Idiopathic Pulmonary Fibrosis: Epithelial-mesenchymal interactions and emerging therapeutic targets

Abstract

Idiopathic pulmonary fibrosis (IPF) is a chronic fibrotic disease of the lung that is marked by progressive decline in pulmonary function and ultimately respiratory failure. Genetic and environmental risk factors have been identified that indicate injury to, and dysfunction of the lung epithelium is central to initiating the pathogenic process. Following injury to the lung epithelium, growth factors, matrikines and extracellular matrix driven signaling together activate a variety of repair pathways that lead to inflammatory cell recruitment, fibroblast proliferation and expansion of the extracellular matrix, culminating in tissue fibrosis. This tissue fibrosis then leads to changes in the biochemical and biomechanical properties of the extracellular matrix, which potentiate profibrotic mechanisms through a “feed-forward cycle.” This review provides an overview of the interactions of the pathogenic mechanisms of IPF with a focus on epithelial-mesenchymal crosstalk and the extracellular matrix as a therapeutic target for idiopathic pulmonary fibrosis.

Introduction

Idiopathic pulmonary fibrosis (IPF) is a chronic progressive fibrotic disease of unknown etiology that is marked by progressive deposition of extracellular matrix proteins and ultimately failure of the respiratory system and death. It is the most common of the idiopathic interstitial pneumonias (IIP’s) accounting for 25–30% of patients diagnosed with interstitial lung disease with an incidence of 3–18/100,000 persons per year in North America and Europe; the incidence and prevalence of IPF appear to be increasing [1–6]. The disease is more common among males [7, 8] and among those with a history of tobacco use [9–12]. IPF is a disease of older adults with a median age of approximately 65 years at the time of diagnosis. In multiple cohorts, median survival averages 3–5 years from the time of diagnosis, though the disease course varies significantly in individuals [2, 13]. Treatment options for patients with IPF remain limited. Lung transplantation remains the only treatment that clearly improves survival in carefully selected patients. After more than a decade of unsuccessful clinical trials, in 2014, the first two pharmacologic treatments for IPF, pirfenidone [14] and nintedanib [15], received FDA approval. These medications were shown to slow the decline in lung function among IPF patients; but did not improve survival or quality of life [14, 15]. Thus, while the success stories of pirfenidone and nintedanib represent important first steps in the care of IPF patients, there is a clear need for new and more efficacious treatments. Through the past decade, there have been substantial insights made into the fundamental pathophysiology of IPF that have revealed promising new therapeutic targets. In this review, we will summarize current understanding of the mechanisms of IPF pathogenesis with a focus on the extracellular matrix as a potential target for disease modifying interventions.

Mechanisms of Disease

Although reports of fibrotic lung disease date back to the 1800’s, the disease currently classified as IPF was formally established in the late 1960’s by Liebow and Carrington who described “usual interstitial pneumonia” (UIP) as a distinct histopathologic form of diffuse lung disease[16]. The clinical correlate for UIP histopathology was initially termed “cryptogenic fibrosing alveolitis”[17]. Consistent with this terminology, based on early work studying bronchoalveolar lavage fluid from IPF patients, the prevailing paradigm through the 1960’s–1980’s suggested that chronic inflammation in the alveolar compartment was the primary mechanism driving fibrotic remodeling in the lung [18]. In the 1980’–1990’s, there was increased focus on the role of secreted factors in the alveolar compartments (i.e. growth factors) on fibroblast activation and proliferation [18]. In parallel, careful pathologic studies by Katzenstein and others [19, 20] began to question whether inflammatory cells were primary effectors of disease. Elegant histopathologic studies and electron microscopy demonstrated relative paucity of inflammatory cells, but evidence of epithelial damage/injury nearby what were termed “fibroblastic foci”[20]. This led to the modern conceptual framework for IPF pathogenesis (Figure 1), in which epithelial injury and cross-talk between a dysfunctional alveolar epithelium and the adjacent mesenchymal compartment leads to aberrant or persistent activation of tissue repair pathways, akin to a “wound-healing response.”

Figure 1.

