Adjuvanted nanoliposomes displaying six hemagglutinins and neuraminidases as an influenza virus vaccine

Abstract

Inclusion of defined quantities of the two major surface proteins of influenza virus, hemagglutinin (HA) and neuraminidase (NA), could benefit seasonal influenza vaccines. Recombinant HA and NA multimeric proteins derived from three influenza serotypes, H1N1, H3N2, and type B, are surface displayed on nanoliposomes co-loaded with immunostimulatory adjuvants, generating "hexaplex" particles that are used to immunize mice. Protective immune responses to hexaplex liposomes involve functional antibody elicitation against each included antigen, comparable to vaccination with monovalent antigen particles. When compared to contemporary recombinant or adjuvanted influenza virus vaccines, hexaplex liposomes perform favorably in many areas, including antibody production, T cell activation, protection from lethal virus challenge, and protection following passive sera transfer. Based on these results, hexaplex liposomes warrant further investigation as an adjuvanted recombinant influenza vaccine formulation.

Keywords: adjuvant; antigen display; influenza virus; liposome; subunit vaccine.

Graphical abstract

Figure 1

Design of a hexaplex liposomal vaccine (A) Indicated influenza antigens are surface-displayed on CoPoP liposomes containing PHAD and QS21 (CPQ) via His-tag cobalt interaction to produce surface-decorated hexaplex particles. (B) Recombinant antigens are designed with a trimerizing domain for hemagglutinin and a tetramerizing domain for neuraminidase, producing polymerized antigens that replicate the quaternary conformation and membrane orientation on influenza viruses. Nickel-NTA competition assay was performed using individual antigens as well as multiplex mixture. (C) Each individual antigen shows high binding to liposomes in electrophoresis gel, and a solution of all six proteins shows high simultaneous binding affinity to CoPoP liposomes. Similar antigen sizes for HA and NA antigens result in bands overlapping. (D) Monoclonal antibody reactivity of liposome-displayed antigens in monovalent or hexaplex formulations by slot-blot assay.

Figure 2

Functional antibodies are induced by liposome-displayed antigens Outbred CD-1 mice were vaccinated with 1.8 μg of total antigen displayed on CoPoP liposomes on days 0 and 21, with serum collected at day 42 for analysis. IgG antibody binding titers were determined by ELISA against recombinant His-tagged antigens. (A) Antibody quantities elicited by hexaplex mixture are comparable to single-antigen vaccines with equal total protein. Viral HA and NA inhibition assays were against mouse-adapted A/California/04/2009 H1N1, mouse-adapted A/Hong Kong/1/1968 H3N2, mouse-adapted B/Malaysia/2506/2004, and human B/Phuket/3073/2013. (B) Hemagglutinin inhibition by antibodies assessed by red blood cell agglutination in vitro. Hexaplex vaccine achieves inhibition against all virus serotypes in quadrivalent formulation. (C) Neuraminidase inhibition assessed by colorimetric enzyme-linked lectin assay in vitro. Biological replicates of n = 5 were used.

Figure 3

HA, NA, and hexaplex liposome vaccines are protective against viral challenge in mice Viral challenge against mouse-adapted A/California/04/2009 H1N1 (top), mouse-adapted A/Hong Kong/1/1968 H3N2 (middle), and mouse-adapted B/Malaysia/2506/2004 (bottom) with hexaplex particle vaccines, 1.8 μg total antigen. BALB/c mice were vaccinated on days 0 and 21 and challenged on day 42. Measurements for body weight loss from the day of viral challenge (A), clinical score assessment (B), survival above 25% weight loss threshold (C), and lung viral load in mice sacrificed on day 4 post challenge (D) were recorded. Dotted lines in (D) indicate lower limit of detection. Statistical analysis for (A), (B), and (D) was performed by one-way ANOVA with Tukey’s post hoc multiple comparisons. Statistical analysis for (C) was performed by log-rank test. Asterisks on survival curves indicate significance relative to control curves. Viral load analysis was performed on log-transformed data. Asterisks indicate ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.005, and ∗∗∗∗p < 0.001. Color-matched asterisks compare corresponding colored data point to the hexaplex data at that point in time. Biological replicates of n = 3 for viral load assays and n = 6 for survival challenges were used.

