|Home | About | Journals | Submit | Contact Us | Français|
Rationale: The hallmarks of allergic asthma are airway inflammation, obstruction, and remodeling. Airway remodeling may lead to irreversible airflow obstruction with increased morbidity and mortality. Despite advances in the treatment of asthma, the mechanisms underlying airway remodeling are still poorly understood. We reported that insulin-like growth factor (IGF) binding proteins (IGFBPs) contribute to extracellular matrix deposition in idiopathic pulmonary fibrosis; however, their contribution to airway remodeling in asthma has not been established.
Objectives: We hypothesized that IGFBP-3 is overexpressed in asthma and contributes to airway remodeling.
Methods: We evaluated levels of IGFBP-3 in tissues and bronchoalveolar lavage fluid from patients with asthma at baseline and 48 hours after allergen challenge, in reparative epithelium in an in vitro wounding assay, and in conditioned media from cytokine- and growth factor–stimulated primary epithelial cells.
Measurements and Main Results: IGFBP-3 levels and distribution were evaluated by Western blot, ELISA, and immunofluorescence. IGFBP-3 is increased in vivo in the airway epithelium of patients with asthma compared with normal control subjects. The concentration of IGFBP-3 is increased in the bronchoalveolar lavage fluid of patients with asthma after allergen challenge, its levels are increased in reparative epithelium in an in vitro wounding assay and in the conditioned medium of primary airway epithelial cell cultures stimulated with IGF-I.
Conclusions: Our results suggest that one mechanism of allergic airway remodeling is through the secretion of the profibrotic IGFBP-3 from IGF-I–stimulated airway epithelial cells during allergic inflammation.
Allergic airway remodeling causes significant morbidity and mortality among patients with asthma and currently there are no effective therapies available to prevent or reverse this process.
Our results suggest that one mechanism of allergic airway remodeling is through the secretion of profibrotic insulin-like growth factor binding protein-3 (IGFBP-3) from insulin-like growth factor-1 (IGF-1) stimulated airway epithelial cells during allergic inflammation.
Allergic asthma is a chronic inflammatory disease characterized by bronchoconstriction and airway remodeling. Classical teaching is that patients with asthma have preserved lung function between exacerbations, and during attacks respond well to bronchodilator and corticosteroid therapy (reversibility). Although this is true for a majority of patients with asthma, there is a subgroup of patients that will progress to irreversible airflow obstruction. It is generally believed that airway remodeling is a result of chronic inflammation, but the exact relationships between inflammatory mediators and the observed structural lung changes are unclear.
As a protective barrier, the epithelium is the first layer to be exposed to allergen, and can play a unique role in initiating the inflammatory process and regulating the injury–repair cycle. Epithelial cell sloughing and denudation, seen in asthmatic airway biopsies, suggest that, in chronic asthma, the epithelium is in a state of constant repair. Subepithelial fibrosis, a component of airway remodeling, could result from a chronic or abnormal injury–repair process. Multiple growth factors, such as transforming growth factor (TGF)-β, epidermal growth factor, insulin-like growth factor (IGF), and vascular endothelial growth factor, are up-regulated during airway inflammation (1), and induce the proliferation and differentiation of nearby mesenchymal cells toward increased collagen secretion, angiogenesis, and smooth muscle hyperplasia—hallmarks of remodeling.
The precise role of IGF-I and the associated IGF binding proteins (IGFBPs) in airway remodeling has not been elucidated. IGF-I is a potent mitogen and stimulant of collagen secretion in fibroblasts (2), and has been implicated in the pathogenesis of fibrosis in a number of lung disorders (3–5). IGF-I is up-regulated in the airways of mice after allergen challenge (6), and IGF-I levels in biopsies of patients with asthma correlate with subepithelial fibrosis (7). The regulation of IGF-I is mediated, in part, by a system of six IGFBPs (8).