Summary of the pathologic features of IPF disease initiation. A variety of genetic and environmental factors have been implicated as sources of injury to the lung and/or alveolar epithelium. Following epithelial injury, activation of canonical injury-repair programs leads to inflammatory cell and fibroblast recruitment, production of collagen and ECM which undergoes pathologic remodeling. This in turn leads to abnormal repair of the alveolar epithelium and failure to coordinate regeneration of functional alveoli.

Disease initiation: A prominent role for the lung epithelium

Several independent lines of evidence support the hypothesis that injury to and/or dysfunction of the alveolar epithelium is central to disease risk initiation. First, established genetic risk factors for pulmonary fibrosis implicate the lung epithelium. Mutations in lung epithelial restricted genes (SFTPC, SFTPA2 and ABCA3) [21–30] have been implicated in familial forms of pulmonary fibrosis (Familial Interstitial Pneumonia, FIP) [31]. Alveolar epithelial dysfunction has been shown to underlie the genetic risk for Hermansky-Pudlak Syndrome[32], a lysosomal trafficking disorder with highly penetrant pulmonary fibrosis in several subtypes[33]. Mutations in genes related to telomere biology [34–41] are found in at least 20% of families with pulmonary fibrosis [31] and perhaps a similar proportion of sporadic IPF [42], and are believed to promote stem cell failure and/or senescence within the alveolar epithelium[43]. In addition, a common polymorphism in the promoter of the gene encoding MUCIN 5B (MUC5B), an airway mucin, is found in more than 2/3 of pulmonary fibrosis patients and carriers of the risk allele have 4–6-fold increased risk for disease [44–48], an unusually large effect size for a common genetic variant. Evidence of alveolar epithelial injury and dysfunction has been reported in asymptomatic family members of patients with pulmonary fibrosis[49]. Further, environmental exposures such as gastroesophageal reflux/chronic aspiration [50], tobacco smoke exposure[9], respiratory viruses [51, 52], wood dust, metal dust, asbestos, and other inhalants [53] have been identified as risk factors for pulmonary fibrosis and may act as sources of recurrent injury to the alveolar epithelium. Finally, studies using transgenic animal models that induce alveolar epithelial cell damage by diphtheria toxin [54] or deletion of shelterin complex components [55–57] develop some degree of spontaneous lung fibrosis. Together, these studies suggest that the lung epithelium plays a prominent role in driving early disease pathogenesis. The mechanisms and mediators linking epithelial cell dysfunction to tissue fibrosis have not yet been comprehensively determined, however it is clear that crosstalk between epithelial cells, the extracellular matrix, and nearby mesenchymal and inflammatory cell populations is central to this process.

The ECM in lung epithelial homeostasis and repair

The pulmonary extracellular matrix is a three-dimensional meshwork consisting of proteins (predominantly collagen and elastin), glycoproteins, and proteoglycans. In addition to these structural elements, there are abundant signaling molecules present within the matrix including transforming growth factor β (TGF-β), fibulin, osteopontin, periostin, connective tissue growth factor (CTGF), and fibronectin [58, 59]. Degradation products of the ECM (i.e. matrikines) also play a signaling role in addition to structural functions. These signaling molecules have been shown to play a central role in the fibrotic response in IPF by modulating epithelial repair, promoting fibroblast recruitment and myofibroblast activation.

As described above, injury to and aberrant repair of the alveolar epithelium is central to the initiation of pulmonary fibrosis. While most work has focused on the role of the ECM and mesenchymal cell interactions in the context of matrix remodeling and disease progression, more recent studies suggests an important role in alveolar homeostasis and coordination of injury-repair. There has been extensive work through the past decade identifying a hierarchy of distinct progenitor (i.e. “stem”) cell niches in the lung epithelium which appear to contribute to epithelial repair in different conditions and experimental models (Reviewed in [60, 61]), including basal cells [62], club/Clara cells [63, 64], bronchoalveolar stem cells (BASCs), lineage-negative epithelial progenitor cells (LNEPs) [65] type II alveolar epithelial cells (AEC2s) [66], and subsets of AEC2s [67].

