Brain-derived neurotrophic factor and airway fibrosis in asthma

Abstract

Airway remodeling in asthma driven by inflammation involves proliferation of epithelial cells and airway smooth muscle (ASM), as well as enhanced extracellular matrix (ECM) generation and deposition, i.e., fibrosis. Accordingly, understanding profibrotic mechanisms is important for developing novel therapeutic strategies in asthma. Recent studies, including our own, have suggested a role for locally produced growth factors such as brain-derived neurotrophic factor (BDNF) in mediating and modulating inflammation effects. In this study, we explored the profibrotic influence of BDNF in the context of asthma by examining expression, activity, and deposition of ECM proteins in primary ASM cells isolated from asthmatic vs. nonasthmatic patients. Basal BDNF expression and secretion, and levels of the high-affinity BDNF receptor TrkB, were higher in asthmatic ASM. Exogenous BDNF significantly increased ECM production and deposition, especially of collagen-1 and collagen-3 (less so fibronectin) and the activity of matrix metalloproteinases (MMP-2, MMP-9). Exposure to the proinflammatory cytokine TNFα significantly increased BDNF secretion, particularly in asthmatic ASM, whereas no significant changes were observed with IL-13. Chelation of BDNF using TrkB-Fc reversed TNFα-induced increase in ECM deposition. Conditioned media from asthmatic ASM enhanced ECM generation in nonasthmatic ASM, which was blunted by BDNF chelation. Inflammation-induced changes in MMP-2, MMP-9, and tissue inhibitor metalloproteinases (TIMP-1, TIMP-2) were reversed in the presence of TrkB-Fc. These novel data suggest ASM as an inflammation-sensitive source of BDNF within human airways, with autocrine effects on fibrosis relevant to asthma.

airway inflammation, hyperresponsiveness, and remodeling are significant features of asthma, resulting in exaggerated airway narrowing, airflow restriction, and dyspnea. Airway remodeling involves epithelial thickening, increased airway smooth muscle (ASM) mass, and increased fibrosis, all leading to thicker, stiffer airways (10, 27, 28, 45, 51, 52). Fibrosis is characterized by modifications in the composition, expression, and deposition of extracellular matrix (ECM) proteins in the airway, with ECM proteins such as collagen-1, collagen-3, laminin, tenascin, fibronectin, and elastin, and the proteolytic activity of matrix metalloproteinases (MMPs) and their inhibitors (TIMPs), being considered important in the airway (15, 32, 37). Accordingly, understanding the mechanisms that contribute to airway fibrosis in asthma is relevant to developing novel strategies to limit a major aspect of this disease not addressed by current therapies.

Although proliferative and fibrotic processes in the asthmatic airway are likely influenced by multiple mechanisms, there is increasing recognition that resident cells including ASM can secrete growth factors and cytokines that have autocrine/paracrine effects on ECM composition and production, as well as cell proliferation (2, 16, 20–22, 24, 25, 28, 36). In this regard, there is increasing interest in neurotrophins as a noncanonical local factor in the airway. Although well documented in the nervous system (12, 26, 49, 54), evidence from recent studies, including our own, show that neurotrophins and their receptors are expressed by structural cells of the lung, more specifically in ASM (43, 44, 47, 50, 59, 61, 64), and thus have the potential to affect all characteristics of airway structure and function (43, 44, 47, 50, 64). In previous studies, we highlighted the role of exogenous BDNF in ASM intracellular calcium ([Ca²⁺]_i) regulation and contractility (1, 44, 47), and cell proliferation (3), where BDNF effects involve the high-affinity neurotrophin receptor TrkB. Relevant to asthma, both sputum and BAL from patients with asthma and allergic rhinitis show elevated BDNF (5, 7, 13, 29, 41, 56, 62), with a recent study suggesting the epithelium as one BDNF source (62). Importantly, we recently showed that ASM is a not only a target, but also a source of BDNF, demonstrating increased BDNF secretion under conditions of inflammation (59) and oxidative insults (53).What is not known is the local production and secretion of BDNF by asthmatic ASM, and the specific role of locally secreted BDNF in airway remodeling. In this regard, the effects of BDNF on ECM components, particularly MMPs, are relevant. There is a complex relationship between neurotrophins and MMPs where MMPs are necessary for extracellular cleavage of secreted pro-neurotrophins (23, 63). Conversely, neurotrophins can regulate MMP expression and function (23, 40, 42). The relationship between MMPs and neurotrophins in asthma is not well understood.

