Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Asthma is associated with a Th2 phenotype in which interleukin (IL)-4, IL-5, and IL-13 are predominant and are produced by different cell types. In murine models of asthma, these cytokines are also upregulated and accompanied by airway eosinophilia, bronchopulmonary hyperreactivity (BHR), and excessive mucus accumulation. Mucins bind allergens, bacteria, viruses, toxins, or particles, which are naturally removed by mucociliary transport (3), and thus protect the epithelium. In airway obstructive diseases, mucus is not only abundantly produced and secreted, but is poorly cleared, causing airway plugging and bronchial obstruction. Mucin composition and concentration are disturbed, with increased gel viscosity (3), which contributes to the morbidity and mortality of asthma or other hypersecretory diseases (3, 4). Mucins are a mixture of mucin gene products present in the respiratory tract; at least seven MUC genes are expressed in the human airways, but three generally predominate: MUC5B, MUC1, and MUC5AC, whose products are thought to be the main gel-forming mucin in respiratory secretions (3).

We have developed a model (1, 2) using BP2 mice (5), which readily exhibit BHR upon a single Ova challenge. This model has not been fully exploited with respect to correlations between eosinophilia, BHR, mucus gene expression, and accumulation. Because of the reported involvement of IL-13 with mucosal metaplasia and BHR (6, 7), here we compared Ova and rmIL-13 for their ability to induce MUC5AC expression and mucus accumulation. In addition, mucus secretion and release into the bronchoalveolar lavage fluid (BALF) was studied after methacholine challenge, and also BHR in immunized/nonimmunized BP2 mice. The role of granulocytes was investigated after their depletion by the myelotoxic drug vinblastine or the antigranulocyte antibody RB6-8C5 (9). These effects were compared with those of dexamethasone. These investigations have allowed us to draw several conclusions about the role of IL-13 and granulocytes in BHR and mucus accumulation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals, Immunization, and Materials

BP2 male mice (Centre d'Elevage R. Janvier, France), aged 6-8 wk, were immunized subcutaneously twice at weekly intervals with 1 µg of Ova (Immunobiologicals, Lisle, IL) and 1.6 mg of aluminium hydroxide (Merck, Darmstadt, Germany) in 0.4 ml 0.9% NaCl (saline). One week after the second immunization, mice were anesthetized with xylazine 12% (20 mg/kg) and ketamine 500 (45 mg/kg) and groups of five were intratracheally challenged with Ova (10 µg) or rm-IL13 (1 or 4 µg/d for 3 d, or 4 µg) in 50 µl of endotoxin-free 0.9% NaCl, or with saline as a control. LPS content was below 1.4 pg per mouse according to previous studies (8). Mucus secretion was triggered with methacholine (Acetyl--methylcholine chloride, 60 mM for 90 sec., or a 12 mg/ml solution; Aldrich Chemical Co., Stoneheim, Germany), thus allowing the evaluation of BHR, or with pilocarpine (Sigma, St. Louis, MO) (150 mg/kg, intraperitoneally). Mice were treated with dexamethasone (sodium phosphate; Sigma) (1.25 mg/kg, intraperitoneally) at 18 h, 1 h, (t₀: challenge), t₀ + 6 h, t₀ + 24 h, t₀ + 48 h, and t₀ + 72 h, to cover the whole duration of the experiment (Figure 1B). Granulocytes were depleted using vinblastine, a vinca alcaloid (5 mg/kg, intravenously.; Velbé, Lilly, France), which induces granulocytopenia 72-120 h after treatment (9). A second injection (4 mg/kg) was performed 48 h after the first one for longer experiments (Figure 1A). Independantly, an anti- granulocyte rat antibody RB6-8C5, (200 µg/ mouse, intraperitoneally) was also used 24 h before challenge to deplete the granulocytes (2, 9). Mice were killed 72 or 96 h after Ova or rmIL-13.

View larger version (24K): [in this window] [in a new window]   Figure 1.   (A) Protocol used for the study of the effects of granulocyte depletion with vinblastine (VIN) (5 mg/kg intravenously). Granulocytes were depleted during 72 to 120 h after a single injection. For longer experiments (144 h), a second dose of 4 mg/kg was injected 48 h after the first one to cover a period of up to 160 h. The Ova or rmIL-13 challenge was performed 72 h after the first injection, when depletion was optimal. (B) Protocol for the repeated treatment of mice by dexamethasone at the arrows (1.25 mg/kg, intraperitoneally).

