INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Bronchial asthma is characterized by inflammation of the airways associated with structural epithelial changes and thickening of the reticular layer of the subepithelial basement membrane (BM) (1). Little information exists on the reversibility of subepithelial fibrosis in BM in asthma. There are suggestions that subepithelial fibrosis may induce asthma to become a more chronic disease. In order to better evaluate the treatment effects at the tissue level, the molecular composition of the BM should be better known (2, 4).

The BM is a thin layer of specialized extracellular matrix (ECM) between the epithelium and the stroma, not only providing mechanical support but influencing epithelial cell- specific functions (7). Heterogeneity in BM structure and functions occurs not only between different tissues but also during the development and in response to different physiologic conditions (7). The major components that, with some exceptions, form all BMs are type IV collagen, laminin, nidogen, and some proteoglycans (7). Several collageneous components such as collagen type I, III, and V (2) have been described to be present in the bronchial BM zone, and in asthma higher levels have been reported for type III collagen and fibronectin (2).

The glycoproteins fibronectin (Fn) and tenascin (Tn) occur in some BMs, but their role is unclear. Because they may be expressed at this location during organogenesis, but usually not in adults, it remains unknown whether or not they represent integral components of the BM (7, 10). Tn is expressed during embryogenesis, usually in a spatially and temporally restricted manner, at the epithelial-mesenchymal border (13- 15). The original Tn, now called Tn-C, was the first member of a family of proteins coded by related genes and now including at least Tn-C, -R, and -X (13). Tn-C has been extensively studied in various organs (13) and has been described as variably expressed in the BM zone of bronchi (16).

Epithelial damage and inflammation in asthma may be associated with a wound-repair process in the airways. Wound healing is characterized by the appearance of many ECM components in a spatial and sequential pattern (10, 19). Many studies have shown that Tn, which is usually scarce in adult tissues but is abundant during embryogenesis and in tumor tissues, is strongly expressed in healing wounds (13, 19). Such repair processes and also developmental processes share common features, indicating that during inflammation and wound healing, mechanisms functioning during embryogenesis may be recapitulated (19, 20).

In the present study we have analyzed whether occurrence of the developmentally important glycoproteins Tn and Fn is enhanced in the inflamed airways of asthmatic patients in comparison with levels in normal control subjects. In addition we selected a group of patients with seasonal birch-pollen-sensitive asthma whose disease was activated during the study by exposure to natural birch-pollen antigen to test the effect of inhaled corticosteroid on ECM composition and inflammatory cells in the airways of these patients.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

The study population consisted of three groups (Table 1): two groups of asthmatic patients, including patients with seasonal and those with chronic asthma, with normal control subjects included for comparison. Any subjects having had respiratory tract infections within 6 wk of the study were excluded. Treatment with any form of glucocorticoids was ceased at least 12 wk before and treatment with inhaled nedocromil sodium 8 wk before the study. None of the asthmatics or control subjects had smoked during the 5 yr preceding the study.

Patients with Chronic Asthma

For the 32 patients with chronic asthma, clinical severity varied from mild to severe (2) according to the Aas 5 score scale (22), for number and duration of asthma episodes, of symptom-free intervals, and the requirement for medication during the previous year. The duration of asthma ranged from 1 to 44 yr (9.3 ± 1.7, mean ± SEM). The patients had daily asthmatic symptoms despite antiasthma treatment and displayed at least 20% variability between morning and evening peak expiratory flow rates (PEFR). Their airway obstruction was reversible, with an increase in FEV₁ from 15 to 104% (29.6 ± 3.3%) after inhalation of 200 mg salbutamol (Glaxo Operations UK Ltd, London, UK). Of these 32 patients, 16 were atopic as determined by their clinical history and skin-prick testing performed with 12 ALK allergens (Allergologisk Laboratorium A/S, Horsholm, Denmark) generally used on the skin. All patients were tested to reveal nonspecific airway hyperreactivity (23). The provocation dose causing a reduction of 15% in FEV₁ (PD₁₅FEV₁) ranged from 0.025 to 0.52 mg histamine (0.096 ± 0.023).

