Low-inflammatory lipid nanoparticle-based mRNA vaccine elicits protective immunity against H5N1 influenza virus with reduced adverse reactions

Abstract

Messenger RNA vaccines based on lipid nanoparticles (mRNA-LNPs) are promising vaccine modalities. However, mRNA-LNP vaccines frequently cause adverse reactions such as swelling and fever in humans, partly due to the inflammatory nature of LNP. Modification of the ionizable lipids used in LNPs is one approach to avoid these adverse reactions. Here, we report the development of mRNA-LNP vaccines with better protective immunity and reduced adverse reactions using LNPs, which contain a disulfide (SS)-cleavable bond and pH-activated lipid-like materials with oleic acid (ssPalmO) as an ionizable lipid (LNP_ssPalmO). We used mRNA expressing H5N1 subtype high-pathogenicity avian influenza virus-derived hemagglutinin or neuraminidase to generate mRNA-LNP vaccines against H5N1 influenza. Compared with conventional LNPs, mRNA-LNP_ssPalmO induced comparable antigen-specific antibodies and better interferon-γ (IFN-γ)-producing T helper type 1 responses in mice. Both mRNA-LNP_ssPalmO and conventional mRNA-LNPs conferred strong protection against homologous H5N1 virus challenge. In addition, mRNA-LNP_ssPalmO showed better cross-protection against heterologous H5N1 virus challenge compared with conventional mRNA-LNPs. Furthermore, we observed that mRNA-LNP_ssPalmO induced less-inflammatory responses (e.g., inflammatory cytokine production, vascular hyperpermeability) and fewer adverse reactions (e.g., weight loss, fever) compared with conventional mRNA-LNPs. These results suggest that mRNA-LNP_ssPalmO would be a safe alternative to conventional vaccines to overcome mRNA-LNP vaccine hesitancy.

Keywords: adverse reaction; cross-protection; high-pathogenicity avian influenza virus; ionizable lipid; lipid nanoparticle; mRNA vaccine.

Graphical abstract

Figure 1

Overview of mRNA vaccine based on low-inflammatory lipid nanoparticle using ssPalmO (A) Experimental schedule of animal experiments. (B) Chemical structure of ssPalmO and SM-102. (C) Lipid composition and characteristics of mRNA-LNP_ssPalmO or LNP_SM-102. (D) Cryo-EM image of mRNA-LNP.

Figure 2

Antibody responses against HA and NA following subcutaneous immunization with mRNA-LNP (A and B) Mice were subcutaneously immunized with HA-LNPs on day 0 (prime) and day 21 (boost). (A) The levels of HA-specific IgG1, IgG2b, and IgG2c in plasma on day 35 were evaluated with ELISA. These data are related to Figure S1. (B) The number of germinal center B cells in dLNs on day 35 was evaluated with flow cytometry. These data are related to Figure S2. (C and D) Hemagglutination inhibition titer of plasma on day 35 was assessed against H5N1 influenza A virus strain (C) A/Viet Nam/1203/2004 and (D) A/Ezo red fox/Hokkaido/1/2022. (E) Mice were subcutaneously immunized with NA-LNPs on day 0 (prime) and day 21 (boost). The levels of NA-specific IgG1, IgG2b, and IgG2c in plasma on day 35 were evaluated with ELISA. These data are related to Figures S5. (A–E) n = 5 per group. Data are means ± SDs. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001; Tukey’s multiple-comparisons test; ns, not statistically significant. (A and E) Tukey’s multiple-comparisons test was performed at a dilution of (A) 10,000 and (E) 2,000.

Figure 3

T cell responses against HA and NA following subcutaneous immunization with mRNA-LNP Mice were subcutaneously immunized with (A–C) HA-LNP or (D–F) NA-LNP on days 0 (prime) and 21 (boost). On day 35, splenocytes from immunized mice were re-stimulated with (A–C) HA or (D–F) NA. Intracellular cytokine levels in (A and D) IFN-γ⁺ CD44^high CD4⁺ T cells, (B and E) IL-13⁺ CD44^high CD4⁺ T cells, and (C and F) IFN-γ⁺ CD44^high CD8⁺ T cells were evaluated. These data are related to Figures S6. (A–F) n = 5 per group. Data are means ± SDs. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001; Tukey’s multiple-comparisons test; ns, not statistically significant.