Evaluation of Bronchopulmonary Hyperreactivity

Airways resistance was assessed in conscious animals by barometric plethysmography (Buxco Electronics Inc., Troy, NY), bronchial hyperreactivity (BHR) being evaluated using the noncumulative methacholine challenges (10). Briefly, mice were placed in a Buxco chamber and respiratory parameters were evaluated after a methacholine aerosol (90 s, 60 mM). Penh (bronchopulmonary resistance, expressed as enhanced pause) = {(expiratory time/relaxation time)  1)} × (peak expiratory flow/ peak inspiratory flow) according to the manufacturer's recommendations (Version 1.5.7; BioSystem XA, Buxco Electronics). Means of 3 data/min for 10 min were recovered. The data are presented as cumulated "areas under the curve" (A.U.C.), allowing a quantitative representation of the results obtained during the 10 min of the experiment.

Bronchoalveolar Lavage Fluid

Mice were anesthetized i.p. with urethane (45 mg/30 g body wt) and the trachea was cannulated. Bronchoalveolar lavage fluid (BALF) was collected with 0.5 ml followed by 3 × 1 ml of PBS containing PMSF (0.005 M; Sigma), EDTA (sodium salt, 0.005 M; Sigma), and dithtiothreitol (0.005 M). Total cell numbers were determined with a Coulter counter and cytospins were stained with Diff Quick (Baxter Dade AG, Düdingen, Switzerland) for differential cell count.

Lung Preparation

Lungs were isolated and perfused with saline via the pulmonary artery. Dispersion was performed with an Ultraturrax (T25 Janke and Kunkel, IKA_R-Labortechnik, Germany) for 30 sec. in 2 ml ice-cold PBS containing antiproteases (PMSF, 0.005 M, EDTA, 0.005 M, and DTT 0.005 M (Sigma)), then centrifuged at 15 000 g for 30 min. The supernatants were precipitated in 70% ethanol and solubilized by sonication. Protein content was determined by OD (205, 280 nm). For gene expression studies, lungs were homogenized in 2 ml RNazol solution (Bioprobe Systems), then centrifuged. RNA_s were extracted as described (9).

Reverse Transcriptase-Polymerase Chain Reaction

Intron-differential reverse transcriptase-polymerase chain reaction (RT-PCR) was performed for lungs, using specific primers: for IL-4, 5' GAGCCATATCCACGGATGCGACAA and 3' CATGGTGGCTCAGTACTACGAGTA (11); for IL-13, P600 (EMBL M23504) 5' CCAGACTCCCCTGTGCAACGATT3' TGTTATAAAGTGGGCTACTTCGATT; for MUC5A (EMBL 10792), two couples were used (in coding sequence), either 5' GTGCCCCGGCCTTGTCGACC (1219-1238) and 3' GAG GA AACACATTGCACCGATCC (1487-1462), or 5' TCTGTAA GGAAGCCACGCTAAC (1643-1665) and 3' AAAGGGCAG GTCTTCGGTATA (2058-2037), and -actin as a control, 5' ACT CCTATGTGGGTGACGAGG and 3' GGGAGAGCATAGC CCTCGTAGAT (11). The cDNAs were synthesized as described (9, 11). PCR was performed varying cDNA dilutions and cycle number (30) at 58-64°C according to primers. PCR products were resolved on a 1.5% agarose gel containing 0.5 µg/ml ethidium bromide (Sigma) and fluorescence under UV illumination was amplified and recorded using an Ultra-Lum system (Paramount, CA). Semi-quantification was achieved on ImageQuant (Molecular Dynamics, Sunnywale, CA), comparing cytokine to -actin levels. Specificity was checked by restriction digestion of PCR products.

Determination of Acidic Sulfo- and Sialomucins, and Neutral Mucins

Acidic sulfo- and sialomucins were determined by an adaptation of the method using Alcian Blue at pH= 2.5, described for intestinal mucins (12) (OD method). Proteins from the cell-free BALF were precipitated in 70% ethanol, solubilized by sonication, and their concentration determined by OD. Acidic mucins were then coprecipitated in the modified Carnoy's reagent, from samples preincubated with Alcian Blue (AB) for 10 min (12). After centrifugation, the samples were dissolved by sonication and the OD determined at 630 nm. Standard curves for acidic mucins were obtained using bovine submaxillary gland mucins (BSM: 0.1-0.5 mg), and for neutral mucins using porcine stomach mucins (PSM: 0.05-0.20 mg). Results were quantified with respect to the protein content of the samples. The term "mucins" refers to "mucin equivalents," calculated from the standard curves for BSM or PSM.

