Potential Association Between Dietary Fibre and Humoral Response to the Seasonal Influenza Vaccine

Abstract

Influenza vaccination is an effective public health measure to reduce the risk of influenza illness, particularly when the vaccine is well matched to circulating strains. Notwithstanding, the efficacy of influenza vaccination varies greatly among vaccinees due to largely unknown immunological determinants, thereby dampening population-wide protection. Here, we report that dietary fibre may play a significant role in humoral vaccine responses. We found dietary fibre intake and the abundance of fibre-fermenting intestinal bacteria to be positively correlated with humoral influenza vaccine-specific immune responses in human vaccinees, albeit without reaching statistical significance. Importantly, this correlation was largely driven by first-time vaccinees; prior influenza vaccination negatively correlated with vaccine immunogenicity. In support of these observations, dietary fibre consumption significantly enhanced humoral influenza vaccine responses in mice, where the effect was mechanistically linked to short-chain fatty acids, the bacterial fermentation product of dietary fibre. Overall, these findings may bear significant importance for emerging infectious agents, such as COVID-19, and associated de novo vaccinations.

Keywords: fibre; influenza; microbiome; short chain fatty acids (SCFA); vaccine.

Figure 1

Influenza vaccine responses are determined by vaccination history. (A) Schematic overview of sampling timepoints relative to the vaccination event. (B) Histogram summarizes the number of seasonal-influenza vaccines received self-reported by participants at least 4 years prior to the study (2012-2015 inclusive; or from 2011-2015 inclusive). (C) Principal component analysis of both HAI and NAI seroconversion of each participant. Participants are coloured according to their responder category. (D) Heatmap showing relationship between seroconversion and year of most recent seasonal-influenza vaccines received by participants in the 4 years prior to the study (2012-2015), and 2011 or earlier. Hemagglutinin (HAI) seroconversion to each of the influenza subtypes contained in the vaccine is shown. (E) Relationship between seroconversion and vaccine history. Hemagglutinin (HAI) and Neuraminidase (NAI) seroconversions are shown, as is the response to each of the influenza subtypes contained in the vaccine.

Figure 2

A unique microbiome signature is correlated with the vaccine response to each of the influenza strains. (A) Non-metric multidimensional scaling (NMDS) of the microbiome for each participant. NMDS is faceted by vaccination history (naïve vs. previously vaccinated) and by timepoint [pre-vaccine (D0) or 25 - 28 days post vaccine (D28)]. Colour of each point reflects the responder category of the participant. (B) Shannon index of alpha diversity. Plot is faceted by vaccination history (naïve vs. previously vaccinated) and by timepoint (pre-vaccine or 28 days post vaccine). Colour of each point reflects the responder category of the participant. (C) Correlation analysis showing the strength of Spearman correlation of operational taxonomic units (OTUs) with HAI seroconversion in individuals previously naïve to influenza vaccination. Each point is one OTU. Colour of each point represents the class level taxonomic assignment of each OTU. OTUs are organized (x-axis) by genus, or by the lowest level of taxonomic hierarchy determined. Size of each point corresponds to -log (P value). Only OTUs with a correlation coefficient > |0.15| and an adjusted P value < 0.05 are shown. Plot is faceted by influenza strain: H1N1 (top), H3N2 (bottom). (D) Scatter-plot showing the correlation between the OTU with the strongest correlation with H1N1 seroconversion (Faecalibacterium prausnitzii; top) and H3N2 seroconversion (identified to the family level as Lachnospiraceae; bottom).

Figure 3

Fiber intake correlates with influenza vaccine responses. (A) Correlation analysis showing strength of Spearman correlation of operational taxonomic units (OTUs) with average dietary fibre intake. Each point is one OTU. Colour of each point represents the class level taxonomic assignment of each OTU. OTUs are organized (x-axis) by genus, or lowest level of taxonomic hierarchy determined. Size of each point corresponds to -log (P value). Only OTUs with an adjusted P value < 0.05 are shown. (B) Correlation analysis of average dietary fibre intake with Hemagglutinin (HAI) seroconversion. Colour represents the number of self-reported seasonal influenza vaccinations each participant received in the last 5 years. On the left all participants are analysed. On the right only participants naïve to previous influenza vaccination are analysed. Grey shadow represents 95% confidence interval.

Figure 4

Dietary fibre and fibre fermentation are required for optimal antibody response in a mouse model of influenza vaccination. 28 days after receiving the trivalent influenza vaccine (TIV), serum levels of TIV-specific IgG were assessed from mice allocated to various fibre interventions. (A) TIV-specific IgG in the serum from mice fed a control or zero fibre diet. (B) TIV-specific IgG from mice on a control diet, with (abx) or without (control) the addition of an antibiotic cocktail to the drinking water. (C) TIV-specific IgG from mice on a zero-fibre diet, a high fibre diet, or a zero fibre diet supplemented with a cocktail of short chain fatty acids containing butyrate, acetate, and propionate (D) TIV-specific IgG from mice on a control diet supplemented with acetate, butyrate, or propionate. Data from (A, C) are from two independent experiments; samples from replicate groups are combined for analysis [(A) n=5/treatment/experiment (B) n=3-4/treatment/experiment]. Data from (B, D) are representative from 2 independent experiments (n = 5/treatment/experiment). Error bars in all panels show standard error of the mean. *P < 0.05. **P < 0.01. ***P < 0.001. BAP, butyrate, acetate, and propionate.