In asthma, mast cells infiltrate the airway smooth muscle cell layer and secrete proinflammatory and profibrotic agents that contribute to airway remodeling. To study the effects of mast cell activation on smooth muscle cell-dependent matrix contraction, we developed coculture systems of human airway smooth muscle cells (HASM) with primary human mast cells derived from circulating progenitors or with the HMC-1 human mast cell line. Activation of primary human mast cells by IgE receptor cross-linking or activation of HMC-1 cells with C5a stimulated contraction of HASM-embedded collagen gels. Contractile activity could be transferred with conditioned medium from activated mast cells, implicating involvement of soluble factors. Cytokines and proteases are among the agents released by activated mast cells that may promote a contractile response. Both IL-13 and IL-6 enhanced contraction in this model and the activity of IL-13 was ablated under conditions leading to expression of the inhibitory receptor IL-13Rα2 on HASM. In addition to cytokines, matrix metalloproteinases (MMPs), and serine proteases induced matrix contraction. Inhibitor studies suggested that, although IL-13 could contribute to contraction driven by mast cell activation, MMPs were critical mediators of the response. Both MMP-1 and MMP-2 were strongly expressed in this system. Serine proteases also contributed to contraction induced by mast cell-activating agents and IL-13, most likely by mediating the proteolytic activation of MMPs. Hypercontractility is a hallmark of smooth muscle cells in the asthmatic lung. Our findings define novel mechanisms whereby mast cells may modulate HASM-driven contractile responses.
Airway smooth muscle cell (ASM)3 hyperplasia and hypercontractility are hallmarks of asthma and contribute to airway remodeling (1). Infiltration of the smooth muscle layer by mast cells is a characteristic feature of asthma, which does not occur in related lung diseases (2, 3). Once present in the smooth muscle cell layer, activated mast cells may contribute to the hypercontractility of airway smooth muscle (4) and could promote chronic inflammatory and remodeling changes, such as fibrosis and airway narrowing, in the asthmatic airway (5). Mast cells infiltrating the airway smooth muscle in asthma show signs of activation, with evidence of degranulation (6, 7). Numerous mast cell-derived mediators have the potential to contribute to ASM contractility, including cytokines, chemokines, and inflammatory mediators such as histamine, leukotrienes, and PGs (8, 9, 10). Mast cell-derived proteases, including the serine protease tryptase (11), and matrix metalloproteases (MMPs) (12) may also contribute to escalation of inflammatory responses and airway remodeling in asthma. In turn, airway smooth muscle-derived mediators, including TGF-β, PGE2, and soluble and membrane-bound stem cell factor (SCF) (5, 7) can modulate the activation state of infiltrating mast cells.
A system of smooth muscle cells embedded in a collagen gel matrix has been used to model the rapid contraction associated with bronchospasm. This response can be triggered by exposure of ASM-embedded gels to methacholine, ATP, or histamine and occurs within minutes (9, 13, 14). The rapidity of this response contrasts with the more gradual, but ultimately more pronounced, collagen gel contraction induced by the major mast cell serine protease tryptase and by MMPs. This response occurs over several days and models matrix contraction as may occur in processes such as wound healing or tissue remodeling (12, 15, 16). This smooth muscle-driven matrix contraction may involve smooth muscle cell differentiation to a more contractile phenotype, increased cell-adhesive interactions, or reorganization of the collagen matrix itself, but is not a direct measure of smooth muscle cell contraction (15, 16, 17).
We have recently used a contraction model with collagen gels coembedded with human lung fibroblasts and mast cells to study cellular interactions contributing to fibrosis (18). We now adapt this system to model mast cell interactions with human airway smooth muscle cells (HASM) and explore the consequences of human mast cell activation in coculture with smooth muscle cells in the collagen gel matrix. The human mast cell line HMC-1 was used to define the system. HMC-1 can be activated by exposure to the anaphylatoxin C5a to undergo calcium mobilization (19) and chemotaxis (20), but lack appreciable expression of FcεRI. To confirm our observations in a primary system and to extend them to include consequences of IgE-mediated mast cell activation, all key findings were reproduced with primary human mast cells (pHMC) derived from circulating progenitors. pHMC undergo a relatively weak response to C5a (21), but exhibit strong IgE-mediated activation and induced potent contraction in the HASM coculture model.