For neutral mucins, purified ethanol-precipitated samples were treated as described (13). The specificity of the products measured by OD after Alcian Blue (AB) or Periodic acid Schiff (PAS) staining was checked by electrophoresis (12).

MUC5AC Determination by Enzyme-Linked Immunosorbent Assay

A mouse anti-MUC5AC antibody (14, 15) recognizing the peptide QTSSPNTGKTSTISTT of the murine homolog of the MUC5AC gene, was used. This antibody recognizes the lung, gastric and intestinal MUC5AC mucins (15). Serial dilutions of ethanol-precipitated samples, or cell-free BALF, and serial dilutions of the standard which cross hybridized with mouse MUC5AC (porcine stomach mucins PSM: 1-100 ng up to 500 ng, serial dilutions), and of known positive samples (serial dilutions of mouse gastric and intestine preparations) were coated on 96-well plates (Nunc, Maxisorp, Roskilde, Denmark), incubated for 1 h at 37°C then overnight at 4°C in PBS. The first antibody (anti-MUC5AC, chicken polyclonal antibody) was then incubated for 1 h at 37°C in PBS-0.1% Tween-1% BSA, or nonimmune control antibody. After washes, the second antibody (peroxidase-conjugated rabbit anti-chicken IgY; Agro-bio, La Ferté St Aubin, France) was incubated in PBS for 1 h at 37°C. Color was developed with 3,3', 5,5'-tetramethyl-benzidine (TMB), then OD red at 630/450 nm.

IL-13 Determination by Enzyme-Linked Immunosorbent Assay

A Rat IgG2b monoclonal anti-IL-13 capture antibody was coated on 96-well plates overnight at 4°C, then serially diluted samples, or serial dilutions of rmIL-13 used as a standard, lied upon the coating for 1 h at 37°C. Detection was performed using a polyclonal biotinylated goat IgG anti-mouse IL-13 antibody (antibody pairs match for enzyme-linked immunosorbent assay [ELISA], from R&D systems [Minneapolis, MN], according to their instructions) for 1 h at 37°C. Streptavidin-peroxidase conjugated (Extravidin, Sigma) was applied for 30 min at 37°C, then revealed by TMB as above. ELISA were also used for punctual IL-4 and interferon (IFN)- determinations (2).

Determination of Lung Myeloperoxidase and Eosinoperoxidase Activities

MPO activity in lung tissue was determined as described (9). The correlation coefficient between the cell count and MPO activity was 0.989 (equation of the linear regression: y = 1,385x  16). For EPO, lungs were homogenized for 1 min at 4°C in Tris 0.05M/HCl containing 0.1% Triton X100 at pH= 8 and centrifuged at 10,000 × g for 10 min at 4°C. The supernatants containing hemoglobin were discarded. The pellets were resuspended in 250 µl of distilled water and 50 µl were incubated in duplicates with (or without) 3-amino-1.2.4 triazole at 50 mM (AMT; Sigma). Ten nanomolars O-phenylenediamine dihydrochloride (OPD, Sigma) were added in the dark to the buffer containing 4 mM H₂O₂. After 1 h of incubation at 37°C, absorbance was read at 490 nm.

Histology

The lungs were flushed, then inflated with OCT (Optimum Cutting Temperature medium; Sakura Finetck, Torrance, CA), half diluted in saline. They were immersed in 10% formaldehyde in PBS overnight at 4°C, and processed to paraffin wax. Sections (5 µm) were stained either with AB/safranine, PAS/hematoxylin, or with their combination, AB/PAS, in standardized conditions (16). A Harris' hematoxylin/eosin staining was also performed for cell analysis. Semiquantitative 4-points scoring system (+ to ++++) was used to assess the intensity of staining of mucous cells at the same position of the main bronchiole, using a grid for counting the same length, at a ×200 magnification.

Statistical Analysis

Results are presented as means ± standard deviation. Significance levels were calculated using one-way ANOVA followed by Scheffe's test, using the SPSS 6.1 software (*significant difference between data with a threshold of P < 0.05).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of the Intratracheal Challenge with Ova on Th1 and Th2 Cytokines, MUC5AC Expression, and Drug Modulation

The lungs from immunized saline-challenged or non-challenged mice, as those from nonimmunized Ova-challenged mice, showed only traces of mRNA_s for IFN-, IL-13 (Figure 2A), of MUC5AC (see below), and of the respective proteins (Figure 2C), in absence of IL-4 mRNA_s. By contrast, the lungs of Ova-challenged immunized mice accumulated IL-4, IL-13, and MUC5AC, in absence of IFN- (Figures 2A and 2C).

