Asthma is a chronic inflammatory disorder of the airways characterized by bronchial hyperresponsiveness (BHR), air-flow limitation, and airway wall remodeling, which consists of goblet cell hyperplasia (Aikawa et al., 1992), subepithelial fibrosis (Roche et al., 1989), and airway smooth muscle hypertrophy and hyperplasia (Dunnill et al., 1969). Airway wall remodeling may result from an imbalance of tissue regeneration and repair mechanisms. Thus, the airway epithelium may undergo repeated episodes of injury and repair in asthma (Stewart et al., 1993) and shows a high level of expression of growth factors such as transforming growth factor (TGF)-β and epidermal growth factor.

TGF-β, which exists in three isoforms (β1, β2, and β3) (Grande, 1997), is an important fibrogenic and immunomodulatory factor. It is produced in the airways of asthmatic subjects by inflammatory cells such as eosinophils infiltrating the bronchial mucosa (Balzar et al., 2005) as well as by structural cells of the airway wall, including fibroblasts, epithelial, endothelial, and smooth muscle cells (Khalil et al., 1991). TGF-β participates in the initiation and propagation of inflammatory and immune responses in the airways (Duvernelle et al., 2003), but it may also function as a down-regulator of immune responses. For example, regulatory T cells expressing surface TGF-β1 have suppressive functions (Nakamura et al., 2001). TGF-β-producing T cells have been shown to ameliorate allergen-induced BHR and airway inflammation (Hansen et al., 2000). TGF-β also regulates repair responses that lead to matrix deposition and tissue remodeling. Increased expression of TGF-β gene in epithelium and eosinophils of bronchial biopsy sections from patients with asthma and chronic bronchitis has been observed (Ohno et al., 1992), and the expression of TGF-β1 mRNA in the bronchial biopsies was found to correlate to the degree of subepithelial fibrosis (Minshall et al., 1997). In general, TGF-β1 induces the proliferation of mesenchymal cells such as fibroblasts and airway smooth muscle cells (Battegay et al., 1990; Okona-Mensah et al., 1998).

We hypothesized that many of the remodeling features in chronic asthma can be reduced by inhibiting TGF-β function and that TGF-β may be involved in allergic inflammation. The active form of TGF-β interacts with a series of serine/threonine receptors, which are part of a family of related receptor molecules termed activin receptor-like kinase (ALK). TGF-β receptor I kinase (ALK5) acts downstream of the TGF-β type II receptor and interacts with members of Smad, a family of cytoplasmic transducer proteins. ALK5 is specific for active TGF-β and/or activin and phosphorylates Smad2 and Smad3. TGF-β1 induces the phosphorylation of TGF-β type I and type II receptors, leading to phosphorylation and translocation of intracellular Smad2 and Smad3 to the nucleus (Heldin et al., 1997), where they regulate gene transcription. To examine the role of TGF-β in chronic asthma, we used a selective and novel 2,4-disubstituted pteridine-derived TGF-β receptor type I (TGF-βRI) kinase inhibitor, SD-208 (Uller et al., 2001), in a chronic rat model of allergen-induced airway inflammation.

Materials and Methods

The studies were performed in accordance with the UK Home Office procedures, in line with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the National Institutes of Health.

Sensitization and Challenges

Pathogen-free, female Brown-Norway rats weighing 150 to 180 g (Harlan, Bicester, UK) were sensitized on days 1, 2, and 3 using 1 mg kg^–1 (i.p.) injections of ovalbumin in 1 ml of 0.9% sterile saline containing 100 mg of Al(OH)₃ as adjuvant. On days 6, 9, 12, 15, 18, and 21, animals were exposed to either saline or 1% ovalbuminaerosol for 20 min.

Study Design

To examine whether the TGF-βRI kinase inhibitor could either prevent or reverse airway wall inflammation and remodeling, we devised two protocols. In the first protocol (protocol 1: preventive), the inhibitor was administered throughout at the period of exposure but before each exposure, whereas in the second protocol (protocol 2: reversal), it was given during the latter part of the exposure period.

Protocol 1: Preventive. We studied the following groups.

1. Sensitized, vehicle-treated and repeatedly saline-exposed (group saline; n = 8). Animals received 1% methylcellulose as vehicle by oral gavage twice daily.