This model was used to examine the role of serine proteases, MMPs, and cytokines in the collagen gel contractile response and to address potential cooperative interactions between these mediators. Both activated mast cells (22, 23) and smooth muscle cells (14, 24, 25, 26) can produce MMPs, which are released from the cell in inactive form and activated by proteolytic processing. Mast cell-derived tryptase has been implicated in the proteolytic activation of MMPs (27, 28, 29), and mast cell- or smooth muscle cell-derived cytokines, including IL-6 and IL-13, can drive MMP expression (14, 30). Therefore, we propose that cytokines, serine proteases, and MMPs should all contribute to the contraction of HASM-containing collagen gels following activation of coembedded mast cells, with cytokines driving MMP generation, serine proteases promoting MMP activation, and MMPs ultimately driving the contractile response.
Materials and Methods
HASM were purchased from ScienCell Research Laboratories and maintained in a defined smooth muscle growth medium (Lonza). The human mast cell line HMC-1 (gift from Dr. J. Butterfield, Mayo Clinic, Rochester, MN) was cultured in IMEM with 10% defined iron-supplemented calf serum (HyClone) and 1.2 mM α-thioglycerol (Sigma-Aldrich). All cells were maintained at 37°C/5% CO2.
Primary human mast cell generation
To generate primary mast cells, CD34-positive progenitors were purified from heparinized human blood (Massachusetts General Hospital, Boston, MA) using CD34-coated magnetic beads (Miltenyi Biotec). Cells were seeded in IMEM containing l-glutamine (2 mM), 30% charcoal-treated FBS (Sigma-Aldrich), 50 μg/ml iron-saturated holotransferrin (Sigma- Aldrich), 1× penicillin-streptomycin (Sigma-Aldrich), 10−5 4 cells/ml. For the first 2 wk of culture, cells were maintained at a density of 1.5 × 105 +CD14−. Cultures were used for assays at 7–9 wk and were >95% mast cells.
Type I collagen gel contraction
Collagen lattices were prepared by mixing neutralized bovine type I collagen (Organogenesis) with HASM at 2.5 × 105 cells/ml in 24-well plates. HMC-1 (2.5 × 105 cells/ml) or pHMC (1 × 105 cells/ml) were added to the mixture as indicated, and gels were solidified overnight at 37°C/10% CO2
In some cases, collagen gels were incubated with conditioned medium from activated mast cells. To generate conditioned medium, HMC-1 cells were seeded in 24-well plates at 1 × 106 cells/well in 1 ml of serum-free IMEM containing α-thioglycerol, and treated for 24 h as indicated. pHMC were seeded in a round-bottom 96-well plate at 5 × 104 cells/well in 200 μl of serum-free IMEM containing 50 ng/ml SCF and treated for 24 h as indicated. For each gel, 500 μl (HMC-1) or 200 μl (pHMC) of the resulting supernatant was added to the incubation medium.
Mast cell activation
HMC-1 cells were incubated for 24 h at 37°C in medium containing 10 nM C5a (Sigma-Aldrich). Supernatants were assayed for IL-3 content by ELISA (BioSource International). For pHMC activation studies, cells were incubated overnight with 0.1 μg/ml human IgE (Millipore), then treated with 10 μg/ml anti-human IgE (Kirkegaard & Perry Laboratories). Supernatants were collected after 30 min. Degranulation was quantified as histamine, assayed by ELISA (Immunotech), or β-hexosaminidase release, assayed by mixing with an equal volume of 1.3 mg/ml p-nitrophenyl-N-acetyl-β-d glucosaminide (Sigma-Aldrich) in 0.08 M sodium citrate (pH 4.5), with mast cell supernatants, incubating overnight at 37°C, and reading absorbance at 405 nM. Tryptase activity was assayed using a kit from Millipore. Leukotriene C4/D4/E4
Collagen gel contraction data are presented as the mean ± SD of 3–10 experiments, each containing three to four replicate gels. Student’s t test was used to analyze differences between groups. A difference of p < 0.05 was considered statistically significant.