View larger version (32K): [in this window] [in a new window]   Figure 2.   Th1/Th2 cytokine profile and MUC5AC expression. (A and B) Time-dependent mRNAs expression for IL-4, IL-13 (Th2), IFN- (Th1), and MUC5AC in murine lungs after the intratracheal challenge with Ova (A), rmIL-13 (B), or saline ( and B). Modulation by vinblastine and dexamethasone. *Statistical significance between data (with a threshold of P < 0.05; n = 5) is observed between saline (Sal) and ovalbumin (Ova) in (A), for MUC5AC, for IL-13 at all time points, and for IL-4 at 6 h only. Vinblastine (Vin) reduced nonsignificantly (up to 20%) the mRNAs expression; dexamethasone (Dex) was effective against the accumulation of IL-4 at 6 h, against IL-13 at all time points, and against MUC5AC mRNAs at 6, 24, and 72 h, but less so at 96 h (P = 0.2). (B) Statistical significance is observed between saline and rmIL-13 for MUC5AC at all time points, but not for IFN-, IL-4, or IL-13, which are not induced, after Ova or rmIL-13. Standard deviations are not presented for clarity of the figure. (C) Time-dependent expression of proteins (by ELISA) after Ova or saline, for (Ca) the Th1/Th2 cytokines IFN-/IL-4; (Cb) for IL-13 and its modulation by vinblastine and dexamethasone at 96 h after Ova; and (Cc) for MUC5AC, in the lungs of immunized BP2 mice (*significance between data with a threshold of P < 0.05; n = 5). Plus signs, IFN-; triangles, IL-4; circles, IL-13; squares, MUC5AC.

A peak of IL-4 mRNAs was observed at 6 h after Ova (Figure 2A), and of the corresponding protein at 24 h (Figure 2Ca), which later disappeared, whereas IL-13 was expressed earlier (6 h) and for a longer period after challenge (Figures 2A and 2Cb). MUC5AC mRNA_s accumulated progressively from 6 h to 24-72 h (Figure 2Cc).

To determine their involvement in the phenotypes observed, granulocytes were depleted from the blood, lungs, and BALF using vinblastine, or the RB6-8C5 antigranulocyte antibody. Under those conditions, no eosinophilic inflammation was triggered by Ova (Figures 3A, 3B, 3C, and 3D) (2, 9), or by rmIL-13 (Figures 3C and 3D). The Ova-induced increased mRNA_s expression for IL-4, IL-13, and MUC5AC and the titers of IL-13 at 96 h were only slightly, but not significantly, reduced by vinblastine (Figures 2A, 2Cb, and 2Cc). Dexamethasone, daily injected at 1.25 mg/kg (see MATERIALS and METHODS), suppressed the early increase of IL-4 and of IL-13 mRNA_s expression in the lungs as well as IL-13 protein at 96 h (Figure 2D), and IL-4 protein at its peak (24 h, not shown). The expression of mRNA_s for MUC5AC was suppressed at 6-24 h (Figure 2A), but only halved later.

View larger version (36K): [in this window] [in a new window]   Figure 3.   (A and B) Interference of vinblastine (Vin or V) or dexamethasone (Dex or D) with (A) neutrophils in the lungs in terms of myeloperoxidase (a), neutrophil count in the BALF (b), and in the blood (c); and (B) eosinophils in the lungs in terms of eosinoperoxidase (a), eosinophil count in the BALF (b), and in the blood (c), 24 h after Ova or Saline (Sal or S) challenge of immunized BP2 mice. (C and D) Interference of the antigranulocyte antibody RB6-8C5 (0.2 mg/mouse; isotype control: Rat IgG2b), of vinblastine (Vin or V), or of dexamethasone (Dex or D, x injections), with neutrophils (C) (myeloperoxidase activity) or eosinophils (D) (eosinoperoxidase activity) in the lungs of Ova or rmIL-13- or Saline (Sal or S)-challenged BP2 mice. A higher dose of the RB6-8C5 antibody (0.8 mg/mouse, not shown) did not improve the depletion.

Mucus Accumulation and Secretion following Methacholine in Ova-Challenged Mice

As indicated below, no mucus-containing cells were identified in the lungs of control mice, from which only traces of acidic mucins or of MUC5AC could be extracted and detected (Figure 5A, second column; and MUC5AC by ELISA, Figure 6A, second column). Starting 48 h after the Ova challenge, mucus-containing cells were detected in the airways, and more intensively so at 72-96 h (Figure 4A and Figure 5A, second column).