Work from several groups has underscored the central role the mesenchyme plays in regulating alveologenesis [66, 68]. Recent work from two groups using single-cell transcriptomics and linage tracing mouse models demonstrated that a subset of fibroblasts characterized by PDGFRα+ [68] or Lgr5+ [69] with evidence of active Wnt signaling (Axin2+) localize to the alveolar compartment and are involved in alveolar homeostasis. This Axin2+ mesenchymal population appears to robustly support alveolar growth and maturation in organoid assays [68], and to preferentially expand and differentiate into myofibroblasts following bleomycin injury[68], although other populations contribute as well [63].

Less is known about the roles of specific components of the ECM in alveolar repair. A specific role for the matrix GAG hyaluronan was recently identified in regulating AEC2 progenitor function [70]. In this work, AECs isolated from IPF patients were found to have reduced cell surface hyaluronan. Mice deficient for the innate immune receptor toll-like receptor 4 (TLR4) or hyaluronan synthase 2 in AEC2’s developed worse experimental fibrosis and had impaired AEC renewal capacity in organoid assays [70]. Thus, hyaluronan appears to promote AEC progenitor function in a manner that requires TLR4. As described further below, the role of hyaluronan more broadly in lung fibrosis is complex and appears to involve competing effects in the epithelium [70] versus mesenchymal compartments [71].

Disease Progression: Growth Factors and Matrix Remodeling

Following injury to the alveolar epithelium, a variety of signaling pathways that must be carefully coordinated to facilitate effective and functional repair. The injured epithelium and recruited inflammatory cells secrete a variety of mediators including transforming growth factor beta (TGF-β), connective tissue growth factor (CTGF), and others that modulate epithelial repair, promote fibroblast recruitment and myofibroblast activation. In IPF, these activated fibroblasts secrete collagen and other ECM components which then undergo pathologic remodeling and abnormal cross-linking, changing the mechanical properties of the pulmonary ECM. Mechanosensing mechanisms in fibroblasts promote further activation and myofibroblast differentiation, establishing a pathologic “feed-forward” cycle of fibrosis begetting more fibrosis. Several steps in this pathologic cycle have emerged as promising novel therapeutic targets (Figure 2).

Figure 2.

Feed-forward profibrotic epithelial-mesenchymal interactions. After activation from injured AECs, there is elaboration and activation of profibrotic mediators including TGF-B, CTGF, and others that lead to recruitment and activation of fibroblasts, increased deposition of ECM and reduced breakdown of the ECM, ultimately leading to a stiffened and poorly organized ECM. The increased matrix stiffness creates a feedback loop along with autocrine signaling from activated fibroblasts to further activate fibrotic signaling.

TGF-β signaling

TGF-β is an integral mediator in the initiation and progression of tissue fibrosis [72]. Within the cell, TGF-β is synthesized non-covalently bound to a latency associated peptide (LAP) to from a complex known as the small latent complex (SLC) [73, 74]. This complex then links to a latent TGF-β-binding protein (LTBP) in the endoplasmic reticulum forming a complex known as the large latent complex (LLC) [73, 74]. The LLC is secreted from the cell into the extracellular matrix where it resides in its inactive form covalently bound to ECM proteins, such as fibronectin and fibrillin-1, and awaits further activation [75–77]. TGF-β can be activated from within the LLC in various ways including variations in temperatures, acidification, oxidation, proteolytic cleavage, or via integrin interactions allowing the TGF-β molecule to interact with the TGF-β Receptors (TGF-βR) and mediate downstream effects [72, 78–81].

In normal physiology, TGF-β has multiple functions including immune regulation, wound healing, control of cellular proliferation, and as a signal for production of extracellular matrix [82]. TGF-β knockout mice rapidly develop a fatal acute inflammatory response, most prominently in the heart and lungs [83]. This phenotype, with few differences, can be replicated with TGF-βRII knockout and via marrow transplantation from TGF-βRII knockout mice to healthy recipients [84]. Together these studies demonstrate that TGF-β plays an important role in suppressing lung inflammation and promoting lung homeostasis.