2. Sensitized, vehicle-treated, and repeatedly OVA-exposed (OVA; n = 8). The procedures were the same as for group saline, except the aerosol was 1% ovalbumin.

3. Sensitized, TGF-βRI kinase inhibitor (SD-208)-treated and repeatedly OVA-exposed (SD-208; n = 8). The procedures were the same as for group OVA. Rats received 60 mg kg^–1 SD-208 twice daily from day 6 to 21, early morning and late evening.

Protocol 2: Reversal. We studied the following groups.

1. Sensitized, vehicle-treated and repeatedly saline-exposed (group saline; n = 10). Animals received 1% methylcellulose as vehicle by oral gavage twice daily.

2. Sensitized, vehicle-treated and repeatedly OVA-exposed (OVA; n = 10). The procedures were the same as for group saline, except the aerosol was 1% ovalbumin aerosol.

3. Sensitized, TGF-βRI kinase inhibitor (SD-208)-treated and repeatedly OVA-exposed (SD-208; n = 8). The procedures were the same as for group SD-208 in protocol 1, except rats received 60 mg kg^–1 SD-208 twice daily from day 15 to 21.

All rats were studied 18 to 24 h after the final exposure to either 1% ovalbumin or 0.9% saline aerosol. We also determined the effect of three exposures to ovalbumin or of saline alone in sensitized rats on bronchial responsiveness, airway inflammation, and features of airway wall remodeling.

Drugs

SD-208 is a selective 2,4-disubstituted pteridine-derived TGF-β RI kinase inhibitor (Uhl et al., 2004). SD-208 was mixed with the vehicle 1% methylcellulose to the amounts needed, using a Dynal sample mixer (Dynal Biotech, Lake Success, NY) for at least 6 h before administration.

Measurement of Bronchial Responsiveness

Lung resistance was measured in anesthetized and tracheostomized rats using an instantaneous on line system, as described previously (Salmon et al., 1999). Increasing concentrations of acetylcholine were administered, and lung resistance was measured after each concentration. We calculated the provocative concentration of acetylcholine (ACh) needed to increase baseline lung resistance (R_L) by 200% (PC₂₀₀).

Immunohistochemistry of Lung Tissues

5-Bromo-2′-deoxyuridine (BrdU; Sigma Chemical, Poole, Dorset, UK), a marker of cells in S phase, was administered as described previously (Salmon et al., 1999). Its uptake into nuclei of airway smooth muscle cells and airway epithelium was quantified in lung sections using quantitative immunohistochemistry techniques and a specific in-house software system. To identify airway smooth muscle cells, an anti-α-smooth muscle actin staining procedure was applied. Slides were observed using an Axioplan microscope (Carl Zeiss Inc., Hertsfordshire, UK). The results are expressed as percentage of positive BrdU cells per total number of nuclei (for airway smooth muscle) or per unit length (for epithelial cells). ASM thickness expressed as micrometers per unit basement membrane length was calculated.

For the detection of eosinophils and T lymphocytes in tissue sections, staining was performed with an anti-major basic protein (MBP) antibody (Monosan, Uden, The Netherlands) and with an anti-CD2 antibody (pan T-cell markers; PharMingen, Cambridge Bioscience, Cambridge, UK), followed by the alkaline phosphatase-anti-alkaline phosphatase technique to visualize positive cells, as described previously (Salmon et al., 1999).

For goblet cell counts, cryostat sections of lung tissues were stained with Alcian Blue/periodic acid-Schiff and counterstained with hematoxylin. Computerized image analysis was used to assess epithelial coverage, expressed as positively staining cells per unit length of basement membrane. The mean number of Alcian Blue/periodic acid-Schiff-positive goblet cells was determined, and the five largest airways in each lung section were assessed.

