INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Airway morphology in asthma not only displays characteristics of an acute inflammatory process, but also structural changes. These include goblet-cell hyperplasia; changes in extracellular matrix (ECM) composition, with increased subepithelial deposition of collagens I, III, and V, fibronectin, and tenascin; and neovascularization and smooth-muscle-cell hyperplasia and/or hypertrophy (1). These structural changes are considered to play an important role in the pathophysiology of bronchial hyperresponsiveness (BHR), the functional hallmark of asthma (7). The remodeling process that produces these changes is thought to result from the combination of chronic, repetitive injury to the airway wall, and the ensuing tissue repair process. A variety of proinflammatory mediators, enzymes, cytokines, and growth factors have been implicated in the pathophysiology of airway remodeling (10); however, both the pathogenesis and the functional consequences of such remodeling must be further established.

In vivo animal models might provide interesting information in this respect. We and others have previously shown that exposure to aerosolized allergen in Brown Norway (BN) rats results in eosinophilic airway inflammation, IgE production, and an increase in airway responsiveness (19). However, in these models, airway remodeling was either not assessed or was revealed to be quite limited (20). The aim of the present study was to evaluate whether longer exposure of sensitized animals to inhaled allergen could result in more pronounced structural airway changes.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Specific pathogen-free, male BN rats were obtained from Harlan CPB (Zeist, The Netherlands). They were housed in the animal research facilities of the Department of Respiratory Diseases of the University Hospital, Gent, for 1 to 3 wk before testing, and received food and water ad libitum. The rats were aged 2 to 5 mo and weighed between 250 and 350 g at the time of testing.

Immunization Procedures and Exposure

All animals were actively sensitized by intraperitoneal injection of 1 mg of ovalbumin (OA) (Grade III; Sigma Chemical Co., Poole, UK) together with 200 µg Al(OH)₃ in 0.5 ml 0.9% NaCl on the first day of the study (Day 0). Different groups of rats (each group containing eight to 10 animals) were treated as follows: (1) Intraperitoneal injection of 1 mg OA and 200 µg Al(OH)₃ in 0.5 ml 0.9 % NaCl on Day 7, followed from Day 14 to 28 by thrice weekly exposure to aerosolized OA or PBS (2 wk). (2) Intraperitoneal injection of 1 mg OA and 200 µg Al(OH)₃ in 0.5 ml 0.9% NaCl on Day 7, followed from Day 14 to 42 by thrice weekly exposure to aerosolized OA or PBS (4 wk). (3) Thrice weekly exposure to aerosolized OA or PBS from Days 14 to 98 (12 wk).

The aerosol was generated with an ultrasonic nebulizer (Sirius Nova; Carl Heyer GmbH, Bad Ems, Germany) and was drawn into an exposure chamber containing awake animals. The output of the nebulizer was 3 ml/min and the median particle size was 3.2 µm according to specifications by the manufacturer. The concentration of OA in the nebulizer was 1% (wt/vol), and the duration of exposure was 30 min. Control animals were exposed to aerosolized, sterile phosphate-buffered saline (PBS).

Measurement of Bronchial Responsiveness

At 24 h after the last exposure to OA or PBS, rats were anaesthetized by the intraperitoneal injection of 60 mg/kg pentobarbital (Sanofi, Libourne, France). The trachea was cannulated with a Teflon cannula (35 mm long and 1.67 mm internal diameter [I.D]). The femoral artery was cannulated with a polyethylene catheter (1.77 mm I.D., 2.80 mm outer diameter; Intramedic; Clay Adams, Parsippany, NJ) to monitor blood pressure and heart rate with a pressure transducer (Celesco Transducer Products Inc., Canoga Park, CA) throughout the experiment. A polyethylene catheter was inserted in the external jugular vein for administration of intravenous drugs and fluids.