Patients with Seasonal Asthma

Seventeen patients with mild seasonal asthma were also included. The duration of symptomatic seasonal asthma had been between 2 and 12 yr (5.3 ± 0.7). The diagnosis was based on a clinical history of seasonal symptoms occurring during the birch pollen season and on a skin-prick test showing a positive reaction > 3 mm in diameter for birch pollen (ALK reagents). The birch-pollen-specific serum IgE, measured by radioallergosorbent test (RAST), was present in all these patients. The FEV₁ was more than 70% of that predicted for patients with seasonal asthma outside of the pollen season.

Control Subjects

Twelve control subjects displayed no history of any respiratory or systemic disease. Clinical examination revealed no current medical symptoms. None had shown any abnormalities in routine hematology. None of the control subjects was allergic as judged by history and by negative results of skin-prick tests done with a range of 12 commonly used allergens (ALK). Lung-function tests were normal in all control subjects, and no significant diurnal variability appeared in PEFR tested during a 2-wk period. There was less than a 15% increase in FEV₁ in a bronchodilator test performed three times with 200 µg inhaled rimiterol (3M Health Care Ltd, Loughborough, UK).

The study was approved by local ethics committees, and written informed consent was obtained from each participant.

Bronchoscopy and Bronchial Biopsies

Bronchoscopic examination and biopsy taking conformed to international guidelines (24). Premedication was either atropine (0.5 to 1 mg) or scopolamine (0.3 to 0.4 mg) and droperidol (5 to 10 mg). For local anesthesia, 2% lignocaine spray or lignocaine solution was used before the procedure, followed by instillation of 2% lignocaine solution through the bronchoscope into the bronchial tree. The fiberoptic bronchoscopes were Olympus BF-20 or BT-IT-20D (Olympus Optical Co., Tokyo, Japan). Bronchial biopsies were taken from the right upper and middle lobe bronchi. In control subjects and patients with chronic asthma, the specimens were taken with Olympus FB-19C forceps, and in patients with seasonal asthma, Olympus FB-15K forceps were used. All specimens were coded before analysis so that their origin remained unknown to the observer.

Design of Treatment Study during Pollen Season

In addition to studying airway ECM and inflammatory cells in patients with chronic and those with seasonal asthma, we also investigated the effect of inhaled corticosteroids. The study was scheduled to include the birch pollen season in Sweden. The first biopsies were taken before the start of medication and before the birch pollen season had begun in Sweden in April. The second biopsies were obtained after 4 to 6 wk of treatment. The patients were randomized in double-blind fashion to two parallel treatment groups, similar with regard to demographic and clinical data. Each group received either budesonide (by Turbuhaler^®, 400 µg/dose; Astra Draco AB, Lund, Sweden) twice daily or placebo (lactose, 200 µg/dose, delivered by Turbuhaler) according to similar protocols. The treatment period started at Visit 1 and continued for 4 to 6 wk until Visit 2. Those who previously had been prescribed a ₂-agonist other than the Bricanyl Turbuhaler^® (500 µg/dose, Astra Draco) were allowed during the study to use the same ₂-agonist and the same dosage. Treatment with a regular nebulized ₂-agonist and/or disodium cromoglycate was, however, discontinued when the patients entered the study. During exacerbations, nebulized ₂-agonists were allowed at the clinic, though for only 24 h. The patients were withdrawn from the trial if another antiasthmatic medication was used during the study period. Lung function and laboratory data were measured, with FVC and FEV₁ measurements performed by VICA TEST 5 (Siemens-Elema AB, Solna, Sweden) and PEFR registered with Mini-Wright peak-flow meters (Clement Clarke International, London, UK) three times for each patient, and the highest values were registered. Use of ₂-agonists was avoided within 6 h before spirometry and before PEFR measurements when possible.

Sample Processing and Immunohistochemistry

Biopsy specimens were snap-frozen in liquid nitrogen and stored at 70° C until sectioning. The samples were embedded in Tissue Tek, ornityl carbamyl transferase (O.C.T.), medium (Miles Inc., Elkhart, IN) and 5-µm serial sections were cut on a Leitz 1720 Digital Cryostat (Ernst Leitz GmbH, Wetzlar, Germany). The sections were air-dried, fixed in acetone, and cooled to 20° C for 10 min.