There is abundant evidence that TGF-β and related signaling play a central role in IPF pathogenesis [74]. TGF-β levels have been shown to be upregulated in mouse models of experimental pulmonary fibrosis and in BAL fluid of patients with IPF [85, 86]. In addition, TGF-βRI expression is downregulated in the airway cells lining honeycomb cysts of IPF lungs while TGF-βRI and II are upregulated in fibroblasts [87]. It is likely that TGF-β is produced by multiple cell types relevant in the pathogenesis of pulmonary fibrosis. After an injury event, activated TGF-β is released from alveolar epithelial cells, macrophages, platelet granules, and can be activated from LLCs in the extracellular matrix [88, 89]. In addition, TGF-β is also produced by infiltrating regulatory T cells (Tregs) [90] that arrive later in the injury/repair process and have been associated with IPF progression[91].

Studies using transgenic animals and have demonstrated that overexpression of TGF-β₁ lead to development of fibrotic lung disease [92, 93]. There are multiple mechanisms through which TGF-β has been shown contribute to lung fibrosis. First, TGF-β has diverging effects on the epithelial and fibroblast proliferative responses; TGF- β generally promotes proliferation of fibroblasts and other mesenchymal cells [73] while suppressing epithelial cell proliferation [72]. TGF- β promotes epithelial cells to undergo a phenotype shift characterized by adoption of mesenchymal features [94, 95]. In addition, active TGF-β signaling plays a central role in mediating myofibroblast activation and subsequent production of collagen and other ECM components [74].

Convincing evidence that TGF-β signaling in the lung epithelium plays a profibrotic role emerged as reports from two groups demonstrated that deletion of TGF-βRII in AECs is protective against bleomycin-induced lung fibrosis [96] despite increased inflammation [97]. TGF-βRII knockout in macrophages was also shown to decrease transition to the M2 phenotype which is associated with wound healing and tissue repair [98]. There has been growing interest in attempting to specifically target mechanisms of TGF- β activation and signaling as a therapeutic strategy.

Integrins and TGF-β activation

Integrins are transmembrane proteins that play a pivotal role in the maintenance of cell-ECM adhesion, cell-cell adhesion, and transduction of signals from the ECM to the intracellular compartment [99]. The integrin family is composed of one of 18 unique α subunits and 8 β subunits that pair to create one of 24 identified integrins [99]. Integrins in the lungs have been associated with asthma, COPD, non-small cell lung cancer, sarcoidosis, bronchopulmonary dysplasia [100] and pulmonary fibrosis [101].

The αvβ6 integrin is the most well-studied in the field of IPF. It is primarily found in the lungs and skin and is upregulated in the lungs of murine bleomycin models of IPF [102–104]. The αvβ6 integrin has been shown to bind to the TGF-β LLC where it leads to extracellular activation of the TGF-β signaling [81]. This occurs via binding of αvβ6to LTBPs where it interacts with an RGD domain on the hinge regions of LTBP-1 and LTBP-3[105]. Through interactions with intracellular actin, αvβ6 exhibits mechanical force on the LLC thus releasing active TGF-β thus propagating its downstream fibrotic potential.

Knockout mouse models of the β6 component have shown similar effects to the TGF -β knockout mouse with inflammation limited to the lung and the skin [104] which can be abrogated by restitution of the β6 gene to type 2 AECs and bronchial epithelial cells [106]. Interestingly, partial inhibition of αvβ6, reduced pulmonary fibrosis in a bleomycin model without increased baseline inflammation in the lungs or skin [107]. Early human studies targeting αvβ6 in IPF are underway. Recently, a monoclonal antibody targeted at the αvβ6 integrin (BG00011, formerly STX-100) was tested in a phase two clinical trial (Clinicaltrials.gov, NCT01371305). Data from this study are anticipated soon.