The current study aims to shed light on the role of ASM-derived BDNF as a contributor to ECM protein composition and its dysregulation leading to remodeling during asthma. We hypothesize that BDNF mediates and modulates inflammation effects on fibrosis, thereby promoting airway remodeling.

MATERIALS AND METHODS

Human ASM tissue and cells.

The techniques for isolating human ASM cells from lung samples incidental to patient surgery have been previously described (47, 48). Briefly, pathologically normal areas of 3rd- to 6th-generation bronchi were collected from nonasthmatic and mild-to-moderate asthmatic patients undergoing lung surgery for focal lung disease (typically nondisseminated cancers; infections and emphysema cases excluded). The experimental protocol was approved by the Mayo Clinic Institutional Review Board. Asthma diagnoses and severity were based on standard clinical criteria and pulmonary function testing, as noted in patients’ clinical records. Nonasthmatics were defined as patients with no prior history of obstructive lung disease (including COPD) and were clinically deemed to have otherwise normal lung function. For this study, samples from both male and female adults were utilized, with ages ranging from 45 to 80 yr.

The ASM layer was dissected, and ASM cells were either enzymatically dissociated, or ASM tissue per se utilized for experiments. Cells were subcultured under standard conditions in phenol red-free DMEM/F-12 (Invitrogen) supplemented with 10% FBS and 1% ABAM (GIBCO 15240–062). Prior to experimentation, cells were serum starved for 72 h. Experiments were limited to passages 1–5 of subculture to ensure maintenance of ASM phenotype, which was assessed periodically by Western analysis for smooth muscle actin, myosin, and smooth muscle specific agonist receptors (3).

For ASM tissues, the ASM layer was weighed and resuspended in 6× volume (i.e., 50 mg tissue in 300 μl volume) of sucrose buffer of the following composition (final concentrations in mM): 250 sucrose, 40 Tris·HCl, pH 7.2. Tissues were homogenized on ice using a glass Potter-Elvehjem tissue grinder and centrifuged at 2,000 g to remove cellular debris. Protein concentrations of the homogenized tissues were determined and equalized before analysis. For Western analysis, 25 μg protein was loaded on each gel and protein expression normalized to β-actin levels.

Cell exposures.

Depending on the experimental protocol, cells were exposed to 1 nM recombinant human BDNF (R&D Systems, Minneapolis, MN), 20 ng/ml TNFα (R&D), 50 ng/ml IL-13 (Sigma-Aldrich, St. Louis, MO), and/or 10 µg/ml TrkB-Fc (R&D).

Conditioned media.

Cells were plated at comparable density, grown to ~80% confluence, and serum starved for 72 h. Media removed from these cells was considered conditioned medium for use in experiments. In a subset of studies (media swap experiments), serum-starved ASM cells of nonasthmatic vs. asthmatic patients were exposed to conditioned medium from asthmatic vs. nonasthmatic, ASM respectively, in the presence or absence of an extracellular BDNF chelator, TrkB-Fc (10 µg/ml), for 72 h and harvested for protein expression and enzyme activity.

Western blot analysis.