For the detection of Smad2/3 in tissue sections of the preventive group, staining was performed with a rabbit anti-Smad2/3 antibody (Upstate Biotechnology, Hampshire, UK) in a dilution of 1:100, and sections were incubated overnight at 4°C followed by incubation with a tetramethylrhodamine isothiocyanate-conjugated donkey anti-rabbit secondary antibody (Jackson ImmunoResearch Laboratories, Cambridgeshire, UK) at a dilution of 1:100 for 1 h at room temperature. Nuclei were stained with 4,6-diamidino-2-phenylindole (DAPI) dihydrochloride (Molecular Probes, Leiden, the Netherlands). Sections were examined under confocal microscope (TCS-SP; Leica, Bucks, UK) under the same conditions and settings. Six pictures were taken along the airway wall of the biggest airway. Digitalized images were analyzed with the program ImageJ (National Institutes of Health, Bethesda, MD). There were three-color components on the image: the red component of tetramethylrhodamine isothiocyanate-conjugated second antibody, the blue component (DAPI-stained nuclei), and the green component (autofluorescence on tissues). The structures of the airway wall were recognized by adjusting the balance of the green component. The regions of interest were marked by using a drawing tablet (Wacom Technology, Vancouver, WA). Color balance was reset back to the original condition before measurement. The red component (tetramethylrhodamine isothiocyanate-conjugated second antibody expression) was isolated and converted to an 8-bit gray color picture where the mean gray value was measured. The mean value obtained from six pictures was derived. The mean expression index of Smad2/3 on epithelial cells were calculated as the total pixel gray value (product of the total pixel count and the mean gray value on each pixel) divided by the number of epithelial nuclei. The mean expression of Smad2/3 on ASM was similarly calculated as the mean gray value per square micrometer of muscle.

Measurement of OVA-Specific IgE

OVA-specific IgE titers were measured by enzyme-linked immunosorbent assay (Nonaka et al., 2000). A standard curve was constructed using doubling dilutions of standards (rat serum identified to be of high IgE titer), and samples were added and incubated for 2 h. Biotinylated-OVA (prepared using EZ-Link Sulfo-N-hydroxysuccinimide-LC-biotinylation kit) was added, followed by incubation for 45 min with 1 μg/ml streptavidin-alkaline phosphatase. The p-nitrophenyl phosphate enzyme substrate was used for color development, and the plates were read spectrophotometrically at 405 nm with an automated enzyme-linked immunosorbent assay plate reader (Automated Microplate Reader EL311; Bio-Tek Instruments, Winooski, VT). Titers were expressed as optical density and were converted to nanograms per milliliter.

Data Analysis

Mean indices were statistically analyzed after logarithmic transformation by one-way analysis of variance, followed by nonparametric t tests to evaluate significant differences between groups. Values are expressed as means ± S.E.M., with p values of less than 0.05 considered significant.

Results

Bronchial Responsiveness to Acetylcholine. We found a significant increase in bronchial responsiveness (represented by the PC₂₀₀ value) in the OVA-exposed groups compared with the saline-exposed groups (preventive study: OVA, 1.97 ± 0.052 versus saline, 1.67 ± 0.07; p < 0.05; and reversal study: OVA, 2.0 ± 0.0.07 versus saline, 1.53 ± 0.08; p < 0.01). There was a leftward shift of the lung resistance concentration-response curves after OVA exposure (Fig. 1, A and B). SD-208 did not significantly prevent PC₂₀₀ induced by allergen exposure (SD-208, 1.71 ± 0.14 versus OVA, 1.97 ± 0.052), but it significantly reversed it (SD-208, 1.41 ± 0.15 versus OVA, 2.0 ± 0.0.07; p < 0.001; Fig. 1, A and B).

Airway Inflammation. Repeated allergen exposure caused a significant increase in MBP⁺ cells (19.7 ± 2.43 cell/mm basement membrane versus 12.2 ± 0.97; p < 0.01) and CD2⁺ T cells (20.1 ± 1.14 cell/mm basement membrane versus 12.2 ± 1.5; p < 0.001) compared with saline control (Fig. 2, A and B). SD-208 prevented (MBP, 9.4 ± 1.5; p < 0.001; and CD2, 10.0 ± 1.1; p < 0.001) and also reversed (MBP, 6.3 ± 0.7; p < 0.001; CD2, 9.5 ± 1.2; p < 0.001) the allergen-induced increases in MBP⁺ and CD2⁺cells compared with OVA control group (Fig. 2, A and B).