Animals were placed on a heating pad (37° C), were connected with a Palmer animal respirator (BioScience, Sheerness, UK), and were ventilated with oxygen-enriched air at a rate of 75 strokes/min and a stroke volume of 2 to 2.5 ml. Animals were injected with pancuronium bromide (Organon Teknika B.V., Boxtel, The Netherlands), at 0.2 mg/kg body weight to interrupt spontaneous respiratory movements. The animals' arterial blood gases were measured after 10 min of ventilation, and the stroke volume was adapted to keep pH, PO₂ and PCO₂ within the normal ranges.

The animals' transpulmonary pressure (Ppt) was measured with a differential pressure transducer (Celesco Transducer Products), of which one end was attached to an air-filled catheter inserted into the pleural cavity and the other end was attached to a catheter connected to a sideport of the outlet of the tracheal cannula. The airflow at the outlet of the intratracheal cannula was measured with a pneumotachograph (Type 00000; Fleisch, Geneva, Switzerland). The tidal volume (VT) was obtained by integration of the flow. Lung resistance (RL) was continuously calculated from VT, air flow, and Ppt (PRS800; Mumed, London, UK).

Carbachol (carbamoylcholine hydrochloride; Merck, Darmstadt, Germany) was nebulized with an ultrasonic nebulizer (Model 2511; Pulmo-Sonic, DeVilbiss Co., Somerset, PA). The mean particle size of the aerosol was 3.8 µm and the output was 1.0 ml/min. The aerosol was led into the inspiratory tube of the respiratory pump. Increasing concentrations of carbachol were administered over a period of 30 s at 5 min intervals, allowing RL to return to baseline between administration of each concentration. Carbachol-induced bronchoconstriction was measured as the percent increase in RL, comparing the peak of the reaction with baseline RL.

Bronchoalveolar Lavage

After determination of bronchial responsiveness, lungs were lavaged via the tracheal cannula with six volumes of 6 ml each of Hanks' balanced salt solution (HBSS), free of ionized calcium and magnesium and supplemented with 0.05 mM sodium ethylene diamine tetraacetic acid (EDTA). Lavage fluid was recovered by gentle manual aspiration with a syringe. The lavage fluid was centrifuged and the cell pellet was washed in HBSS and incubated for 7 min in 1 ml of NH₄Cl to remove red blood cells. The cell pellet was then washed twice and resuspended in 1 ml HBSS. Total numbers of leukocytes in the bronchoalveolar lavage fluid (BALF) were determined with a Coulter counter (Coulter Electronics Ltd., Harpeneden, UK). A differential cell count was performed on cytocentrifuged preparations (Cytospin 2; Shandon Ltd., Runcorn, UK) stained with May-Grünwald-Giemsa, and was based on standard morphologic criteria of at least 300 cells.

Histologic Analysis of Lung Tissue

After being lavaged, lungs were fixed with 4% paraformaldehyde via the tracheal cannula. The lungs were removed and slices from different lobes were embedded in paraffin and cut in 2-µm-thick sections. Histologic changes were evaluated on sections stained with hematoxylin and eosin (H&E). Additionally, Congo Red and Sirius Red stains were performed for specific assessment of eosinophil number and collagen deposition respectively.

Immunohistochemical Staining for Fibronectin

The detection of fibronectin was done as previously described (24), with a slight modification. Sections were deparaffinized, rehydrated, and incubated in 3% H₂O₂ in methanol for 10 min at room temperature to block endogenous peroxidase activity. The sections were then washed with 0.25% Brij 35 (Merck) in Tris-buffered saline (TBS) (0.15 M NaCl, 0.01 M Tris-HCl; pH 7.4), incubated in 0.4% pepsin (Sigma Chemical) in 0.01 N HCl at 37° C for 15 min, and rinsed with TBS. Nonspecific binding sites were blocked with 1% blocking reagent (Boehringer Mannheim GmbH, Mannheim, Germany) in PBS for 30 min. Excess reagent was removed and the sections were incubated for 1 h with goat antirat fibronectin antibody (Calbiochem-Novabiochem GmbH, Bad Soden, Germany) at a dilution of 1:800 in TBS. After incubation with the primary antibody, the sections were rinsed with 0.25% Brij 35 in TBS (pH 7.4) and incubated for 30 min with biotinylated rabbit antigoat IgG (DAKO A/S, Glostrup, Denmark) diluted 1:200 in TBS (pH 7.4). The primary antibody-secondary antibody complex was detected by incubating the sections for 30 min with streptavidin-biotinylated horseradish peroxidase complex (Nycomed Amersham Plc., Buckinghamshire, UK) diluted 1:200 in TBS (pH 7.4). The chromogenic substrate for the peroxidase was 3,3'-diaminobenzidine (DAB) (DAKO), giving a brown reaction product. Sections were counterstained with hematoxylin and mounted in Paramount. Two control procedures were performed for fibronectin detection: (1) normal goat serum was substituted for the primary antibody; and (2) goat antirat fibronectin antibody (diluted 1:800) was adsorbed overnight at 4° C with 100 µg fibronectin, and the supernatant of the antibody- fibronectin complex was substituted for the primary antibody.