Activation of HMC-1 induces contraction of HASM-embedded collagen gels
HASM cells were embedded in collagen gels in the presence or absence of mast cells. In asthma, mast cell infiltration has been quantitated at a mean of 5–10 mast cells/mm2 of smooth muscle (range, 0–33) or approximately one mast cell per four smooth muscle cells (3, 31). In the collagen gels, HMC-1 and HASM were coembedded at a cell number ratio of 1:1, which would fall within the range of mast cell infiltration in ASM. The HMC-1 cell activation response to the anaphylatoxin C5a was confirmed by assaying production of IL-3 (Fig. 1⇓a). When HASM were treated with C5a in the absence of mast cells, no significant collagen gel contraction was observed (Fig. 1⇓b). When HMC-1 were embedded in the collagen gels along with HASM, a detectable contractile response was observed (Fig. 1⇓b; p < 0.001). Significant contraction above this level was seen in response to mast cell activation with C5a (Fig. 1⇓b; p < 0.005 for treatment with C5a compared with medium). Collagen gels embedded with mast cells alone, in the absence of HASM, did not undergo detectable contraction under these conditions.
To address whether this increased contraction upon mast cell activation involved generation of soluble mediators, HMC-1 cells were exposed to C5a for 24 h in the absence of HASM and the supernatant was collected. HASM embedded in collagen gels were incubated with the supernatant of resting or C5a-activated HMC-1 cells. Increased contraction was observed upon exposure to supernatants of activated HMC-1 cells (Fig. 1⇑c; p < 0.005 for supernatants of C5a-activated HMC-1 compared with control supernatants). These observations demonstrate that the contractile response could be at least partially transferred with agents secreted by activated mast cells.
Activation of primary human mast cells induces contraction of HASM-embedded collagen gels
A major mechanism for mast cell activation is cross-linking of the cell surface FcεRI by IgE-bound allergen. HMC-1 do not express appreciable FcεRI and lack this response. To confirm that IgE-mediated mast cell activation could influence contractility of HASM-embedded collagen gels, we investigated pHMC derived from progenitors in peripheral blood (Fig. 2⇓a). pHMC responded to IgE receptor cross-linking with cytokine production (Fig. 2⇓b), degranulation (Fig. 2⇓c), leukotriene synthesis (Fig. 2⇓d), and tryptase release (Fig. 2⇓e). To investigate the effects on contraction, pHMC were coembedded with HASM in collagen gels. Because of their limited availability, the pHMC studies were done with a ratio of 1 pHMC per 2.5 HASM cells, which is lower than the 1:1 ratio used with HMC-1, but adequate to induce a robust signal. In the absence of pHMC, HASM embedded in collagen gels were exposed to IgE alone or to IgE plus anti-IgE, and no contractile response was observed. In contrast, when pHMC were coembedded in the collagen gels, treatment with IgE plus anti-IgE or with the conditioned medium of IgE-activated pHMC resulted in increased gel contraction (Fig. 2⇓f; p < 0.005 for IgE-activated compared with resting pHMC or conditioned medium). Although monomeric IgE has been reported to induce some activation responses in mast cells (32), cross-linking was required for induction of HASM contraction in this system (Fig. 2⇓f).
Cytokines promote mast cell-dependent HASM gel contraction by a mechanism involving protease activation
In a coculture system of mast cells and smooth muscle cells, several agents have the potential to influence contractility. We first examined the contraction-inducing activity of cytokines. In the absence of mast cells, TGF-β directly induced contraction of HASM-containing gels over 48 h, but the cytokines IL-6 and IL-13 had no direct effect (Fig. 3⇓a). In the presence of HMC-1 (Fig. 3⇓a; p < 0.0005) or pHMC (Fig. 3⇓b; p < 0.05), however, both IL-6 and IL-13 produced significant increases in gel contraction over 48 h. These findings show that, in the presence of mast cells, IL-6, IL-13, and mast cell-activating agents induced contraction of HASM-containing collagen gels.
Mast cell-mediated contraction of HASM-containing gels is MMP dependent
We next addressed the role of MMPs and serine proteases in promoting this mast cell-dependent response. Both the broad spectrum MMP inhibitor batimastat and the serine protease inhibitor leupeptin blocked HMC-1-dependent contraction of HASM-containing gels induced by IL-6, IL-13, or C5a (Fig. 3⇑c) and blocked the pHMC-dependent contraction induced by IgE receptor cross-linking (Fig. 3⇑d). The inhibitory effects of batimastat and leupeptin on contraction did not result from a shift in kinetics, as inhibition was seen at every time point measured for the contraction driven by pHMC activation (Fig. 3⇑e). In contrast to cytokines, MMPs, and serine proteases, neither histamine nor cysteinyl leukotrienes promoted contraction of HASM-containing collagen gels under these experimental conditions (Fig. 3⇑f).
