Peripheral airways type 2 inflammation, neutrophilia and microbial dysbiosis in severe asthma

Abstract Background IL‐13 is considered an archetypal T2 cytokine central to the clinical disease expression of asthma. The IL‐13 response genes, which are upregulated in central airway bronchial epithelial of asthma patients, can be normalized by high‐dose inhaled steroid therapy in severe asthma. However, this is not the case within the peripheral airways. We have sought to further understand IL‐13 in the peripheral airways in severe asthma through bronchoalveolar analysis. Methods Bronchoalveolar lavage samples were collected from 203 asthmatic and healthy volunteers, including 78 with severe asthma. Inflammatory mediators were measured using a multiple cytokine immunoassay platform. This analysis was replicated in a further 59 volunteers, in whom 16S rRNA analysis of BAL samples was undertaken by terminal restriction fragment length polymorphism. Results Severe asthma patients with high BAL IL‐13, despite treatment with high‐dose inhaled corticosteroids, had more severe lung function and significantly higher BAL neutrophil percentages, but not BAL eosinophils than those with normal BAL‐13 concentrations. This finding was replicated in the second cohort, which further associated BAL IL‐13 and neutrophilia with a greater abundance of potentially pathogenic bacteria in the peripheral airways. Conclusion Our findings demonstrate a steroid unresponsive source of IL‐13 that is associated with BAL neutrophilia and bacterial dysbiosis in severe asthma. Our findings highlight the biological complexity of severe asthma and the importance of a greater understanding of the innate and adaptive immune responses in the peripheral airways in this disease.


| INTRODUC TI ON
Asthma is a highly prevalent airway disease that affects more than 300 million people across the world. 1 In the majority of patients, low-dose inhaled corticosteroid therapies are able to reduce airway inflammation and successfully control symptoms and exacerbations.
However, approximately 10% of patients with asthma continue to struggle, despite treatment with high-dosage inhaled corticosteroids (ICS) plus a second controller and/or systemic corticosteroids 2 : these patients with persistent uncontrolled asthma suffer from high rates of morbidity 3 and account for a disproportionate use of asthma healthcare resources. 4 In this difficult to control asthma population, the phenotypic heterogeneity 5,6 and varied response to currently available treatments [7][8][9] testify to the complexity and variability in the underlying disease-related biology. The pleiotropic cytokine interleukin-13 (IL-13) is increased in bronchoalveolar lavage fluid and overexpressed in sputum and bronchial biopsy specimens of severe asthma patients. 10 Produced by type-2 helper T-cells (Th2), mast cells, basophils and ILC2 cells, IL-13 is considered an archetypal T2 cytokine central to asthma pathophysiology. The IL-13 response genes, CLCA1 (chloride channel, calcium-activated family member (a), POSTN (periostin) and SERPINB2 (serine peptidase inhibitor clade B member (b) have been shown to be upregulated in central airway bronchial epithelial cells in non-steroid treated asthma 11 and normalized in the central airways by high-dose inhaled steroid therapy in those with severe asthma. 12 However, this is not the case within the peripheral airways, 12 which represent the majority of airway luminal surface area, and are increasingly appreciated to contribute to the clinical expression of severe asthma. 13 We have thus sought to gain insight into persistent IL-13 activity in the peripheral airways of severe asthma through bronchoalveolar lavage analysis.

| Study populations
Bronchoalveolar lavage samples were collected from a large cohort of patients and findings validated in a second smaller cohort. All patients provided informed consent to the study, which was conducted in accordance with the Helsinki Declaration. Independent ethics committee approval was obtained (MREC No. 05/Q1702/165). All severe asthma participants were biologic therapy naïve but fulfilled the following severity criteria: maintenance treatment with high-dose inhaled steroids plus at least 2 add on maintenance therapies (including long-acting beta-agonists, leukotriene receptor antagonists or oral steroids) and treatment at GINA/British Thoracic Society asthma management steps 4 or 5. Detailed clinico-physiological characterization was undertaken. There were two comparator groups. Healthy volunteers, all of whom had no history of lower airways disease, had normal bronchial reactivity on histamine inhalation challenge and were on no respiratory medications. Those labelled mild asthma had a diagnosis of asthma, had abnormal airway hyperreactivity and were either not receiving any maintenance asthma therapy or were managed with low-dose inhaled corticosteroids (GINA step 1 and 2).