View larger version (32K): [in this window] [in a new window]   Figure 5.   Interference of vinblastine and dexamethasone with the accumulation and secretion of acidic mucins into the BALF, 96 h after the intratracheal challenge with Ova (A), rmIL-13 (B), or saline, and 1 h after the inhalation of methacholine (Mch) or of saline. The basal level (saline controls) is indicated by a dotted line. *Significant difference between data with a threshold of P < 0.05; n = 5. Open bars, saline; thinly striped bars, vinblastine; thickly striped bars, dexamethasone.

View larger version (31K): [in this window] [in a new window]   Figure 6.   Interference of vinblastine and dexamethasone with the accumulation of MUC5AC by the lungs and secretion into the BALF, 96 h after the intratracheal challenge with Ova (A), rmIL-13 (B), or saline, and 1 h after methacholine (Mch). (*Significance between data with a threshold of P < 0.05; n = 5). Controls and treatments are the same as described for acidic mucins, Figure 5. Results represent MUC5AC "equivalent to porcine stomach MUC5AC mucin." Open bars, saline; thinly striped bars, vinblastine; thickly striped bars, dexamethasone.

View larger version (125K): [in this window] [in a new window]   View larger version (160K): [in this window] [in a new window]   Figure 4.   Histochemistry of lung sections of immunized BP2 mice (n = 5). The bar at each magnification represents 100 µm. Mucins (glycoproteins) in airways epithelium appear in pink with PAS (neutral mucins). Hematoxylin is used as a counterstain (light purple). (A-D) Ova-challenged mice; (E-H) rmIL-13 (4 µg/d for 3 d) challenged mice. (A) ×400. Ova-instilled mice: metaplasia and hyperplasia of non-ciliated epithelial cells (arrow a). This corresponds to the "full lung" and "empty BALF" states, as determined for mucin content in Figure 5A (second column). Inflammatory cells, essentially eosinophils (arrow b) are observed in the mucosa and submucosa. Lumen obstruction (lower part of the micrograph) occurs with mucus, epithelial thickening, and cell debris. (B) ×400. Ova-instilled mice exposed to methacholine: mucins are secreted from the granules of epithelial cells (arrow a) into the lumen of bronchioles (arrow b) by exocytosis, corresponding to the "empty lung" state in Figure 5A (third column, lower). Cells (arrow c) and debris are associated with mucus. The bronchoalveolar lavage clears out the mucus from the lumen, corresponding to "empty lung" and "full BALF" states (Figure 5A, third column, higher). Mucins are not completely released from the epithelium (50-70% release [3]). (C) ×200. Ova-instilled mice pretreated with vinblastine: though metaplasia and hyperplasia of the mucus cells are observed, as in Ova controls (arrow a), inflammatory cells are absent (arrow b). (D) ×400. Ova-instilled mice treated with dexamethasone show no inflammation, edema, metaplasia, or hyperplasia of epithelial cells. As a consequence, a low mucus accumulation is observed in the lungs ("empty lung") and no secretion is induced ("empty BALF" as in Figure 5A, fifth column). (E) ×400. rmIL-13 instilled mice (daily administration): metaplasia and hyperplasia of epithelial cells are observed. It corresponds to the "full lung" and "empty BALF " states in Figure 5A (second column). (F) ×400. rmIL-13-instilled mice exposed to methacholine, followed by a BAL: methacholine causes the regulated secretion of mucins by exocytosis, from the granules of epithelial cells into the lumen of bronchioles, corresponding to an "empty lung" state in Figure 5A (third column, lower) and to a "full BALF" state (Figure 5A, third column, higher). (G) ×200. Vinblastine-rmIL-13-instilled mice: mucin-accumulating cells are observed, the intensity of staining being near to rmIL-13 alone. No inflammation is detectable. (H) ×400. Dexamethasone-treated and rmIL-13-instilled mice. Traces of mucins are observed. Inflammation is suppressed. (I) ×200. rmIL-13-instilled mice. The lung sections are double-stained (AB/PAS); metaplasia and hyperplasia of epithelial cells are observed. Acidic mucins (in blue) dominate on the apical pole of the cell, and the basal pole is occupied by neutral mucins (in pink), as in case of Ova (see K). (J) ×100. rmIL-13-instilled mice (4 µg/d for 3 d) exposed to methacholine. The lung sections are double-stained with AB/PAS. Secreted mucins in the lumen of the bronchiole show acidic properties (blue staining prevails), though the remaining mucins in epithelial cells are rather neutral (stained in pink by PAS). (K) ×200. Ova-instilled mice; double staining (AB/PAS). Comments as for J. (L) ×200. Control Saline/Ova/Saline on non- immunized mice: only rare traces of mucus are observed after PAS staining. (M) ×400. Control Vinblastine/Saline/Methacholine on immunized or nonimmunized mice: no or only rare traces of mucus. Similar profiles are observed on the controls: Saline/Saline/Saline or Dexamethasone/Saline/Saline on immunized or nonimmunized mice, with no mucus. (N) ×100. rmIL-13-instilled mice as in J. Main bronchiole stained by Alcian blue, specific for acidic mucins (sialo- and sulfomucins). Thickening of the epithelium (arrow a), with doubling of the height of the affected zone as compared with a nonaffected zone (arrow b). Arrow C shows an inflammatory infiltrate around the main bronchioles.

