Gamma Irradiation-Inactivated Respiratory Syncytial Virus Vaccine Provides Protection but Exacerbates Pulmonary Inflammation by Switching from Prefusion to Postfusion F Protein

Abstract

Respiratory syncytial virus (RSV) is a common respiratory pathogen that causes lower respiratory diseases among infants and elderly people. Moreover, formalin-inactivated RSV (FI-RSV) vaccine induces serious enhanced respiratory disease (ERD). Radiation has been investigated as an alternative approach for producing inactivated or live-attenuated vaccines, which enhance the antigenicity and heterogeneous protective effects of vaccines compared with conventional formalin inactivation. In this study, we developed an RSV vaccine using gamma irradiation and analyzed its efficacy against RSV vaccine-induced ERD in a mouse model. Although gamma irradiation-inactivated RSV (RI-RSV) carbonylation was lower than FI-RSV carbonylation and RI-RSV showed a significant antibody production and viral clearance, RI-RSV caused more obvious body weight loss, pulmonary eosinophil infiltration, and pulmonary mucus secretion. Further, the conversion of prefusion F (pre-F) to postfusion F (post-F) was significant for both RI-RSV and FI-RSV, while that of RI-RSV was significantly higher than that of FI-RSV. We found that the conversion from pre- to post-F during radiation was caused by radiation-induced reactive oxygen species. Although we could not propose an effective RSV vaccine manufacturing method, we found that ERD was induced by RSV vaccine by various biochemical effects that affect antigen modification during RSV vaccine manufacturing, rather than simply by the combination of formalin and alum. Therefore, these biochemical actions should be considered in future developments of RSV vaccine. IMPORTANCE Radiation inactivation for viral vaccine production has been known to elicit a better immune response than other inactivation methods due to less surface protein damage. However, we found in this study that radiation-inactivated RSV (RI-RSV) vaccine induced a level of immune response similar to that induced by formalin-inactivated RSV (FI-RSV). Although RI-RSV vaccine showed less carbonylation than FI-RSV, it induced more conformational changes from pre-F to post-F due to the gamma radiation-induced reactive oxygen species response, which may be a key factor in RI-RSV-induced ERD. Therefore, ERD induced by RSV vaccine may be due to pre-F to post-F denaturation by random protein modifications caused by external stress. Our findings provide new ideas for inactivated vaccines for RSV and other viruses and confirm the importance of pre-F in RSV vaccines.

Keywords: enhanced respiratory disease; gamma irradiation inactivation; inactivated vaccine; postfusion F; prefusion F; respiratory syncytial virus.

FIG 1

Gamma irradiation-inactivated RSV vaccine and its immunogenicity. (A and B) Gamma irradiation-inactivated RSV vaccine production. Live RSV (2 × 10⁶ PFU/mL) was irradiated with 0-, 2.5-, 5-, 10-, 20-, and 30-kGy gamma rays, and viral viability was calculated by plaque assay (left panel) and TCID₅₀ assay (right panel) (A). Protein carbonylation of live RSV, RI-RSV, and FI-RSV was measured by DNPH immunoassay (B). (C to E) Immunogenicity of live RSV, FI-RSV, and RI-RSV vaccines. Experimental vaccination scheme. BALB/c mice (n = 5 per group) were immunized twice intramuscularly with 100 μL of live-RSV, RI-RSV or FI-RSV (5 × 10⁵ PFU) absorbed with 100 μL of aluminum hydroxide. Three weeks after the last immunization (day 42), mouse serum, BALF and spleen were collected (C). RSV-specific IgG levels in sera or BALF were assessed using RSV (2 × 10⁵ PFU/mL, 100 μL)-immobilized ELISA plates. Endpoints were 1.58 for BALF and 8.23 for serum (D). RSV neutralization antibody titers were calculated by plaque reduction neutralization assay (E). Data are presented as mean values ± SD of triplicate samples. The asterisks indicate significant differences compared with the FI-RSV group (***, P < 0.001).