G R A P H I C A L A B S T R A C T
In this study, we stratified biologic naïve severe asthma patients treated with high-dose inhaled corticosteroid therapy by their bronchoalveolar lavage IL-13. Patients with high bronchoalveolar lavage IL-13, despite steroid therapy, have a higher percentage of bronchoalveolar lavage neutrophils. We replicated these findings in a second smaller cohort, which also associated high bronchoalveolar lavage IL-13 with bacterial dysbiosis.

| Bronchoscopy and BAL processing
Flexible bronchoscopy was performed as described in concordance with established guidelines. 14 Bronchoalveolar lavage (BAL) was undertaken by instilling six sequential 20-mL aliquots of prewarmed (37°C) normal saline into a subsegmental bronchus of the right upper lobe followed by gentle suction 10 seconds after each instillation. BAL fluid was filtered (BD Falcon Cell Strainer) and then centrifuged at 800 g for 10 minutes at 4°C. Cell pellets were resuspended in phosphate-buffered saline for cytospins, and the supernatant was stored at -80°C for later analysis. Cells were stained by using a rapid Romanowsky stain (Raymond Lamb Ltd) to distinguish between macrophages, neutrophils and eosinophils, and 400 cells were counted blind by using coded samples. BAL supernatant concentrations of various cytokines were measured by electrochemiluminescence immunoassay.

| Inflammatory cell cut-offs
A 2% cut-off was used for eosinophilic and non-eosinohilic; 7.5% for neutrophilic and non-neutrophilic. These were used to define four inflammatory phenotypes: eosinophilic, neutrophilic, mixed granular and paucigranular.

| Cytokine analysis
Inflammatory mediators were measured using a V-plex multiple cytokine immunoassay platform (Meso Scale Discovery, MSD) as per the manufacturer's instructions. The assay use SUFO-Tag labelled Detection Antibody for electroemiluminescence.

| Microbial analysis
16S rRNA analysis of BAL samples was undertaken by T-RFLP as described previously. 15 Briefly, nucleic acid was extracted directly from BAL samples and a 927 base fragment of the 16S rRNA genes amplified with a 5′ IRD700-tagged primer. Subsequently, the amplified 16S rRNA genes were digested with CfoI and resolved on a IR2 automated DNA sequencer (LI-COR Biosciences). T-RFLP profiles were analysed using Phoretix one-dimensional advanced software, v.5.10 (Nonlinear Dynamics). T-RF band sizes were determined by comparison with MicroSTEP-15a (700 nm) size marker (Microzone).
T-RF band volume was determined and expressed as a percentage of the total volume of bands detected in each electrophoretic profile.

| Statistical analysis
Due to the non-normal distribution of data, descriptive variables are presented as median (interquartile range) and binary variables as prevalence (%). A Kruskal-Wallis or chi-square test was used for between group comparisons of continuous and categorical data, respectively. Correlations were assessed using Spearman's rank tests.
All tests were two-tailed, and a p-value <0.05 was considered significant. Dunn's method was used to correct for multiple comparisons using statistical hypothesis testing. Statistical analysis was performed with the SPSS 26.0 software and Graphpad Prism 8. Severe asthma patients were stratified into tertiles, based on their BAL IL-13 concentrations, to facilitate an extreme group analysis.

| Clinical characteristics of the bronchoscopy patient cohort
Bronchoalveolar lavage (BAL) samples were obtained at flexible bronchoscopy from 203 volunteers; 45 non-asthmatic healthy volunteers, 80 non-steroid treated asthmatics and 78 asthmatics with severe disease (Table 1). Compared to patients with mild asthma, patients with severe asthma were older, less atopic, more likely to have a smoking history and have poorer lung function despite highdose inhaled corticosteroid therapy (Table 1). Pairwise comparisons between groups, corrected by Dunn's test for multiple comparisons, demonstrate mild asthma (p < 0.001) and severe asthma (p = 0.004) to have higher BAL eosinophils than healthy controls; there was no statistically significant difference between Mild and severe asthma.
In contrast, there was no statistically significant difference in BAL neutrophils between mild asthma and healthy controls, but these were higher in severe asthma compared to mild asthma (p < 0.001) and healthy controls (p < 0.001).

| Extreme group analysis of severe asthma BAL IL-13
In the BAL IL-13 tertiles, severe asthma patients with high BAL IL-13 and those with low BAL IL-13 were similar in terms of demographics and were treated with similar doses of inhaled corticosteroids (Table 2). However, high BAL IL-13 patients were less often atopic and had poorer lung function. They were also distinct in term of inflammatory cell counts: high BAL IL-13 patients showed no difference in terms of BAL eosinophils but did have increased BAL neutrophils ( Figure 2 and Table S2). In patients with severe asthma, IL-13 correlated, by Spearman rank, moderately with BAL neutrophilia, 0.580, p < 0.001 but only very weakly with BAL Eosinophilia, 0.271, p = 0.017. BAL IL-4 and BAL IL-5 were both increased in the high BAL IL-13 group compared with the low IL-13 severe asthma group.

