Ethics All studies we have referred to have been reviewed by ethics committee and have been approved.
Conflicts of interest This article forms part of a supplement sponsored by GlaxoSmithKline. OH has received two research grants from AstraZeneca 2006 and 2008. GMV, KL, LB, KN, AAS and GWT have no competing interests.
Authorship GWT has participated in the design, carried out the review and written and revised the article draft. KL, KN, AAS, OH, GMV and LB have participated in the design, commented on drafts versions of the article, and approved the final version.
Prof. Gunilla Westergren-Thorsson, Department of Experimental Medical Science, Division of Vascular and Airway Research, Lund University, BMC D12, S-22184 Lund, Sweden. Tel: +46 46 222 33 14 Fax: +46 46 222 31 28 email: email@example.com
Introduction: Airway remodelling refers to a wide pattern of patophysiological mechanisms involving smooth muscle cell hyperplasia, increase of activated fibroblasts and myofibroblasts with deposition of extracellular matrix. In asthma, it includes alterations of the epithelial cell layer with goblet cell hyperplasia, thickening of basement membranes, peri-bronchial and peri-broncheolar fibrosis. Moreover, airway remodelling occurs not only in asthma but also in several pulmonary disorders such as chronic obstructive pulmonary disease, idiopathic pulmonary fibrosis and systemic sclerosis. Asthma treatment with inhaled corticosteroids does not fully prevent airway remodelling and thus have restricted influence on the natural course of the disease.
Objectives: This review highlights the role of different fibroblast phenotypes and potential origins of these cells in airway remodelling.
Results: During inflammatory conditions, such as asthma, fibroblasts can differentiate into an active, more contractile phenotype termed myofibroblast, with expression of stress fibres and alpha-smooth muscle actin. The origin of myofibroblasts has lately been debated, and three sources have been identified: recruitment and differentiation of resident tissue fibroblasts; fibrocytes – circulating progenitor cells; and epithelial–mesenchymal transition.
Conclusion: It is clear that airway mesenchymal cells, including fibroblasts/myofibroblasts, are more dynamic in terms of differentiation and origin than has previously been recognised. Considering that these cells are key players in the remodelling process, it is of utmost importance to characterise specific markers for the various fibroblast phenotypes and to explore factors that drive the differentiation to develop future diagnostic and therapeutic tools for asthma patients.
Please cite this paper as: Westergren-Thorsson G, Larsen K, Nihlberg K, Andersson-Sjöland A, Hallgren O, Marko-Varga G and Bjermer L. Pathological airway remodelling in inflammation. Clin Respir J 2010; 4 (Suppl. 1): 1–8.
Asthma has been recognised as an inflammatory airway disorder for more than a century. Around 1890, William Osler (1849–1919) described asthma as an inflammatory disease based on several pathological changes within the asthmatic airways including oedema, gelatinous mucus and Charcot–Leyden crystals (‘asthma crystals’) in the sputum (1). However, it took nearly another 80 years until this knowledge was fully implemented into clinical practice. The biopsy studies by Laitinen et al. contributed substantially to the shift of paradigm, changing focus from relieving bronchial smooth muscle spasm into viewing asthma as mainly central airway inflammatory disorder (2). More recently, it has been proposed that asthma is a disease that also affects lower airways (3). Another important ‘historical’ observation was the recognition of the eosinophil as a central cell in the asthmatic inflammation. In a case report on a patient with previously untreated asthma, it was shown that an acute inflammatory picture could be turned into a microscopically normal mucosa after a few months of inhaled corticosteroid (ICS) treatment (3). This observation created a lot of optimism, and it was believed that some asthmatics, provided corticosteroid treatment was initiated early enough and in sufficient doses, could be cured from their disease (4). Unfortunately, today, we know that in most cases, this is not true. The remission of treated or untreated adult asthma is low, less than 1% (5). Moreover, it has, in several studies, been shown that ICS treatment per se does not control all aspects of the asthmatic inflammation, and it has been clearly shown that many patients with asthma develop structural changes in the airways (airway remodelling) despite being treated with corticosteroid therapy for longer periods of time (6).
To get one step further with the aim of treating and possibly curing the disease, it is of utmost importance to understand the underlying mechanisms behind chronic inflammation and development of structural changes, i.e. remodelling of the airways. Airway remodelling refers to a wide pattern of pathophysiological mechanisms involving smooth muscle cell hyperplasia (7, 8), increase of activated fibroblasts and myofibroblasts with deposition of extracellular matrix (ECM) components such as collagens, fibronectins and proteoglycans in and around the epithelium and vessels (3, 9–11). Moreover, airway remodelling occurs not only in asthma but also in several pulmonary disorders such as chronic obstructive pulmonary disease, idiopathic pulmonary fibrosis (IPF), and systemic sclerosis (SSc). This review will focus on the role of different fibroblast phenotypes and potential origins of these cells in airway remodelling.
