Non-neutralizing antibodies induced by seasonal influenza vaccine prevent, not exacerbate A(H1N1)pdm09 disease

Abstract

The association of seasonal trivalent influenza vaccine (TIV) with increased infection by 2009 pandemic H1N1 (A(H1N1)pdm09) virus, initially observed in Canada, has elicited numerous investigations on the possibility of vaccine-associated enhanced disease, but the potential mechanisms remain largely unresolved. Here, we investigated if prior immunization with TIV enhanced disease upon A(H1N1)pdm09 infection in mice. We found that A(H1N1)pdm09 infection in TIV-immunized mice did not enhance the disease, as measured by morbidity and mortality. Instead, TIV-immunized mice cleared A(H1N1)pdm09 virus and recovered at an accelerated rate compared to control mice. Prior TIV immunization was associated with potent inflammatory mediators and virus-specific CD8 T cell activation, but efficient immune regulation, partially mediated by IL-10R-signaling, prevented enhanced disease. Furthermore, in contrast to suggested pathological roles, pre-existing non-neutralizing antibodies (NNAbs) were not associated with enhanced virus replication, but rather with promoted antigen presentation through FcR-bearing cells that led to potent activation of virus-specific CD8 T cells. These findings provide new insights into interactions between pre-existing immunity and pandemic viruses.

Figure 1. TIV-immunization induced non-neutralizing yet cross-reactive, weak-binding Abs that attenuated Cal08 infection.

(A–E) Balb/c mice (5 mice/group) were i.m. immunized with 9 μg TIV/100 μl or PBS. At d30 post-immunization, the sera, BAL and nasal washes were collected for characterization of TIV-induced Abs. (A) Functional neutralizing activity of sera was measured by HI and MN titers against wild-type Bris59 or Cal08 virus. (B,C) Virus-binding ability of sera was measured by ELISA against WIV Bris59 or Cal08 virus or His-tagged rHA Bris59 or Cal07 proteins. (D) Sera Ab binding strength for Bris59 or Cal08 virus were measured by biolayer interferometry. Student t-test was used for comparison of titers between the 2 groups (P < 0.05) (E) Mucosal surface (BAL and nasal washes) Ab responses were measured by ELISA against WIV Bris59 or Cal08 virus. (F,G) TIV or mock (PBS)-immunized mice (5 mice/group) were infected with 5 × 10⁶pfu Cal08 virus (TIV/Cal08, PBS/Cal08 group) or Bris59 virus (TIV/Bris59 group) at d > 30 post-immunization. Control mice were TIV-immunized, mock-infected (TIV/PBS group). All mice were sacrificed at d5 post-Cal08 infection. (F) Lung virus titers at 5 dpi were assessed via plaque assay on MDCK cells. Unpaired t test with Welch’s correction was used to compare TIV/Cal08 vs. PBS/Cal08 groups (P < 0.05), as data points in TIV/PBS and TIV/Bris59 could not be transformed for one-way ANOVA. (G) BW was monitored for 5 days post-infection. Two-way ANOVA with Bonferroni’s multiple comparison test was used for P values. *Comparison between TIV/Cal08 vs. PBS/Cal08 groups (P < 0.001), ! comparison between TIV/Cal08 vs. TIV/PBS groups (P < 0.001), ^†Comparison between TIV/Cal08 vs. TIV/Bris59 groups (P < 0.001). The error bars represent standard error of the mean (SEM). The data are a representative of 5 experiments.

Figure 2. Pro-inflammatory mediators were significantly elevated in TIV/Cal08 mice.

