Introduction
Asthma is a chronic inflammatory airway disease with increasing prevalence affecting over 300 million people worldwide. Severe asthma only affects approximately 5 to 10% of asthmatic patients. However, this proportion of asthmatics represents a significant healthcare challenge and contributes to up to half of direct asthma related costs. They respond poorly to standard asthma treatments and are more exposed to oral corticosteroids to achieve acceptable control with significant adverse outcomes associated with deterioration in quality of life [
1].
According to current clinical standards, biologic treatment will be considered in such patients (GINA). Nevertheless, most of currently available biologics target T2 inflammation and are more effective in patients harboring features of such inflammation [
2]. Even though, some patients might not be eligible or might not benefit from biologics therapy. For instance, a study evaluating response to mepolizumab (anti-IL5) of patients with high blood eosinophils reported cessation of treatment in about 14% of patients mainly because of failure to improve asthma control or based on clinician’s decision [
3].
An alternative to biologics therapy lays in bronchial thermoplasty (BT) which has been preconized in severe asthma patients unresponsive or ineligible to biologics, and therefore has been utilized in patients harboring features of T2-high and T2-low inflammation [
2,
4,
5]. This approach consists in a single delivery of radiofrequency thermal energy in airways of 3–10 mm in diameter. Clinical trials showed that BT is safe and effective to reduce asthma exacerbations, emergency department visits and hospital admissions and improves quality of life for up to 10 years in patients with severe asthma treated with BT [
6‐
10]. Observational and meta-analysis studies have further showed that BT induces clinical improvements comparable to biologics [
11‐
13].
BT is the only treatment approach known to modulate airway remodeling, a characteristic asthma-related feature. BT induces airway smooth muscle (ASM) ablation occurring the first weeks after treatment and persisting for over a decade [
14‐
20]. Subepithelial membrane thickening, and airway associated nerve fibers are similarly reduced [
14‐
17,
20]. We and others have further showed that airway epithelium may be a target of BT. For instance, we showed that the structure of airway epithelium was improved following BT treatment and that airway epithelium regeneration in response to BT-induced epithelial injury decreased mucin expression, improved goblet cell metaplasia and regenerated the ciliated cell layer, providing a durable improvement of airway epithelium integrity [
21,
22]. BT also seems to modulate epithelial metabolic profile leading to a shift from a glycolysis-biased gene expression profile to an oxidative phosphorylation-based profile [
23]. BT further seems to modulate the inflammatory profile of the airway epithelium [
17,
21,
24].
Despite these observations, biological mechanisms explaining asthma control improvement following BT are still not well understood. This along with recent success of Tezepelumab targeting the epithelial derived thymic stromal lymphopoietin (TSLP) to improve asthma control regardless of T2 inflammation reemphasizes the cardinal role of the airway epithelium in asthma pathophysiology [
25]. In this study, we assess whether BT affects gene signatures relevant to severe asthma pathophysiology in bronchial epithelial cells (BECs). To do so, we evaluate the transcriptome of BECs obtained pre- and post-BT treatment. We further validate protein expression of differentially expressed genes in bronchial biopsies pre- and post-BT treatment.
Discussion
This study showed that transcriptomic analysis of BECs identified a downregulation of inflammation with the modulation of asthma-related relevant canonical pathways such as IL-13 and S100A alarmins downstream of IL-17 signaling as potential gene family regulated by BT treatment. We found that BT downregulated S100A7, S100A8 and S100A9 gene expressions of cultured BECs. This downregulation was further validated at gene and protein levels in BECs and bronchial biopsies of the same patients. Moreover, this downregulation of S100A family 1-year or more post-BT paralleled the reduction of asthma exacerbations and better asthma control. We showed that S100A family receptors; RAGE, TLR4 and CD36, are upregulated in severe asthmatic BECs and that BT induces a downregulation of TLR4 and CD36 expression. Finally, gene expressions of TSLP and hBD2, are downregulated post-BT in BECs.