Detection of tenascin and fibronectin isoforms. Three different mAbs against Tn were used: the mouse mAb 100EB2, which recognizes the fourth and fifth fibronectinlike domains in the Tn-C molecule, the mouse mAb BC-2, which recognizes the high molecular-weight isoform of Tn-C, and the mouse mAb BC-4, which recognizes all Tn isoforms (25). The following mAbs were used to study the expression of specific isoforms of Fn: BF12, recognizing all Fn isoforms (12), 52DHI (12) recognizing EDA-Fn, and BC-1 (26) recognizing EDB-Fn, and FDC-6 (27) recognizing oncofetal fibronectin (Onc-Fn). The mAbs BC-1, BC-2, and BC-4 were kindly provided by Prof. L. Zardi (Genoa, Italy) and FDC-6 by Prof S.-I. Hakomori (Seattle, WA). In the indirect immunofluorescence technique, the specimens were first exposed to the primary mAb at room temperature for 30 min after being washed in phosphate-buffered saline, the sections were incubated with fluorescein isothiocyanate, FITC-coupled sheep antimouse IgG (1:50; Jackson Immunosearch Laboratories, West Grove, PA) for 30 min. The sections were mounted in sodium veronal-buffered glycerol (1:1 at pH 8.4) and examined under a Leitz Aristoplan fluorescence microscope equipped with an appropriate filter system for FITC fluorescence. Negative controls were obtained by omission of the primary antibody or by replacement with a buffer or an irrelevant mAb. For this purpose we used mAb against -amino-butyric acid produced in our laboratory.

To study the relationship of Tn immunoreactivity to BM selected samples were examined by double-label immunostaining by exposure of specimens first to Mab 100EB2 followed by the FITC-conjugate, then to rabbit antiserum to laminin followed by the tetramethyl rhodamine isothiocyanate-coupled goat anti-rabbit IgG antibody (Jackson Laboratories). Polyclonal antiserum against EHS-laminin was from Dr. P. Liesi (Helsinki, Finland).

Detection of inflammatory cells. Eosinophils were stained with the mAb EG2 (Kabi Pharmacia Diagnostics AB, Uppsala, Sweden), which recognizes the cleaved form of eosinophil cationic protein (ECP). The mAb AA1 (M 7052; Dako, A/S, Glostrup, Denmark) developed against human mast cell tryptase was applied for detection of mast cells. T-lymphocytes were marked with an anti-CD3 mAb (M 0756; Dako) and macrophages with mAb Ber-MAC3 (M 0794; Dako). The dilutions used were 1:50, 1:500, 1:1,000, and 1:50, respectively. The alkaline phosphatase, antialkaline phosphatase (APAAP) technique was used in accordance with the manufacturer's instructions (Dako). The color reaction was developed by means of a substrate system based on New Fuchsin (Sigma, St. Louis, MO). Endogenous alkaline phosphate activity was blocked by 1 M levamisole (Sigma). The sections were briefly counterstained with Mayer's hemalan (Merck, Darmstadt, Germany), mounted, and examined under a Leitz Dialux 22 EB light microscope.

Quantification of Tn and inflammatory cells. All quantitative measurements of Tn were done with Mab 100EB2-induced immunofluorescence. One specimen from each patient was examined morphometrically. The positively stained cross sections of BM areas were photographed at preliminary magnification ×80 with Kodak T-MAX 400/800 black and white film (Eastman Kodak Company, Rochester, NY). Sections where the BM zone was not cut crosswise were excluded. Paper photocopies were made at a magnification ×643. For quantification, a 42 × 60-inch Kurta IS/THREE digitizing tablet (Kurta Corp., Phoenix, AZ) linked to a computer was applied. With a pointing device, the upper and lower margins of the immunostained BM area were pointed out in the photographs from 152 to 460 points in the specimen; mean, 306 points. Data were computed by AutoCad program, version 10.1 (Autodesk Inc., Sausalito, CA). Minimal distances between each point on the superficial limit and the closest to each on the deeper border of the stained BM were calculated by the AutoCad program, and the mean value was considered to be the thickness of the Tn staining. Thickness was also calculated by the area and length of the specifically stained BM portion. There was no significant difference between values obtained by either method.