We have previously demonstrated that IGFBP-3 plays a role in fibrosis and extracellular matrix (ECM) deposition in idiopathic pulmonary fibrosis (4). To investigate the hypothesis that IGFBP-3 is also overexpressed in asthma, and may contribute to airway remodeling, we examined IGFBP-3 levels in bronchoalveolar lavage (BAL) fluid of patients with mild asthma before and after allergic challenge, as well as by immunofluorescent staining of epithelial cells in patients with mild and severe asthma. Finally, we evaluated the influence of various cytokines and growth factors on IGFBP-3 secretion by primary airway epithelial cells.
Patients with asthma of varying severity were recruited as part of ongoing studies of severe asthma, some of which were from the Severe Asthma Research Program. The patients with severe asthma all met the American Thoracic Society workshop definition of severe asthma (9). This definition includes the continuous use of high-dose inhaled and/or oral steroids for greater than 50% of the previous year, along with two of seven minor criteria. The patients with mild asthma were on low-dose or no inhaled corticosteroids (ICS), and had an FEV1 of more than 75% predicted, with no urgent care visits or oral prednisone in the previous year. Normal subjects had no history of respiratory disease, and were not taking any medications. No subjects had more than a 5-pack-year history of cigarette smoking, and none had smoked in the previous year. Atopy was identified as the presence of one or more positive skin prick test (wheal ≥ histamine control) to local allergens. A history of mild but untreated seasonal allergic rhinitis was allowed in the normal control subjects. Informed consent was obtained from each subject before participation. The studies were approved by the National Jewish Medical and Research Center and the University of Pittsburgh (S.E.W.) institutional review boards (Table 1).
A total of 10 subjects with allergy with physician-diagnosed asthma were studied. All subjects were nonsmokers on short-acting β-agonists only; none were using ICS or systemic corticosteroids. All had a positive skin prick test to one or more aeroallergens, and did not have a respiratory infection or asthma exacerbation within 30 days of study, and had not received long-acting β-agonists within 2 days, antihistamines or leukotriene antagonists within 7 days, or corticosteroids within 30 days of study enrollment. Informed consent was obtained from each subject before participation. The study was approved by the University of Wisconsin, Madison (N.N.J.) Center for Health Sciences Human Subjects Committee (Tables 1 and and22).
Bronchoscopy, BAL, and segmental bronchoprovocation with allergen (SBP-AG) were performed in two segments, as previously described (10). BAL (160 ml/segment) was performed before and 48 hours after SBP-AG. Allergen dose for the two segments was 10 and 20% of the dose of allergen causing a 20% fall in FEV1 (Ag-PD20). BAL fluid at baseline (0 h) and 48 hours after SBP-AG was analyzed for IGFBP-3 by Western blot and ELISA. Additional details are included in the online supplement. As a quality control measure to ensure that segmental allergen challenge does not cause a nonspecific increase in IGFBP-3 due to activation of airway epithelial cells by the allergen or the bronchoscopy procedure, a retrospective analysis was performed on BAL fluid obtained from nine normal subjects challenged with allergen (1,000 protein nitrogen units of ragweed). See the online supplement for details.
Endobronchial biopsies were taken from the 3rd–4th level subcarinae of the lower lobes, as previously described (11). Sections (6 μm) were deparaffinized, antigen-retrieved, and incubated with a polyclonal rabbit anti-IGFBP-3 antibody (GroPep, Adelaide, SA, Australia) and fluorochrome-tagged secondary antibody (see the online supplement for a detailed protocol). Fluorescence intensity was analyzed using MetaMorph (Molecular Devices, Sunnyvale, CA). Subepithelial membrane (SBM) thickness was measured on formalin-fixed tissue sections stained with Movat's pentachrome stain, which differentially stains tissue structural elements, including collagen. Pictures of tissue sections were taken at 400× magnification. SBM thickness was measured in areas of good epithelial orientation using Image-Pro software (Media Cybernetics, Inc., Bethesda, MD).