FIG 2

Lightly carbonylated RI-RSV induced severe ERD in vivo. (A) Experimental vaccination and challenge scheme. BALB/c mice (n = 5 per group) were immunized intramuscularly twice at a 4-week interval with RI-RSV (5 × 10⁵ PFU) or FI-RSV (5 × 10⁵ PFU) with alum (20 μg) adjuvant or intranasally (i.n.) with live RSV (5 × 10⁵ PFU). At 4 weeks after the second immunization, the mice were challenged i.n. with live RSV A2 strain (2 × 10⁶ PFU). (B and C) Protective response by vaccination. Mouse lungs were homogenized with 2-mm beads at 5 dpc (day 61), and RSV titers were measured from the supernatant using plaque assay (B) and mouse body weight was monitored daily for 5 days from the time of the RSV challenge (day 56) (C). (D and E) Histopathological changes in the lung tissue infected with live RSV at 5 dpc (day 61). Representative histological images of H&E- or PAS-stained lung tissue sections are shown (D). The perivascular aggregates of leukocytes (PVA) and mucus secretion scores were calculated (E). (F) The level of IgE in the supernatant of lung homogenates was measured by ELISA. Data are presented as mean values ± SD of five samples. The asterisks indicate significant differences compared with the PBS-immunized groups (**, P < 0.01; ***, P < 0.001).

FIG 3

Th2-biased immune responses by FI- or RI-RSV immunization. (A to C) Production of RSV-specific IgG1 and IgG2a in the BALF (A) and sera (B) and the calculated IgG1/IgG2a ratios. Prefusion F protein-specific IgG1 and IgG2a levels in sera from immunized mice were assessed using prefusion F protein (300 ng/mL, 100 μL, DS-Cav1)-immobilized ELISA plates. The production of prefusion F protein-specific IgG1 and IgG2a in the sera was detected, and IgG1/IgG2a ratio was calculated (C). Endpoints were 1.58 for BALF and 8.23 for serum. (D to F) Th1 and Th2 cytokine production by splenocytes. BALB/c mice (n = 5 per group) were immunized twice intramuscularly with 100 μL of live RSV, RI-RSV, or FI-RSV (5 × 10⁵ PFU) absorbed with 100 μL of aluminum hydroxide. Three weeks after the last immunization (days 42), mouse splenocytes were collected and stimulated with FI-RSV or RI-RSV (D). Splenocytes from immunized mice were stimulated with RI-RSV (1 × 10⁵ or 3 × 10⁵ PFU/mL equivalent with 1 or 3 μg of protein) or FI-RSV (1 × 10⁵ or 3 × 10⁵ PFU/mL) for 3 days, followed by examination of the levels of IL-5 (E) and IFN-γ (F) production in the culture supernatants. Data are presented as mean values ± SD of five samples. The asterisks indicate significant differences compared with the live-RSV-immunized groups (*, P < 0.05; **, P < 0.01; ***, P < 0.001).

FIG 4

Oxidative stress induces the conversion of prefusion F protein to postfusion F protein in RSV. (A) Live RSV (5 × 10⁷ PFU/mL), RI-RSV (5 × 10⁷ PFU/mL), and FI-RSV (5 × 10⁷ PFU/mL) were prepared. (B) RSV (5 × 10⁷ PFU/mL) was gamma irradiated with 0, 2.5, 5, 10, or 15 kGy. (C) RSV (5 × 10⁷ PFU/mL) was treated with 0, 1, 3, 10, or 30 mM H₂O₂. (D) RSV (5 × 10⁷ PFU/mL) was gamma irradiated with 10 kGy in the presence or absence of NAC (0, 3, or 10 mM). (E) RSV (5 × 10⁷ PFU/mL) was treated with 30 mM H₂O₂ in the presence or absence of NAC (0, 3, or 10 mM). Samples were blotted onto a nitrocellulose membrane and detected by prefusion F protein- or total F protein-IgG, followed by addition of HRP-conjugated secondary antibodies and visualization with chemiluminescence. One of three similar results is shown in the top portion. In the bottom portion, the densities of proteins, as quantified by densitometry analysis using ImageJ software, are shown. Data are presented as mean values ± SD of triplicate samples. The asterisks indicate significant differences compared with live RSV (*, P < 0.05; **, P < 0.01; ***, P < 0.001). #, P < 0.05 compared to gamma irradiation or H₂O₂ treatment group.