| Replication of bronchoscopy findings
Bronchoalveolar lavage (BAL) samples were obtained in a second smaller study from 59 volunteers: 7 non-asthmatic healthy volunteers, 25 non steroid treated asthmatics and 27 asthmatics with  severe disease. Extreme group analysis of the severe asthma patients, defined by BAL IL-13 tertiles, once again demonstrate worse disease in high BAL IL-13 patients with worse lung function. In this smaller cohort, they also had worse self-report asthma control and required higher doses of inhaled corticosteroid (Table 3). We were able to replicate our finding that there are no differences in BAL eosinophils between these groups but an increase in BAL neutrophils (Table 3 and Figure S1).

| DISCUSS ION
We have extended our bronchial brush findings of persistent IL-13 gene expression in the peripheral airways of patients with severe asthma, 12 by demonstrating elevated levels of BAL IL-13 in patients with severe asthma compared to healthy volunteers and those with mild asthma. Interleukin-13 is commonly associated with T2 disease processes that are linked to eosinophilia and are corticosteroid sensitive 10,16 ; accordingly, it might be anticipated that both BAL eosinophils and BAL IL-13 would be suppressed by the high dose inhaled corticosteroids therapy used in these patients. Our finding demonstrates this to be true for some severe asthma patients, identified as the low BAL IL-13 group. However, we also identify a number of severe asthma patients, who despite, their therapy have elevated BAL IL-13. These patients can be distinguished from healthy controls, mild asthmatics and severe asthmatics with low BAL IL-13 concentrations, by having particularly severe lung function and airway neutrophilia. We have replicated these findings in a second cohort, which further associates BAL IL-13 and BAL neutrophils to the abundance of potentially pathogenic microbes.
Though arguably an oversimplistic paradigm, 17 IL-13 is typically considered to be reciprocally regulated 18 to one of the central mediators for airway neutrophil recruitment, IL17A. 19 Neutrophils are usually the first non-resident immune cells to arrive at a site of inflammation 20 but one of the functions of the T2 cytokine milieu is to cause an IL-4R mediated inhibition of neutrophil activity. [21][22][23] In helminth infections, failure of this regulatory T2 response leads to IL-17 mediated lung tissue damage. 24 Division of the subjects into inflammatory phenotypes (Table S3) identified higher IL-13 BAL levels in both neutrophilic and mixed inflammatory phenotypes. The parallel increase of BAL IL-13 and BAL neutrophilia in the high tertile population could all represent a mixed inflammatory phenotype, as IL-5 concentrations were higher in this sub-group than in the low IL-13 tertile. Arguing against such a consideration was the finding that eosinophil proportions (and numbers - Table S2) were no different from the low IL-13 tertile, which would be the standard way to define inflammatory phenotypes. However, we have previously generally not recognized to generate IL-4, as was seen in our analy- sis. An alternative source of IL-4 and IL-13 are mast cells, a cell population increased within the peripheral airways of severe asthma and whose activation is poorly supressed by steroid therapy. 12,26 Traditionally, mast cells are linked to allergy but are also recog- pathogens or if these represent two separate co-incidental events.
The replication of the findings makes the co-incidental consideration less plausible. Moreover, where therapeutic strategies have broadly been unsuccessful for neutrophilic asthma, 19 it is notable that azithromycin, which leads to exacerbation reduction for eosinophilic and non-eosinophilic patients, 30 was most effective in those patients colonised by H. influenzae. 31 There is a wealth of data supporting the observation that human microbiota influences the maturation and function of the host immune system 32     asthmatics have a history of recurrent severe exacerbations and to limit the impact of this, no bronchoscopies were undertaken within 8 weeks of an exacerbation. Nevertheless, this will be a population that has exposure to oral steroid bursts and antibiotic therapy.
This steroid unresponsive source of IL13 is associated with BAL neutrophilia and bacterial dysbiosis, highlighting the importance of a greater understanding of the innate and adaptive immune responses in the peripheral airways of severe asthma and how they relate to clinical disease expression. Better appreciation and understanding of the different endotypic mechanisms leading to IL13 production is critical in being able to offer targeted therapies.

ACK N OWLED G EM ENTS
The authors wish to thank the volunteers who participated in this study. They also wish to acknowledge the facility and staff support

CO N FLI C T O F I NTE R E S T S
Dr. Azim has nothing to disclose. Dr. Green has nothing to disclose.
Dr. Lau has nothing to disclose. Dr. Rupani has nothing to disclose.