Cell Proliferation in the Airways

To detect proliferating cells in lung tissue, animals were injected intraperitoneally twice weekly from Day 14 onward with bromodeoxyuridine (BrdU) (20 mg/kg) (Nycomed Amersham). BrdU is a thymidine analogue and is incorporated during the S phase of the cell cycle. BrdU was detected immunohistochemically in 2-µm sections that were deparaffinized and rehydrated. Endogenous peroxidase activity was blocked with 3% H₂O₂ in methanol for 10 min, after which the sections were incubated at 37° C for 15 min in 0.1% trypsin (Sigma Chemical) rinsed in tap water, and incubated for a further 15 min at 37° C with 2 M HCl. Nonspecific binding sites were blocked with 5% normal sheep serum (NSS) for 30 min. Excess reagent was removed and the sections were incubated for 2 h with mouse anti-BrdU antibody (Nycomed Amersham) diluted 1:75 in 1% NSS. After incubation with the primary antibody, the sections were rinsed with TBS (pH 7.5) and incubated for 30 min with biotinylated sheep antimouse IgG (Nycomed Amersham) diluted 1:200 in TBS (pH 7.5). The primary antibody-secondary antibody complex was detected by incubating the sections for 30 min with streptavidin-biotinylated horseradish peroxidase complex (Nycomed Amersham) diluted 1:200 in TBS (pH 7.5). The sections were then incubated for 10 min with DAB, rinsed in tap water, and counterstained with hematoxylin. As a negative control, 1% NSS was substituted for the primary antibody. Sections of the duodenum were used as a positive control.

Morphometric Analysis of the Airways

H&E-stained tissue sections were examined light-microscopically, and morphometric measurements were made with a Zeiss Axiophot microscope (Zeiss, Oberkochen, Germany) at magnifications between ×200 and ×400. For each treatment group, four stained lung sections of from six to 10 rats each were analyzed. The following morphometric parameters (as previously described by Bai and colleagues [25]) were measured in all airways cut in reasonable cross-sections (defined by a ratio of minimal I.D. to maximal I.D. of less than 0.5): the length of the basement membrane of the epithelium (Pbm), the perimeter of the outer boundary of the smooth muscle (Pmo), and the outer adventitial perimeter (Po); the areas defined by these parameters (Abm, Amo and Ao) and the area of smooth muscle (WAm). The following parameters were calculated based on the measured values: the total bronchial wall area (WAt = Ao  Abm), the inner wall area (WAi = Amo  Abm), and the outer wall area (WAo = Ao  Amo). Furthermore, the number of epithelial cells, the number of BrdU-positive cells, and the number of goblet cells along the basement membrane were counted. The total number of eosinophils and the number of BrdU-stained smooth-muscle cells in the bronchial wall were determined.

