Type I conventional dendritic cells relate to disease severity in virus‐induced asthma exacerbations

Abstract Rationale Rhinoviruses are the major precipitant of asthma exacerbations and individuals with asthma experience more severe/prolonged rhinovirus infections. Concurrent viral infection and allergen exposure synergistically increase exacerbation risk. Although dendritic cells orchestrate immune responses to both virus and allergen, little is known about their role in viral asthma exacerbations. Objectives To characterize dendritic cell populations present in the lower airways, and to assess whether their numbers are altered in asthma compared to healthy subjects prior to infection and during rhinovirus‐16 infection. Methods Moderately‐severe atopic asthmatic patients and healthy controls were experimentally infected with rhinovirus‐16. Bronchoalveolar lavage was collected at baseline, day 3 and day 8 post infection and dendritic cells isolated using fluorescence activated cell sorting. Measurements and Main Results Numbers of type I conventional dendritic cells, which cross prime CD8+ T helper cells and produce innate interferons, were significantly reduced in the lower airways of asthma patients compared to healthy controls at baseline. This reduction was associated serum IgE at baseline and with reduced numbers of CD8+ T helper cells and with increased viral replication, airway eosinophils and reduced lung function during infection. IgE receptor expression on lower airway plasmacytoid dendritic cells was significantly increased in asthma, consistent with a reduced capacity to produce innate interferons. Conclusions Reduced numbers of anti‐viral type I conventional dendritic cells in asthma are associated with adverse outcomes during rhinovirus infection. This, with increased FcεR1α expression on lower airway plasmacytoid DCs could mediate the more permissive respiratory viral infection observed in asthma patients.


| INTRODUC TI ON
Respiratory viruses, especially rhinoviruses (RV), are the major precipitant of asthma exacerbations. 1,2 Concurrent viral infection and allergen exposure synergistically increase risk of asthma exacerbation, [3][4][5] with individuals with asthma experiencing more severe and more prolonged rhinovirus infections. 6,7 This increased disease may be due, at least in part, to impaired innate interferon (IFN) responses identified in airway cells from asthma patients. [8][9][10] T helper type 2 (Th2) cytokines, which along with pro-type 2 mediators IL-25 and IL-33, are induced in the lower airways of asthma patients during experimental RV-16 infection 7,[11][12][13] and have been shown to impair innate IFN responses. 14,15 A recent study used machine learning to cluster asthma characteristics based on immunophenotypes and demonstrated that children with the lowest ex vivo IFN responses to RV not only had a high risk of allergic sensitization but were at the greatest risk of hospitalization due to lower respiratory tract infections. 16 These studies together emphasize the importance of understanding the interplay between type 2 and antiviral responses in viral asthma exacerbations.
The immune response to both virus and allergen is mediated by dendritic cells (DCs) which are airways sentinels which sample the lumen, sensing and interpreting signals to orchestrate downstream immune responses. Human DCs are grouped into conventional (c) DCs, the major antigen-presenting subset and plasmacytoid (p)DCs, which are potent producers of type I IFNs following viral recognition. 17 Human cDCs are further subtyped into type I and type II cDCs. Type II cDCs present antigen to CD4 + T cells and following viral recognition, prime Th1 cells. Human type I cDCs are reported to cross-present viral antigen to CD8 + T cells, 18 polarize Th1 cells 19,20 and produce IFNβ and IFNλ. 21 Most of our understanding of human DCs comes from studies in peripheral blood. Blood type II cDCs from asthma patients have an increased Th2 priming following allergen exposure, and reduced capacity to prime Th1 cells following LPS stimulation. 22 Blood pDCs from asthma patients have impaired type I and type III IFN responses to viral infections, 23 with crosslinking and signalling via the high-affinity IgE receptor FcεRIα implicated as a mechanism. [24][25][26] FcεRIα expression has been assessed on blood DCs in asthma, 23,24 however, differences in receptor expression between asthmatic and healthy subjects have not been characterized on lower airway DC populations.
Whilst a wealth of understanding has been gained from complex mouse models interrogating the role of DCs in airway disease, 27 the limited studies that assessed airway DCs in asthma did not include comparisons with healthy volunteers, instead comparisons were made with lung tissue from non-atopic patients undergoing lung resection due to cancer or other illness, with close proximity to diseased tissue likely to affect the phenotype and function of DCs. 28

What this study adds to the field
We present the novel observation that baseline numbers of type I conventional (c)DCs are reduced in the lower airways of people with asthma compared to healthy controls.
Reduced numbers of type I cDCs were associated with greater baseline total and house dust mite-specific serum IgE, as well as to adverse outcomes during infection including increased viral replication, greater falls in lung function, increased numbers of eosinophils, and reduced CD8 + T cell airway numbers.

