Early mucosal events promote distinct mucosal and systemic antibody responses to live attenuated influenza vaccine

Abstract

Compared to intramuscular vaccines, nasally administered vaccines have the advantage of inducing local mucosal immune responses that may block infection and interrupt transmission of respiratory pathogens. Live attenuated influenza vaccine (LAIV) is effective in preventing influenza in children, but a correlate of protection for LAIV remains unclear. Studying young adult volunteers, we observe that LAIV induces distinct, compartmentalized, antibody responses in the mucosa and blood. Seeking immunologic correlates of these distinct antibody responses we find associations with mucosal IL-33 release in the first 8 hours post-inoculation and divergent CD8⁺ and circulating T follicular helper (cTfh) T cell responses 7 days post-inoculation. Mucosal antibodies are induced separately from blood antibodies, are associated with distinct immune responses early post-inoculation, and may provide a correlate of protection for mucosal vaccination. This study was registered as NCT04110366 and reports primary (mucosal antibody) and secondary (blood antibody, and nasal viral load and cytokine) endpoint data.

Fig. 1. Vaccine shedding following LAIV inoculation.

a Study outline, whereby n = 40 young adult volunteers were vaccinated with live attenuated influenza vaccine (LAIV). Mucosal samples were collected by nasosorption for analyses of antibodies and mucosal cytokine and chemokine immune mediators. Nasal transcriptomic responses to LAIV utilised nasal curettage sampling, while blood samples were utilised for peripheral cytometry and antibody analyses. Shedding of vaccine viruses was quantified from nasosorption samples for b B/Yamagata and c B/Victoria. d Heatmap of area under curve (AUC) of vaccine virus shedding in all 40 volunteers. Pfu/ml=plaque forming units per millilitre. Panel a was created with Biorender.com. Source data are provided as a Source Data file.

Fig. 2. Mucosal cytokine and chemokine responses following LAIV inoculation.

Immune mediators in the nasal site of vaccination were quantified from nasosorption samples collected over a time course post-vaccination for a IFNλ/IL-29, b IFNγ, c CXCL10, d IL-6, and e IL-33. f IL-33 levels over the first 8 hours post-LAIV (red, n = 40) and post vehicle control (blue, n = 8). g Correlation matrix of nasal mediator responses to LAIV at the peak time point for each mediator. Correlations between nasal IL-33 responses, measured as area-under-curve (AUC) of 0-8 h post-LAIV, versus later (1-7 days post-LAIV) AUC of vaccine shedding for h B/Victoria, and i B/Yamagata. Panels a–f show median and interquartile range values. In a–d and f, significance was tested between each time point and baseline (0 h) using two-sided Friedman tests with Dunn’s multiple comparison tests. Correlations in g–i used two-sided Spearman R-values and P-value adjustment for multiple testing where the significance of correlations is denoted by asterisks in g. *P < 0.05, **P < 0.01, ***P < 0.001 ****P < 0.0001. Source data are provided as a Source Data file.

Fig. 3. Robust mucosal transcriptomic anti-viral response to vaccination irrespective of vaccine shedding.

RNA-Sequencing was conducted on nasal tissue samples collected at baseline (SD0, n = 34) and at 72 h post-vaccination (SD3, n = 26). a Volcano plot of differentially expressed genes (DEGs) between SD0 and SD3 in those participants where vaccine shedding was evident in at least one time point (“Shedders”, n = 20). b Biological processes enriched in DEGs identified in a. c Volcano plot of DEGs between SD0 and SD3 in those participants where vaccine shedding was not detected at any time point (“Non-Shedders”, n = 5). d Biological processes enriched in DEGs identified in c. Levels of key DEGs in Shedders and Non-Shedders at SD0 and SD3; e CXCL9, f CXCL10, g SOCS1, h CXCR1. Median nasal cytokine kinetics in Shedders and Non-shedders over the study time course for i CXCL10, j CXCL8, and k IL-33 proteins. Panels a–d P-values attained by the Wald test were corrected for multiple testing using the Benjamini and Hochberg method. Panels e–k show median and interquartile range values.

Fig. 4. Antibody responses to LAIV in the blood and nasal mucosa are independent.

Multiplex immunoassays were utilised to measure antigen-specific antibody titres against each LAIV haemagglutinin (HA) at SD0 and SD28 in participants (n = 40); a Serum IgG, b Nasal IgA. c The frequency of conversions (≥4-fold titre rises) in serum IgG and nasal IgA between SD0 and SD28. d Serum HA inhibition (HAI) assay titres at SD0 and SD28 for each vaccine HA. Antibody dependent phagocytosis assays results using opsonised HA coated beads (or HIV gp120 as negative control) for e serum and f nasal fluid at SD0 (light colors) and SD28 (dark colors). g Hierarchically clustered heatmap of Z-scored serum IgG and nasal IgA changes between SD0 and SD28 for each participant for all 4 vaccine HA antigens. Participant clusters are labelled with their predominant humoral response; largely nasal IgA (“Naso-Ab”, n = 16), largely serum IgG (“Sero-Ab”, n = 12), or a mixed/absent humoral response (“Mix-Ab”, n = 12). h Correlation matrix of nasal and serum IgG and IgA changes between SD0 and SD28 in all participants. Significance was tested between groups using two-sided Wilcoxon tests or Mann-Whitney U tests. Correlations used Spearman R-values, where blank spaces denote non-significant correlations after P-value adjustment for multiple testing. Panels a, b represent antibody binding titre data in arbitrary units (AU). Lines and dotted lines in a, b, e, and f show medians and interquartile ranges, respectively. Source data are provided as a Source Data file.