Primary human airway epithelial cells (HBEs) were cultured from excess pathological tissue after lung transplantation and organ donation under a protocol approved by the University of Pittsburgh Institutional Review Board. HBEs were cultured on human placental collagen–coated Costar Transwell filters (Catalog no. 3470, 0.33 cm2, 0.4 μm pore), as previously described (12), and used for experimentation after 4 to 6 weeks of culture in an air–liquid interface.
Polyclonal goat anti-human IGFBP-3 antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) was used for Western blot analyses. Signals were detected after incubation with horseradish peroxidase–conjugated anti-goat secondary antibody (Santa Cruz Biotechnology, Inc.) and chemiluminescence (Perkin Elmer Life Sciences, Inc., Boston, MA). IGFBP-3 levels in undiluted BAL fluid were quantified using a commercially available ELISA (R&D Systems, Minneapolis, MN). The following cytokines were used for stimulation assays: IL-4, IL-13, and TNF-α (Leinco Technologies, St. Louis, MO); IGF-I (Sigma-Aldrich, St. Louis, MO); and TGF-β1 (R&D Systems). Purified, low-endotoxin, natural Felis domesticus allergen 1 (nFel d 1) was purchased from Indoor Biotechnologies (Charlottesville, VA).
Mechanical scrape injury was induced by creating a line with a pipette tip across the surface of a confluent HBE cell layer, as previously described (13). At 18 hours after epithelial injury, IGFBP-3 expression was evaluated by immunofluorescence using polyclonal rabbit anti–IGFBP-3 (GroPep) and secondary antibody conjugated to Alexa-488 (Jackson Research, West Grove, PA). DRAQ5 (Alexis Corporation, Lausen, Switzerland) and fluorescently labeled phalloidin (Molecular Probes, Carlsbad, CA) were used as nuclear and cytoskeletal markers, respectively. Detailed methods appear in the online supplement.
Differentiated primary HBEs were transiently cultured in Dulbecco's modified Eagles medium in the presence of the indicated cytokines, growth factors, purified allergen, or dexamethasone for 24 or 48 hours. Equivalent surface area of differentiated HBEs (0.33 cm2) and media volumes were used per experimental condition, and equivalent apical and basolateral media volumes were used per well. The apical and basolateral media were collected and centrifuged for 10 minutes at 4°C to remove cell debris. IGFBP-3 levels in equivalent volumes of the sample supernatants were analyzed by Western blot.
Data were assessed for normalcy of distribution. All data were normally distributed and therefore analyzed by analysis of variance or Student's t test. When an overall difference among the groups was observed, intergroup differences were analyzed by Tukey-Kramer testing. A P value less than 0.05 was considered significant for all results.
IGFBP-3 levels were examined by Western blot of equal volumes of BAL fluid from 10 patients with mild asthma before (0 h) and 48 hours after segmental allergen bronchoprovocation (Figure 1A). The qualitative increase in IGFBP-3 was quantified by ELISA (Figure 1B). IGFBP-3 levels increased variably between a minimum of 2.5-fold (Subject 2) to greater than 50-fold (Subject 10), with a mean baseline concentration of 0.27 (±0.63) ng/ml, and a mean post-challenge concentration of 14.85 (±13.64) ng/ml. Increased levels of IGFBP-3 were induced by challenge with three different allergens (house dust mite, ragweed, and cat). After allergen challenge, total numbers of BAL cells and percentage of BAL eosinophils were significantly increased (169 ± 231 × 106 vs. 32 ± 5 × 106 total cells, P < 0.001; 44 ± 19% eosinophils vs. 1.7 ± 2.5% eosinophils, P < 0.001) (Table 2), and correlated with levels of IGFBP-3 (Spearman's correlation coefficient > 0.7; P ≤ 0.004). BAL fluid recovery was not significantly different after challenge (73 ± 5% at baseline vs. 71 ± 5% after challenge) (Table 2). Allergen challenge of normal, nonallergic subjects with no airway disease did not induce airway eosinophilia or marked changes in BAL fluid concentrations of IGFBP-3. BAL eosinophils were 0.3 (±0.7)% at baseline, 0.7 (±0.3)% 48 hours after saline challenge, and 0.6 (±0.7)% 48 hours after allergen challenge. Concentrations of IGFBP-3 in BAL fluid were 0.9 (±0.1) ng/ml at baseline, 2.0 (±1.4) ng/ml 48 hours after saline challenge, and 1.9 (±0.9) ng/ml 48 hours after allergen challenge (see Figure E1 in the online supplement). The lack of a marked increase in IGFBP-3 after allergen or saline challenge of normal subjects suggests that neither the allergen nor the bronchoscopic procedure provide sufficient signals to induce IGFBP-3 secretion by airway epithelium.
