Fig 2: BAFF and APRIL expression is enhanced by IFN-α in NALT upon primary infection with RSV. Gene expression in NALT 7 days post infection with RSV from RSV-infected adult (AR) and neonatal mice receiving IFN-α (INR) or placebo (NR) 16 h before RSV infection. Controls were inoculated with media. (a) Tnfsf13b (BAFF); (b) Tnfsf13 (APRIL); (c) Tnfrsf13c (BAFF-R); (d) Tnfrsf13b (TACI) and (e) Tnfrsf17 (BCMA). n = 3–5; N = 2 *P < 0.05, **P < 0.01.

Fig 3: Lack of IFN-I signalling in DCs impairs Tfh cell differentiation.(A–B) Quantitation of median fluorescence intensity (MFI) levels of CD80 and CD86 on GFP+ CD11b+ cDC2s in mice treated with anti-IFNAR1 blocking or isotype control antibodies (n = 7 per group/experiment). (C-D) Representative flow cytometric histograms (C) and quantitation (D) of divided CTV+CD45.1+CD4+, which had been transferred into Ifnar1-/- and Ifnar1+/+ mice and immunised with ovalbumin (OVA) in Alum three days earlier (n = 6 per group/experiment). (E-H) Flow cytometric analysis (E, G) and quantitation (F, H) of total (E-F) and 1W1K-I-Ab+ (G-H) CXCR5hiPD-1hiFoxp3-CD4+ T follicular helper (Tfh) cells isolated from Ifnar1fl/fl:Itgaxcre/+ or Ifnar1fl/fl:Itgax+/+ control mice seven days after immunisation with NP-1W1K in Alum (n = 8–9 per group/experiment). Bar graphs show the results of one of two experimental repeats. Bar height corresponds to the median, and each circle represents one biological replicate. P-values were determined using the Mann-Whitney test. Supporting data is shown in Figure 6—figure supplement 1.Figure 6—source data 1.Lack of IFN-I signalling in DCs impairs Tfh cell differentiation.

Fig 4: Poor vaccine responses in older persons correlate with attenuated IFN-I signalling.(A) Time course of the expression of IFN-I-stimulated genes in whole PBMC after influenza vaccination. Median expression of IFN-I-stimulated genes were calculated for each sample at each day and normalised by the day 0 value. The mean and range for five individuals are shown. (B) Heatmap of expression values of a curated list of significant probe sets for IFN-I-stimulated genes (ISGs) induced with a log2-fold change of >0.5 determined using the dataset in (A), and applied to days 0 and 1 in 18–36 year-old (18–36 yo; n = 19) or 65–86 year-old (65–85 yo; n = 39) individuals. (C) Principal component analysis (PCA) of the differentially expressed ISGs from (B) with data from younger and older individuals plotted in blue and pink, respectively. Open circles represent data from day 0, closed circles represent data from day 1 after vaccination, and boxplots represent the distribution of PC1 coordinates for each group. (D-F) The distance from baseline to day 1 after vaccination on the PCA plot in (C) (‘IFN-I distance’) correlates with log2 fold-changes (day 28/day 0) of HAI titres for A/Perth 2009 (D), B/Brisbane 2008 (E) and A/California 2009 (F). In (F-H), Spearman’s correlation coefficients (rho) and their p-values are shown and younger and older individuals are plotted in blue or pink, respectively. Data are reanalysed from publicly available datasets (Franco et al., 2013; Henn et al., 2013; Nakaya et al., 2015).Figure 5—source data 1.Poor vaccine responses in older persons correlate with attenuated IFN-I signalling.