To determine whether the accumulated mucus could be mobilized, mice were exposed to the secretagogue methacholine 72-96 h after the intratracheal provocation with Ova. Under those conditions, acidic mucins left the epithelium and enriched the BALF (Figure 5A, third column). Mucus secretion occurred as early as 15 min after the methacholine inhalation, with a maximal release after 30- 60 min, to progressively decrease at 2, 3, and 6 h (not shown). Accordingly, a 1-h time-point after methacholine was retained for further studies. Pilocarpine (150 mg/kg, intraperitoneally) failed, to a greater extent than did methacholine, to stimulate the mucus release into the BALF (not shown).

The amounts of secreted acidic mucins were increased 2- to 4-fold in the BALF of Ova-challenged mice 72-96 h after exposure to the secretagogue methacholine (Figures 5A and 6A). Similarly, the amount of MUC5AC protein was increased 2-fold (Figure 6A), but neutral mucins were less affected (not shown).

Vinblastine reduced by 20% (nonsignificantly) the expression of the MUC5AC gene (Figure 2A), and slightly affected MUC5AC accumulation and secretion caused by methacholine after Ova (Figure 6A). The global acidic mucin accumulation in the lungs was slightly modulated by vinblastine (Figures 5A). By contrast, dexamethasone drastically prevented the lung accumulation of mucus (Figure 4D) and, as a consequence, its secretion (Figures 5A and 6A, last columns).

BHR

In immunized mice, bronchopulmonary responses to methacholine were increased 2-fold 96 h after Ova (Figure 7A) and 3-fold after 72 h (Figure 8B). As expected, vinblastine, the RB6-8C5 anti-granulocyte antibody (not shown), and dexamethasone all prevented Ova-induced BHR (Figure 7A). In nonimmunized mice, Ova failed to induce BHR after methacholine (Figure 8B).

View larger version (32K): [in this window] [in a new window]   Figure 7.   Interference of vinblastine (V) and dexamethasone (D) with the BHR to methacholine (M) 96 h after the intratracheal challenge with Ova (O), rmIL-13 (13), or saline (S), in immunized BP2 mice. The data are presented as cumulated "areas under the curve" (A.U.C.). Controls were performed without methacholine and correspond to: SSS, DSS, VSS. When methacholine was used, groups were: VSM, DSM, and SSM: Saline/Saline/ Methacholine-treated mice show the basal level with methacholine. SOS: Saline/Ovalbumin/Saline- or S13S: Saline/rmIL-13/Saline-treated mice show the basal Penh level after challenge, in absence of methacholine. Methacholine was used to express BHR and groups are SOM: Saline/Ovalbumin/Methacholine- or S13M: Saline/ rmIL-13/Methacholine-treated animals. VOM: Vinblastine/Ovalbumin/Methacholine- or V13M: Vinblastine/rmIL-13/Methacholine-treated animals show inhibition of BHR by vinblastine after Ova (VOM), but not after rmIL-13 (V13M). DOM: Dexamethasone/Ovalbumin/Methacholine- or D13M: Dexamethasone/rmIL-13/Methacholine-treated animals show the inhibition of BHR by dexamethasone, after Ova or rmIL-13. The arrow indicates the basal resistance. Horizontal dotted line is the basal bronchoconstriction by methacholine (*Significance between data with a threshold of P < 0.05; n = 5). Open bars, saline; thinly striped bars, vinblastine; thickly striped bars, dexamethasone.