In addition to αvβ6, other integrins in the αv family have been shown to activate latent TGF-β as well. The αvβ1, αvβ3, αvβ5, αvβ8 integrins have all been shown to contribute to TGF-β activation, activation of fibroblasts and myofibroblasts, and increased ECM production [108–110]. These αv integrin containing complexes have also emerged as potential therapeutic targets [111], in particular αvβ1 integrin. αvβ1 integrin is expressed on fibroblasts, where it has been shown to mediate adhesion of the TGF- β latency associated peptic (LAP) [112]. The LAPs bind secreted TGF- β and maintain it in an inactive state; binding of αvβ1 integrin to LAP leads to TGF- β activation. A small molecule inhibitor of αvβ1 integrin was shown to be protective against bleomycin-induced fibrosis [112].

Growth Factors

A variety of other secreted factors have also been implicated in pulmonary fibrosis pathogenesis, including connective tissue growth factor (CTGF), platelet-derived growth factor (PDGF) and fibroblast growth factors (FGFs).

CTGF is small peptide that has a variety of stimulatory effects on fibroblasts. It is synthesized at low levels in the healthy lung, but has been shown to be upregulated under several models of pulmonary stress including bleomycin and radiation induced fibrosis [113, 114]. CTGF is upregulated in both Type 2 AECs and pulmonary fibroblasts under stress [115–117]. CTGF secreted under stimulation by TGF-β and has downstream effects via the MAP kinase pathway including stimulation of fibroblast migration and increased collagen production via upregulation of the Col1A promoter [116, 118, 119]. In fact, several studies of TGF-β interactions with CTGF suggest that many aspects of the downstream fibrotic activity of TGF-β are dependent on the presence of CTGF [118, 120–123]. Interestingly, CTGF has been shown to increase TGF-β production in fibrotic systems, indicating a potential internal positive feedback loop [124].

PDGF is a common growth factor produced in many cells including macrophages, platelets, endothelial cells, and fibroblasts and is upregulated under stimulation by inflammatory cytokines[125]. PDGF receptors (PDGFRs) are expressed in two forms the α form and the β form which bind to different forms of PDGF and have differed downstream effects [125]. PDGF gene expression is upregulated in bleomycin induced pulmonary fibrosis [126, 127] and introduction of a PDGF-B gene into the mouse lung via adenovirus induces severe lung fibrosis [128]. Inhibition of PDGF via anti-PDGF antibodies or the PDGFR via small molecule inhibition is able to inhibit the progression of lung fibrosis in the murine model [127, 129].

Fibroblast growth factors (FGFs) are a family of 22 proteins with 5 associated FGF-receptors (FGFRs) which have pleiotropic effects on the deposition of fibrotic tissue within the pulmonary parenchyma. Several FGFs have been shown to be upregulated in IPF, making them of interest in the modulation of fibrosis [130, 131], although mechanistic studies to date have not yielded a consensus mechanism. For example, FGF-1 has been shown to have anti-apoptotic and pro-migratory effects on fibroblasts in one study [132], but was shown to have pro-apoptotic effects in another [133]. Interestingly, when fibroblasts harvested from IPF lungs were exposed to FGF-1, there was minimal effect [132]. FGF-2 can be released by macrophages, AECs, fibroblasts, T-lymphocytes and mast cells under signaling from TGF-β [134]. FGF2 levels are upregulated in the IPF lung, and promotes the recruitment of myofibroblasts and stimulation of the production of extracellular matrix proteins [135]. Inhibition of FGF-2 decreased cellular DNA production genes suggesting this is one of the ways in which TGF-β induces the fibrotic response [136]. Importantly, inducible inactivation of FGFRs in murine models attenuated the fibrotic response to bleomycin and reduced fibroblast migration into fibrotic areas suggesting signaling from these receptors is involved in the development of fibrosis [137].

Therapies targeting these molecules and their receptors have been among the most successful antifibrotic treatments to date. Nintedanib, one of the two FDA-approved treatments for IPF, inhibits PDGFRs and FGFRs, in addition to VEGFRs [15, 138]. A monoclonal antibody targeted to CTGF has shown promise in murine and early human studies. Within murine models of fibrosis, including radiation and bleomycin, inhibition of CTGF can both prevent lung fibrosis, as well as, reverse already established lung fibrosis [114, 116, 139–141]. Recent Phase II trials of anti-CTGF monoclonal antibodies in humans showed no particular side effects, and there was no evidence of progression of disease via pulmonary function testing in the patients receiving the drug [142]. Phase III clinical trials of the anti-CTGF monoclonal antibody Pamrevlumab (formerly F-3019) in IPF are currently underway.

