The airway inflammation in asthma is typically eosinophilic and accompanied by elevation of Th₂ cytokines. Eosinophils are a key feature of Th₂ inflammation and are a useful biomarker in guiding treatment [1]. However, Th₂ inflammation alone cannot explain all features of asthma. For example airway hyperresponsiveness and tissue remodelling are not entirely linked to this inflammation [2]. There are a number of asthmatic patients in whom anti-inflammatory therapy does not lead to symptom control and who are considered treatment resistant. Furthermore whilst recognized to modify eosinophilic inflammation, inhaled corticosteroids treatment in atopic children with recurrent wheezing has shown to have no effect on declining in lung function and the natural history of asthma [3-5]. This irreversible airflow obstruction has been shown to develop despite appropriate use of inhaled corticosteroids, as advocated by international disease management guidelines [6].

Pathological repair of the airways leads to structural changes that are called airway remodelling. Airway remodelling which has been proposed to result in lower lung function is characterised by subepithelial thickening from collagen deposition, epithelial denudation with goblet cell metaplasia, increased airway smooth muscle mass, angiogenesis and alterations in the extracellular matrix components (ECM) such as collagens, proteoglycans and glycoproteins throughout the airway wall [7] (Fig. 1). Structural remodelling of the airways has been found in children with recurrent wheezing regardless their atopic status [8]. It has also been reported that airway epithelial cells in asthmatic children express makers of injury, such as the epidermal growth factor receptor (EGFR), even in the absence of significant eosinophilic inflammation [9]. In paediatric severe therapy resistant asthma, it was shown the increased subepithelial layer thickness without the evidence of mucosal Th₂ cytokine cell [10]. These studies suggest that remodelling can occur independently of Th₂ inflammation. Furthermore, evidence of airway remodelling, such as epithelial layer damage, thickening of basement membrane, angiogenesis has been demonstrated in children as early as 4 years of age in asthmatic subjects [8, 11, 12]. It is thus an early feature of the disease and not only a marker of long standing chronic disease. However, the subepithelial thickening was not demonstrated in wheezer infants [13]. These indicate that airway thickening begins early in the development of asthma and may play role in the disease progression in some patients.

Human airway development requires a mechanical environment to promote proliferation and airway elongation. The major structural cells of the airways (epithelial cells, fibroblasts, and smooth muscle cells) are responsible for these mechanical environments. During in utero development, mechanical stress results from epithelial fluid secretion to the airways, peristaltic movement of fluid and intermittent foetal breathing. At the time of birth, the mechanical environment alters suddenly as the air-liquid interface is a novel factor contributing to the dynamic balance between muscle contraction, airway lumen patency and wall structure [14]. Airway smooth muscle contraction produces mechanical force from compressive stress on the airway epithelium, fibroblasts and smooth muscle itself. Therefore, abnormal mechanical loading conditions may result in altered cellular activations and modify the composition of ECM leading to fibrosis in the airways [15].

Recent studies have shown that mechanical forces activates epithelial cells causing release of factors that are involved in airway remodelling. Savla and Waters [16] have shown that mechanical forces from cyclical mechanical strain and compressive stress to human and cat airway epithelium cells inhibited epithelial layer repair after wounded by scraping. A model of compressive stress on differentiated normal human bronchial epithelial cells cultured at air-liquid interface has been shown to promote airway remodelling by increasing the gene expression of transforming growth factor (TGF)-β, endothelin-1, and plasminogen activator gene [17, 18], enhancing the release of profibrotic cytokines: TGF-β2 and endothelin [17], increasing intracellular mucin-5AC (MUC5AC) levels [19], increasing expression of EGFR and EGFR ligand [20], enhancing the production of matrix metalloproteinase (MMP)-2 and MMP-9 [18], YKL-40 [21], a chitinase like protein which was recently shown to be associated with airway remodelling in children as well as tissue factor, a coagulation factor that was shown to enhance angiogenesis [22]. Cyclical mechanical strain of airway epithelium cells has also been shown to increase the production of reactive oxygen species (ROS) [23], and to down-regulate prostaglandin E₂ synthesis (PGE₂) [24]. PGE₂ was found to inhibit fibroblast proliferation and collagen production in vitro [25, 26]. Comparative studies have shown that mechanical strain enhances DNA synthesis in rat foetal epithelial cells and fibroblasts when cultured in three-dimensional (3D) organotypic cultures and in this respect has greater influence than monolayer mechanical strain [27]. Mechanical injury to guinea-pig epithelial cells, co-cultured with fibroblasts in the human amnion chamber, results in fibroblast differentiation to myofibroblast and the expression of procollagen I and III [28]. Compressive mechanical stress of human epithelial cell with fibroblasts has been shown to have a greater effect on increasing the MMP-9/tissue inhibitors of metalloproteinase (TIMP)-1 ratio than when epithelial cells or fibroblasts are cultured alone [29]. Similar effects have been seen on collagen production, where mechanical stress applied to 3D epithelial cells co-cultured with fibroblasts causes enhanced collagen expression more than if fibroblasts are cultured alone [30]. Application of 3D dynamic lateral compressive stress to foetal rat lungs cells in organotypic cultures has also been shown to increase the production of fibronectin in the culture supernatants [31]. These studies highlight the importance of epithelial-mesenchymal cross talk in airway remodelling.