Standard techniques were applied. Nitrocellulose membranes were blocked with Odyssey Li-Cor blocking buffer and incubated at room temperature for 2 h or overnight at 4°C with specific primary antibodies. Fluorescent secondary antibodies IR Dye 680 nm and 800 nm (Li-Cor NB; cat. nos. 926-32211, 926-32210, 926-32214, 926-68021, 926-68020) were added and imaged using a Li-Cor Odyssey imaging system (Li-Cor Systems, Omaha, NE). Blots were quantified in Image Studio program as relative fluorescent units. Primary antibodies used were MMP-9 sc-6840 (Santa Cruz, Dallas, TX), MMP-2 (8B4) sc-13595 (Santa Cruz), TIMP2 (3A-4) sc-21735 (Santa Cruz), or TIMP2 (D18B7) 57385 (Cell Signaling, Danvers, MA), collagen-1 AB758 (Millipore, Darmstadt, Germany) or ab34710 (Abcam, Cambridge, MA), TrkB ab33655 (Abcam), BDNF (H-117) sc-20981 (Santa Cruz), GAPDH 2118L (Cell Signaling), β-actin A2228 (Sigma-Aldrich).

qPCR.

ASM cells were treated in serum-free media with and without treatment for the specified duration (see results for individual protocols). Cells were rinsed with RNA-grade PBS, trypsinized, and harvested. RNA isolation was completed using an RNeasy minikit (Qiagen, Valencia, CA), and complementary DNA was synthesized using Transcription kit (Roche, Indianapolis, IN). Standard qPCR techniques were used (optimized for Roche LC480 Light Cycler). Primers were purchased from Qiagen: TIMP2 (QT00017759), TIMP1 (QT00084168), MMP-2 (QT00088396), MMP-9 (QT00040040), fibronectin (QT00038024), collagen-1 (QT00072058), collagen-3 (QT00058233), BDNF (QT00235368), and TrkB primer (forward ACT ACT ACA GGG TCG G; reverse CCC TAG CCT AGA ATG TCC). mRNA expression is represented as a fold change from vehicle in treated experiments, or nonasthmatic vehicle.

BDNF ELISA.

ASM cells were grown to confluence in 60-mm plates, then serum starved for 48 h. Cells were treated in fresh serum-free media with TNFα or IL-13 for 72 h. Conditioned medium was collected following treatment and concentrated using Amicon filtration and centrifuged at 4,000 rpm for 30 min (Amicon Ultra-UFC900324). The concentrated conditioned medium was normalized to cell lysate protein concentration and adjusted for total media concentration based on start and end volumes. This conditioned medium was used to measure BDNF by ELISA (R&D).

ECM deposition.

ASM cells were grown to confluence in 96-well black-walled clear-bottom plates, and then serum-starved for 48 h. Cells were treated in fresh serum-free media with BDNF or in a subset of studies with TNFα in the presence or absence of TrkB-Fc for 72 h. In separate experiments, conditioned medium from asthmatic and nonasthmatic cells in the presence or absence of TrkB-Fc (72 h) was collected. Following the treatment period, wells were decellularized using 0.016 N NH₄OH for 30 min, and cellular removal was confirmed visually, ensuring that only ECM remained. ECM deposition was then measured using a semiquantitative immunofluorescence Li-Cor In-Cell Western technique for collagen-1 (ab34710), collagen-3 (ab7778), and fibronectin (sc-9068) (58). ECM intensity was normalized to cell number via MTS assay (Promega, Madison, WI) before cell removal.

Gelatin zymography.

Matrix metalloproteinase (MMP-2, MMP-9) activity was measured using gelatin zymography (17). Concentrated conditioned media were normalized to cell lysate protein concentration, then diluted with Bio-Rad zymogram sample buffer (Bio-Rad, cat. no. 161-0764) and loaded on a 10% gelatin SDS-PAGE gel (cat. no. 345-0079, Bio-Rad, Hercules, CA) for 3 h. Following electrophoresis, the zymogram gel was incubated at room temperature for 1 h on a shaker in renaturing buffer (cat. no. 161-0765, Bio-Rad) and then incubated in development buffer for 36 h in a 37°C water bath. Gels were then stained with Coomassie Blue R-250 for 1 h at room temperature, followed by destaining for 30 min with 50% deionized water, 40% methanol, and 10% glacial acetic acid solution. MMP activity was quantified using the Li-Cor imaging system and data presented as percent increase from vehicle.