TIV or mock-immunized mice (4–5 mice/group) were infected with 5 × 10⁶pfu Cal08, or Bris59 virus or mock-infected. Lung lysates were collected at d5 post-infection. (A) Lung lysates of TIV/Cal08, PBS/Cal08, TIV/PBS or TIV/Bris59 mice were analyzed for the level of pro-inflammatory mediators (IL-1β, Mip-1β, IL-6, MCP-1) by Bio-Plex^TM. (B) Whole lung tissues from a representative mouse of each group were fixed with 10% formalin for histology (H&E staining). (C) Lung single cell suspensions were analyzed for the frequency of CD11c⁺ and their subsets (SSC^hi, SSC^lo/CD11b⁺, SSC^lo/CD103⁺), and CD11b⁺ and their subsets (inflammatory monocytes: Gr1⁺ CD11b⁺ SSC^int and neutrophils:CD11c⁻CD11b⁺ Gr1^hi) by flow cytometry and their numbers were calculated from lung cell counts. (D) IL-12, TNFα, IFNγ in lung lysates were analyzed by Bio-Plex^TM and the frequency of CD8 T cells was measured by flow cytometry. The error bars represent standard error of the mean (SEM). One-way ANOVA with Bonferroni post-test was used to calculate the P values (*P < 0.05; **P < 0.01; ***P < 0.001). The data are a representative of 3 experiments.

Figure 3. Virus-specific CD8 T cells were significantly induced in TIV/Cal08 mice.

TIV or mock-immunized mice (4–5 mice/group) were infected with 5 × 10⁶pfu Cal08, or Bris59 virus or mock-infected. All mice were sacrificed at d5 post-infection. (A) Sections of formalin-fixed lung tissues at 5 dpi were stained via immunochemistry for NP signals. (B) Lung single cell suspensions at 5 dpi were intracellularly stained for NP via flow cytometry and the %CD11c^hi cells that stained for NP are shown. (C) Virus-specific CD8 T cells were stained with NP- or HA-pentamers for flow cytometry and their cell numbers were calculated from total lung cell counts. (D) CD8 or CD4 T cells secreting IFNγ, granB, or IL-4 ex vivo were analyzed by intracellular staining of cytokines without in vitro stimulation. (E) Virus-specific CD8 or CD4 T cell responses were assessed by in vitro stimulation with Cal08 or Bris59 virus at a multiplicity of infection (MOI) of 1 overnight and subsequent surface and intracellular cytokine staining for flow cytometry. The error bars represent standard error of the mean (SEM). One-way ANOVA with Bonferroni post-test was used to calculate the P values (*P < 0.05; **P < 0.01; ***P < 0.001). The data are a representative of 3 experiments.

Figure 4. IL-10R signaling prevented disease exacerbation in TIV/Cal08 mice.

TIV or mock-immunized mice (4–5 mice/group) were infected with 5 × 10⁶pfu Cal08 or Bris59 virus or mock-infected. All mice were sacrificed at d5 post-infection. (A) IL-10 in lung lysates was measured by Bio-Plex^TM. (B) Lung single cell suspensions were ex vivo stained for CD4⁺ CD25⁺ Foxp3⁺ cells secreting both IL-10 and TGFβ. Cells were also in vitro-stimulated with Cal08 or Bris59 virus at MOI 1 overnight for measuring IL-10-secreting CD4 T cells by flow cytometry. (C,D) TIV-mice (4–5 mice/group) were injected with 0.5 mg/mouse IL-10R blocking Ab (αIL-10R) or isotype control Abs on days −1, +1 and +3 post-infection (arrows). Mock-immunized mice were control. All mice were infected with 5 × 10⁶pfu/mouse Cal08 virus and sacrificed at d5 post-infection. (C) BWs were monitored for 5 days post-infection and lung virus titers at 5 dpi were measured via plaque assay. Two-way ANOVA with Bonferroni’s multiple comparison test was used for P values. **Comparison between TIV/Cal08 + αIL-10R Ab vs. TIV/Cal08 + iso Ab groups (P < 0.01), ! comparison between TIV/Cal08 αIL-10R Ab vs. PBS/Cal08 groups (P < 0.01). (D) Lung cells were stained with NP pentamers for assessment of virus-specific CD8 T cells or a high level of CD16/32 expressing CD11c^hi cells via flow cytometry. The error bars represent standard error of the mean (SEM). All comparisons except (C) were done by One-way ANOVA with Bonferroni post-test. *P < 0.05; **P < 0.01; ***P < 0.001. The data (A,B) are a representative of 3 experiments and data (C,D) are a representative of 2 experiments.