A key player in fibrosis – the activated fibroblast
The fibroblast is considered to be an important player in airway remodelling because of its ability to produce new ECM and regulate the normal turnover of the matrix by secreting components that stimulate the breakdown of ECM (9–11). The fibroblast is an elongated, spindle-like shaped cell with an irregularly shaped nucleus, abundant endoplasmic reticulum and intracellular filaments that facilitate the intracellular trafficking in these cells, which include the synthesis and secretion of ECM components (12). Under non-pathological conditions, fibroblasts are sparsely distributed in the connective tissue. Since fibroblasts attach to the surrounding structural network of ECM components, these cells may be controlled by molecules bound to the matrix such as cytokines and growth factors (13). Other studies have shown that components of the ECM itself may affect the activity of fibroblasts (14). Collectively, these sources of potential stimuli may thus control the fate of the fibroblasts in both normal and pathological conditions.
It is generally accepted that airway remodelling in asthma and other disorders can be described as an aberrant wound healing mechanism (Fig. 1). During the initial steps of this process, fibroblasts migrate to the injured sites where they produce an inflammatory matrix. Important chemotactic factors include alternatively spliced fibronectin (ED-A fibronectin), platelet-derived growth factor (PDGF) (15) and transforming growth factor-β (TGF-β) (16), which are produced and secreted by epithelial, inflammatory and surrounding mesenchymal cells (17).
In the inflammatory matrix, the fibroblasts change into an activated phenotype, termed myofibroblast, that is classically characterised by the expression of α-smooth muscle actin (α-SMA) (18, 19). These cells may express stress fibres with non-muscle actins and smooth muscle proteins such as heavy-chain myosin and calponin. However, there are several situations in animal models where these cells do not express α-SMA, which make the definition of the myofibroblast rather complex (19, 20). Different subclasses of myofibroblasts have therefore been proposed, including protomyofibroblasts (expression of stress fibres with cytoplasmic actins and increased motility) and the more smooth muscle-like differentiated myofibroblasts (i.e. expression of stress fibres with α-SMA and less motility). The increased expression of components associated with the contractile apparatus of the myofibroblasts enable the cell to exert mechanical contraction, thereby participating in wound closure. The differentiation of fibroblast into protomyofibroblasts is facilitated by mechanical tension within the wound, accompanied by secreted ED-A fibronectin (21) and PDGF that induce the formation of stress fibres and motility (15). It has been suggested that TGF-β participates in the differentiation process by inducing the expression of α-SMA in cultured fibroblasts, thereby stimulating the differentiation of fibroblasts and protomyofibroblasts into myofibroblasts (22).
Under non-pathological conditions, the myofibroblasts undergo apoptosis or revert back to a fibroblast phenotype after wound closure. However, in asthma and other fibrotic disorders, the wound healing process persists and the myofibroblasts are continuously abundant in tissue. It has been proposed that TGF-β plays an important role here by inhibiting myofibroblast apoptosis (23).
Fibroblasts isolated from patients with asthma, SSc and control subjects display distinct phenotypes and have recently been isolated from bronchoalveolar lavage (BAL) fluid (24, 25). BAL fluid (BALF) fibroblasts display features of both protomyofibroblasts and differentiated myofibroblasts, which make them possible key cells in the remodelling process. This finding is interesting since BALF samples (mostly mirroring lower airways) are believed primarily to cover areas of the airway lumen, and fibroblasts are associated with mesenchymal areas beneath the bronchial epithelium where fibroblasts can attach to an appropriate structural network of ECM components, a prerequisite for fibroblast survival. It is therefore not clear what origin these BAL fluid fibroblasts have but recent data suggest that these cells may differentiate from circulating fibroblast progenitor cells (26).
The formation of the ECM matrix is a dynamic process that is tightly regulated under non-pathological conditions. In patients with asthma, the loss of this regulation leads to an increase in the deposition of predominant ECM components such as collagen I, III, V and ED-A fibronectin in the thickened lamina reticularis. The proteoglycans versican, biglycan and decorin were significantly abundant in the submucosa beneath the epithelium in bronchial biopsies from patients with asthma (27) and in post-mortem studies of patients where asthma was the primary cause of death (28). Collectively, these observations underline the fatal consequences when the tightly controlled interplay between the ECM and the fibroblasts is unbalanced, which is characteristic for the aberrant wound healing process in airway fibrosis.