Statistical analysis.

All experiments were performed in ASM cells or tissues from five or more patients, although not all protocols were performed on each sample. Statistical differences between experimental groups were analyzed using Student’s t-test or one-way ANOVA followed by post hoc analysis. Data are presented as means ± SE, and statistical significance was established at P < 0.05.

RESULTS

BDNF expression in human ASM.

BDNF mRNA and protein expression were compared between nonasthmatic and asthmatic ASM cells or tissues. Compared with nonasthmatic ASM cells, asthmatic ASM cells showed significantly higher BDNF mRNA (Fig. 1A; P < 0.05) and protein (Fig. 1C; P < 0.05). Similarly, ASM tissue homogenates from asthmatic patients demonstrated a higher baseline expression of BDNF compared with those from nonasthmatic patients (Fig. 1E; P < 0.05). In addition, expression of the full-length (functional) isoform of the high-affinity BDNF receptor TrkB-FL (mRNA and protein) was significantly higher in both cells and tissue homogenates of ASM from asthmatic patients (Fig. 1, B and D, for mRNA and protein from cells, respectively; Fig. 1F for tissue protein; P < 0.05). Overall, these data support the idea that the BDNF/TrkB system is upregulated in asthmatic ASM.

Fig. 1.

Brain-derived neurotrophic factor (BDNF) expression in nonasthmatic and asthmatic airway smooth muscle (ASM). Gene expression of BDNF (A) and the full-length version of its high-affinity receptor tropomyosin-related kinase (TrkB-FL; B) was significantly increased in ASM cells obtained from asthmatic patients compared with cells from nonasthmatic patients. Protein expression of BDNF (C) and TrkB-FL (D) was correspondingly increased in whole cell lysates of asthmatic ASM. Epithelium-denuded ASM tissues from asthmatic patients also demonstrated significant increase in BDNF (E) and TrkB-FL (F) expression compared with ASM tissue from nonasthmatic patients. Means ± SD with individual patient data points [n values (patient numbers)] specified in individual panels. *Significant difference between nonasthmatics and asthmatics (P < 0.05). Asthmatics indicated in gray for clarity.

BDNF and ECM.

In nonasthmatic ASM cells, exogenous BDNF significantly increased collagen-1, collagen-3, and fibronectin (intracellular) expression compared with vehicle (Fig. 2A; P < 0.05 for each protein), with effects on the collagens being more prominent. In separate studies, ASM cells from nonasthmatic vs. asthmatic patients were seeded at comparable densities on 96-well plates. After 72 h of BDNF treatment, ECM deposition was determined via modified In-Cell Western technique (58) (Fig. 2B). Collagen-1 and collagen-3 deposition was higher in ASM from asthmatic patients compared with nonasthmatics (Fig. 2B; P < 0.05) and BDNF significantly increased deposition of these collagens in both groups (Fig. 2B; P < 0.05), with a greater effect in ASM from asthmatics, particularly for collagen-1. Fibronectin deposition was higher in ASM from asthmatics compared with ASM from nonasthmatics (Fig. 2B; P < 0.05), although BDNF per se only trended toward increasing fibronectin deposition in either group. These findings were generally consistent with expression of functional TrkB-FL receptors and an effect of BDNF on ECM production/deposition by ASM of both nonasthmatics and asthmatics.

Fig. 2.

BDNF increases extracellular matrix (ECM) expression and deposition by ASM: Forty-eight-hour exposure to BDNF significantly increased expression of major ECM proteins collagen 1, -3, and fibronectin in nonasthmatic ASM cells compared with vehicle alone (A). ECM protein deposition was measured via modified In-Cell Western technique (B; essentially a semiquantitative immunofluorescence measurement; see materials and methods). BDNF exposure induced increases in collagen-1, -3, and fibronectin deposition in both nonasthmatic and asthmatic ASM cells, with a relatively greater effect of BDNF in asthmatic ASM. Means ± SD with individual patient data points [n values (patient numbers)] specified in individual panels. #Significant BDNF effect; *significant difference between nonasthmatics and asthmatics (P < 0.05). Asthmatics indicated in gray for clarity.

