The droplet size of emulsion adjuvants has significant impact on their potency, due to differences in immune cell-recruitment and -activation

Abstract

Self-emulsification is routinely used for oral delivery of lipophilic drugs in vivo, with the emulsion forming in vivo. We modified this technique to prepare novel oil-in-water emulsions of varying droplet size and composition on bench to enable adjuvanted vaccine delivery. We used these formulations to show that smaller droplets (20 nm) were much less effective as adjuvants for an influenza vaccine in mice than the emulsion droplet size of commercial influenza vaccine adjuvants (~160 nm). This was unexpected, given the many claims in the literature of the advantages of smaller particulates. We also undertook cell-recruitment mechanistic studies at site of injection and draining lymph nodes to directly address the question of why the smaller droplets were less effective. We discovered that emulsion droplet size and composition have a considerable impact on the ability to recruit immune cells to the injection site. We believe that further work is warranted to more extensively explore the question of whether, the smaller is not 'better', is a more common observation for particulate adjuvants.

Figure 1

Antibodies against influenza antigen induced by vaccination with different adjuvants. Trivalent inactivated influenza virus antigens: H1N1 A/California/7/09, H3N2 A/Texas/50/2012 and B/Massachusetts/2/2012 were administered twice at 0.1 µg three weeks apart. Animals were divided into 10 animals per group and were administered PBS, non-adjuvanted TIV and TIV adjuvanted with SEA20, MFA90, MFA160, SEA160, MF59, diluted SEA160 and diluted MF59. Sera from 2wp2 were analyzed by ELISA (A) to assess IgG for each antigen individually. One-way ANOVA with post hoc analysis by Dunnett’s multiple comparison using MF59 for comparative purposes showed that adjuvants having 160 nm droplet size, especially SEA160, diluted SEA160 and diluted MF59 were not statistically different than MF59. One-way ANOVA with post hoc analysis Kruskall Wallis’s multiple comparison using PBS for comparative purposes showed that 160 nm adjuvanted groups - SEA160, diluted SEA160, MF59 and diluted MF59 were statistically higher than PBS. (B) Functional hemagglutination inhibition (HI) titers were analyzed for each antigen individually. One-way ANOVA with post hoc analysis by Dunnett’s multiple comparison using MF59 for comparative purposes showed that for H1N1 and B/Massachusetts adjuvants with droplet size of 160 nm, especially SEA160, diluted SEA160 and diluted MF59 were not statistically different than MF59. One-way ANOVA with post hoc analysis Kruskall Wallis’s multiple comparison using PBS for comparative purposes showed that for H1N1 and H3N2 antigens 160 nm adjuvanted groups - SEA160, diluted SEA160, MF59 and diluted MF59 were statistically higher than PBS.

Figure 2

T cell responses assessed for flu antigen adjuvanted with novel emulsions Balb/c mice at 4wp2. Spleens were harvested from the treated animals and single cell suspension was generated. CD4+ positive T cells were re-stimulated in vitro with the protein used for immunization and were analysed by flow cytometry. T-helper subsets (Th0, Th1 or Th2) were identified as described in Material and Methods.

Figure 3

Cell recruitment induced by selected adjuvants at different timepoints post-vaccination. Mice were immunized IM with fluorescent OVA-AF647 together with the adjuvants of interest. At various timepoints animals were euthanized and organs harvested for analysis of immune cell composition at SOI. Once the muscle cells were homogenized into a single cell suspension and stained with the antibody mixture, they were analyzed by multi-color flow cytometry. Number of animals per time-point were: 3 animals for 6 h and 72 h, and 6 animals for 24 h and 48 h.

Figure 4

Antigen positive immune cells at the dLNs of same mice shown in Fig. 3. Once the dLNs were homogenized into a single cell suspension and stained with the antibody mixture, they were analyzed by flow cytometry. Numbers of animals per time-point were: 3 animals for 6 h and 72 h, and 6 animals for 24 h and 48 h. Each animal was  injected in both thighs; so each animal indicates two replicates since dLNs near both SOI were harvested.