The origin of the fibroblast in airway remodelling is not fully understood, and the recent discoveries in epithelial mesenchymal transition and the plasticity of circulating fibroblast progenitor cells have made it even more complex. Therefore, the possible fibroblast origins to date can be divided into the following three categories that will be further discussed: recruitment and differentiation of resident cells, circulating fibroblast progenitor cells and epithelial–mesenchymal transition (EMT).
During airway remodelling, fibroblasts are likely to be derived simultaneously from these sources.
Recruitment and differentiation of resident cells
Since differentiated myofibroblasts are similar to airway smooth muscle cells with respect to the expression of α-SMA, it has been suggested that these cells act as potential sources of the myofibroblasts in airway remodelling. In granulation tissue, there is evidence that residing local fibroblasts are the main source since myofibroblasts were seen to revert back to the original fibroblast phenotype after wound closure (29). In addition, different fibroblast clones have been observed to accumulate within fibrotic tissue with distinct ECM production (30). Other studies have shown that smooth muscle cells may acquire a phenotype that is similar to a myofibroblast with respect to morphology and functional characteristic (31). It may therefore seem that myofibroblasts are derived from pre-existing resident lung fibroblasts, under the influence of factors secreted from surrounding lung cells and ECM.
Circulating progenitor cells: the fibrocyte
The term ‘fibrocyte’ was first used to define a circulating progenitor cell in 1994 by Bucala and coworkers (32). The fibrocyte cell type was characterised by expressing collagen I, vimentin and the surface molecule CD34. In these earlier studies, it was suggested (and has recently been demonstrated) that fibrocytes cultured from peripheral blood originate from the bone marrow (33–35). Other surface markers used to characterise fibrocytes include the pan-leukocyte antigen CD45, the chemokine receptor CXCR4 and HLA-DR, which reflects the antigen presenting ability of these cells (32, 36). Moreover, studies have shown that fibrocytes secrete further ECM components implicated in wound repair and remodelling such as fibronectin and collagen III (37). These characteristics make the fibrocyte unique in being a matrix-producing cell of the peripheral blood. It was suggested that these cells were involved in wound healing because of their rapid entry from the blood circulation into wounded tissue. In this process, fibrocytes may play a role in the deposition of ECM components and because of the ability of these cells to present antigens to inflammatory cells, they may also participate in modulating the immune response (38). Furthermore, studies have shown that fibrocytes may differentiate into α-SMA expressing cells, exhibiting a contractile force, after stimulation with TGF-β(39). These findings suggest that fibrocytes may differentiate into myofibroblast-like cells under the influence of TGF-β, which further establish the recruitment and participation of fibroblasts as a potential key event in remodelling.
Fibrocytes express several chemokine receptors, such as CCR3, CCR5, CCR7 and CXCR4, which are believed to play an important role in the recruitment and homing mechanism to the wounded tissue. However, this field is in need of more research to clarify this matter. In an animal model with bleomycin-induced pulmonary fibrosis, Phillips and coworkers showed that fibrocytes migrate to fibrotic tissue in response to stromal-derived factor (defined as SDF-1 or CXCL12), which is the ligand to the receptor CXCR4 (40). Inhibition of this chemokine resulted in a decrease of collagen deposition. This was the first study to demonstrate that fibrocytes actually contribute to the pathogenesis of pulmonary fibrosis. Another chemokine receptor that has gained attention recently due to its plausible role in fibrocyte recruitment is the CCR2 to which soluble monocyte chemoattractant protein-1 binds (41). Fibrocyte recruitment was decreased in CCR2-knockout mice, and importantly, CCR2 expression was lost when fibrocytes were transferred into fibroblasts (42). This observed transition was associated with increased production of collagen I. It is important to mention that CCR2 is not expressed in human fibrocytes, but the observation in this animal model may be an interesting parallel to fibrocyte recruitment in humans (43). Factors proposed to be involved in regulating fibrocyte recruitment include serum amyloid protein (SAP), where inhibition of fibrocyte outgrowth from peripheral blood was observed in vitro under the influence of SAP (43). This observation has been supported in patients with SSc, a disease characterised by fibrotic lesions to the skin and internal organs where decreased levels of SAP were observed (44). Another regulating factor in the recruitment of fibrocytes may be the transcription factor hypoxia-inducible factor-1 (HIF-1). Evidence has shown that the induced CXCL12 levels, which promote progenitor cell trafficking, are regulated by HIF-1 (45).