To explore the mechanisms underlying BDNF effects on fibrosis, MMP-2 and -9 as well as TIMP-1 and -2 expression and activity were measured. The relevance of MMPs and TIMPs further lies in their ability to cleave extracellular pro-BDNF (the predominantly secreted form). Expression of MMP-2 and -9 and TIMP-1 and -2 was determined via Western blot of conditioned media from BDNF-treated ASM cells (Fig. 3, A and B). Here, ASM cells were grown to confluence and then serum-deprived for 72 h, and the culture medium was concentrated and considered as conditioned medium. Western analysis found no significant pattern of changes in intracellular MMP-2, MMP-9, TIMP-1, or TIMP-2 protein expression per se (Fig. 3B), although there was a trend toward decreased TIMPs and increased MMPs. However, the ratios of MMP to TIMP (MMP-9/TIMP-1 and MMP-2/TIMP-2) were significantly increased by BDNF exposure (Fig. 3B; P < 0.05), suggesting overall increased MMP enzymatic activity. Indeed, gelatin zymography showed significant increases in MMP-2 and MMP-9 activities in conditioned media following BDNF exposure (Fig. 3, C and D; P < 0.05).

Fig. 3.

BDNF effects on ECM regulatory enzymes. Nonasthmatic human ASM cells treated with BDNF showed no significant increase in matrix metalloproteinases-2 or -9 (MMP-2, MMP-9) levels or decreases in corresponding tissue inhibitor of metalloproteinases-1 or -2 (TIMP-1, TIMP-2) in conditioned medium (A, left panel of B). However, MMP-9/TIMP-1 and MMP-2/TIMP-2 ratios significantly increased with BDNF treatment (right panel of B), suggesting increased overall enzyme activity. Gelatin zymography correspondingly demonstrated an increase in MMP-2 and MMP-9 activity with BDNF treatment in conditioned media (C, D). Means ± SD with individual patient data points [n values (patient numbers)] specified in individual panels. #Significant BDNF effect (P < 0.05).

Inflammation induces BDNF secretion in ASM.

Concentrated conditioned medium from nonasthmatic and asthmatic ASM cells were used to quantify BDNF release with ELISA. ASM from asthmatics showed increased BDNF secretion at baseline compared with nonasthmatics (Fig. 4; P < 0.05). Exposure of TNFα significantly increased BDNF secretion in ASM of both asthmatics and nonasthmatics, whereas IL-13 did not significantly increase BDNF secretion in either group (Fig. 4; P < 0.05).

Fig. 4.

Proinflammatory cytokines induces BDNF secretion in ASM. Asthmatic ASM secrete more BDNF at baseline compared with nonasthmatic ASM measured from concentrated conditioned medium using ELISA and normalized to cell lysate protein concentration. Furthermore, exposure of TNFα significantly increased BDNF secretion in ASM, whereas IL-13 did not significantly increase BDNF secretion in nonasthmatic or asthmatic ASM. Means ± SD with individual patient data points [n values (patient numbers)] specified in individual panels. #Significant TNFα effect; *significant difference between nonasthmatics and asthmatics (P < 0.05). Asthmatics indicated in gray for clarity.

BDNF mediates inflammation-induced ECM production in ASM.

The above protocols demonstrated the effect of exogenous BDNF on ASM-derived ECM, and that ASM is a potential source of BDNF as well as a target. Accordingly, we next tested whether ASM-derived BDNF can have autocrine effects in the context of inflammation. Human ASM cells were treated with TNFα in the presence or absence of the extracellular BDNF chelator TrkB-Fc for 72 h, and ECM deposition of collagen-1, collagen-3, and fibronectin was quantified using the In-Cell Western technique. Both nonasthmatic and asthmatic cells significantly deposited more collagen-1, collagen-3, and fibronectin in response to TNFα treatment, and such significant increases were attenuated in the presence of TrkB-Fc (Fig. 5, A–F; P < 0.05 for TrkB-Fc effect). TrkB-Fc by itself had no consistent influence on ECM protein production.

