Using Dual Toll-like Receptor Agonism to Drive Th1-Biased Response in a Squalene- and α-Tocopherol-Containing Emulsion for a More Effective SARS-CoV-2 Vaccine

Abstract

A diversity of vaccines is necessary to reduce the mortality and morbidity of SARS-CoV-2. Vaccines must be efficacious, easy to manufacture, and stable within the existing cold chain to improve their availability around the world. Recombinant protein subunit vaccines adjuvanted with squalene-based emulsions such as AS03™ and MF59™ have a long and robust history of safe, efficacious use with straightforward production and distribution. Here, subunit vaccines were made with squalene-based emulsions containing novel, synthetic toll-like receptor (TLR) agonists, INI-2002 (TLR4 agonist) and INI-4001 (TLR7/8 agonist), using the recombinant receptor-binding domain (RBD) of SARS-CoV-2 S protein as an antigen. The addition of the TLR4 and TLR7/8 agonists, alone or in combination, maintained the formulation characteristics of squalene-based emulsions, including a sterile filterable droplet size (<220 nm), high homogeneity, and colloidal stability after months of storage at 4, 25, and 40 °C. Furthermore, the addition of the TLR agonists skewed the immune response from Th2 towards Th1 in immunized C57BL/6 mice, resulting in an increased production of IgG2c antibodies and a lower antigen-specific production of IL-5 with a higher production of IFNγ by lymphocytes. As such, incorporating TLR4 and TLR7/8 agonists into emulsions leveraged the desirable formulation and stability characteristics of emulsions and can induce Th1-type humoral and cell-mediated immune responses to combat the continued threat of SARS-CoV-2.

Keywords: AS03; COVID-19; MF59; SARS-CoV-2; TLR4; TLR7/8; adjuvant; subunit vaccine; toll-like receptor (TLR) agonist.

Figure 1

Structures of compounds incorporated in a squalene-based emulsion. INI-4001, a lipidated oxoadenine and TLR7/8 agonist, and/or INI-2002, a synthetic TLR4 agonist, were incorporated into an AddaS03-like emulsion. Lipid moieties present in the compounds allowed incorporation into the oil–water interface, along with the surfactant Tween-80. The oil phase was comprised of squalene and DL-α-tocopherol.

Figure 2

Cryo-EM images show no qualitative difference between AddaS03-like emulsions and those containing INI-4001 or INI-2002 alone or in combination. Cryo-EM was used to determine qualitative differences between (A) AddaS03-like emulsions and AddaS03-like emulsions containing (B) INI-4001 at 1.75 mM, (C) INI-2002 at 1.75 mM, and combinations of INI-4001 and INI-2002 at (D) 1.75 mM and 0.175 mM, respectively, and (E) 1.75 mM and 1.75 mM, respectively. Images were taken at 22,000× magnification with scale bars indicating 100 nm. Images are representative of each emulsion.

Figure 3

Effect of incorporation of INI-4001 and INI-2002 on hydrodynamic diameter, polydispersity (PDI), and zeta potential in AddaS03-like emulsions. INI-4001 or/and INI-2002 were incorporated alone or in combination into AddaS03-like emulsions at “low” (L) 0.175 mM or “high” (H) 1.75 mM target concentrations. Dynamic light scattering was used to measure (A) hydrodynamic diameter (Z.Avg, nm), (B) PDI, and (C) zeta potential (mV) at a pH of 6.7–6.8. Physical properties are given as mean ± SEM with black dots representing each replicate emulsion (n = 4). A one-way ANOVA with a post hoc Tukey’s test for multiple comparisons demonstrated significant differences between group means within compound concentrations (**** p < 0.0001) or when compared or to the AddaS03-like vehicle (# p < 0.05, #### p < 0.0001).

Figure 4

AddaS03-like emulsions maintained high colloidal stability at various temperatures with the addition of TLR4 and/or lipidated TLR7/8 agonists. An AddaS03-like emulsion was used to incorporate INI-4001 (TLR7/8 agonist) and/or INI-2002 (TLR4 agonist). Agonists were incorporated at low (0.175 mM) or high (1.75 mM) target concentrations, singly or coencapsulated. Dynamic light scattering was used to measure (A) hydrodynamic diameter (Z.Avg, nm), (B) polydispersity (PDI), and (C) zeta potential (mV) after storage at 4, 25, or 40 °C. Physical properties are given as mean ± SEM for replicate emulsions (n = 4).

Figure 5

Addition of TLR4 and/or TLR7/8 agonists to AddaS03-like emulsions increases RBD-specific IgG2c antibody titers without reducing total IgG or IgG1 titers. (A) C57BL/6 mice were immunized twice with recombinant SARS-CoV-2 RBD admixed with emulsions containing INI-4001 (TLR7/8 agonist) or INI-2002 (TLR4 agonist), singly or dually encapsulated. Agonists were administered at a low dose (L, 0.876 nmol) or high dose (H, 8.76 nmol). Fourteen days after the booster immunization, anti-RBD serum antibody titers were measured for (B) total IgG, (C) IgG1, (D) IgG2c, and (E) the ratio of IgG2c/IgG1. Results are presented as box-and whisker plots representing the group median with quartiles (box) and range (whiskers) overlaid with data from individual animals. A Kruskal–Wallis test with a post hoc Dunn’s multiple comparisons test was used to determine significant differences in group mean rank when compared to AddaS03-like (* p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001) or MF59-like (# p < 0.05, ## p < 0.01) emulsions alone.

Figure 6

AddaS03-like emulsions, with or without toll-like receptor (TLR) agonists, result in neutralization of SARS-CoV-2 variants. A surrogate virus neutralization test (sVNT) was used to test the neutralization potential of serum taken fourteen days after the booster immunization. The percent reduction in the binding of recombinant ACE2 to the indicated SARS-CoV-2 protein is given for (A) RBD (A/Wuhan), (B) spike (A/Wuhan), (C) spike (D614G mutation), and spike from the variants of concern: (D) Alpha (B.1.1.7), (E) Beta (B.1.351), and (F) Gamma (P.1). Results are presented as box-and whisker plots representing the median and quartiles (box) and range (whiskers) overlaid with data from individual animals. A one-way ANOVA with a post hoc Tukey’s test for multiple comparisons demonstrated significant differences between group means (* p < 0.05 and **** p < 0.0001) when compared to unadjuvanted RBD (n = 6–8).

Figure 7

Addition of TLR4 or TLR7/8 agonists to AddaS03-like emulsions significantly skewed away from Th2-type cytokine profiles in the draining lymph node (DLN). Fourteen days post-secondary immunization, lymphocytes were isolated from the DLN, grown ex vivo, and stimulated with SARS-CoV-2 RBD for 72 h. RBD-specific (A) IFNγ, (B) IL-17A, and (C) IL-5 were measured in the cell supernatant for each group. A one-way ANOVA with a post hoc Tukey’s test for multiple comparisons demonstrated significant differences in group means when compared to the AddaS03-like (** p < 0.01, **** p < 0.0001) or MF59-like (# p < 0.05, ### p < 0.001, and #### p < 0.0001) emulsions alone.