| Study participants
Non-smoking volunteers with moderately-severe asthma that was not well-controlled (with an Asthma Control Questionnaire (ACQ) score greater than 0.75 33 ) and non-smoking, non-atopic, healthy volunteers were recruited as described in  Table 1 reports clinical characteristics of both subject groups, full inclusion and exclusion criteria are displayed in the online supplement (See Table S1)

| Study design
Subjects were experimentally infected on day (D)0 with rhinovirus (RV-A16) intra-nasally at 100 tissue culture 50% infective dose (TCID 50 ) as previously described. 7 Subjects underwent bronchoscopies at baseline (approximately day −14) and on D3 and D8 post infection ( Figure 1). Virus load was determined in nasal lavage and induced sputum samples to confirm infection as described in Dhariwal et al. (2021). Spirometry was performed at baseline, D3 and D8, and daily symptom scores were assessed at home (see Table S2 for details of data and sample collection). The clinical out-   (Table S3). Live, single, HLA-DR + lineage − were sorted into type I cDCs (CD11c + BDCA3 + ), type II cDCs (CD11c + BDCA1 + ) and pDCs (CD123 + BDCA2 + BDCA4 + ), see Figure S1 for the gat-   Flow cytometry isolated CD3 + cells were stained with the T cell panel (Table S3) and acquired using a Becton Dickinson FACS Aria II. Frequencies of CD8 + T cells were determined using the following gating strategy: single, live, CD3 + , CD8 + CD4 − cells. BAL eosinophils were determined using differential cell counts on cytospins. 11

| Serum IgE measurement using ImmunoCAP
Total IgE levels were measured in serum using a total IgE ImmunoCAP system on a Phadia 100 reader (Thermo Fisher). Levels of Der p1 and Der p2-specific IgE in human serum were measured using the ImmunoCAP Der p1 and Der p2-specific IgE protocols (Thermo Fisher).

| Statistical analysis
Statistical analysis was performed using Prism 6 (GraphPad Software).
Data are presented as median (interquartile range), Kruskal-Wallis tests were first carried out, and where statistical significance was observed followed by either Wilcoxon's signed rank test to determine statistical differences within groups or Mann-Whitney U tests to determine differences between groups. Correlations were investigated using Spearman's correlation coefficient. Differences were considered statistically significant at p values <.05. All p values are two-sided.

| RE SULTS
Representative  and pDCs at baseline tended to be lower in asthma patients compared to healthy controls but there were no statistically significant differences observed in numbers of type II cDCs and pDCs between subject groups (Figure 2A).

| Enumeration of DC populations in the lower airways during experimental RV-16 infection
There were no statistically significant changes in DC populations within the airways during rhinovirus infection, in either healthy subjects or asthma patients, (Figure 2B-D Figure 2B).

| Reduction of airway type I cDCs is associated with house dust mite-specific IgE
We next sought to determine whether there was evidence to suggest that impaired DC numbers in asthma patients were associated with their atopic status. We identified a trend towards a negative correlation between the numbers of baseline type I cDCs in BAL and total serum IgE at baseline (p = .051, r = −.41) ( Figure 3A and Table S4)

| Reduction of airway type I cDCs is associated with worse disease outcomes during infection
To determine whether the reduced numbers of type I cDCs iden-  Table S4.

| Airway type I cDCs are associated with CD8 + T cell numbers during infection
Since type I cDCs are reported to cross-present viral antigens to CD8 + T cells 19 we also assessed relationships between these two cell types during these RV infections and observed positive correlations between these cell populations in BAL at both D3 (p = .015, r = .502) and D8 post-infection (p = .001, r = .642) ( Figure 5A,B).

| Increased FcεRIα expression on lower airway pDCs from asthma patients
Expression of the high-affinity IgE receptor FcεRIα on peripheral blood pDCs has previously been shown be increased in asthma, and to inversely correlate with their production of IFNα in response to influenza virus infection. 24 Increased FcεRIα expression has also been demonstrated on blood type II cDCs and pDCs from children with asthma compared to healthy children, 23  We did not identify a significant difference in the expression of FcεRIα on type II cDCs between asthma patients and healthy controls at baseline or following infection, though numbers tended to be higher in asthma patients at each time point ( Figure 6C). (p = .068) ( Figure 6D).

| DISCUSS ION
To date, there has been only limited characterization of human airway DC populations. 28,29,34 We have very little information regard- We also present the novel finding that type I cDCs in the lower airways express FcϵRIα. We identified a significant reduction in FcϵRIα expression on type I cDCs during the course of the infection in patients with asthma. Down-regulation of expression has been documented in blood pDCs following ex vivo viral and CpG-ODNstimulation, 24,39 and there was a trend towards reduction in the expression on airway pDCs during the course of the infection in patients with asthma in this study ( Figure 6D), we can speculate this may suggest a shared mechanism of counter regulation between FcϵRIα and IFN production in both DC populations.
It is difficult to rule out the possibility of contamination with in- Another limitation is that all asthma patients in this study were on ICS treatment for their disease, thus further studies would be required to determine the effect of corticosteroid usage on dendritic cell biology within the airways.
This study provides the first data detailing the complete enumeration of DC populations within the lower airways of asthma patients during steady-state disease and further, provides the first documentation of the kinetics of airway DC migration following viral infection in man. Herein, we highlight deficiency in airway type I cDC numbers in asthma, which correlated with atopic status of the individual and was associated with reduced CD8 + cytotoxic T cell recruitment and worse clinical, virologic and pathologic outcomes during infection. Further this study provides a basis for the understanding of DC kinetics in the airways. More work is needed to address the questions raised, however, the expression of FcϵRIα on type I cDCs may provide a link as to why we see greater impairment in the most severely atopic patients.

ACK N OWLED G EM ENTS
The authors thank the St Mary's Flow Cytometry Core Facility, National Heart and Lung Institute, Imperial College London for support and access to equipment.