Fig. 5. Peripheral blood lymphocyte activation following live attenuated influenza vaccination.

Phenotypic analysis of peripheral blood mononuclear cells (PBMCs) following live attenuated influenza vaccine (LAIV) inoculation of young adults (n = 23). a CD4⁺ T cell and b CD8⁺ T cell subtype frequencies at SD0, SD7, and SD28 based on expression of CCR7 and CD45RA (T_EM = T effector memory, T_EFF = T effector, T_CM = T central memory). c CD8⁺ subpopulation frequency fold-changes between SD7 and SD0. d Fold-change of CXCR3⁺ CD8⁺ T_EFF frequency at SD7 and SD28, relative to SD0. e Activated circulating T follicular helper (cTfh) (ICOS⁺ CD38⁺ of total CD4⁺ CXCR5⁺ T cells) and antibody secreting cells (ASCs; CD27⁺ CD38⁺ of total CD19⁺ B cells) frequencies at SD0 and SD7. f Fold-change between SD7 and SD0 in the frequency of activated (ICOS⁺ PD-1⁺) total cTfh and cTfh1-like (CXCR3⁺) and cTfh-2/17-like (CXCR3^-) subpopulations (dashed line denotes fold-change of 1, indicating no change between SD7 and SD0). g Frequency over time of alternatively activated (ICOS⁺ CD38⁺) cTfh. h Correlation matrix of key PBMC changes between SD7 and SD0. i Correlation between serum IgG anti-B/Yamagata titre fold-changes between SD28 and SD0 versus cTfh1-like cell activation (ICOS⁺ PD-1⁺) and alternatively activated (ICOS⁺ CD38⁺) total cTfh at SD7. j Correlation between nasal IgA anti-B/Yamagata titre fold-changes between SD28 and SD0 versus naïve and T_EFF CD8⁺ T cell fold-changes at SD7 relative to SD0. Panels a and b represent mean + standard deviation frequencies. Panels c and f represents median and interquartile ranges. Statistical testing in g used two-sided ANOVA with correction for multiple testing. Panel h assessed correlation using two-sided Pearson’s methods for parametrically distributed data where numbers denote Pearson R values and asterisks denote significant correlations. Panels i and j utilized two-sided Spearman’s correlations for testing non-parametrically distributed data. *P < 0.05, **P < 0.01. Source data are provided as a Source Data file.

Fig. 6. Divergent mucosal responses to LAIV between nasal and blood antibody response groups.

Participants (n = 40) were segregated into groups of predominantly nasal (“Naso-Ab”, n = 16), blood (“Sero-Ab”, n = 12) or mixed (“Mix-Ab”, n = 12) antibody responses to LAIV, based on clustering in Fig. 3g. Viral load between Naso-Ab, Mix-Ab and Sero-Ab groups for a B/Yamagata and b B/Victoria. c Differentially expressed genes (DEGs) in nasal tissue samples at SD3 between Naso-Ab and Sero-Ab response groups. d Biological processes enriched in the DEGs from c. Expression of e CXCL10, f DDX58, g OASL, and h Ly9 at SD0 and SD3 in the Mix-Ab, Naso-Ab, and Sero-Ab response groups. Nasal protein levels of i IFNγ, j CXCL10, and k IL-6 at baseline (0 h) and SD3 (72 h) post-vaccination. Panels e–k represent medians and interquartile ranges. Panels e–h represent P-values attained by the Wald test corrected for multiple testing using the Benjamini and Hochberg method. In i–k, significance was tested between baseline (0 h) and 72 h post-inoculation within each group using Wilcoxon tests.

Fig. 7. Schematic summary of mucosal and systemic immune responses to LAIV.

Following inoculation with live attenuated influenza vaccine (LAIV), the earliest observed mucosal immune response was an elevation of IL-33 within 5-8 h post-inoculation. The scale of this IL-33 release was inversely associated with the scale of LAIV replication in the nose, which was observed in ~50% of volunteers for each of the Influenza B virus constituents. By 72 h (SD3) post-inoculation, the transcriptional response to LAIV in the airway is dominated by interferon responses, also observed as protein level increases. Transcription of interferon stimulated genes like OASL and DDX58 is greatest in the “Naso-Ab” group while protein levels of IL-6 are highest in the ‘Sero-Ab’ group at SD3. At 7 days post inoculation (SD7), CD4 and CD8 T cell activation is evident, alongside activation of B cell populations. CD8 T cell activation appeared distinct from T follicular helper cell (cTfh) and B cell activation. The scale of these lymphocyte responses was in turn associated with divergent mucosal IgA and peripheral blood IgG responses to LAIV antigens. Mucosal IgA (akin to the Naso-Ab group) responses were stronger in participants that typically had higher CD8 T cell activation, while peripheral blood IgG responses (akin to the Sero-Ab group) were associated with cTfh and B cell activation. Created with Biorender.com.