Deposition of collagen and fibronectin was examined by light microscopy, and quantitative measurements of each were made with a computerized image analysis system (LEICA Q500MC; Leica Cambridge Ltd., Cambridge, UK). A camera sampled the image of the stained sections and generated an electronic signal proportional to the intensity of illumination, which was then digitized into picture elements or pixels. The digital representation of the airways was analyzed with Qwin software (Leica Cambridge). Each pixel in a color image was divided into three color components (hue, saturation, and intensity). The threshold for each color component for the red stain of collagen and the brown stain of fibronectin was defined and kept constant throughout the analysis. In different fields around an airway, the area covered by the stain was determined by the software, and its value was calculated. For each experimental group, three lung sections from 10 rats each, stained for collagen, were examined. Fibronectin deposition was measured in three lung sections from five rats exposed to OA or PBS for 2 or 12 wk. The amount of protein deposited in the various components of the total airway wall was measured for all airways cut in reasonable cross-sections.

Measurement of OA-specific IgE

OA-specific IgE was measured with an enzyme-linked immunosorbent assay (ELISA). Blood was obtained by cardiac puncture at the end of the exposure period. Serum was added to OA (Grade V; Sigma Chemical) -coated microwell plates, and OA-specific IgE was detected with mouse antirat IgE antibody (H. Bazin, Experimental Immunology Unit, Brussels, Belgium) followed by horseradish peroxidase-conjugated goat antimouse immunoglobulins (DAKO). After addition of the substrate (o-phenylenediamine dihydrochloride; Sigma Chemical), plates were read in a plate reader (Titertek Multiskan MCC; Flow Laboratories, Irvine, Scotland) at 492 nm. The results are expressed as nanograms of OA bound to IgE per milliliter of serum.

Data Analysis

Reported values are expressed as mean ± SEM. For measurements of bronchial responsiveness, cumulative dose-response curves for the changes in RL with increasing doses of carbachol were constructed. The changes in RL were expressed as percent increases in RL. The concentration of carbachol causing a 50% increase in baseline RL (PC₅₀ RL) was calculated by log-linear interpolation of the dose-response curve. The cumulative dose-response curves were compared through analysis of variance (ANOVA). PC₅₀RL values, cellular composition of the BALF, OA-specific IgE, BrdU-positive cells and goblet cells in the epithelium, and peribronchial eosinophils for the different experimental groups were compared through the Kruskal-Wallis test for multiple comparisons. When significant differences were observed, pairwise comparisons were made, using the Mann-Whitney U test with Bonferroni's corrections.

For analysis of the morphometric measurements and the measurements of collagen and fibronectin deposition, the data for the rats in a particular experimental group were pooled, and the values of WAt, WAi, WAo, area of collagen deposition (WC), and area of fibronectin deposition (WF) were normalized to Pbm. Thereafter, the airways were divided into three categories according to the length of their basement membrane, as previously described by Sapienza and colleagues (20): Pbm  1 mm (small), Pbm > 1 mm and  2 mm (medium), and Pbm > 2 mm (large). To ensure that a similar range of airway sizes was being compared, we compared the frequency distribution of Pbm for the PBS and OA experimental groups for each category of airways, using the Kolmogorov-Smirnov test. The mean values of WAt, WAi, WAo, WC, and WF normalized to Pbm for each category of airways were compared for the experimental groups through an unpaired t test. Values of p < 0.05 were regarded as significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Airway Responsiveness

A 2-wk exposure to aerosolized OA induced an increase in airway responsiveness to inhaled carbachol as compared with PBS-exposed animals. This was reflected by a significant leftward shift of the dose-response curve (ANOVA, p < 0.05) (Figure 1) and a significant decrease in the PC₅₀RL for carbachol (p < 0.01) (Table 1). More prolonged exposure to OA resulted in a loss of airway hyperresponsiveness (AHR). After 4 wk of OA exposure, the PC₅₀RL and dose-response curve for carbachol were no longer significantly different from those of control animals (Table 1). The PC₅₀RL for rats exposed to OA for 12 wk was significantly greater than that of control animals (Table 1), although the dose-response curve for these animals was not significantly different from that of control animals (Figure 1). The dose-response curves for carbachol of control animals exposed for 2, 4, or 12 wk to PBS were not significantly different from one another. A leftward shift of the dose-response curves for animals exposed to OA for 2 wk was observed as compared with the curves for animals exposed to OA for 4 and 12 wk (ANOVA; p < 0.05). No significant difference was found in the dose-response curves for animals exposed to OA for 4 and 12 wk.