Collagen and matrix remodeling

In addition to increased production of ECM components, aberrant matrix remodeling is also a well-recognized feature of pulmonary fibrosis [59]. Collagen is a linear protein formed into cross linked fibrils that is responsible for the strength of the ECM and makes up about 20% of the dry weight of the normal human lung [143]. The predominant type of collagen within both the normal and IPF lung is Type 1 collagen which is deposited in higher amounts within the ECM compared to other collagen subtypes [58, 144]. Collagen is secreted as individual fibers by the fibroblasts and myofibroblasts under signaling primarily from TGF-β, and also stimulated by several other cytokines and via hypoxia [145, 146]. Collagen fibrils are then crosslinked by lysyl oxidase (LOX), lysyl oxidase like proteins (LOXLs), tissue transglutaminase (TTG) and other enzymes to form covalent bonds that solidify the structure of the collagen and increase the tensile strength of the extracellular matrix [147, 148].

Matrix crosslinking appears to be vital to the progression of pulmonary fibrosis. Gene expression of LOX, LOXL and TTG are each upregulated in IPF lungs, and increase the crosslinking between the collagen fibrils. Increase in LOXL2 and TTG activity leads to an overall increased crosslinking of collagen fibrils and increased stability and stiffness of the extracellular matrix [147, 148]. Inhibition of TTG2 was able to increase ECM turnover in vitro and was associated with decreased collagen accumulation in bleomycin treated mice [147, 149]. LOXL2 has been shown to correlate with increased disease progression and mortality in IPF [150]. In addition, LOXL2 was shown along with LOXL3 to be involved in fibroblast transition to myofibroblasts [146]. LOXL2 has the strongest effect on fibrotic tissue remodeling and some data suggests it may play a role in release of TGF-β from the ECM [151]. In addition, inhibition of LOX and the LOXL family has been shown to reduce fibrosis in bleomycin treated mice [152, 153]. Ultimately, increased expression/activity of crosslinking proteins leads to an increase in the thickness and tensile strength of the extracellular matrix and subsequent fibroblast activation [153, 154]. Early phase clinical trials targeting this pathway have been initiated; results to date are limited but one study was terminated at the second interim analysis for lack of efficacy [155].

Matrix metalloproteinases (MMPs) along with their inhibitors, tissue inhibitors of metalloproteinases (TIMPs), are secreted by multiple cell types in the lung including AECs and Fibroblasts. MMPs are responsible for the degradation of collagen and other matrix proteins. MMP-1, MMP-2, MMP-7, MMP-8, and MMP-9 have been shown to be upregulated in IPF [156–160]. MMP-7 has been associated with increasing fibrosis, and is predominantly produced in activated alveolar epithelial cells and bronchiolar cells [160]. Knockout of MMP-7 reduced pulmonary fibrosis in bleomycin treated mice [160]. MMP-9 is also of significant interest. Immunohistochemical staining of human lungs showed evidence of MMP-9 within the ECM in areas of dense scar in UIP, as well as, neutrophils and myofibroblasts [160, 161]. In addition, MMP-9 has been shown to activate TGF-β from the extracellular matrix [78]. In animal models, MMP-9 over-expression within alveolar macrophages attenuated bleomycin induced fibrosis [162], while knockout did not seem to have effect on total fibrosis but did decrease bronchiolization of alveoli [163]. MMP-3 has been shown to have pro-fibrotic effects with overexpression leading to fibrosis and MMP-3 null mice having reduced fibrosis in response to bleomycin [164]. Several other MMPs have been implicated in the pathogenesis of pulmonary fibrosis, and studies are ongoing [165].

