Reduced Type-I Interferon by Plasmacytoid Dendritic Cells and Asthma in School-Aged Children

Abstract

Background: Allergic sensitization and reduced ability to respond to viral infections may contribute to virus-induced wheeze and asthma development in young children. Plasmacytoid dendritic cells (pDC) are rare immune cells that produce type I interferons (IFN-I) and play a key role in orchestrating immune responses against viruses.

Objective: To further evaluate the function of pDC in children with asthma.

Methods: This study was based on a subset of 71 children from the Early Life Lung Function (ELLF) cohort at the age of 7 years. As part of the ELLF study, participants were characterized for atopic sensitization, viral infection history, and lung function testing. pDC responses to a TLR7/8 agonist were assessed in the presence or absence of anti-IgE using an in vitro assay. Responses were evaluated utilizing flow cytometry, multiplexed cytokine assays, and transcriptional analysis of isolated pDC.

Results: pDC responses varied considerably across individuals, and those who responded with IFN-I following stimulation showed a lower proportion of asthma compared to those who responded with TNF-only. A TNF-only response was associated with increased atopy and reduced upregulation of IFN-associated genes. Anti-IgE stimulation reduced pDC activation, and the reduction was associated with baseline expression of the IgE receptor (FcεR1). A reduction in a gene module centralized around genes such as TPM2, LILRA4, and CLEC4C was also observed.

Conclusion: Together, these findings suggest that pDC responses are variable, associated with asthma, and appear influenced by environmental stimuli. This response thus appears to be an important aspect of asthma pathology in children.

Keywords: IgE; childhood asthma; interferon response; plasmacytoid dendritic cells.

FIGURE 1

Immune cell abundance in children with asthma (A, B). Flow‐SOM clusters projected onto a UMAP of immune cell populations identified using a general immune panel (A) or a pDC/B cell panel (B). (C) Cell abundance of relevant immune cell subsets from samples with ex vivo data available. (D) Abundance of immune cell subsets in children with asthma or with no asthma. (E, F) Immune cell subsets identified to be increased (E) or decreased (F) in children with asthma. The significance of the difference between children with asthma and with no asthma was calculated using the Mann–Whitney test; p‐values are displayed as *p < 0.05.

FIGURE 2

pDC response to stimulation and association with asthma. (A) Clustering of pDC responses based on TNF, IFNα, and IP‐10 expression following R848‐stimulation. (B) Prevalence of asthma in the two main response clusters, IFN^hi and IFN^low. (C) Associations of pDC responses and clinical parameters. (D) Comparison of average wheal size in IFN^hi and IFN^low individuals. Statistical significance of differences was calculated using Fisher's Exact test for (B, C). The Mann–Whitney test was used for (D). p‐values are displayed as a number or as *p < 0.05, **p < 0.01.

FIGURE 3

pDC response clusters and association to early life virus infection. (A, B). Time from birth to first detection of RSV (A) or HRV (B) in individuals with an IFN^hi or IFN^low response. (C, D). Proportion of RSV (C) and HRV (D) positive tests returned from weekly swabs in IFN^hi or IFN^low individuals. (E) Proportion of HRV positive tests in children with or without asthma split across IFN^hi or IFN^low responses. (F) Average C _t‐values during HRV+ swabs as a proxy for viral load in IFN^hi and IFN^low individuals. (G) Correlation between the proportion of positive HRV tests and the proportion of days with a lower respiratory tract infection (LRI) coinciding with HRV infection in children with IFN^hi or IFN^low responses. Statistical significance of differences in (A, B) was calculated using a parametric survival fit adjusted for the season of birth. In (C, F), a Mann–Whitney test was used, and Spearman's correlation was used to evaluate associations in (G). p‐values are displayed as *p < 0.05.

FIGURE 4

pDC responses are reflected in pDC and immune cell phenotype. (A) Ratio and significance of immune cell subset abundance in children with an IFN^hi or IFN^low response. (B–D) Correlation of TNF⁺ [IFN⁺/IP‐10⁺] (B), TNF⁻ [IFN⁺/IP‐10⁺] (C) or TNF⁻ [IFN⁺ or IP‐10⁺] (D) responses in pDC and Monocytes following R848 stimulation. (E, F) Cytokine concentration in culture supernatant in unstimulated (unstim) and R848 stimulated samples from a subset of children. (G) Association with pDC or monocyte activation and cytokine concentration in culture supernatant. (H, I) IFNα production in IFN^hi and IFN^low individuals (H) or in children with or without asthma (I) following stimulation. Significance of difference in cell abundance between individuals with an IFN^hi or IFN^low pDC or with or without asthma was calculated using the Mann–Whitney test and adjusted for multiple comparisons, if relevant, in (A, H, and I). Spearman's correlation was used to evaluate the association between cytokine production in pDC and monocytes in (B–D) or between cytokine levels and pDC or monocyte activation in (F). A Wilcoxon paired rank test, following adjustment for multiple comparisons, if relevant, was used to evaluate cytokine responses to R848 in (E, F). p‐values are displayed as *p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 5

pDC specific WGCNA modules and response to stimulation. (A) Multidimensional scaling (MDS) plot of unstimulated (unstim) and R848 stimulated samples. (B) Gene clustering based on Topographical Overlap Matrix (TOM)‐ based dissimilarity to create WGCNA modules. (C) Comparison of eigengene values in unstim and R848 stimulated samples. (D) Overrepresented Reactome pathways in each module. (E, F) Fold change and baseline levels in module eigengene values in annotated modules following R848 stimulation (E) or at baseline (F) in children with an IFN^hi or IFN^low pDC response. Significance of difference following stimulation in (C) or between children with an IFN^hi or IFN^low pDC response in (E, F) was calculated using Mann–Whitney tests. p‐values are displayed as *p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 6

Impact of IgE crosslinking on pDC activation. (A, B) Cytokine concentration in culture supernatant in R848 stimulated samples with or without prior anti‐IgE crosslinking. (C) Total pDC activation, based on cytokine production, in R848 stimulated samples with or without prior anti‐IgE crosslinking. (D) Change in proportion of specific cytokine producing pDC following anti‐IgE crosslinking and subsequent R848 stimulation, displayed as fold change to R848 stimulation only. (E) Association of ex vivo expression of pDC markers and change in proportion of specific cytokine producing pDC following anti‐IgE crosslinking. (F) Fold change to unstimulated, in annotated WGCNA modules following R848 stimulation with or without prior anti‐IgE crosslinking. (G) Genes within the blue module with high intramodular connectivity and module membership. (H) Difference in gene expression of identified hub genes in R848 stimulated samples with or without prior anti‐IgE crosslinking. (A) Wilcoxon paired rank test was used to evaluate cytokine responses in (A, B), significance of difference in pDC activation in (C), and fold change following anti‐IgE crosslinking in (D). Spearman's correlation was used to evaluate association between marker expression and IgE‐induced change in pDC activation in (E). Differences in module eigengene values in (F) or gene expression levels in (H) were calculated using the Wilcoxon paired rank test. p‐values are displayed as *p < 0.05, **p < 0.01, ***p < 0.001.