Fig 5: Reduced type I interferon (IFN-I) signalling in cDC2s from aged mice.(A) Principal component analysis (PCA) of the 1000 genes with the largest variance in sorted GFP+CD11b+ cDC2s cells from adult 2-month-old (blue) and aged 23-month-old (pink) mice (n = 6 per group). (B) Functional categories significantly affected by age as determined by Gene Ontology Analysis using Seqmonk. From a publicly available list of gene sets (Merico et al., 2010), significantly different gene ontology terms are shown (Kolmogorov-Smirnov test, p<0.05, average absolute z-score >1, multiple testing correction). Bars are labelled with the number of genes in each set (inside) and the adjusted p-value (outside). (C) Average RPKM (read per kilobase million) expression of IFN-I-responsive genes in GFP+ CD11b+ cDC2s as determined by RNA sequencing. RPKM expression values of the same IFN-I stimulated gene in cDC2s from 2-month-old and 23-month-old mice are connected by lines. (D-E) 2–3 month-old and 22–24 month-old C57BL/6 mice were subcutaneously immunised with Ea-GFP in IFA. 22 hr later, Ifit1 (D) and Mx1 (E) mRNA levels were determined in sorted GFP+ CD11b+ cDC2s by RT-qPCR (n = 4–6 per group/experiment). (F) 2–3 month-old and 22–24 month-old C57BL/6 mice were subcutaneously immunised with Ea-GFP in IFA. 6 hr later, Ifnb1 mRNA levels were determined in cells isolated from the draining lymph nodes (LN) by RT-qPCR (n = 5 per group/experiment). (G-H) Representative flow cytometric plot (G) and quantitation (H) of STAT1 phosphorylation in CD11b+ cDC2s (CD11b+CD11c+MHCII+CD172a+CD8-CD4-B220- cells) from 2 to 3 month-old and 22–24 month-old C57BL/6 mice upon ex vivo treatment of LN cells with 50 U murine IFNa for 30 min, or no cytokine controls (n = 4–7 per group/experiment). (I-J) Ifit1 (I) and Mx1 (J) mRNA levels were determined in sorted GFP+ CD11b+ cDC2s by RT-qPCR upon ex vivo treatment of LN cells with 50 U murine IFNa for 3 hr. Expression levels were normalised to age-matched no cytokine controls (n = 5 per group/experiment). (K-M) Bone marrow (BM) cells from 3-month-old adult or 23-month-old aged mice were transferred into 2-month-old irradiated CD45.1+ C57BL/6 recipient mice. 8 weeks later, these BM chimeras were immunised subcutaneously with Ea-GFP in IFA. 22 hr later, DC populations were analysed by flow cytometry (n = 3–8 per group/experiment). (K) Quantitation of GFP+CD11b+ cDC2 cells in the draining LNs of BM chimeras. (L-M) Quantitation of median fluorescence intensity (MFI) levels of CD86 (L) and CD80 (M) on the surface of GFP+CD11b+ cDC2s in BM chimeras. (N) Ifnb1 expression in different cell populations FACS-sorted from the draining LNs of unimmunised adult mice or mice immunised with EaGFP 16 hr earlier as determined by RT-qPCR. Cell populations were defined as follows: CD3+ T cells, CD19+ B cells, macrophages (CD11c+CD64+F4/80+), pDC (CD3-CD19-B220+CD11c+PDCA1+), cDC1 (CD3-CD19-CD64-F4/80-MHC-II+CD11c+Xcr1+) and a population including cDC2s and Langerhans (Lhs) cells (CD3-CD19-CD64-F4/80-MHC-II+CD11c+Xcr1-). (O) Flow cytometric quantitation of the number of plasmacytoid DCs (pDCs; defined as B220+CD11c+CD11b-PDCA1+ cells) in the draining LNs of 2–3 month-old and 22–24 month-old C57BL/6 mice 22 hr after immunisation with Ea-GFP/IFA (n = 5–6 per group/experiment). (P-S) Bone marrow (BM) cells from 2-month-old CD45.1+ C57BL/6 SJL mice were transferred into 2-month-old or 21-month-old irradiated C57BL/6 recipient mice. 8 weeks later, these BM chimeras were immunised subcutaneously with Ea-GFP in IFA. 22 hr later, DC populations were analysed by flow cytometry (n = 2–5 per group/experiment). (P-Q) Quantitation of CD45.1+ pDCs (P) and GFP+CD11b+ cDC2s (Q) in the draining LNs of BM chimeras. (R-S) Quantitation of median fluorescence intensity (MFI) levels of CD86 (R) and CD80 (S) on the surface of CD45.1+GFP+CD11b+cDC2s in BM chimeras. Bar graphs show the pooled results from at least two experiments except in (I-J), which show the results of one of two experimental repeats. Bar heights correspond to the median, and each circle represents one biological replicate. In (N) bar heights correspond to the mean and error bars represent standard deviation. P-values were determined using the Mann-Whitney test.Figure 4—source data 1.Reduced type I interferon (IFN-I) signalling in cDC2s from aged mice.