To further examine IGFBP-3 expression levels in the airways of patients with asthma, and to determine the source of IGFBP-3 in vivo, airway tissue sections were analyzed using immunofluorescence to evaluate the amount and distribution of cell-associated IGFBP-3 protein. As shown in Figure 2A, IGFBP-3 was highly expressed in the epithelial cell layer in patients with mild and severe asthma compared with healthy donors. To quantify the expression of IGFBP-3 in vivo, we analyzed the mean fluorescence intensity in the airway epithelial layer in tissues from eight patients with mild asthma, seven patients with severe asthma, and seven healthy control subjects. As shown in Figure 2B, the intensity was 2.2 times (2.2 ± 0.8) higher in mild and 4.4 times (4.4 ± 1.7) higher in severe asthmatic airways compared with healthy control airways (P < 0.0146 and P < 0.0004, respectively). Epithelial cell expression of IGFBP-3 correlated negatively with FEV1 % predicted (r = −0.54; P = 0.01), but did not correlate with SBM thickness (data not shown).
Based on the high expression of IGFBP-3 in the epithelial cell layer of patients with asthma, we hypothesized that up-regulation of IGFBP-3 may result from epithelial cell injury and repair. To test this hypothesis, we evaluated IGFBP-3 levels by immunofluorescence in an in vitro model of epithelial cell wound repair. As demonstrated in Figure 3, IGFBP-3 levels are significantly increased at the leading edge of wound repair, and subsequently return to baseline levels in regions where the epithelial monolayer has been successfully restored to a normal phenotype.
Because IGFBP-3 was markedly up-regulated in BAL fluid of patients with asthma after challenge, as well as in airway epithelial cells of patients with asthma, we examined potential inducers of IGFBP-3. Primary airway epithelial cells were stimulated for 24 and 48 hours with cytokines and growth factors, as listed in Figure 4. At 24 hours, IGF-I induced a 300-fold increase in IGFBP-3 secretion at the apical surface, and a 10-fold increase from the basolateral surface. In comparison, TGF-β1 alone induced a more modest increase in IGFBP-3 (83-fold and 5-fold, respectively). When combined with IGF-I, TGF-β1 inhibited IGF-I stimulation of IGFBP-3 at the apical and basolateral surfaces (60 and 10%, respectively). Of note, IGFBP-3 exists as two predominant glycosylation variants, with approximate molecular masses of 41 and 45 kD, that can be reduced to a single approximately 36-kD band by deglycosylation; all three of which have the capacity to bind IGF ligands (14).
To ensure that exposure to allergen does not result in a nonspecific increase in IGFBP-3 production by airway epithelial cells, we examined the effect of nFel d 1 on IGFBP-3 secretion and degradation. In vitro, nFel d 1 does not increase IGFBP-3 secretion or degradation in primary airway epithelial cells in either the presence or absence of IGF-I (Figure E2).
Because a subset of patients with asthma received corticosteroids, we evaluated whether exposure to corticosteroids alone results in increased IGFBP-3. Primary airway epithelial cells were incubated for 48 hours with dexamethasone in either the presence or absence of IGF-I. Dexamethasone alone does not result in increased IGFBP-3 secretion or detection of degraded IGFBP-3, and, in fact, demonstrates a trend toward diminished IGFBP-3 secretion in the presence of IGF-I (Figure E3).