MMPs and serine proteases induce contraction of HASM-containing gels
To further explore the ability of MMPs and serine proteases to contribute to the contraction response, collagen gels embedded with HASM in the presence or absence of HMC-1 were treated directly with catalytically active MMPs or with tryptase. MMPs 1, 2, 9, and 12 all directly induced contraction of collagen gels embedded with HASM, even in the absence of mast cells (Fig. 4⇓, a–d; p < 0.002 for treatment of HASM-containing gels with MMP compared with medium). These responses were blocked by the MMP inhibitor batimastat, but not by the serine protease inhibitor leupeptin. When resting HMC-1 were coembedded with HASM, the gels underwent contraction resulting from the presence of HMC-1. There was a further response to MMPs 1, 2, 9, and 12 above this level, resulting in additive contractile effects due to the presence of HMC-1 and treatment with MMPs (Fig. 4⇓, a–d; p < 0.002 for treatment of HASM plus HMC-1-containing gels with MMP compared with medium). As for HASM alone, these responses were blocked by batimastat, but not by leupeptin. These observations support direct contraction of HASM-containing gels in response to MMPs 1, 2, 9, and 12 and indicate that mast cells may enhance, but are not required for, HASM contractile responses to catalytically active MMPs.
Because HMC-1-dependent contraction responses to IL-6, IL-13, and C5a were blocked by leupeptin in addition to batimastat (Fig. 3⇑c), we also examined the ability of the major mast cell serine protease tryptase to directly induce contraction of HASM-containing gels. Human lung tryptase promoted contraction of collagen gels embedded with HASM alone (Fig. 4⇑e; p < 0.002). As expected, this response was blocked by the serine protease inhibitor leupeptin. Surprisingly, however, it was also blocked by the MMP inhibitor batimastat (Fig. 4⇑e). The ability of batimastat to block this response is consistent with an MMP dependence of the contractile activity induced by tryptase.
As noted above, the presence of HMC-1 coembedded with HASM produced an appreciable gel contraction, even in the absence of additional stimulation. In contrast to observations with MMPs 1, 2, 9, and 12, tryptase did not produce any additional contractile response above this level in gels coembedded with HMC-1 (Fig. 4⇑e). Thus, exogenously added tryptase induced contraction of HASM-containing gels in the absence of HMC-1, but had no further contractile activity in the presence of HMC-1 (Fig. 4⇑e).
Tryptase activation of pro-MMPs induces contraction
In collagen gels embedded with HASM alone, tryptase-mediated contraction was inhibitable by batimastat (Fig. 4⇑e), suggesting that tryptase activity involved MMP activation. To address whether serine protease activity could promote MMP-dependent gel contraction, HASM-embedded collagen gels were treated with the inactive precursor (pro) forms of MMP-1 and MMP-2. A partial contractile response was observed comparable to that produced by tryptase alone (Fig. 5⇓a; p < 0.005 for pro-MMP compared with medium). Tryptase enhanced the contraction in response to either pro-MMP-1 or pro-MMP-2 (Fig. 5⇓a; p < 0.005 for pro-MMP and tryptase compared with pro-MMP alone). In accordance with the increased contraction, the processed forms of MMP-1 and MMP-2 were detected following exposure of pro-MMP-1 or pro-MMP-2 to tryptase and were reduced by addition of leupeptin (Fig. 5⇓b).