Airway Smooth Muscle and Epithelial BrdU Indices. After repeated allergen exposure, there was an increase in ASM (preventive study, 2.9 ± 0.3%; p < 0.001; and reversal study, 2.7 ± 0.2%; p < 0.01) and epithelial cells (preventive study, 24.5 ± 7.0%; p < 0.05; and reversal study, 16.8 ± 2.4%; p < 0.05) expressing BrdU compared with the saline exposure (ASM, 1.4 ± 0.1%; epithelium, 9.7 ± 1.7%; Fig. 3, A and B). SD-208 prevented the increases in BrdU-positive ASM (1.1 ± 0.2%; p < 0.001) and epithelial cells (8.7 ± 1.8%; p < 0.05) and reversed these increases in ASM (1.3 ± 0.2%; p < 0.01) and epithelial cells (7.1 ± 1.4%; p < 0.05) expressing BrdU induced by OVA exposure (Fig. 3, A and B).

Goblet Cell Count. Repeated allergen-exposure caused an increase in goblet cell number in rat airways (29.4 ± 2.4 versus 14.9 ± 1.8 cells/mm basement membrane; p < 0.01) compared with saline control (Fig. 4A). SD-208 prevented the increase in the number of goblet cells (11 ± 1.1; p < 0.01) and reversed the increases (13.8 ± 0.1; p < 0.05) induced by allergen exposure compared with the OVA control group (Fig. 4, A and B).

Airway Smooth Muscle Thickness. There were significant changes in ASM thickness observed between OVA-exposed and saline-exposed groups (9.1 ± 1.0 mm) in both preventive (17.7 ± 4.1 μm; p < 0.05) and reversal (13.4 ± 1.1 mm; p < 0.05) treatment groups. Both SD-208 preventive and reversal treatments did not lead to a significant decrease in the increase in muscle thickness induced by repeated allergen exposure (Fig. 5, A and B).

Effect of Three OVA Challenges Compared with Six OVA Challenges. We showed the development of BHR, airway inflammation, and airway wall remodeling after three successive exposures, except that the airway wall muscle thickness did not increase significantly (Table 1). Table 1 compares the effect of three successive exposures with that of six successive exposures (data obtained from protocol 1). Compared with the six successive exposures, there was no further increase in BHR, mucosal T cells, and goblet cell hyperplasia, and no further changes in ASM thickness. However, eosinophils decreased (p < 0.001), and the increase in goblet cells only achieved a p < 0.063.

Smad2/3 Expression. Differences in intensity of the Smad2/3 expression on airway wall tissues among groups were observed under confocal microscope examination (Fig. 6, top). Smad2/3 expression index of epithelial cells was significantly increased after OVA exposure (260 ± 16.3% of saline control; p < 0.001) compared with saline control (100 ± 5.79% of saline control). Preventive treatment with SD-208 decreased the Smad2/3 expression index increase induced by OVA exposure (87.9 ± 6.03% of saline control; p < 0.001) (Fig. 6A, bottom). Similar increases in the mean gray value of Smad2/3 on ASM were also observed in OVA-exposed group (287 ± 15.6% of saline control; p < 0.001) compared with saline control group (100 ± 5.55% of saline control). SD-208 preventive treatment significantly decreased this increment induced by OVA (111 ± 8.23% of saline control; p < 0.001) (Fig. 6B, bottom).

Serum IgE. Twenty-four hours after repeated OVA exposure, there was no significant change in the serum levels of OVA-specific IgE compared with the sensitized, saline-exposed group. SD-208 preventive treatment decreased both serum levels of total IgE (2136 ± 237 ng/ml; p < 0.01) and OVA-specific IgE (176.8 ± 17.2 arbitrary unit; p < 0.001) compared with untreated allergen-exposed control group (total IgE, 3773 ± 345 ng/ml; OVA-specific IgE, 5760 ± 854 arbitrary unit) but not in the reversal treatment protocol (Fig. 7, A and B).