We previously demonstrated that IGFBP-3 is associated with fibrosis and ECM deposition in IPF (4). We therefore tested the hypothesis that IGF and its binding proteins are also involved in the remodeling process of asthma. In this study, we demonstrate an increase in levels of IGFBP-3 protein in BAL fluid of patients with asthma after allergen challenge that correlates with percentages of BAL eosinophils—a hallmark of the inflammatory response in asthma. We show that baseline cell-associated IGFBP-3 levels in airway epithelial cells are higher in patients with asthma versus healthy donors, and that levels increase significantly with disease severity and worsening lung function, although they do not correlate directly with SBM thickness. Taken together, these observations demonstrate that IGFBP-3 is increased in the airway epithelium of patients with asthma, and suggest that it is released into the lumen (where it is detectable by BAL) in significant quantities only in response to irritant stimuli, such as inhaled allergen.
We have demonstrated that intracellular IGFBP-3 expression is increased at the leading edge of epithelial wound repair in an in vitro system, and that IGF-I is sufficient to dramatically up-regulate secreted IGFBP-3 levels in nondisrupted, differentiated primary airway epithelial cell culture. These observations support our hypothesis that increased epithelial cell–associated IGFBP-3 in asthma is associated with the cyclic damage and repair of the epithelium that characterizes allergic asthma. Injured airway epithelial cells are therefore one source of IGFBP-3 in post-challenge BAL, and, based upon our previous observations (4), we hypothesize that secreted IGFBP-3 is involved in the development and/or perpetuation of subepithelial fibrosis and airway remodeling in asthma.
Damage to the epithelium can range from loss of important components, such as cilia, to shedding of the epithelium and denudation of the basement membrane. Epithelial cells have been shown to release multiple profibrogenic factors in response to exogenous chemical or mechanical stimuli in an in vitro coculture model with myofibroblasts, including TGF-β, fibroblast growth factor-2, endothelin-1, platelet-derived growth factor-BB, and IGF-I (15, 16). In this model, collagen gene expression is increased and myofibroblast proliferation is enhanced (16). Our data identify IGFBP-3 as a product of airway epithelial cells induced in response to injury and to IGF-I elaboration. Taken together with our previous findings that IGFBP-3 induces a profibrotic phenotype in vitro (4), our results suggest that IGFBP-3 is another mediator of subepithelial fibrosis and allergic airway remodeling. The activity of IGF-I is modulated by IGFBPs in a cell- and tissue-specific manner (2), and IGF-I can also modulate IGFBP expression (8). Our data suggest that the latter may be the case in allergic airway inflammation. This is supported by two recent studies implicating IGF-I in airway subepithelial fibrosis in human and murine asthma models (6, 7). Hoshino and colleagues (7) reported a significant correlation between IGF-I levels in patients with asthma and collagen thickening of the lamina reticularis of the basement membrane. Furthermore, Yamashita and colleagues (6) found that IGF-I was up-regulated in the BAL of mice after allergen challenge, with a concomitant increase in IGFBP-3 mRNA in lung tissue. Treatment with neutralizing antibody against IGF-I attenuated airway hyperresponsiveness and subepithelial fibrosis, and resulted in a significant decrease in lung tissue levels of IGFBP-3 mRNA (6).
We recognize that there are certain limitations to studies of this nature. First, our sample size may be too small to detect a significant correlation between IGFBP-3 expression and SBM thickness; alternatively, our observations may be reflective of a necessary, but not sufficient, role for IGFBP-3 in the complex mechanism of chronic airway remodeling and subepithelial fibrosis. Second, all subjects in the severe asthma cohort were on ICSs, which may affect IGFBP-3 expression independent of disease severity. However, in vitro exposure of cultured primary airway epithelial cells to corticosteroids does not result in increased IGFBP-3 levels (Figure E3), and IGFBP-3 expression was also significantly increased in non–steroid-treated subjects with mild asthma compared with normal control subjects (Figure 2); thus, increased IGFBP-3 in asthmatic airways is at least partially independent of steroid exposure. Lastly, it is possible that allergen directly activates airway epithelial cell release of IGFBP-3; however, this is not likely, based on the observation that SBP-AG did not induce IGFBP-3 in normal subjects without allergy (Figure E1). Furthermore, addition of allergen to ex vivo cultures of airway epithelial cells did not increase IGFBP-3 levels (Figure E2). Thus, allergen appears to have an indirect effect on epithelial cell production of IGFBP-3.