Inflammatory cells were counted throughout the entire section area of the bronchial biopsies by means of the digitizing table and the pointing device described above. For each cell type, the complete sections were photographed on Kodak Ektachrome^® EPN 100 color slide film at an original magnification of ×16. The slides were projected onto the digitizing tablet to reach a final magnification of ×695. For proper calibration of the tablet, a microscopic scale was photographed at the same original magnification. The area of both the epithelium and lamina propria as well as the number of specifically stained cells were transferred into the computer. Cell densities were calculated with the AutoCad program.

Sodium dodecyl-sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting. For the detection of Tn and Fn isoforms, bronchial biopsy tissue pieces were homogenized in a buffer containing 0.5% sodium deoxycholate (E. Merck, Darmstadt, Germany), 150 mM-NaCl, and 1 mM phenylmethylsulfonyl fluoride (Sigma) in 10 mM TRIS-HCl at pH 8.0 at 0° C for 3 × 10 min to achieve enrichment of ECM material as described earlier (28). After being washed in the buffer, the centrifuged pellets were lysed in electrophoresis sample buffer by being boiled for 5 min. The polypeptides were separated by SDS-PAGE gel electrophoresis in 6.5% slab gels under reducing conditions. Thereafter, the proteins were transferred to nitrocellulose sheets (Millipore, Bedford, MA). The following monoclonal antibodies were used: mAb BF12 to total-Fn, 52DH1 to EDA-Fn, BC-1 to EDB-Fn, FDC-6 to onc-Fn, 100EB2 to all Tn isoforms, and BC-2 to high Mr Tn isoform. After exposure to the mAbs the specimens were exposed to peroxidase-coupled rabbit antimouse IgG antiserum (Dako) and underwent a reaction to peroxidase.

Statistical Analysis

The results were analyzed by the Kruskal-Wallis one-way analysis of variance (ANOVA) by ranks. Where appropriate, Dunn's test for multiple comparisons was used to study the difference between control and patient groups. The correlation between the expression of Tn and cell counts, as well as clinical parameters, was analyzed using Spearman's rank correlation test. The statistical significance was defined as p < 0.05. In the treatment study, the Mann-Whitney U test with a normal two-tail approximation was applied to the difference between the budesonide and the placebo groups. The changes within both treatment groups were subjected to Wilcoxon's signed rank test for matched pairs.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Tn and Fn in the BM

Tn immunoreactivity was absent in half of the control subjects. When Tn immunoreactivity could be detected in association with the BM, it was seen as a thin interrupted line located beneath the epithelium. The BM layer of every patient with asthma showed a strong immunoreactivity for Tn as a broad and continuous band that was significantly thicker than that of the control subjects (Figure 1). The thickness of this Tn immunoreactivity band in BM is shown in Figure 2. All three mAbs to Tn gave similar staining results. In both patients with seasonal and those with chronic asthma, the Tn staining was localized in the BM layer, although at a lesser staining intensity, some immunoreactivity could be seen spreading to the stroma. Other structures such as smooth muscle, bronchial glands, and blood vessels stained both in control and in patient groups. In double-immunostaining experiments, an unaltered laminin immunoreactivity was visible just above the broad Tn-reactive zone in the asthmatic patients (Figure 3).

View larger version (76K): [in this window] [in a new window]   Figure 1.   Light microscopy, indirect immunofluorescence method. Bronchial biopsy specimens stained with mAb 100EB2 to detect tenascin. Top panel: Airway mucosa from a control subject with intact airway epithelium (E). No immunoreactivity to Tn detectable in the BM (arrow). Middle panel: Strong Tn immunoreactivity visible in subepithelial BM layer (arrowheads) in a patient with chronic asthma. Bottom panel: Tn immunoreactivity in bronchial BM layer (arrowheads) of seasonal asthmatic out of birch-pollen season. In the lamina propria, bronchial smooth muscle (SM) stains positive for Tn. E = epithelium. Original magnification: ×250.

