from Yam88 B and challenged with the homologous Yam88 B virus or the heterologous influenza B virus strains Vic87 and Mal04. The mice in the N2 control group were primed and boosted with rNA from HK68/X-31. (A to F) Weight loss and survival after homologous challenge with Yam88 B (A and D) or heterologous challenge with Vic87 (B and E) or Mal04 (C and F). (G to I) Seroreactivities of influenza B virus Yam88 B NA-vaccinated mice to Yam88 B (G), Vic87 (H), or Mal04 (I) virus. (J to L) The same sera used in the experiment whose results are shown in panels G to I were tested in triplicate for NI activity against the respective challenge viruses.

FIG 5

Vaccination with rNA does not induce heterosubtypic immunity in mice. To test the possibility of NA-induced heterosubtypic cross-protection, a sizeable challenge study was performed in which mice were separated into groups (n = 5) and primed and boosted with representative rNAs from subtypes N3 to N9. Similar to the experiment whose results are shown in Fig. 1, animals received identical primes and boosts of 10 µg rNA (5 µg delivered i.m. and 5 µg delivered i.n.) adjuvanted with poly(I ⋅ C). Negative-control mice were primed and boosted with 10 µg BSA (5 µg delivered i.m. and 5 µg delivered i.n.) adjuvanted with poly(I ⋅ C). No reduction in weight loss was observed upon lethal (5 LD₅₀) challenge with PR8 (A) or HK68/X-31 (B). (C and D) Survival curves from the challenge experiments whose results are shown in panels A and B. No appreciable protection from mortality was observed.

FIG 6

Seasonal IIV vaccination is inefficient at inducing NA reactive antibodies in humans. HA and NA reactivities of human pre- and postvaccination sera from 12 individuals who received the 2004-2005 inactivated seasonal vaccine were determined. (A and B) The geometric mean H1 titer was relatively high at baseline (~1,600) and was induced approximately 24-fold upon vaccination (P < 0.0001) (A), while the geometric mean N1 baseline titer was low (~200) and did not increase upon vaccination (B). (C and D) The geometric mean H3 baseline titer (~800) was lower than that of H1 and vaccination induced a 6.4-fold induction (P = 0.0003) (C), while the geometric mean N2 baseline titer was higher than that of N1 and increased 2-fold upon vaccination (P = 0.0230) (D). (E) IIV induced significantly higher endpoint titers against HA than against NA for both influenza A virus subtypes included in the vaccine (P = 0.0003 for H1N1, and P = 0.0240 for H3N2).

FIG 7

The amounts of Cal09 NA contained in seasonal IIVs from the 2013-2014 influenza season varied. (A) Five-fold serial dilutions of 4 IIVs recommended for the 2013-2014 influenza season were analyzed for N1 NA content via Western blot assay. Membranes were blotted with 4A5 (an MAb specific for N1 NA). Each panel represents a separately run Western blot of a unique vaccine brand. Dilutions of baculovirus-expressed Cal09 rN1 (left blot in each panel) of known concentrations were run alongside each vaccine sample on the same gel. Dilutions of vaccines and amounts of standard are displayed above the gels, and the names of the vaccines are displayed below. (B) Quantities of N1 NA per adult vaccine dose (0.5 ml) as measured by ELISA. Bar graphs show the mean values and standard deviation (SD).