Fibroblasts are the major cell that responds to mechanical signals, translating them into biological events especially in expression of ECM genes. As a result, fibroblasts play a pivotal role in tissue remodelling and wound healing [32]. Previous studies have highlighted the role of airway fibroblasts in the production of ECM in response to mechanical stress, including up-regulation of versican and decorin mRNA expression [33, 34], as well as up-regulation of procollagen mRNA expression [35]. Airway fibroblasts from normal and asthmatic subject have been shown to respond to mechanical stimuli differently. Mechanical strain increased versican mRNA expression only asthmatic bronchial fibroblasts but not in normal bronchial fibroblasts [33]. In contrast, Ludwig et al. [34] found that mechanical strain up-regulated versican mRNA expression both in normal and asthmatic bronchial fibroblasts, but decorin was up-regulated only in asthmatic bronchial fibroblasts. However, these investigators reported the mRNA expression using northern blot analysis without showing the house keeping gene, so the difference in gene expression may have been due to the difference in RNA content [34]. Le Bellego et al. [33] have reported that asthmatic bronchial fibroblasts secrete more IL-6 than fibroblasts from normal controls after 24 hours of mechanical strain. A recent study has shown that mechanical strain promoted airway fibroblasts to secret more soluble collagen [36]. Furthermore, mechanical strain has been found to up-regulate IL-8 mRNA expression and enhanced the secretion of IL-8 in culture supernatants in both normal and asthmatic fibroblasts [33, 36]. The impact of mechanical strain on fibroblast proliferation is controversial. Whilst Bishop et al. [37] reported an increase in foetal lung fibroblast cell numbers after mechanical strain, Sanchez-Esteban et al. [38] found that mechanical strain led to both an increase in apoptosis and a decrease in cell proliferation. In a study of the effect of mechanical stress on foetal rat lung fibroblasts, it has been found that a 3D model promoted more DNA synthesis than a monolayer model [27]. Therefore, the mechanical conditions are essential to the cellular responses.

Mechanical strain has been reported to play a critical role in airway smooth muscle (ASM) proliferation and migration [39, 40], increase in stiffness and contractile function [41, 42], induction vascular endothelial growth factor (VEGF) expression and release [43] and ECM deposition [40]. These studies underline the possibility that mechanical stress to ASM participates in the pathogenesis of airway remodelling.

Human airways are exposed to a range of mechanical forces that may potentially arise in several ways, such as during inspiration-expiration, cough and bronchoconstriction from airway smooth muscle contraction during asthma exacerbation. The major structural cells of the airways (epithelial cells. fibroblasts, and smooth muscle cells) are responsible for these physical stimulations. Airway smooth muscle contraction in response to stimuli such as allergen produces a compressive stress on the airway epithelium, fibroblasts and smooth muscle itself. Previous reports have shown that tidal breathing produces a 4% strain of ASM and a deep inspiration causes a 25%-30% strain [44]. Therefore, abnormal mechanical loading to the airways may result in altered cellular activations and modify the composition of ECMs leading to airway structural changes or airway remodelling. Airway wall thickening as demonstrated by computed tomography (CT) [45] and bronchial biopsy [46] has been observed in patients with cough variant asthma and non-asthmatic chronic cough. A recent in vivo study which has shown increases in collagen deposition in the subepithelial layer, mucus secreting goblet cells and cell proliferation in both subepithelial layer and submucosal layer after bronchoconstriction using methacholine challenge, a stimulus that did not affect airway inflammation [47]. Formoterol-budesonide, a treatment targeting both airway inflammation and bronchoconstriction, has been shown to decrease subepithelial layer thickness in asthmatic subjects as assessed by high resolution computed tomography (HRCT) [48] and airway biopsy [49]. However, bronchial hyperresponsiveness has been shown to be inversely related with the airway wall thickness [50, 51]. It was also shown that asthmatic patients who have highly variable airway obstruction showed less airway wall thickening, while those who had less variable or fixed airway obstruction exhibited more thickened airways [52]. Thus the thickening with deposition of the matrix proteins may be a protective mechanism by increasing the stiffness of the airways to attenuate the force from smooth muscle contraction [53].