Fig. 5.

BDNF mediates inflammation-induced ECM deposition in ASM. Exposure to TNFα for 72 h significantly increased ECM protein (collagen-1, -3, and fibronectin) deposition in both nonasthmatic (A, C, E) and asthmatic (B, D, F) ASM cells. Attenuation of BDNF through chelation of extracellular BDNF using the chimeric protein TrkB-Fc blunted TNFα-mediated increase in collagen-1 (A, B), collagen-3 (C, D), and fibronectin deposition (E, F). Means ± SD with individual patient data points [n values (patient numbers)] specified in individual panels. #Significant TNFα effect; %significant TrkB-Fc effect (P < 0.05). Asthmatics indicated in gray for clarity.

TNFα exposure also resulted in significant increases in MMP-9 to TIMP-1 ratio, an effect attenuated in the presence of TrkB-Fc (Fig. 6A; P < 0.05). Increased MMP-9 activity with TNFα exposure (as detected by gelatin zymography) was also blunted by TrkB-Fc (Fig. 6B; P < 0.05). Similar blunting effects of TrkB-Fc treatment were noted for MMP-2 to TIMP-2 ratio (Fig. 6C; P < 0.05) and MMP-2 activity (Fig. 6D; P < 0.05). Overall, these data were consistent with the idea of an autocrine BDNF role in inflammation-induced enhancement of ECM.

Fig. 6.

BDNF mediates inflammation-induced MMP regulation in ASM: MMP and TIMP expression in conditioned medium from nonasthmatic human ASM cells was significantly altered by TNFα exposure. Furthermore, MMP/TIMP ratios also increased significantly with TNFα exposure (A, C). Gelatin zymography of conditioned media was used to determine MMP activity. TNFα exposure significantly increased MMP-9 and MMP-2 activity (B, D). BDNF chelation via TrkB-Fc significantly reduced TNFα-mediated effect on MMP activity. Means ± SD with individual patient data points [n values (patient numbers)] specified in individual panels. #Significant TNFα effect; %significant TrkB-Fc effect (P < 0.05).

To further investigate the role of endogenous BDNF in regulating ECM protein expression, asthmatic vs. nonasthmatic conditioned supernatant medium (ACM vs. NCM, respectively) was obtained from ASM samples in the presence vs. absence of TrkB-Fc. The logic was that if BDNF has an autocrine effect, and asthmatic ASM produce more BDNF (and/or are more sensitive to BDNF due to increased TrkB-FL as suggested by the data above), then conditioned medium from asthmatic ASM, i.e., ACM, should stimulate ECM production, particularly in naive nonasthmatic ASM, and conversely, asthmatic ASM should respond to NCM (which contain some BDNF) compared with vehicle alone. However, the presence of BDNF-chelating TrkB-Fc should blunt the effects of conditioned media. Therefore, nonasthmatic cells were treated with ACM, and conversely, asthmatic ASM cells were treated with NCM. Parallel sets of exposures included TrkB-Fc. Expression of collagen-1, -3, and fibronectin by ASM cells was measured.

The presence of TrkB-Fc significantly attenuated the enhanced expression of ECM proteins by nonasthmatic ASM cells exposed to ACM (Fig. 7, A, C, and E; P < 0.05 for TrkB-Fc effect). Here, the enhancing effect of ACM was compared with baseline ECM production by nonasthmatic cells (dotted black line in Fig. 7, A, C, and E). Similarly, asthmatic cells treated with NCM showed increased ECM production compared with baseline production by asthmatic cells (dotted gray line in Fig. 7, B, D, and F). The presence of TrkB-Fc resulted in a significant decrease in ECM protein expression compared with NCM treatment alone (Fig. 7, B, D, and F; P < 0.05).