View larger version (32K): [in this window] [in a new window]   Figure 8.   Interference of immunization with (A) the accumulation and release of acidic mucins in the BALF of BP2 mice, 72 h after Ova, rmIL-13 or saline, and 1 h after methacholine (B) with Penh values just after methacholine inhalation (*Significance between data with a threshold of P < 0.05; n = 5). Open bars, OVA; shaded bars, saline; thinly striped bars, rmIL-13: 1 × 4 µg; thickly striped bars, rmIL-13: 3 × 1 µg.

Inflammatory Cell Recruitment; Drug Modulation

Saline-saline-, dexamethasone-saline-, and vinblastine-saline- challenged animals showed abundant macrophages in the BALF, and no eosinophils or neutrophils. Accordingly, neither blood macrophage nor lymphocyte counts were affected after dexamethasone-saline or vinblastine-saline treatments. Eosinophils and neutrophils were recruited in the lungs and BALF after the intratracheal challenge with Ova. Eosinophil counts were elevated from 24 h (Figure 3) to 72 h in the BALF, whereas neutrophil counts peaked at 24 h and then vaned (2). Accordingly, controls for depletion were performed at the moment of the challenge (not shown) and 24 and 72 h later. The critical time point was 24 h after the challenge, when both types of cells were consistently depleted (Figure 3). Both vinblastine and dexamethasone suppressed the recruitment of eosinophils and neutrophils into the BALF and into the lungs at 24 h after challenge (Figure 3A-B). At 72h the inhibition was still complete (not shown).

Histochemistry of the Lungs after Ova Challenge; Drug Modulation

The intratracheal instillation of Ova (or rmIL-13) preferentially places Ova (or rmIL-13) onto the larger bronchioles. Accordingly, the effects on mucosal metaplasia were always evaluated at the same position of the major bronchiole in longitudinal sections. Nonimmunized saline- or Ova-challenged mice, as well as immunized and saline-challenged mice, showed no mucus-containing cells in the respiratory epithelium (Figure 4, micrographs L and M). Mucus appeared in secretory granules, at the apical pole of the cytoplasm of nonciliated epithelial Clara cells from the lungs of Ova-challenged animals (Figure 4A) (2). Ninety-six hours after challenge, these cells were intensively stained for mucus, scoring 4+, and the height of the epithelial cells doubled. The ratio between cells producing mucus/ciliated cells raised from 0 up to 1-1.4, corresponding to 60-80% of cells colored for mucus, the counts being performed at the same location of the major bronchiole (Figure 4 and Table 1, n = 5).

Smooth muscle thickening, inflation of the subepithelial areas, cell rearrangements, and edema (Figure 4), as well as lymphoid islets or eosinophil granuloma around the main bronchioles, were also observed (2). The subepithelial thickening after Ova, as after rmIL-13 (see below), was associated with larger epithelial cells, mucus accumulation, and plugs, which led to obstruction.

As observed with double staining (AB/PAS) of the lungs from animals not exposed to methacholine, an intense double staining was noted in the majority of the cells if Alcian Blue is applied for a long time (40 min), suggesting that mucins, which represent different gene products (3), are either superimposed on the same site, or have acidic and basic residues on the same molecule (16) due to their high degree of glycosylations. However, when staining was applied for a shorter time (20 min), the blue/purple double staining was clearly observed at the apices of the cell, though the base remained pink-stained by PAS. After secretion induced by methacholine, residual neutral mucins were preferentially located at the base of the cell, though acidic mucins were clearly released from the cell, possibly by exocytosis (Figure 4K for Ova, and 4I and 4J for rmIL-13). Globally, acidic mucins were predominant in the secretions after Ova provocation.

Even though vinblastine and dexamethasone virtually suppressed inflammation in the lungs of Ova-challenged mice (Figures 4C and 4D), vinblastine failed to significantly affect mucus cell metaplasia in the larger bronchioles, as shown by cell ratios and by the intensity of mucus staining (Figure 4C and Table 1). In parallel, the RB6-8C5 anti-granulocyte antibody slightly decreased the intensity of mucus staining after Ova (not shown). By contrast, dexamethasone markedly downregulated the accumulation of acidic and neutral mucin in the airways (Figures 4D and 5A), and, accordingly, prevented airways obstruction. The intensity of cell staining was reduced (from 4+ [with Ova] to 0.5 to 2+ [with dexamethasone plus Ova]) in the lungs of animals not exposed to methacholine. The presence of morphologic changes such as subepithelial or smooth muscle thickening indicated remodeling, despite dexamethasone.

rmIL-13 Imitates Ova Challenge in Immunized, but Differences Appear in Nonimmunized, Mice