View larger version (9K): [in this window] [in a new window]   Figure 1.   Bronchial responsiveness to aerosolized carbachol in sensitized BN rats exposed for (A) 2 wk (B) 4 wk, and (C ) 12 wk to OVA (closed squares) or PBS (open squares) (n = 10 for each group) (ANOVA for OA versus PBS, 2 wk, p < 0.05).

OA-Specific IgE

Serum levels of OA-specific IgE were consistently increased in sensitized rats exposed to OA for 2, 4, and 12 wk as compared with those of control animals (Table 1). Peak levels of OA-specific IgE were reached after 4 wk of OA exposure (Table 1).

BALF

Total cell counts were significantly increased in BALF obtained from OA-exposed animals (Table 2). The total cell number increased at up to 4 wk and remained stable afterwards. OA-exposed groups also had significantly greater numbers of eosinophils and neutrophils in their BALF than did controls. The number of eosinophils was not significantly different at 2, 4, or 12 wk, whereas the number of neutrophils increased further at 4 and 12 wk (Table 2).

Histology

Inflammatory cell infiltrates, including eosinophils, were observed in the peribronchiolar area of sensitized and OA- exposed BN rats (Figure 2). After 2 wk of OA exposure, the number of eosinophils around the airways was significantly increased over that of control animals (Table 3). The number of infiltrating eosinophils remained increased after 4 and 12 wk of OA exposure (Table 3).

View larger version (132K): [in this window] [in a new window]   View larger version (125K): [in this window] [in a new window]   Figure 2.   Airway sections of sensitized BN rats exposed to PBS (A) and OA (B) for 2 wk (Congo Red; original magnification ×400).

Light-microscopic analysis did not reveal epithelial desquamation in any of the BN rats. The ratio of goblet cells to epithelial cells was increased in all sensitized and antigen-exposed rats (Table 3). The maximal increase in goblet cells was observed after 2 wk of exposure. The number of BrdU-positive cells in the epithelium was measured in sensitized rats exposed to PBS or OA for 2 or 4 wk. An increase in proliferating epithelial cells was observed after 2 wk (6.0 ± 1.5 BrdU-positive cells/mm Pbm [mean ± SEM] versus 1.9 ± 0.6 BrdU-positive cells/mm Pbm in PBS-exposed animals; p < 0.05) and after 4 wk (11.1 ± 0.9 BrdU-positive cells/mm Pbm versus 2.3 ± 0.9 BrdU-positive cells/mm Pbm in PBS-exposed animals; p < 0.001) of OA exposure. No BrdU-positive cells were found in the airway smooth muscle.

Morphometric Measurements

For the morphometric measurements, a mean of 54 airways (range: 45 to 68) per experimental group were analyzed. The average number of airways analyzed within each category (large, medium, and small) was 20 (range: 12 to 33). The total airway wall area of small, medium, and large airways was increased in sensitized rats exposed to OA for 2 wk, as compared with the areas in control animals (Figure 3). This increase disappeared upon prolonged exposure. Similarly, the inner wall area of small and large airways increased significantly in rats exposed to OA for 2 wk, but not for 4 or 12 wk (Figure 3). As compared with that of control animals, the outer wall area was increased in small (p < 0.001), medium (p < 0.01), and large (p < 0.05) airways after 2 wk of OA exposure, and in small (p < 0.01), and medium (p < 0.05) airways after 4 wk of OA exposure (data not shown). The area of the smooth-muscle layer around the airways did not change after OA exposure.

View larger version (17K): [in this window] [in a new window]   Figure 3.   Effect of exposure to OA (solid bars) or PBS (open bars) for 2 wk (A and B), 4 wk (C and D), and 12 wk (E and F  ) on WAt (A, C, and E ) and WAi (B, D, and F  ) in small, medium, and large airways.