View larger version (18K): [in this window] [in a new window]   Figure 2.   Individual data in thicknesses of Tn immunoreactivity in the subepithelial airway BMs in control subjects, and in patients with seasonal or chronic asthma, in micrometers (µm). *p < 0.01; **p < 0.0001.

View larger version (116K): [in this window] [in a new window]   Figure 3.   Light microscopy. Bronchial biopsy specimen from a patient with seasonal asthma. Double immunostaining of the same section using different filters. Top panel: Immunoreactivity against laminin is demonstrated as a thin line (arrowheads) beneath the epithelium (E). Long arrows mark the nonstained lower border of the lamina reticularis. Bottom panel: Tn immunoreactivity (arrows) in the lamina reticularis is shown. L = airway lumen, E = epithelium. Original magnification: ×400.

Total-Fn and EDA-Fn immunoreactivity was visible diffusely in the lamina propria of both the control and patient groups without clear accumulation in the BM layer. Because of their diffuse staining patterns, however, no quantitation of their expression could be made. The EDB-Fn and onc-Fn in the mucosa both of the control subjects and of the patients were negative.

Inflammatory Cells

The greatest number of EG2-positive eosinophils in the bronchial mucosa was observed in patients with chronic asthma (Table 2). Mucosal mast cells were present in the entire study population, with the greatest number of mast cells visible in the patients with chronic asthma, differing significantly from those of the control subjects. CD3+ T-lymphocytes were detected in all patients but in only eight of the 12 control subjects, with their highest densities detected in the patients with chronic asthma, which demonstrated a significant difference from the control subjects. The highest densities of macrophages were found in patients with chronic asthma, significantly higher than among the control subjects (Table 2). In the patients with chronic asthma, the numbers of mast cells, T-lymphocytes, and macrophages in the bronchial mucosa were significantly higher than those in the patients with seasonal asthma (p < 0.01, p < 0.01, and p < 0.001, respectively).

Among the patients with chronic and seasonal asthma, Tn expression did not correlate with the number of any cell type in the bronchial mucosa. Nor was any correlation found between Tn expression and any of the clinical and lung function parameters.

Western Blotting

In Western blotting, distinct polypeptides of Mr  250,000 and  190,000 were seen in biopsy specimens from both patients with chronic and those with seasonal asthma with mAb 100EB2 (Figure 4). In control specimens, however, reactivity was either absent or too low to be detected by Western blotting. For both control subjects and asthmatics a different distinct polypeptide of Mr  240,000 was found in all specimens with mAb 52DH1 reacting with EDA-Fn (Figure 4), and even more distinct polypeptides of similar molecular weight were found with mAb 52BF12, also reacting with Fn deposited from plasma (Figure 4). Instead, neither EDB-Fn nor onc-Fn (Figure 4) were detectable in Western blotting.

View larger version (54K): [in this window] [in a new window]   Figure 4.   Western blotting of SDS-PAGE separated polypeptides of enriched extracellular matrix of biopsy specimen from patient with asthma stained for Tn and Fn isoforms. Molecular weight markers are shown in kilodaltons in lane 1. The mAb 100EB2 to tenascin reveals polypeptides of Mr 250,000 and 190,000 (lane 2), mAb BC-2 reveals the polypeptide of Mr 250,000 corresponding to the high molecular form of tenascin (lane 3), mAb 52DHI to EDA-Fn reveals a faint polypeptide of Mr 240,000 (lane 4), mAb 52BF12 to total Fn shows a distinct protein band of Mr 240,000. mAb FDC-6 to onc-Fn is negative.

Effect of Inhaled Budesonide Treatment

When subject characteristics were compared between the two treatment groups (Table 3), no statistically significant difference appeared in any parameter. No major adverse events occurred during the study. There were no significant changes in FVC or FEV₁ among the budesonide-treated or placebo-treated patients. Differences between changes caused by the two treatments did not reach significance.