Figure 5. TIV-immunized mice cleared Cal08 virus and recovered faster than controls following infection.

(A–C) TIV or mock-immunized mice (4–5 mice/group were infected with 5 × 10⁶pfu Cal08 virus. Four or five mice were sacrificed at d3, 5, 7 and 10 post-infection. (A) Lung virus titers at days 3, 5, 7 and 10 post-infection were assessed via plaque assay. (B) Lung cells were stained with NP pentamers to assess virus-specific CD8 T cells via flow cytometry at days 3, 5, 7 and 10 post-infection. (C) CD11c^hi cells with intracellular NP signal or expressing high levels of MHC class I (H-2^d) were measured via flow cytometry at days 3, 5, 7 and 10 post-infection. (D,E) TIV or mock-immunized mice were infected with 5 × 10⁶pfu Cal08 or Bris59 virus or mock-infected. (D) BW was monitored until 21 dpi. Two-way ANOVA with Bonferroni’s multiple comparison test was used to compare multiple groups. Comparison between TIV/Cal08 vs. PBS/Cal08 mice is shown. (E) At 21 dpi, mice were sacrificed and splenocytes were stimulated in vitro with Cal08 or Bris59 virus at MOI 1 for measuring IL-10-secreting CD4 T cells, or placed onto ELISPOT plates to measure ASCs specific to Cal08 or Bris59 virus. N/D; not detectable. The error bars represent standard error of the mean (SEM). All comparisons except (D) were done by One-way ANOVA with Bonferroni post-test. *P < 0.05; **P < 0.01; ***P < 0.001. The data are a representative of 2 experiments.

Figure 6. NNAbs were not associated with enhanced disease.

TIV- or mock-immunized Balb/c mice (4–5 mice/group) were infected with 5 × 10³-5 × 10⁶pfu/mouse Cal08 virus in 1 log increments at d > 30 post-immunization. (A) Lung virus titers at 5 dpi were analyzed via plaque assay. (B) Lung lysates were analyzed for the level of the pro-inflammatory mediator, MCP-1 via Bio-Plex^TM. (C) Lung cells were stained with NP pentamers to measure virus-specific CD8 T cells via flow cytometry. (D) Lung CD8 T cells secreting granB ex vivo were measured via flow cytometry. The error bars represent standard error of the mean (SEM). One-way ANOVA with Bonferroni post-test was used to calculate the P values (*P < 0.05; **P < 0.01; ***P < 0.001). The data are a representative of 2 experiments.

Figure 7. NNAbs potentiated recruitment and activation of memory CD8 T cells in TIV/Cal08 mice.

Splenocytes and sera were collected from TIV (9 μg)-immunized Balb/c mice (10 mice/group) at d > 30 post-immunization and pooled. CD8 T cells were isolated via MACS and adoptively transferred to naïve Balb/c mice (1 × 10⁷ cells/mouse) with or without 200 μl sera (4–5 mice/recipient group). One day following adoptive transfer, recipients were infected with 5 × 10⁶pfu/mouse Cal08 virus. Control mice were PBS-transferred and then infected. All mice were sacrificed at d5 post-infection. (A) Lung virus titers were analyzed via plaque assay. (B) Lung cells were stained with NP pentamers to measure virus-specific CD8 T cells or in vitro stimulated with Cal08 or Bris59 virus at MOI 1 overnight to measure granB-secreting CD8 T cells. (C) Pro-inflammatory mediators (Mip-1β, IL-6, MCP-1) in lung lysates were measured via Bio-Plex^TM. (D) Lung cells were intracellularly stained with NP to measure CD11c^hi cells containing NP signals. Lung cells were in vitro stimulated with Cal08 or Bris59 virus at MOI 1 overnight to measure IFNγ-secreting CD4 T cells via flow cytometry. The error bars represent standard error of the mean (SEM). One-way ANOVA with Bonferroni post-test was used to calculate the P values. *P < 0.05; **P < 0.01; ***P < 0.001. The data are a representative of 2 experiments.