Induction of IL-13Rα2 blocks HASM contraction responses
In the presence of mast cells, IL-13-induced contraction of HASM was dependent on both MMPs and serine proteases (Fig. 3⇑c). In the next series of experiments, we further explored the effects of IL-13 on mast cell-dependent contraction of HASM. IL-13 responses are mediated through a receptor consisting of IL-13Rα1/IL-4Rα chains. An additional binding chain, IL-13Rα2, is inducible on the surface of several cell types, including fibroblasts and smooth muscle cells, and antagonizes IL-13 responses on these cell types (33, 34). Recently, however, it has been proposed that, under certain conditions, IL-13Rα2 may act as an agonist, rather than antagonist, of IL-13 responses (35). To address whether IL-13Rα2 expression would be activating or inhibitory in this system, we treated HASM with a combination of IL-13 and TNF-α. Under these conditions, IL-13Rα2 could be detected on the cell surface by flow cytometry (Fig. 6⇓a) and increased IL-13Rα2 transcript was found by real-time PCR (Fig. 6⇓b). IL-13Rα1 transcript was induced by TNF-α, but was not further increased by the combination of IL-13 and TNF-α (Fig. 6⇓c), and there was constitutive expression of IL-4Rα under all treatment conditions (Fig. 6⇓d). Under conditions associated with induction of IL-13Rα2 on the cell surface (IL-13 plus TNF-α), HMC-1-dependent contraction of HASM-containing gels was significantly reduced compared with that seen in the presence of IL-13 alone (Fig. 6⇓e; p < 0.001 for IL-13 plus TNF-α compared with IL-13 alone). Thus, induction of IL-13Rα2 expression was associated with reduced IL-13- dependent HASM gel contractile responses in this system, consistent with antagonist function.
Mast cells are a major source of IL-13. To determine whether IL-13 played a role in the contraction-inducing activity of C5a-activated HMC-1 in this system, sIL-13Rα2-Fc was added to collagen gels coembedded with HASM and HMC-1. The potent contractile response to C5a was not affected by control human Ig, but was significantly antagonized by sIL-13Rα2-Fc (Fig. 6⇑f; p < 0.01 for C5a plus sIL-13Rα2-Fc compared with C5a plus control Ig or C5a alone).
Cytokine effects on MMP content of HASM-HMC-1 cocultures
Both mast cells and smooth muscle cells can produce MMPs. To examine the range of MMPs produced in this system, gels were seeded with HASM in the presence or absence of HMC-1. The medium surrounding the gels was collected, concentrated 10-fold, and MMP content examined by the Fluorokine multiplex assay. High levels of MMP-1 and MMP-2 were seen, but no detectable MMP-3, MMP-7, MMP-8, MMP-9, MMP-12, or MMP-13 (data not shown).
Gelatin zymography confirmed the strong expression of pro-MMP-2 in the HASM cultures. The MMP-2 content was marginally increased following IL-13 treatment of gels containing HASM and HMC-1 cells, but no bands corresponding to the low molecular weight processed forms of MMP-2 were seen (Fig. 7⇓a). Strong expression of MMP-1 was confirmed by Western blot (Fig. 7⇓b). Both the proform and the proteolytically processed form, indicative of MMP-1 activation, increased in gels embedded with HASM and HMC-1 upon treatment with IL-13, TNF-α, or both (Fig. 7⇓, b and c).
MMP content of HASM-HMC-1 cocultures following mast cell activation
Following C5a activation of HMC-1, MMP-2 expression was increased in cocultures of HASM and HMC-1 and expression of MMP-9 was marginally detectable (Fig. 8⇓a). The processed form of MMP-1 was readily detectable in medium surrounding collagen gels coembedded with HASM and HMC-1 (Fig. 8⇓, b and c). Both the proform and processed MMP-1 were also found in extracts of the gels themselves (Fig. 8⇓, b and c), indicating association with the collagen gel matrix. In gels seeded with HASM only, pro-MMP-1 could be found, but the processed form was not apparent (Fig. 8⇓, b and c).
In cultures containing pHMC only, high levels of MMP-9 were found by zymography, but little MMP-2 was expressed (Fig. 9⇓a). In contrast, cultures of HASM only contained high levels of MMP-2, but no detectable MMP-9. In medium surrounding collagen gels coembedded with HASM and pHMC, trace amounts of MMP-9 were seen, but pro-MMP-2 was abundantly expressed and the processed form was readily detectable (Fig. 9⇓a). Neither MMP-2 nor MMP-9 expression was modulated by IgE receptor cross-linking. MMP-1 was also expressed in the cocultures of HASM and pHMC (Fig. 9⇓b). The processed form of MMP-1 was readily detectable and extracts of gels coembedded with HASM and pHMC revealed processed MMP-1 as the predominant form (Fig. 9⇓, b and c). In both the gels and surrounding medium, processed MMP-1 increased following IgE-dependent mast cell activation (Fig. 9⇓, b and c).