From a total of 14 asthmatic patients, bronchial biopsies were obtained both before and after budesonide treatment. The mean thickness of the expression of Tn in the BM of these 14 before treatment was 6.1 ± 0.51 µm (range, 1.7 to 8.2). There was no significant difference between the two treatment groups before the beginning of the treatment (p = 0.097). Budesonide (n = 7) reduced significantly the thickness of subepithelial Tn staining, from 6.7 ± 0.7 µm (range, 3.9 to 8.2) to 4.1 ± 0.6 µm (range, 2.3 to 7.1) (p = 0.009). Such reduction in Tn thickness occurred in all seven patients receiving budesonide (Figure 5). The average decrease in thickness of Tn expression was 2.68 ± 0.53 µm in budesonide treatment, whereas the placebo group showed an increase of 1.21 ± 1.02 µm. The difference between budesonide and placebo treatments was significant (p = 0.013) (Figure 6).

View larger version (103K): [in this window] [in a new window]   Figure 5.   Effect of treatment on Tn immunoreactivity in bronchial BM. Top panel: Airway mucosa of a seasonal asthmatic after treatment with inhaled budesonide is shown. Only faint Tn immunoreactivity (arrowheads) was detected under the epithelium (E). Bottom panel: In contrast, the placebo-treated asthmatic shows strong immunoreactivity against Tn in BM layer (arrowheads). Original magnification: ×250.

View larger version (21K): [in this window] [in a new window]   Figure 6.   Effect of budesonide and placebo treatment on tenascin expression in the BM of patients with birch-pollen allergic asthma during the pollen season. Individual responses are given. The difference between the treatments was significant (p = 0.01).

The two treatment groups were similar with regard to the density of all inflammatory cells (p > 0.05). There was a significant reduction in CD3-positive T-lymphocytes in the bronchial mucosa in the budesonide-treated group (p = 0.03), but the tendency towards a decrease in the density of EG2-positive eosinophils, mast cells, and macrophages in the lamina propria did not reach statistical significance (Table 4). In the placebo group, all cell types showed an increase in density in the bronchial mucosa, but only the increase in the number of mast cells was statistically significant (p = 0.03). The two treatment groups differed significantly from each other only by the numbers of mast cells in the lamina propria (Table 4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study showed, in the BM layer of normal human bronchi, a minimal or an absence of expression of Tn isoforms. In contrast, asthmatic bronchi showed accumulation of Tn in the BM zone. Western blotting technique demonstrated a typical doublet of Mr 250,000 and Mr 190,000 Tn polypeptides. In double immunostaining, Tn immunoreactivity was visible as a thick band beneath immunoreactivity for laminin, suggesting that it is accumulated in the BM zone beneath the basal lamina. Thus, Tn may contribute to formation of the thickened BM often described as a characteristic feature of asthma (3, 6). In previous studies a variable Tn immunoreaction has been reported in human bronchi (16). Tn immunoreactivity was observed as a narrow variable band under the airway epithelium both in biopsied control subjects and in smokers undergoing resection for lung cancer (16). Similar to this, our study showed a faint focal Tn immunoreactivity in the subepithelial BM in control subjects. However, of 12 control subjects, Tn immunostaining in the BM area was totally negative in six. In double-blind morphometric analyses a much thicker BM-confined Tn-immunoreactive band was detected in asthmatics than in control subjects, suggesting that in asthmatic airways a pathologic process leads either to increased production of Tn or to its decreased degradation.

Epithelial-stromal interactions are important for Tn production. During rat airway branching morphogenesis it has been shown based on in situ hybridization experiments that both epithelial and stromal cells may serve as a source of Tn (14, 15). In the present study the increased Tn immunoreactivity was concentrated close to the epithelial cells, suggesting an abnormality in this region. We have previously reported epithelial cell damage in asthmatic airways (1, 29). Epithelial cell damage may reflect altered turnover of the cells and cause altered signaling between epithelial cell surface receptors and the stroma, perhaps mimicking events during morphogenesis. Accordingly, we have recently shown in the bronchial BM of asthmatics expression of laminin 2 and 2 chains (9), which are normally present only during early embryonic development (8). The marked accumulation of Tn in asthma may indicate that wound-healing and tissue-remodeling processes in asthma are incomplete. This hypothesis is supported by a recent study showing that during cutaneous wound-healing in mice, Tn-C mRNA returned to low levels only when the wound repair was complete (21). In this respect it is notable that Tn content has been demonstrated to be prognostic for the common variety of interstitial pneumonia (18).