Figure 4

Protective effect of mRNA-LNP vaccines against H5N1 influenza A virus challenge Mice were subcutaneously immunized with (A, C, and E) HA-LNPs or (B, D, and F) NA-LNPs on days 0 (prime) and 21 (boost). On day 35, the mice were intranasally challenged with (A and B) H5N1 A/Viet Nam/1203/2004 strain, (C and D) H5N1 A/Ezo red fox/Hokkaido/1/2022 strain, or (E and F) H1N1 A/California/07/2009 strain. Body weight changes and survival rates were monitored for 14 days after virus challenge. Data for the PBS-treated groups are the same. These data are related to Figures S8. (A–F) n = 5 per group. Data are means ± SDs. Body weight: ^#p < 0.05; ^##p < 0.01; ^###p < 0.001; ^####p < 0.0001; Tukey’s multiple comparison test compared to HA-LNP_SM-102 or NA-LNP_SM-102. Survival: ∗p < 0.05; ∗∗p < 0.01; ∗∗∗∗p < 0.0001; log rank test compared with PBS; ^#p < 0.05, log rank test compared with HA-LNP_SM-102 or NA-LNP_SM-102.

Figure 5

Antigen expression following intramuscular immunization with mRNA-LNP (A–F) Mice were intramuscularly injected with luciferase (Luc)-LNP. (A and B) At indicated time points, luminescence was measured using in vivo imaging system. (A) Whole image at 6, 24, 48, and 72 h. (B) Total flux at the injection site of whole images was quantitated. (C–F) At 6 h, (C) muscle, (D) dLN, (E) liver, and (F) spleen were collected from the mice. Luc activity in each tissue homogenate was measured. (G) Mice were intramuscularly injected with Luc-LNPs repeatedly at 3-week intervals. At 6 h after the second dose, Luc activity in the muscle homogenate was measured. (H) Mice were intramuscularly immunized with HA-LNPs on day 0 (prime) and day 21 (boost). At 24 h post-booster immunization, muscles were collected, and the expression of HA in DCs was measured using flow cytometry. These data are related to Figures S11. (A–H) n = 5 per group. Data are means ± SDs. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗∗p < 0.0001; Tukey’s multiple-comparisons test; ns, not statistically significant.

Figure 6

Activation of the antigen-presenting cells following subcutaneous immunization with mRNA-LNP Mice were subcutaneously immunized with HA-LNP. After 24 h, the levels of CD86 in (A) pDCs, (B) cDCs, (C) mDCs, (D) B cells, and (E) macrophages in the dLNs were evaluated by flow cytometry. These data are related to Figures S12. (A–E) n = 5 per group. Data are means ± SDs. ∗p < 0.05; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001; Tukey’s multiple-comparisons test; ns, not statistically significant.

Figure 7

Attenuation of inflammatory cytokine production following subcutaneous immunization with mRNA-LNP_ssPalmO Mice were subcutaneously immunized with HA-LNPs on days 0 (prime) and 21 (boost). At 6 h after primary and booster immunizations, the levels of (A) IFN-α, (B) IFN-β, (C) IFN-γ, (D) CCL2, (E) CXCL10, and (F) IL-6 in the plasma were measured. (A–F) n = 5 per group. Data are means ± SDs. ∗p < 0.05; ∗∗∗∗p < 0.0001; Tukey’s multiple-comparisons test; ns, not statistically significant.

Figure 8

Attenuation of adverse reaction following intramuscular immunization with mRNA-LNP_ssPalmO (A) The leakage of Evans blue at injection site was measured 6 h after intramuscular immunization with HA-LNP. (B) Body weight was measured before and 24 h after intramuscular immunization of HA-LNP. (C and D) Rectal temperature was measured 6 h after (C) primary and (D) booster immunization of HA-LNP. (A–D) n = 5 per group. Data are means ± SDs. ∗∗p < 0.01; ∗∗∗p < 0.001; Tukey’s multiple-comparisons test; ns, not statistically significant.