Mast cell infiltration of airway smooth muscle appears to be unique to asthma and does not occur in related respiratory conditions (2, 3, 5). Elaboration of chemokines by ASM likely initiates the process (36, 37), but the cellular interactions maintaining infiltration and promoting subsequent airway remodeling may involve activation of either cell type. We applied a model system of human mast cell/HASM coculture in a collagen gel matrix to study the consequences of mast cell activation on HASM contractility. Within collagen lattices, C5a activation of HMC-1 cells or IgE-dependent activation of pHMC triggered contraction of HASM-containing gels, which was dependent on the activity of MMPs. Cytokines contributed to MMP generation and serine proteases contributed to MMP activation, leading to mast cell dependence of HASM contractile responses within the gel matrix.
The matrix contraction model used in our study is driven by smooth muscle activation, but is distinct from the immediate-type smooth muscle contraction described by the classic Schultz-Dale model (38). Anaphylactic contractile responses in immunologically sensitized smooth muscle can be triggered by anti-IgE or Ag, are mediated by cysteinyl leukotrienes and histamine, and initiate within seconds of exposure to the stimulus (39, 40, 41). In contrast, the collagen gel contraction system used is our study is analogous to extracellular matrix contraction during processes such as healing of vascular wounds or airway remodeling. In this model, shrinkage of the collagen gel matrix occurs over a period of hours to days and may result from cellular contraction, reorganization of the collagen matrix, or cell adhesive interactions (16, 17).
MMPs are key mediators of contractility in HASM-embedded collagen gels. Both activated mast cells (42, 43) and smooth muscle cells (14, 24, 25, 26) can produce MMPs, including MMP-1, MMP-2, MMP-9, and MMP-12, all of which are expressed in asthmatic airways (44). Although exogenously added, catalytically active MMPs 1, 2, 9, and 12 all triggered contraction of HASM-containing collagen gels, the collagenase MMP-1, and the gelatinase MMP-2 were the predominant forms endogenously produced by HASM in the collagen gel system.
Although activation of HMC-1 with C5a or activation of pHMC with anti-IgE induced expression of MMP-1 in coculture with HASM, mast cells were not required for expression of MMP-1. In gels containing HASM only, the combination of IgE/anti-IgE induced low levels of MMP-1 expression and processing. HASM express FcεRII (CD23) (45), which can stimulate production of IL-1β (46), an activator of MMP-1 expression (47, 48). HASM may also express FcεRI, resulting in IgE-dependent production of IL-13 (49), an additional inducer of MMP-1 (14). Although MMP-1 expression was found in gels containing HASM only, no IgE-dependent contraction was observed in the absence of mast cells, indicating that MMP-1 expression was not sufficient for induction of HASM contraction. Similarly, although MMP-1 expression and activation were increased upon treatment of HASM-embedded collagen gels with TNF-α, this did not result in appreciable contraction. Thus, MMP-1 does not appear to be either sufficient or directly responsible for, HASM contraction in collagen gels.
In addition to MMP-1, the gelatinase MMP-2 was constitutively produced by HASM and increased upon coculture with mast cells. In gels coembedded with HASM and HMC-1, MMP-2 content and collagen gel contraction were both increased following treatment with C5a or IL-13. In contrast, coembedding pHMC with HASM markedly induced the proteolytically processed form of MMP-2, even in the absence of IgE receptor cross-linking and mast cell-dependent contraction. Thus, MMP-2 may contribute to the low level contraction seen in the presence of resting mast cells (50), but expression of the processed form was not modulated upon IgE-dependent activation. As for MMP-1, the presence or activation of MMP-2 did not appear sufficient to account for contractile activity induced by mast cell activation. MMP-9 was produced by cultures of pHMC alone, but was greatly down-modulated under coculture conditions and therefore also not associated with the activation response. Additional proteases, cytokines, or other agents present in the coculture system may influence contractile responses to MMPs.