TIMPs are also upregulated in the presence of pulmonary fibrosis and spread throughout the pulmonary parenchyma [161]. TIMPs inhibit MMPs in a myriad of ways and are their upregulation is thought to decrease MMP induced degradation of the ECM and thus a pro-fibrotic microenvironment throughout the extracellular matrix [161, 166]. MMP-generated collagen neoepitopes have been identified in serum from IPF patients and have been associated with disease progression, suggesting monitoring this axis may be a promising biomarker strategy [167]. Further investigations are underway to better understand and potentially modulate the MMP/TIMP system [165].

Matrikines

In addition to their structural roles, ECM components or their breakdown products have an increasingly recognized role as signaling mediators known as matrikines. While much work has focused on matrikines in the context of inflammatory signaling, a direct role in regulating epithelial and mesenchymal injury responses is emerging.

Hyaluronan (HA) is a non-sulfated glycosaminoglycan that is formed as a linear chain of repeated sugars, specifically D-glucuronic acid and N-acetylglucosamine [168]. The long chain allows for binding of salts and water and creates a visco-elastic substance that provides structure to the ECM [169, 170]. Within the pulmonary system, HA is primarily localized to the airways and interstitial spaces and has been shown to be induced by inflammatory signaling from cytokines including IL-1β and TNF-α [171, 172].

The role of HA has proven to be complex. As discussed above, in the alveolar epithelium HA promotes AEC homeostasis and progenitor cell function through an interaction with TLR4 [70]. Interestingly, HA secretion has been shown to be induced by endoplasmic reticulum stress in AECs[173]; whether this is adaptive or pathologic is less clear. In the mesenchymal compartment, HA confers a profibrotic phenotype [71]. HA is produced primarily by myofibroblasts and has been shown to be increased within the lungs of patients with IPF compared to normal controls [174]. In mouse models of bleomycin injury, HA has been shown to accumulate in the fibrosing lung. Further analysis of HA action in the IPF lung suggest that it contributes to inflammation, mediated by interaction with CD44 on a variety of cells. The CD44/HA axis has also been linked to mesothelial cell EMT[175], which has been implicated as a potential source of myofibroblasts in IPF[176, 177].

CD44/HA interactions have been shown to be important in fibroblast recruitment and upregulation, recruitment of inflammatory cells from the peripheral blood, and signaling for its own degradation. Knockout of the hyaluronan 2 synthetase (HAS2) gene was embryonic lethal secondary to cardiac malformations [178]. Conversely, overexpression of HAS2 within myofibroblasts was shown to increase these cells invasiveness into the ECM and augmented their fibrotic potential [71], mediated in part through HAS2 regulation of fibroblast senescence[179]. Conditional deletion of HAS2 in mesenchymal cells, as well as a more specific loss in fibroblasts, lead to decreased hyaluronan and collagen deposition and decreased accumulation of myofibroblasts within sites of lung remodeling [71]. The CD44 receptor has been shown to be upregulated in the IPF lung and in the lungs of bleomycin challenged mice [180]. CD44 is involved in the wound resolution response by inducing macrophage clearance of ECM proteins, especially hyaluronan, and is under the control of several cytokines including IL-1, TNF-α, and LPS [181–183]. CD44 knockout mice are not resistant to bleomycin induced fibrosis, however, they are resistant to development of fibrosis induced by overexpression of hyaluronan. Similarly, inhibition of CD44 by a monoclonal antibody abrogated the fibrotic response to hyaluronan overexpression, but not bleomycin induced fibrosis in the murine model. Evaluation of fibroblasts treated with anti-CD44 monoclonal antibody in vitro and in vivo models showed that inhibition of CD44 reduced myofibroblast invasiveness [71, 184].

The collagen breakdown proline-glycine-proline (PGP) motif has been well established as a neutrophil chemoattractant and driver of COPD pathogenesis [185] and regulator of acute airway inflammation[186]. While acetylated PGP is resistant to degradation and promotes sustained neutrophilic inflammation, it surprisingly appears to protect against fibrosis [187] in a manner dependent on angiotensin-converting enzyme (ACE)-mediated degradation. This provides at least one potential mechanism explaining the paradox of why some smokers develop emphysema while others develop pulmonary fibrosis. Conversely, studies suggest that collagen V, tenascin C, and elastin deposited into the ECM under fibrotic conditions can stimulate fibroblasts to increase ECM remodeling, thus creating a positive feedback loop that increases pulmonary fibrosis [188]. With further study, it seems likely that increasingly complex signaling roles for ECM components and matrikines will be discovered and present new therapeutic targets.