In the present study, Tn immunoreactivity in the BM region was thicker in patients with chronic asthma whose disease was more severe than in the BM patients with seasonal asthma and symptoms related only to the pollen season. Patients with chronic asthma had the most EG2-positive eosinophils, T-lymphocytes, and macrophages in the airway mucosa. Cytokines such as TGF- and IL-1 are thus far the major known upregulators of Tn and Fn synthesis and of their integrin receptors in various cultured cell models (28, 30, 31). Although the inflammatory cells in the airways may contribute to Tn production by releasing TGF- (32), the numbers of lymphocytes or EG2-positive eosinophils in the present study did not correlate with level of Tn expression. At the cell-culture level, TGF- induces in bronchial epithelial cell as well as in other cell-culture models a concomitant production of both Tn and Fn (28). As no clear upregulation of Fn isoforms was found in our study, we suggest that mechanisms other than inflammatory ones may be involved in enhanced Tn expression and that still unknown mechanisms cause activation of epithelial or some mesenchymal cells to produce Tn in asthma.

The restricted distribution of Tn in adult tissues and its higher expression in pathological situations (13, 16) each suggests mechanism to downregulate and upregulate Tn expression in adult organs. In the present study we were able to show an increased Tn expression during natural birch pollen exposure in the group receiving placebo, whereas inhaled budesonide treatment decreased bronchial BM-associated Tn. Such a decrease in thickness of the subepithelial Tn band could be related to the effects of corticosteroids reported earlier at the morphologic level (33). Inhaled steroids have been shown to ameliorate inflammation and restore a normal ciliated epithelium (33), findings in agreement with those of long-term clinical studies showing that inhaled steroids have clear effects on clinical parameters of asthma (34). Although in the present study Tn immunostaining decreased after steroid treatment, it remains unknown if these changes were accompanied by a decrease in the overall thickness of the BM visible in routine hematoxylin-eosin or otherwise stained sections. Earlier, dexamethasone has been shown to downregulate Tn mRNA expression in bone marrow culture (35). Recently, in mice, dexamethasone was shown to diminish, both in vivo and in cell culture, Tn-C expression during wound-healing. The expressions of nidogen and fibulin, other extracellular matrix proteins, were, however, not diminished, suggesting that differing regulatory patterns exist, and the involvement of these proteins during would repair may also differ (21). Such direct down regulatory effects of glucocorticoids on the production of Tn remain to be shown in human airway cell cultures, although in some other cell models they appear to increase ECM production. Thus, it is possible that corticosteroids not only downregulate inflammation but have other as-yet-unknown effects on epithelial cells and ECM protein synthesis and regulation.

We conclude that Tn is accumulated in the BM layer of asthmatic bronchi. The inflammatory cells may not directly influence Tn production because the best-known upregulator of Tn synthesis, TGF-, would have increased local production of Fn as well. However, airway epithelial cell damage and inflammation may be important pathologic components leading to a remodeling process in the airways, which induces Tn synthesis as a step into the healing cascade. The cause of any selective Tn accumulation is unknown. It may be related to the type of interactive functions between epithelium and stroma, which are similar to those during development of the airways reflecting alterations in the homeostasis between the epithelium and its cell-surface receptors as well as their ligand molecules in the stroma. Tn is not specific for asthma, but it has been described in conjunction with many other diseases involving ongoing remodeling processes. Our results showed that Tn expression was higher in the patients with chronic and severe asthma. Corticosteroid treatment decreased Tn expression in the birch-pollen-sensitive asthmatics. Tn accumulation in asthmatic bronchi may reflect incomplete healing and remodeling in the airways rather than injury itself, and thus it may serve as a marker to detect disease activity in asthma.