The major mast cell serine protease tryptase may modulate MMP activation. pHMC contained tryptase that was released upon activation. Although HMC-1 cells are sparsely granulated, they also contain tryptase (51), as was confirmed here (data not shown). Tryptase induces a range of responses in smooth muscle cells, including cytokine and chemokine synthesis (52), TGF-β production and activation (15, 52, 53), proliferation (54), direct contraction (15), hyperresponsiveness to histamine (11), and differentiation to myofibroblasts (15). In our system, tryptase triggered HASM contraction in collagen gels and this activity could be blocked by leupeptin, in agreement with published observations (15). Interestingly, however, this response was also blocked by batimastat, raising the possibility that endogenous MMP(s) mediate contraction in response to tryptase. MMPs are released as latent precursors and are activated extracellularly by proteolytic processing. In accordance with this, we observed direct proteolysis of both pro-MMP-1 and pro-MMP-2 by human lung tryptase, in agreement with observations in other systems (28, 29). Our findings suggest that tryptase-mediated MMP activation may contribute to mast cell-dependent HASM contraction.
In the presence of mast cells, contraction of HASM-containing collagen gels was stimulated by IL-6 or IL-13, both of which may directly induce expression of MMP-1 from smooth muscle cells (14, 30). IL-6, a product of activated HASM (55), drives human lung mast cell proliferation in coculture with HASM and may contribute to mast cell myositis in asthma (31). Because IL-13 is a critical cytokine for generation of asthmatic responses (56), we explored the IL-13-induced contraction further. IL-13 has been localized to mast cells within the smooth muscle layer in asthmatic airways (57, 58). Smooth muscle cells express IL-13Rα1 and IL-4Rα (59) and respond to IL-13 with a range of activities, including proliferation (60), migration (61), chemokine release (62), and contraction (63). A second IL-13 binding chain, IL-13Rα2, has higher affinity than IL-13Rα1, but triggers no known signaling responses and is thought to act as a “decoy” to antagonize IL-13 activity (34). Recent studies, however, suggest that IL-13Rα2 may mediate profibrotic responses to IL-13 by an undefined mechanism (35). Expression of IL-13Rα2 can be inducibly regulated on the surface of various cell types including smooth muscle cells (33). Human cell lines overexpressing IL-13Rα2 have reduced signaling responses to IL-13 (64, 65), consistent with antagonist function, but there have been relatively few studies of IL-13Rα2 effects on IL-13 responses of primary human cells (33, 66, 67). We induced IL-13Rα2 expression on HASM by treatment with IL-13 in combination with TNF-α. Under these conditions, the mast cell-dependent HASM contractile activity of IL-13 was ablated, strongly suggesting that in this system, IL-13Rα2 antagonizes IL-13 functional activity. IL-13 may also promote contraction of HASM-containing gels following C5a-induced mast cell activation, as sIL-13Rα2-Fc, a potent IL-13 antagonist, significantly reduced contraction in response to C5a-activated HMC-1.
Numerous processes may contribute to hypercontractility of ASM in asthma, including increased responsiveness to activating agents or reduced responsiveness to negative regulators (68). In asthma, the airway smooth muscle is infiltrated by mast cells, which may show signs of activation and degranulation (2, 6) and contribute to airway remodeling. Our findings with IgE-dependent activation of primary mast cells, C5a activation of HMC-1, or cytokine treatment support a model by which mast cell-derived cytokines induce MMP generation and serine proteases mediate MMP activation, leading to HASM-dependent matrix contraction. Such interactions have the potential to exacerbate pathology in asthma and other lung disorders.
Karl H. Nocka, Agnes Brennan, Bijia Deng, Margaret Fleming, Samuel J. Goldman, and Marion T. Kasaian are current employees of Wyeth Research. Alexander Margulis is a past employee of Wyeth Research.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 Current address: Genzyme Corporation, Framingham, MA.
↵2 Address correspondence and reprint requests to Dr. Marion T. Kasaian, Wyeth Research, 200 Cambridge Park Drive, Cambridge, MA 02140. E-mail address:
↵3 Abbreviations used in this paper: ASM, airway smooth muscle cell; MMP, matrix metalloproteinase; HASM, human airway smooth muscle cell; pHMC, primary human mast cell; sIL-13Rα2-Fc, soluble IL-13Rα2-Fc; SCF, stem cell factor.
- Received November 25, 2008.
- Accepted May 27, 2009.
- Copyright © 2009 by The American Association of Immunologists, Inc.