The S100A family proteins of interest here are immunomodulatory, antioxidant and calcium/zinc binding proteins mainly expressed by leucocytes of myeloid origin (neutrophils, monocytes) and/or epithelial cells (basal respiratory cells, keratinocytes) [
29,
30]. Higher expressions of S100A8, S100A9 and/or calprotectin (S100A8/A9 complex) were reported in sputum, bronchoalveolar lavage (BAL) and serum of asthmatic patients when compared to healthy controls [
31‐
37]. Decaesteker et al. reported that there was no difference in serum calprotectin levels between asthmatics patients with high sputum eosinophilia and patients with sputum neutrophilia. However, severe asthmatic patients seemed to have higher calprotectin levels than healthy controls. The authors suggest that calprotectin may be a marker of disease severity rather than a marker for specific inflammatory subtypes [
34]. Lee and colleagues found that S100A9 levels were higher in sputum from patients with severe uncontrolled asthma with neutrophilic inflammation than in sputum from eosinophilic and paucigranulocytic groups [
37]. Together, these studies show that the role of S100A family proteins may vary according to the inflammatory context in the asthmatic airways. The role of S100A family is not clear in T2-high setting as it may favor or regulate T2 inflammation depending on the context [
38‐
40]. Nevertheless, S100A molecules appear of cardinal importance in T17-high setting [
33,
35,
37,
41]. Thus, BT-induced downregulation of S100A family proteins could imply modulation of more than one inflammatory pathway further supporting its ability to act on broader pathophysiological mechanisms rather than a specific endotype. Östling et al. reported an enrichment of canonical pathways associated to T17 biology (Role of IL-17A in psoriasis, role of IL-17F in allergic inflammatory airway diseases), TLR signaling, iNOS signaling and inflammasome in bronchial epithelial brushings of a subset of asthmatics. This was characterized by an upregulation of
S100A8 and
S100A9 amongst other genes [
41]. Therefore, there are similarities in the gene signature identified by Ostling et al. in a subset of asthmatics and the gene signature we identified to be modulated by BT in our population. S100A7 is upregulated in response to IL-22, a Th17 cytokine in BECs [
42]. S100A8 and S100A9 can promote mucus hypersecretion typical of asthma in BECs, which is consistent with our previously published works documenting reduced MUC5AC post-BT along with IL-13+ cells in bronchial biopsies collected from a similar cohort [
21,
43]. Further, we showed that S100A7 and S100A8 expressions might be indicators of disease severity as measured by lower ACSS score and that patients with a greater S100A protein decrease had more of these proteins in their airway tissues prior to BT. This goes along with the lines of studies reporting higher S100A9 or calprotectin expression in serum with lower FEV
1/FVC ratio, increased airway hyper-reactivity, or uncontrolled asthma [
32,
34,
35]. Our results are consistent with current literature.
S100A family proteins are ligands of receptor for advanced glycation end products (RAGE) and of toll-like receptor 4 (TLR4). They are believed to promote inflammation through these pathways [
30]. A house dust mite (HDM) sensitized RAGE-KO mouse model showed that airway structural cell expression of RAGE contributes to IL-33 accumulation and further ILC2 accumulation [
44]. In other assays conducted in mice, calprotectin secretion was largely dependent upon RAGE signaling when stimulated with T2 cytokines (IL-4, 5 and 13) [
45]. TLR4 signaling has also been shown to contribute to T2 inflammation. By selectively inactivating TLR4 on airway structural cells, a study group showed that airway epithelium of HDM stimulated mice produced more T2 alarmins (e.g. IL-25, IL-33, TSLP) and contributed to dendritic cells activation if they expressed TLR4 [
46]. Exposition of normal BECs to
Alternaria allergens induced TSLP and IL-25 secretion. Adding calprotectin further enhanced secretions of these cytokines [
38]. A study using an asthma murine model challenged with
A. fumigatus mixed with OVA showed that neutralizing antibodies against S100A8 and S100A9 alleviated airway inflammation and eosinophil recruitment [
39].
On the other hand, some studies have reported beneficial effects for S100A8 and/or S100A9 expressions. For instance, a study using a calprotectin-deficient mouse model sensitized to
Alternaria alternata showed that allergen exposition did increase S100A9 expression in the lung of wild type mice though the deficient mice displayed a worsened T2 inflammation and bronchial hyper-responsiveness [
40]. Administering recombinant S100A8 to ovalbumin sensitized rats reduced pulmonary resistance and bronchial hyper-responsiveness. Further, in vitro administration of recombinant S100A8 on isolated tracheal rings from the same rat model reduced airway smooth muscle contractility [
47].