Fibronectin and Collagen Deposition

Fibronectin deposition was analyzed in a mean of 59 airways (range: 53 to 63) per experimental group, and collagen deposition was analyzed in a mean of 43 airways (range: 39 to 51) per experimental group. The average number of airways studied for fibronectin deposition in each category was 20 (range: 13 to 25), and the average number studied for collagen was 21 (range: 11 to 20). At 2 wk, increased fibronectin deposition was observed in large and medium airways of OA-exposed animals. This was further increased after 12 wk exposure (Figure 4). Fibronectin was deposited in the outer airway wall adjacent to the smooth-muscle layer (Figures 4 and 5). Large airways of rats exposed to OA for 12 wk also showed increased fibronectin deposition in the inner airway wall (p < 0.01) (data not shown). The amount of collagen was increased in large airways of animals exposed to OA for 4 and 12 wks (Figure 6). After 4 wk, an increased amount of collagen was deposited in the outer airway wall (p < 0.05) (data not shown); after 12 wk, an increase was observed in the inner (Figure 6) as well as in the outer airway wall (p < 0.01) (data not shown).

View larger version (18K): [in this window] [in a new window]   Figure 4.   Effect of exposure to OA (solid bars) or PBS (open bars) for 2 wk (A and B) and 12 wk (C and D) on the amount of fibronectin deposition in the total airway wall (WFt) (A and C ) and in the outer airway wall (WFo) (B and D) of small, medium, and large sized airways.

View larger version (93K): [in this window] [in a new window]   View larger version (89K): [in this window] [in a new window]   Figure 5.   Airway sections of sensitized BN rats exposed to PBS (A and C) and OA (B and D), stained for fibronectin after an exposition period of 2 wk (A and B) and 12 wk (C and D) (hematoxylin counterstain; original magnification ×400).

View larger version (15K): [in this window] [in a new window]   Figure 6.   Effect of exposure to OA (solid bars) or PBS (open bars) for 2 wk (A and B), 4 wk (C and D), and 12 wk (E and F  ) on the amount of collagen deposition in the total airway wall (WCt) (A, C, and E ) and in the inner airway wall (WCi ) (B, D, and F  ) of small, medium, and large airways.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Airway remodeling encompasses the structural changes observed in asthmatic airways. The pathophysiology of remodeling and its effect on airway physiology remain to be fully established. In vivo animal models might provide useful information in this respect. However, most of the currently developed animal models of allergic airway inflammation have been restricted to acute inflammatory changes following relatively short periods of allergen exposure. In the present study we investigated the effect of more prolonged OA exposure on the airways of sensitized BN rats. In addition to producing an increase in OA- specific IgE and persistent eosinophilic airway inflammation, repeated exposure to aerosolized OA over a period of up to 12 wk resulted in greater structural airway changes. These included goblet-cell hyperplasia, increased mitogenic activity in the epithelium, and changes in ECM components in the form of increased collagen and fibronectin deposition. The dynamics of these structural changes are clearly different, since goblet-cell hyperplasia and fibronectin deposition could be observed from 2 wk of exposure onward, whereas increased collagen deposition occurred only after more prolonged exposure. Other models have also shown goblet-cell hyperplasia after relative short exposure periods (26, 27). However, to the best of our knowledge, ours is the first model to show increased mucosal collagen deposition following repeated allergen exposure, thus mimicking the subepithelial fibrosis from increased deposition of collagen I, III, and V observed in human asthma (3, 4).

Airway remodeling is thought to have a profound effect on airway responsiveness. Different mathematical models describe the potential effect of structural changes on the function of asthmatic airways (7). Thickness of the inner airway wall, smooth-muscle mass, and adventitial structure are considered to be the main determinants in the airway response to a bronchoconstrictor stimulus (28). Increased airway wall thickness has little effect on baseline resistance, but enhances the effect of a given smooth-muscle contraction, following Poiseuille's law for laminar flow, thus contributing to increased airway sensitivity (28). It has been shown in rats that changes in pulmonary resistance after methacholine inhalation are mainly due to changes at the large airway level (20). In our model an increase in inner wall area at the large airway level was observed only in rats exposed to aerosolized OA for 2 wk. The observed leftward shift of the dose-response curve of RL for aerosolized carbachol could therefore result from this increase in inner airway wall thickness.