Fig. 7.

Conditioned medium from ASM potentiates BDNF-induced ECM expression. ECM protein (collagen-1, -3, and fibronectin) expression was measured from nonasthmatic ASM cells treated for 72 h with asthmatic conditioned medium (ACM) with and without TrkB-Fc, with the idea that if asthmatic ASM produces/secretes more BDNF that modulates ECM, then BDNF chelation with TrkB-Fc should attenuate the effects of the ACM on production of ECM proteins. Indeed, on the one hand ACM increased ECM production in nonasthmatic ASM cells compared with baseline levels (black dotted lines in A, C, E). Conversely, TrkB-Fc attenuated such increases in ECM protein expression in nonasthmatic ASM induced by the ACM. In a parallel set of studies using asthmatic ASM cells, treatment with nonasthmatic conditioned media (NCM) also enhanced ECM production (B, D, F) compared with baseline production by such asthmatic ASM cells (gray dotted line), supporting the idea of nonasthmatic cells also producing BDNF that can act on TrkB-FL in asthmatic ASM cells (which are higher compared with nonasthmatic cells; Fig. 2). Indeed, TrkB-Fc also attenuated NCM-induced increases in production of collagens and fibronectin. Means ± SD with individual patient data points [n values (patient numbers] specified in individual panels. @Significant ACM or NCM effect; %significant TrkB-Fc effect (P < 0.05). Asthmatics indicated in gray for clarity. Note different y-axis scales for the different panels.

DISCUSSION

The asthmatic airway demonstrates hyperactivity and remodeling with the latter showing thickening of ASM and epithelial layers and altered as well as increased ECM protein within the airway layers (8, 27, 28, 45, 51, 52). Several studies have explored mechanisms that contribute to increased cell proliferation in this context (reviewed in 45, 46), although fewer have explored mechanisms that regulate ECM. Although multiple inflammatory mediators (derived from immune cells, epithelium, and even ASM) could regulate ECM in the airway, there is increasing interest in the role of locally produced factors with autocrine/paracrine effects, particularly those enhanced in the presence of inflammation. Emerging data from studies including our own have highlighted such a potential role for exogenous BDNF in the context of [Ca²⁺]_i regulation and cell proliferation (3, 53, 62). The present study determined that ASM-derived BDNF per se contributes to airway remodeling by enhancing ECM production and deposition (Fig. 8).

Fig. 8.

ASM-derived BDNF mediates/modulates inflammation effects on airway fibrosis. ASM is a potentially major source of BDNF in the airway. Inflammation and asthma enhance BDNF release by ASM, which can contribute to stimulating a positive-feedback loop of airway fibrosis via effects on collagen-1, -3, fibronectin, MMPs, and TIMPs. Accordingly, targeting ASM BDNF may be a novel approach attenuating remodeling in diseases such as asthma.

Enhanced ECM production and deposition around the ASM suggest a role for ASM-derived ECM proteins per se in the remodeling process, as demonstrated in several in vitro studies using human ASM cells (8, 30, 57). In this regard, our approach using human ASM cells from asthmatics vs. nonasthmatics to examine the role of BDNF in modulating ECM secretion and deposition is similar to previous models. Furthermore, if remodeling in asthma involves enhanced ASM proliferation, then the potential exists for even greater ASM-derived ECM formation. Here, ASM has been shown to produce ECM proteins such as fibronectin, tenascin, and collagens, as well as matrix regulatory factors such as MMPs and their inhibitors (TIMPs) (9, 22, 45, 46). In turn, ECM proteins have also been shown to modulate cell behaviors such as survival, differentiation, proliferation, and migration (8, 46). Thus the relevance of ECM protein in the airway lies beyond stiffness alone.

