A VLP vaccine induces broad-spectrum cross-protective antibody immunity against H5N1 and H1N1 subtypes of influenza A virus

Abstract

The recent threats of influenza epidemics and pandemics have prioritized the development of a universal vaccine that offers protection against a wider variety of influenza infections. Here, we demonstrate a genetically modified virus-like particle (VLP) vaccine, referred to as H5M2eN1-VLP, that increased the antigenic content of NA and induced rapid recall of antibody against HA(2) after viral infection. As a result, H5M2eN1-VLP vaccination elicited a broad humoral immune response against multiple viral proteins and caused significant protection against homologous RG-14 (H5N1) and heterologous A/California/07/2009 H1N1 (CA/07) and A/PR/8/34 H1N1 (PR8) viral lethal challenges. Moreover, the N1-VLP (lacking HA) induced production of a strong NA antibody that also conferred significant cross protection against H5N1 and heterologous CA/07 but not PR8, suggesting the protection against N1-serotyped viruses can be extended from avian-origin to CA/07 strain isolated in humans, but not to evolutionally distant strains of human-derived. By comparative vaccine study of an HA-based VLP (H5N1-VLP) and NA-based VLPs, we found that H5N1-VLP vaccination induced specific and strong protective antibodies against the HA(1) subunit of H5, thus restricting the breadth of cross-protection. In summary, we present a feasible example of direction of VLP vaccine immunity toward NA and HA(2), which resulted in cross protection against both seasonal and pandemic influenza strains, that could form the basis for future design of a better universal vaccine.

Figure 1. H5N1-VLP conferred protection against homologous and heterologous influenza viruses.

Serum samples were collected at 2 weeks after a boosting vaccination. (A) The hemagglutination inhibition (HAI) titers of each H5N1-VLP vaccinated mouse against 4 HA units of H5N1, CA/07, and PR8 viruses, respectively. Data are presented as scatter plot with mean values ± SEM of the same group. HAI titer of 40 is set as threshold of seroprotection. (B) The miniaturized neuraminidase-inhibition (NAI) assays that measure the titers of neutralizing antibody against the NA of different viruses. NAI titer >1.3 fold was set as significant. (C) Reactivity of mice sera (five-fold or two-fold dilution) with the NA of H5N1 and H1N1 influenza viruses. The percentage of NA activity inhibition in each group is represented. Columns, means; bars, SEM. Comparing the vaccination and PBS control groups, asterisk (***), (**), and (*) indicates a significant difference (p<0.001), (p<0.01), and (p<0.05), by one-way ANOVA/Tukey’s range test. (D) Microneutralization assay to examine the neutralizing antibody response in vaccinated mice against H5N1, CA/07, and PR8 viruses. On day 42, vaccinated mice were challenged intranasally with a lethal dose (10× MLD₅₀) of homologous or heterologous viruses as marked. (E) Survival rate and (F) body weight were monitored for 14 days post challenge.

Figure 2. Comparative characterizations of N1-based VLPs with their prototyped H5N1-VLP.

(A) Secreted VLPs of different designs are as indicated to the left of panels. Purified VLPs were adsorbed onto formvar/carbon-coated nickel grids. NS, negatively stained with 2% uranyl acetate; immunogold stained with specific antibodies as marked on the top. The secondary antibodies were goat anti-rabbit or anti-mouse conjugated to 12 nm gold beads. The grids were observed by TEM at 100,000× magnification. (B) Western blot analysis of virus-specific proteins in VLPs. Equal amount of VLPs were separated on a 7.5–12.5% gradient gel followed by Coomassie blue staining or western blot analysis with specific antibodies. Identity of viral proteins in the VLPs were detected and labeled on the right. The same blot was probed with anti-tubulin antibody as a loading control of VLP preparations. Middle, the HA and NA amount in 0.5 µg of each VLP was marked below. (C) Assessment of HA function by hemagglutination assay. The amounts of VLPs used are indicated, in a two-fold serial dilution. TNE (buffer of VLPs) was used as the negative control. The antibodies used in this study were tubulin (ab6160), N1 (ab21305), and M2e (ab5416) from Abcam (Cambridge, MA). Rabbit polyclonal antibody against H5 was provided by Dr. Che Ma (Genomics Research Center, Academia Sinica).

Figure 3. N1-based VLPs induced cross-reactive humoral immune responses.