Figure 4

Hexaplex composition and immunogenicity compared with contemporary vaccines Serum was analyzed from mice vaccinated on day 0 and day 21 with 50-μL doses containing 0.3 μg of each antigen in the hexaplex vaccine, Fluad diluted to 0.3 μg of each HA, or Flublok diluted to 0.9 μg of each HA. (A) Multiple admixtures of all antigens with binding, non-binding (2HPQ) liposomes, and alum adjuvant-induced binding antibodies in mice. (B) Monoclonal antibodies were able to detect all six antigens in functional conformation in hexaplex, Fluad vaccine (with notable reduction in N1 and N2 binding), and the three HA antigens contained in the Flublok vaccine. (C) Hexaplex liposomes induced greater immune responsiveness in splenocyte cells than Flublok or Fluad. (D) Hemagglutination inhibition differed between formulations, with hexaplex vaccine favoring higher H3 and B Victoria inhibition, while Fluad and Flublok elicited higher H1 and B Yamagata inhibition. (E) Splenocyte activation in response to antigen stimulation was heightened in hexaplex vaccine, owing to QS21 adjuvant incorporation. Statistical analysis was performed by two-way ANOVA with multiple comparisons. Asterisks indicate significance relative to hexaplex; asterisks over hexaplex bars indicate significant increase over all other groups, while asterisks over comparator bars indicate significant increase over hexaplex. Asterisks indicate ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.005, and ∗∗∗∗p < 0.001. Biological replicates of n = 5 were used.

Figure 5

Comparison of contemporary vaccines against viral challenge in mice Viral challenge against mouse-adapted A/California/04/2009 H1N1 (top) and A/Hong Kong/1/1968 H3N2 (bottom) comparing hexaplex nanoparticle vaccine against available Flublok and Fluad formulations. Mice were vaccinated on days 0 and 21 with 0.3 μg of each antigen in hexaplex (a total of 1.8 μg of antigen), Fluad diluted to 0.3 μg of each HA, or Flublok diluted to 0.9 μg of each HA. Hexaplex achieved marginally higher body weight protection (A), reduced clinical scores (B), increased survival (C), and greater overall survival (D) of animals than Flublok or Fluad. Dotted lines in (D) indicate lower detection limit. Statistical analysis for (A), (B), and (D) was performed by one-way ANOVA with Tukey’s post hoc multiple comparisons. Statistical analysis for (C) was performed by log-rank test. Viral load analysis was performed on log-transformed data. Asterisks indicate ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.005, and ∗∗∗∗p < 0.001. Color-matched asterisks compare corresponding colored data point to the hexaplex data at that time. Survival curve asterisks indicate significance relative to control curve. Biological replicates of n = 3 for viral load assays and n = 6 for survival challenges were used.

Figure 6

Serum transfer protects mice from virus challenge Mice were vaccinated as in previous challenges, after which vaccinated mice were sacrificed, serum was collected, pooled, and 300 μL was introduced into naive mice intraperitoneally 2 h prior to challenge with A/California/04/2009 H1N1. Mice experienced similar weight loss (A) and clinical score time courses (B), but only mice transferred serum from hexaplex or H1 liposome vaccinated animals resulted in significant increases in survival (C). Statistical analysis for (A) and (B) was performed by one-way ANOVA with Tukey’s post hoc multiple comparisons. Statistical analysis for (C) was performed by log-rank test. Asterisks indicate ∗p < 0.05 and ∗∗p < 0.01. Asterisks in (B) show comparison of the control group to the hexaplex group. Survival curve asterisks indicate significance relative to control curve. Biological replicates of n = 9 were used.