Matrix Stiffness and Mechanosensing in disease progression

The injury-repair mechanisms described above lead to a variety of alterations in the composition and organization of the pulmonary extracellular matrix. This remodeling leads to changes in the mechanical properties of the ECM, particularly an increase in matrix stiffness[189]. It has become increasingly evident that in the setting of increased matrix stiffness, mechanosensing mechanisms in fibroblasts potentiate profibrotic cellular phenotypes [190]. Normal human lung fibroblasts grown on decellularized lung matrices from IPF lungs show evidence of enhanced TGF-β signaling and differentiate into myofibroblasts [191], and ex-vivo culture experiments have indicated this process is mediated through Rho/ROCK signaling and megakaryoblastic leukemia factor-1 (MKL1) [192].

Several potential sensors of matrix stiffness have been implicated, including the transient receptor potential vanilloid 4 (TRPV4) channel [193] and α6 integrin [194]. The signaling events driving stiffness-mediated fibroblast activation involve accumulation of phosphorylated myosin-2 [195]. The transcriptional control of stiffness-driven myofibroblast differentiation is mediated through the yes-associated protein (YAP) and transcriptional coactivator with PDZ-binding motif (TAZ) [196]. Knockdown of YAP and TAZ specifically abrogate matrix-stiffness driven fibroblast activation, and expression of active YAP or TAZ in normal fibroblasts grown on soft matrix promotes invasiveness and myofibroblast differentiation [196].

Intriguingly, myofibroblast differentiation does not appear to be a terminal cell fate. Mice treated with a Rho kinase inhibitor during the recovery phase of bleomycin injury were protected from lung fibrosis [197]. Further, when grown on physiologic (i.e. normal) stiffness matrix, primary IPF fibroblasts lose their contractile phenotype; this can be recapitulated by Rho kinase inhibition [198].

The role of ECM stiffness on epithelial cells has been less studied, although there are suggestions ECM stiffness regulates epithelial cell phenotypes. Data from isolated rat AECs suggested that matrix stiffness did not alter AEC differentiation or promote adoption of mesenchymal features, but did increase laminin and fibronectin expression [199]. The Hippo-Yap-Taz axis has been implicated in regulating lung epithelial development [200], and in regulating basal cell migration and differentiation following injury[201]; additional study is needed to better elucidate the role of matrix stiffness and mechanosensing on epithelial repair more broadly.

Together, these data indicate that the mechanical properties of the ECM drive a feed-forward cycle in which matrix remodeling drives further fibroblast activation, myofibroblast activation, and additional matrix remodeling both directly and through paracrine signaling [188, 202]. There is extensive ongoing investigation in multiple fields looking to target aspects of mechanosensitive signaling as an antifibrotic strategy [190]. Further work is needed to better understand the endogenous “checks” on this feed-forward system as this may also hold promise as a therapeutic target.

Conclusions

Advances in understanding of the pathogenesis of idiopathic pulmonary fibrosis have highlighted the many roles the extracellular matrix plays in coordinating normal lung homeostasis and mediating pathologic lung remodeling. Available evidence suggests that the ECM plays a central role in driving feed-forward cycles of pathogenic disease signaling mediating by integrins, growth factors, matrikines, and matrix metalloproteinases. In addition, the matrix structure and biomechanical properties have emerged as key drivers of the fibrotic process. A variety of exciting novel therapeutic strategies targeting these ECM-driven processes hold promise as novel treatments for IPF and other chronic fibrotic disorders.

Highlights.

• Lung epithelial injury-related signaling is involved in initiating IPF pathogenesis

• The ECM plays a central role in coordinating injury-repair signaling

• Aberrant and persistent activation of injury-repair signaling may be mediated through changes in the biomechanical properties of the ECM, driving a feed-forward profibrotic cycle