CD36 is a scavenger receptor interacting with a broad range of ligands such as thrombospondin, collagen, long chain fatty acids and bacterial diacylated lipopeptides [
48]. CD36 can associate with TLR2/6 or TLR4/6 complexes and favor immune response in context of infection or sterile inflammation [
49,
50]. Further, calprotectin-arachidonate complex are recognized by CD36 receptor which could further favor arachidonate uptake and eicosanoids production by epithelial cells [
51]. We observed a reduction of S100A family and of two of its receptors (TLR4 and CD36) in BECs post-BT; suggesting that BT benefits might be related to the reduced signaling pathways involving S100A family, TLR4 and CD36. Though, the role of S100A family proteins in TLR and CD36 signaling in the context of severe asthma remains unclear and needs further investigation.
We investigated the effect of BT on the epithelial expression of the trio of alarmins
IL-25 and
IL-33 and
TSLP which are well documented in asthma as well as on
hBD2, an antimicrobial peptide [
52,
53]. Interestingly, we found a downregulation of
TSLP and of
hBD2 in parallel to
S100A family gene expressions. Though TSLP was classically associated to eosinophilic inflammation, it was more recently shown to be involved in non-eosinophilic inflammation and remodeling [
54]. The results of recent clinical trials support this idea. Indeed, treatment with anti-TSLP reduced the exacerbation rate of patients with and without features of eosinophilic inflammation [
25]. Considering that S100A8, S100A9, TSLP and hBD2 can be elicited by stimuli of various origins (allergens, pollutants, bacteria, cytokines), our results suggest that BT might act on mediators involved in broader pathophysiological mechanisms rather than targeting a specific endotype [
38,
53,
54]. This is supported by other studies reporting better asthma control post-BT in patients with and without T2 inflammation [
4,
5]. This might be a downstream effect of the BT-induced renewal of the epithelium and subsequent improved integrity [
21,
22]. The later might improve the ability of the epithelium to orchestrate a proper response to external stimuli.
Severe asthma is associated with greater ASM mass with greater CXCL8 and eotaxin ASM expression [
55,
56]. ASM can further stimulate the epithelium to produce various cytokines [
57]. Many studies showed that BT induced long lasting ASM ablation [
14‐
20]. It is not clear to which mechanism(s) S100A family downregulation may be attributed. It might be related to improvement in epithelial integrity leading to proper response to external stimuli and reduced production of alarmins and their receptors; ASM ablation reducing the need for contractility modulating proteins or ASM ablation leading to reduced production of chemokines and modulation of inflammation resulting in proper epithelial priming.
Taken together our results support the idea that the S100A family proteins and their related signaling pathways play a relevant role in severe asthma pathophysiology and that a reduced expression might be an indicator of response to BT treatment. They also support that BT, through its effects on airway structural cells, contributes to modulate local innate immune response and inflammation. The nature of BT treatment (heat delivery) and current evidence support that its beneficial effects may not be solely attributable to ASM ablation. For instance, we previously showed that BT induced the renewal of the ciliated cell layer, lastingly increased the number of basal progenitor cells and reduced the production of MUC5AC. This is accompanied by an increase in proliferating epithelial cells in the first two months [
4,
21,
58] post-treatment while this goes back to basal level more than 1 year post treatment [
21]. Further, in vitro experiments exposing epithelial cell cultures to heat as a proxy of BT showed that it may increase proliferation of epithelial cells and modulate their expression of differentiation markers [
58,
59]. While in vitro heat treatment may be useful in assessing the effect of BT on epithelial cells, this only captures acute effect of heat which might not reflect long-lasting disease modifying mechanisms. We therefore believe that the current study assessing effect of BT on S100A family is more likely related to lasting clinical benefits.
Though the exact role of S100A family proteins in asthma pathophysiology is still not clear, higher expression is associated with disease. Further studies are needed to better understand the role of S100A family proteins and their receptors in relation to asthma pathophysiology which might be relevant in treatment monitoring.
Western Blots of selected S100A family receptors (RAGE, TLR4, CD36) on top with associated graphical representations below.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.