Based on our findings, we hypothesize the mechanism diagrammed in Figure 5 to incorporate the IGF-I/IGFBP-3 profibrotic pathway into the complex cascade of events leading to subepithelial fibrosis and chronic airway remodeling in allergic asthma. In this model, IGF-I is up-regulated in airway epithelial cells in patients with asthma during allergic inflammatory injury, and is released by injured airway epithelial cells and inflammatory cells (15). IGF-I then stimulates the epithelium to produce IGFBP-3, a potent inducer of fibroblast ECM production (4). This is supported by previous findings that neutralization of IGF-I in mice caused a significant decrease in IGFBP-3 mRNA (6), and our current findings that IGF-I induces production and secretion of IGFBP-3, both apically and basolaterally, by cultured HBEs. IGFBP-3 secreted basolaterally can trigger ECM production by resident fibroblasts, with resulting subepithelial fibrosis, and IGFBP-3 secreted apically into the lumen may cause further injury to airway epithelium and perpetuate the injury–repair cycle, as IGFBP-3 can induce epithelial cell apoptosis (17–19). The injury–repair cascade may be additionally perpetuated by IGF-I through induction of TGF-β secretion by fibroblasts (20), which is a potent stimulus for both ECM production and IGFBP-3 secretion by fibroblasts (4). Additionally, Minshall and colleagues (21) have shown that eosinophils are a source of TGF-β in asthma. Thus, our observed association of IGFBP-3 levels with BAL eosinophil percentages raises the intriguing hypothesis that eosinophils could affect airway remodeling via TGF-β, either directly or indirectly through the induction of IGFBP-3 expression by subepithelial fibroblasts; however, our data do not demonstrate causality. A vicious loop would thereby be established, which serves to augment the response to epithelial injury and perpetuate fibrosis.
Irreversible airway remodeling and resultant loss of lung function is a profound complication of asthma that occurs despite current available therapy, and results in significant morbidity and mortality. This is the first report demonstrating IGF-I induction of IGFBP-3 in human airways and its relationship to airflow limitation. Future studies are warranted to investigate the roles that IGF-I and IGFBP-3 may play in human airway hyperresponsiveness, the induction of proinflammatory mediators, and allergic airway remodeling. Our findings have important clinical implications, as this pathway represents a novel target for the rational design of new pharmacologic therapies.
Supported by National Institutes of Health grants AR050840 (C.A.F.-B.), HL69174, HL56396, and HL088594 (N.N.J.), AI-40600 (S.E.W.), and 1K08HL087932-01A1 (M.M.M.), and an American Lung Association Career Investigator Award (C.A.F.-B.).
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.200810-1555OC on July 16, 2009
Conflict of Interest Statement: K.L.V. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; B.T.G. has served as a speaker from GlaxoSmithKline from 2007 to the present, receiving $4,000 in 2007 and $2,000 in 2008; H.Y. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; M.M.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; E.A.K. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; S.B. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; N.N.J. received $2,000 in 2007–2008 for serving on an advisory board for Genentech and $2,000 in 2007–2008 for serving on an advisory board for GlaxoSmithKline, $4,000 in 2008 for speaking at a conference sponsored by Merck, and works as a consultant for Asthmatix ($8,000 in 2005) and Genentech ($2,000 in 2007–2008). In 2007–2008, N.N.J. received research grants for participating in multicenter clinical trials from GlaxoSmithKline ($150,000), Merck ($494,000), Genentech ($164,000), and MedImmune ($365,000); J.M.P. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; S.E.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; C.A.F-B. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.