Discussion

TGF-β, an important fibrogenic and immunomodulatory factor, may also function either as a pro- or as an anti-inflammatory cytokine and may play an integral role in promoting the structural changes of airway remodeling. We studied the effects of SD-208, which belongs to a family of potent, selective TGF-βRI kinase inhibitors, in a chronic allergen exposure model that we have previously characterized and that demonstrates features of BHR, eosinophilic inflammation, and airway wall remodeling (Salmon et al., 1999). SD-208 effectively prevented and reversed airway eosinophilic inflammation, and proliferation of ASM and epithelial cells and goblet cell hyperplasia induced by repeated allergen challenges; in contrast, it did not significantly prevent BHR but could reverse BHR induced by repeated allergen exposure. In addition, SD-208 prevented but did not reverse the rise in serum total and OVA-specific IgE. ASM thickness was not significantly altered. These data therefore indicate that there is an important role for TGF-β released during chronic allergen exposure, that the role of TGF-β is proinflammatory, and that TGF-β is an important growth factor involved in various aspects of airway wall remodeling.

In vitro data using the incorporation of radiolabeled ATP into a peptide or protein substrate in a cell-free assay showed that SD-208 had a specificity for TGF-βRI kinase of >100-fold compared with TGF-βRII kinase and at least 20-fold over members of a panel of protein kinases such as c-Jun NH₂-terminal kinase, extracellular signal-regulated kinase, and p38 kinase, and epidermal growth factor receptor kinase (Uller et al., 2001). Thus, there was reasonable selectivity against the closely related kinases. SD-208 has been shown to inhibit the release of IL-6 and vascular endothelial growth factor from bone marrow stromal cells, together with an inhibition of the proliferation of these cells, indicating anti-inflammatory and antiproliferative effect of this compound in vitro (Hayashi et al., 2004). SD-208 is also a specific inhibitor of TGF-βRI kinase on primary rat lung fibroblasts stimulated by TGF-β1 in vitro with an IC₅₀ of ∼35 nM; the release of plasminogen activator type 1 protein expression was blocked in a dose-dependent manner from these cells (Bonniaud et al., 2005). In a rat model of TGF-β-mediated lung fibrosis, SD-208 administered orally at doses of 25 to 50 mg/kg inhibited progressive fibrosis (Bonniaud et al., 2005). Data from the current work indicate that at similar doses SD-208 was sufficient to reduce the activation of Smad2/3 in the airway's mucosa.

We conclude that TGF-β enhances allergic airway inflammation. Blockage of the TGF-β functions by inhibiting TGF-βRI kinase activity with SD-208 decreased inflammatory cells such as T cells and eosinophil recruitment into the airway. The results support that TGF-β has a positive role in allergic inflammation. The mechanism is not known but TGF-β may enhance T helper 2 cytokine expression (Wenzel et al., 2002). Our results also demonstrate that blocking of TGF-β function with SD-208 given preventively decreases serum level of total and OVA-specific IgE, indicating that TGF-β positively regulates IgE level in allergic inflammation. In the reversal study, however, there was no effect of SD-208 on serum IgE levels. SD-208 was administered over a shorter period in the reversal study compared with the preventive study, and this may be a possible reason for the lack of effect.

Our results are consistent with a proeosinophilic effect of TGF-β. TGF-β can enhance the effects of IL-13 on activation and survival of eosinophils, and on the surface expression of IL-13Ra1 on eosinophils (Myrtek et al., 2004). In addition, TGF-β has chemoattractant effects for eosinophils (Luttmann et al., 1998). In epithelial cell cultures, TGF-β selectively enhanced granulocyte macrophage–colony-stimulating factor and regulated on activation normal T cell expressed and secreted production, whereas it suppressed IL-8 production, a profile of chemokine response promoted by TGF-β that would favor eosinophil, lymphocyte, and monocyte recruitment, hallmarks of chronic allergic inflammation, over neutrophil sequestration (Jagels and Hugli, 2000). TGF-β also plays a role in recruiting macrophages into the airway epithelium in chronic obstructive pulmonary disease (de Boer et al., 1998). Mast cell function may be up-regulated by members of the TGF-β family, including TGF-β1 and activin A (Funaba et al., 2002), and mast cell chemotaxis can be promoted by TGF-β (Gruber et al., 1994). Our results are consistent with the results of overexpression of TGF-β1 in the lungs of rats using an adenovirus vector, which resulted in severe inflammation with mononuclear cells and to a lesser extent with neutrophils and eosinophils, and persistent fibrosis (Sime et al., 1997).