(A) VLP-induced avN1 antibodies against the NA activity of the homologous H5N1 virus. Sera were collected from mice immunized with 15 µg dosages and types of VLPs as labeled and diluted tenfold for NA inhibition assay. (B) HAI titers and (C) NAI titers induced by N1-based VLPs against homologous and heterologous viruses were determined. The viruses are indicated above. (D) Antisera elicited by different VLPs were fivefold diluted and examined for the inhibition of NA activity derived from PR8 virus. (E) Microneutralization assay to assess the humoral immune responses induced by the VLP vaccines against viruses as labeled. (F) Effect of antisera (tenfold dilution) from N1-VLP vaccinated mice on the NA activity of heterologous strains of influenza virus. The NA antigen derived from N1- or N2-serotyped influenza viruses are labeled at top. CA/07-R is a clinical virus strain with a Tamiflu-resistant phenotype. A/Taiwan/9042/2008(H1N1) (H1N1/08), A/Taiwan/83/2006(H3N2) (H3N2/06), and A/Taiwan/4055/2009(H3N2) (H3N2/09) are clinical virus strains isolated from patients in Taiwan. Comparing the vaccination and PBS control groups, asterisk indicates statistic significance as used in Figure 1.

Figure 4. H5M2eN1-VLP vaccination broadens heterologous protections.

Mice groups (n = 6–10) immunized with various VLP vaccines and infected with a lethal dose (10× MLD₅₀) of reassortant RG-14 (H5N1) (A), CA/07 (B), or PR8 (C) influenza viruses. Body weight changes after infection with H5N1 (D), CA/07 (E), or PR8 (F) were recorded daily and plotted independently. The used dosage and individual VLP vaccines were as indicated.

Figure 5. T helper cell responses in VLP-vaccinated mice.

The level of IL-4 (A) and IFN-γ (B) spot-forming cells/5×10⁵ in spleens of mice was determined by ELISpot assay. Mice groups (n = 5) were immunized as described in the text either with 15 µg of VLPs vaccines or mock-treated (PBS). Splenocytes were stimulated in vitro with BEI-inactivated whole viruses of H5N1 (10 µg/mL), PR8 (4 µg/mL) or alternatively challenged with PR8 virus for four days (post-D4) as labeled on the top of A and B. The cells were stimulated with ConA (2.5 µg/mL, 1×10⁵ splenocytes/well) as a positive control for IFN-γ ELISpot assay or were mock stimulated as negative control (Control) for both ELISpot assays. Bars represent means ± SEM of spot counts in triplicate wells. Comparing the vaccination and PBS control groups, asterisk indicates statistic significance as used in Figure 1. (C) Cytotoxic T lymphocytes (CTL) responses to VLP vaccinations. Data shown were obtained at an effector/target ratio of 100∶1. Each bar represents the percentage of specific lysis (mean ± SEM). The used autologous target cells (4T1) infected with PR8 (PR8-IN) or either transduced with recombinant lentivirus expressing NA of PR8 (PR8/NA) or wild-type lentivirus (vector) were labeled. The splenocyte effectors prepared from the mice vaccinated with VLP vaccines or mock-treated (PBS) were in vitro stimulated with BEI-inactivated PR8 virion (4 µg/mL) for 5 days or in vivo challenged with PR8 virus for 4 days as indicated.

Figure 6. Recall antibody response induced by VLP vaccination to influenza.

(A) Western blot analysis of VLP-induced humoral immunity against PR8 HA. The PR8 HA₀ (11684-V08H, Sino Biological Inc.; expressed by baculovirus system) was treated with TPCK-trypsin (5 µg/mL) at 37°C for 15 min to induce cleavage into HA₁ and HA₂. The cleaved PR8 HA was admixed with purified H3N2-VLP (providing annexin A2 as loading control), serially diluted in two folds and quantified by western blot analysis. The antibody against the annexin A2 (ab41803, Abcam) in H3N2-VLP was used to normalize the loading amount of protein samples and transfer efficiency of western blotting. Mice vaccinated with various VLPs were bled before (Pre-), 4 days (Post-D4) or 14 days (Post-D14) after challenge as indicated. The antisera (500-fold dilution used in this assay) were analyzed by western blotting. HA₀, HA₁, and HA₂ are labeled on the left. (B) Comparative analysis of cross-reactive anti-HA₂ antibody elicited by VLP vaccination before challenge or recalled after homologous and heterologous viral challenges. VLP vaccines and strains of challenge virus are as indicated. (C), (D) The relative folds of antibody response specific to PR8 HA₁ or HA₂ in panel A and B were quantified and summarized correspondingly. Comparing the relative level of HA₂ antibody before and after viral challenge, asterisk indicates statistic significance as used in Figure 1. (E) The HAI titers in vaccinated mice 4 days after PR8 infection were determined by standard methods. Scatter plot with mean values of the same group; bars, SEM. Type of VLP vaccine or PBS control is indicated.