BDNF was initially recognized in the nervous system, but several neurotrophins including BDNF and its receptors (particularly TrkB) have been demonstrated in nonneuronal tissues, including lung (43, 44, 47, 50, 61, 62, 64). What are less clear are the sources vs. targets of BDNF in the lung. Intracellular BDNF does not necessarily reflect extracellular active levels, given that neurotrophins are secreted in a pro-form which in itself can be active, but usually requires extracellular cleavage via MMPs, tPA, and other proteases (23, 63). In this regard, our past studies (59, 61) showing ASM-derived levels of BDNF comparable to circulating and brain levels (34, 55, 65) suggest that ASM is not only a substantial source of BDNF, but given a functional TrkB receptor, also a target.

The influence of BDNF on ASM includes significantly increasing airway contractility and hyperreactivity through intracellular calcium regulatory pathways, as well as increased ASM proliferation (3, 44, 48, 53). Furthermore, BDNF has also been found to be elevated in the sputum and/or bronchoalveolar lavage from patients with allergic rhinitis, chronic cough, or asthma (6, 7, 13, 29, 41). Recently, Watanabe et al. (62) reported the association of BDNF in asthmatic patient sputum to be linked to airway epithelium secretions and asthma severity. In the present study, we establish asthmatic ASM as a potentially major contributor of BDNF expression and secretion in the airway (Fig. 8).

Previously, we have reported the proliferative effect of BDNF in ASM which was comparable to the proinflammatory cytokine-induced effect (3). This forms the basis for the current study in which the role of BDNF in the fibrosis aspect of remodeling was explored. In tumorigenesis, MMP-9, -2, -3. and -14 are overexpressed to allow for significant remodeling by the tumor to occur (33). Similarly, sputum MMP-2 and -9 are increased in severe asthma (4, 11, 14, 35, 39). Consistent with the above findings, a previous study showed that BDNF stimulates MMP-9 secretion (18) which is consistent with our current findings. Additionally, our novel data link BDNF to increased MMP-2 activity and TIMP-2 expression as a contributor to airway fibrosis in the presence of inflammation and in asthma.

In response to insults such as cigarette smoke, and proinflammatory cytokines such as TNFα, ASM has been shown to increase BDNF release (53, 59). The novel data from the current study show significantly increased BDNF release from asthmatic ASM at baseline, that is further stimulated by Th1 signaling (via TNFα) with a lesser extent by Th2 signaling (via IL-13). These differential levels of endogenous production of ASM BDNF by Th1 vs. Th2 cytokine exposure could be important for the facilitation of BDNF-induced airway fibrosis and the role of ASM per se. Chelation of BDNF via TrkB-Fc in the presence of inflammation reduced ECM deposition, indicating the contribution of endogenously produced BDNF in airway fibrosis, and supports the idea that BDNF may contribute to ECM production and deposition: processes relevant to the asthmatic phenotype (19, 27, 31, 45). This concept is further supported by our data showing increased ECM production in the presence of asthmatic conditioned medium, and attenuation of the increase by BDNF chelation. Accordingly, we posit that our data support a potentially novel target for blunting remodeling and fibrosis in the asthmatic airway by interrupting the autocrine cycle involving BDNF.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grants HL-088029 (Y. S. Prakash), HL-088029-S3 (R. D. Britt, Jr.), HL-56470 (Y. S. Prakash), and HL-123494 (V. Sathish).

DISCLOSURES

No conflicts of interest, financial or otherwise are declared by the authors.

AUTHOR CONTRIBUTIONS

M.R.F., V.S., C.M.P., and Y.S.P. conceived and designed research; M.R.F., V.S., L.J.M., S.W., and M.A.T. performed experiments; M.R.F., V.S., L.J.M., S.W., and Y.S.P. analyzed data; M.R.F., L.J.M., and Y.S.P. prepared figures; M.R.F. and Y.S.P. drafted manuscript; M.R.F., V.S., L.J.M., S.W., R.D.B.J., M.A.T., C.M.P., and Y.S.P. approved final version of manuscript; V.S., R.D.B.J., M.A.T., and Y.S.P. interpreted results of experiments; C.M.P. and Y.S.P. edited and revised manuscript.