The role of fibrocytes in the subepithelial fibrosis of asthma is unclear; however, several studies have shown evidence that these cells may play an important role in this disease in a similar manner to pulmonary fibrosis. In patients with mild asthma, the levels of recruited CD34+/CD45+/α-SMA fibrocytes have been shown to correlate to the thickening of the lamina reticularis in these patients (26). This study also reported that recruited fibrocytes were present in areas close to the basement membrane and indicated a link to neutrophilic inflammation. Furthermore, fibrocytes were reported in BALF from a subgroup of these patients, and these cells could differentiate into myofibroblast-like cells with increased ECM production in vitro without additional growth factors. In addition, a specific haptoglobin expression was observed in the BALF from this subgroup of patients, suggesting a specific inflammatory profile during differentiation of these fibroblast progenitor cells (46).
In patients with allergen-induced asthma, endobronchial biopsies showed that airway fibrosis was associated with the presence of CD34+/collagen I+ fibrocytes (47). Furthermore, the same study showed an inverse correlation between CD34+ expression and α-SMA in tissue, and that TGF-β or endothelin-1 decreased the expression of CD34 in fibrocytes, accompanied by an induction of α-SMA, indicating that fibrocytes may acquire a myofibroblast phenotype in fibrotic tissue from patients with asthma.
Interestingly, in patients with IPF, fibrocytes have also been found in increased level in both peripheral blood (48) and in lung tissue (49). The increased level of fibrocytes was especially observed when the combination of the CXCR4/procollagen was analysed. Also, in this group of patients, the number of fibrocytes could be correlated with structural changes in the lung such as the number of fibroblastic foci. Furthermore, the fibrocyte attractant chemokine CXCL12 was increased in plasma vs healthy controls and detectable in the BAL fluid of 40% of the patients but not in controls. In the lungs, CXCL12 was strongly expressed by alveolar epithelial cells (49).
Apart from being a potentially important player in remodelling, fibrocytes have been shown to exert important influences in other areas. For example, fibrocytes have been suggested to induce angiogenesis by secreting vascular endothelial growth factor (VEGF) (50), which has given fibrocyte attention in tumour biology (51) and vascular remodelling occurring in structural changes of the lung (52, 53).
The third possible fibroblast origin that is known today is EMT. This process is self-explanatory: epithelial cells are phenotypically converted into mesenchymal cells (fibroblasts) under the influence of epithelial stress, inflammation or wounding (54). Until recently, EMT was primarily associated with kidney fibrosis and during development, but recent findings suggest that it may be involved in the lungs as well (55). The mechanism of EMT is characterised by induction of components that promote proteolytic digestion of the basement membrane upon which the epithelium reside (56). This process is initiated by metalloproteinases or membrane assembly inhibitors (57). As the epithelial cells lose their adherences junctions, tight junctions, desmosomes and cytokeratin intermediate filaments, the expression of ED-A-fibronectin and α-SMA increases (58, 59).
The conversion of epithelial cells into mesenchymal cells is facilitated by factors such as TGF-β, epidermal growth factor, insulin growth factor II, or fibroblast growth factor-2 (FGF-2) (60–62). Factors that may regulate and counterbalance EMT include BMP-7 (63).
Recent evidence has shown that EMT occurs and is facilitated by TGF-β1 in human alveolar epithelial cells from patients with IPF and in vivo models (59, 64). Furthermore, EMT has been suggested as being a potential fibroblast source in lung transplant recipients (65). Myofibroblasts derived from EMT have been shown to produce ECM components such as collagen I, hereby implicating a role for these cells in fibrosis. Although EMT may be an important factor during pathological remodelling, no strong evidence of this process has so far been linked to human airway disorders.
In summary, it is clear that the airway mesenchymal cell population should be considered as more dynamic than has previously been recognised. The constitutive elements represented by mesenchymal cells and the interplay with the epithelium play a key role in aberrant tissue repair processes and thereby, airway remodelling. This view is also in line with the clinical observation that treatment with ICS does not fully prevent airway remodelling and thus have restricted influence on the natural course of the disease. It is therefore of utmost importance to gain more information about specific markers for the various fibroblast phenotypes as well as knowledge about factors that affect proliferation, recruitment, migration and proliferation of these cells to develop future diagnostic and therapeutic tools.