Increased smooth-muscle mass around the airways could also contribute to BHR (9). Our model showed no changes in the smooth-muscle layer, either by morphometric analysis or BrdU incorporation at 2 and 4 wk. This factor is therefore unlikely to have contributed to the altered airway behavior in our model. This contrasts with previous findings in BN rats of an increase in smooth-muscle thickness and smooth-muscle proliferation following repeated allergen exposure (20, 29). These divergent observations can at least be partly explained by differences in the substrains of BN rats (BN/RyHsd versus BN/SSN Hsd) and in the sensitization and exposure protocols used in these studies and our own.

Another element that could affect airway responsiveness are changes in the composition of the ECM in the airways. It has been calculated that subepithelial collagen deposition in asthmatic airways can influence the buckling pattern of the airway mucosa when the airway narrows (30). A smaller number of folds results in an enhanced increase in intraluminal pressure. In addition, changes in the ECM of the adventitial layer, in combination with fluid fluxes during smooth-muscle contraction, can reduce elastic recoil exerted on the airways by the lung parenchyma, thus contributing to AHR (9). Conversely, increased collagen deposition could increase airway wall stiffness, thus opposing against narrowing of the airway wall (31), and increased collagen deposition within and around the smooth-muscle layer can interfere with smooth-muscle contraction, thus again limiting airway narrowing (32). In the present model, a small increase in fibronectin deposition around the smooth-muscle layer was observed after a 2-wk exposure period. At 12 wk this was far more pronounced. Increased deposition of collagen was observed after 4 wk, but especially after 12 wk exposure. This coincided with a gradual reduction in airway wall thickness, whereas the AHR observed after 2 wk waned and turned into hyporesponsiveness at 12 wk.

It has to be borne in mind that the ECM does not merely constitute a biologically inactive framework within tissues, but consists of a highly regulated network of proteins and proteoglycans that interact profoundly with inflammatory cells and other structural tissue components (33). Among its many effects, fibronectin binds to ₅₁ integrins on the surfaces of cells, thus not only influencing cell trafficking, but also affecting cell function and survival (34). During tissue regeneration, fibronectin production precedes collagen synthesis (35). The ongoing inflammation observed in the airways of rats after prolonged OA exposure might induce a repair process in which the ECM is regenerated. Changes in the deposition of fibronectin start after 2 wk, but the increase in collagen content of the airway wall is observed only after a longer exposure period. Fibronectin attaches collagen fibers to the elastin framework and to fibroblasts, mediating contraction of collagen by fibroblasts (36). The more pronounced increase in fibronectin deposition around the airway smooth-muscle layer after 12 wk, in combination with increased collagen deposition in the outer airway wall, could therefore result in a tighter adhesion between lung parenchyma and the outer airway wall and prevent fluid fluxes into the adventitium during smooth-muscle contraction. The changes in collagen composition in the inner airway wall after 12 wk could also increase the stiffness of the airway wall (31). These observations suggest that by increasing airway wall stiffness, reducing airway wall thickness, and enhancing elastic recoil, changes in the ECM both within and around the smooth-muscle layer might constitute an attempt to oppose BHR.

The present model also adds to previous rat models showing that the presence of eosinophils per se in the airway tissue is insufficient to cause AHR (22, 23). The influx of eosinophils into and around the airways persists with prolonged allergen exposure, as does IgE synthesis, yet the increase in airway responsiveness progressively declines. This suggests that the loss of BHR is probably not due to immunologic tolerance, but is related to the development of compensatory mechanisms, the loss of which could contribute to the persistent BHR observed in human asthma.

In conclusion, we have described an in vivo rat model that displays characteristics of airway remodeling. We postulate that the increase in airway responsiveness is due to increased thickness of the inner airway wall of the large airways. Prolonged exposure results in scarring of the airway wall, with a reduction in airway wall thickness and a loss of AHR. These observations suggest that, depending on the extent and distribution of airway wall fibrosis, airway remodeling might oppose, rather than enhance, BHR.