Some opposite conclusions have been reached in murine models of allergic inflammation. Thus, in mice heterozygous for TGF-β1 gene deletion leading to a reduced expression of TGF-β1, an increased eosinophilic inflammation and mucus secretion in response to ovalbumin sensitization have been reported (Scherf et al., 2005). Moreover, these mice develop significantly enhanced T helper 2-cytokine levels, decreased interferon-γ production, and increased levels of OVA-specific IgE in serum. However, these mice are not completely devoid of TGF-β, because they express ∼30% of wild-type TGF-β protein, such that this amount of TGF-β may still be involved in the proallergic inflammatory responses observed. In another murine allergic sensitization model, a blocking antibody to TGF-β did not affect ovalbumin-induced airway eosinophilic inflammation, but it reduced extracellular deposition, airway smooth muscle proliferation, and mucus production (McMillan et al., 2005), indicating a role for TGF-β solely in airway wall remodeling. Whether these differences observed are species-dependent is a possibility.

There has been conflicting data regarding the effect of TGF-β1 on the proliferation of various cell types in vitro. TGF-β1 inhibited mitogen-induced human ASM cell proliferation, whereas in other reports it stimulated it (Okona-Mensah et al., 1998; Chen and Khalil, 2006). The effect of TGF-β is likely to be dependent on the proliferative state of the cells and on the concentration of TGF-β used. TGF-β1 stimulates the proliferation of ASM cells at confluence, whereas it inhibits proliferation at subconfluence (Okona-Mensah et al., 1998). There is also synergy between TGF-β and cysteinyl leukotrienes in ASM proliferation (Espinosa et al., 2003). Our data indicate that TGF-β1 is involved in the proliferation of ASM cells in the allergic model. In preliminary studies, we have shown that SD208 inhibits the translocation of Smad2/3 in airway smooth muscle cells in vitro and TGF-β-induced proliferation (Xie and Chung, 2006).

We showed that SD-208 inhibited the increase of Smad2/3 expression induced by allergen exposure, indicating that the Smad pathway may be important in mediating some of the features of allergen-induced airway inflammation and airway wall remodeling. Our data are concordant with the results of McMillan et al. (2005) who demonstrated in a mouse model of allergen exposure that an anti-TGF-β antibody caused a reduction in the expression of phospho-Smad2 and the up-regulation of Smad7. The expression of the inhibitor Smad7 in bronchial epithelial cells has been shown to correlate inversely to basement membrane thickness and airway hyperresponsiveness in patients with asthma (Nakao et al., 2002), and the role of Smads as intracellular mediators of airway inflammation has been proposed (Groneberg et al., 2004). Indeed, in the small airways of patients with chronic obstructive pulmonary disease, a decrease in Smad6 and Smad7 expression has been found (Springer et al., 2004). These data indicate that the inhibitor of TGF-βRI kinase inhibitor, SD-208, may also be used to treat chronic obstructive pulmonary disease.

We have also shown that epithelial cell proliferation after allergen challenge is caused by TGF-β. The localization of TGF-β expression in the epithelium in a mouse allergic model (Kelly et al., 2005) indicates that this may be an autocrine effect of TGF-β. Endogenous TGF-β2 production by bronchial epithelial cells may also increase airway mucin expression. IL-13-induced mucin expression may also occur in part through TGF-β2 up-regulation (Chu et al., 2004). By using SD-208 to inhibit the effects of all TGF-β isoforms, we demonstrated that bronchial epithelial cell and ASM cell proliferation and goblet cell hyperplasia induced by allergen exposure, measured as BrdU-positive cells, were inhibited. Thus, TGF-β promotes structural cell proliferation and airway remodeling in allergic airway inflammation. However, an increase in ASM thickness was not detected in our model.

Our study using the specific TGF-βRI kinase inhibitor SD-208 in studying the functions of TGF-β demonstrates that TGF-β enhances allergen-induced airway inflammation, goblet cell and ASM cell hyperplasia, and BHR. Taken together, these findings support important role for TGF-β in a chronic rat allergic model of asthma and point to the potential of SD-208 in the treatment of chronic asthma.