Because of the involvement of IL-13 with mucosal metaplasia and BHR (6, 7), we compared Ova and rmIL-13 for their ability to induce mucins and MUC5AC expression, BHR, and mucus accumulation after challenge and mucus release into the BALF after methacholine. Instilled intratracheally, rmIL-13 failed to induce IL-4, IL-13, or IFN- mRNA_s, or the corresponding proteins (Figures 2B, 2Ca, and 2Cb). As Ova, rmIL-13 induced the expression of MUC5AC mRNA_s, with a signal 3-fold increased in the immunized (Figure 2B), and less so in the nonimmunized, mice (not shown). Vinblastine only slightly decreased this effect (Figure 2B). Dexamethasone inhibited MUC5AC mRNA_s at 6-24 h but was less efficient at 72-96 h, because mRNA_s reappeared at half their maximal levels at 72-96 h (Figure 2B).

Induction of BHR by rmIL-13

As with Ova treatment, rmIL-13 promoted BHR to methacholine (6, 7). Thus, bronchoconstriction in response to methacholine was increased 3-fold, 72 h after the first (or 24 h after the third) daily rmIL-13 administration (3 × 1 µg), both on immunized and on nonimmunized mice (Figure 8B). Seventy-two hours after the single administration of rmIL-13 (5 µg), BHR was still marked (not shown), being less apparent at 96 h. Contrary to Ova (Figure 7A), vinblastine failed to prevent BHR induced by rmIL-13, under conditions in which dexamethasone did so (Figure 7B).

Inflammatory Cell Recruitment; Drug Modulation

The number of nucleated cells recruited into the BALF increased by 50% 72 h after rmIL-13 in immunized, and by 25% only in nonimmunized, mice (not shown); eosinophil and neutrophil numbers also increased by 10-20%. The presence of enlarged alveolar macrophages containing vacuoles suggested activation. As with Ova and dexamethasone, vinblastine and the RB6-8C5 antibody depleted the granulocytes from the BALF and the lungs. Both vinblastine and dexamethasone suppressed the inflammatory cell recruitment (particularly eosinophils and neutrophils) induced by rmIL-13 in the BALF and in the lungs (Figures 3Cc, 3Cd, 3Dc, and 3Dd).

rmIL-13 Promotes Mucin Accumulation, Acidic Mucins, and MUC5AC Protein, Which Are Releasable by Methacholine; Drug Modulation

rmIL-13 induced mucin-containing cells in the airways of immunized mice (Figure 4E). As in case of Ova, the inhalation of methacholine by rmIL-13-treated mice, in which mucosal metaplasia had occurred, was followed by a reduction of mucus content in the lung and by the corresponding enrichment of the BALF (Figures 4E, 4F, 5B, and 6B).Acidic mucins were increased 2-fold in the BALF of immunized BP2 mice 72-96 h after repeated or single instillation of 1 or 4 µg of rmIL-13 (Figures 5B and 8A), as compared with saline, as were neutral mucins, but to a lesser extent (Figures 4E). Vinblastine slightly modified acidic (Figure 5B) or neutral mucin accumulation (Figure 4G) after rmIL-13.

MUC5AC in the BALF, evaluated by ELISA, increased by approximatively 2-fold 96 h after rmIL-13 (Figure 6B). Vinblastine failed to significantly decrease MUC5AC accumulation, but dexamethasone did so (Figure 6B). The latter also downregulated acidic and neutral mucins, as well as MUC5AC synthesis and accumulation in the lungs and, as a consequence, their methacholine-induced release into the BALF (Figures 5B and 6B). In nonimmunized mice, the increase of acidic mucins in the lungs and release into the BALF after methacholine was reduced (Figure 8A) when compared with immunized mice, but a clear metaplasia was observed after AB staining (not shown).

rmIL-13 Induces Mucosal Hyperplasia and Metaplasia of Epithelial Cells; Drug Modulation

rmIL-13 induced mucus accumulation, cell hyperplasia, and metaplasia (Figure 4E) in immunized or in naive mice, but to a lesser extent than in immunized mice (Figure 8A). A lower accumulation of mucins was observed after 1 or 4 µg of rmIL-13 (repeated doses), as compared with 10 µg of Ova, which agrees with the results using the OD method described above (Figures 8A and 5A/B), and with histochemistry (ratio: cells containing mucins/ciliated cells = 0.8-1 only; Table 1).

Acidic mucins also prevailed at the apical pole of the cell and/or in secretions (Figures 4E and 4F). Obstruction of some bronchioles was noted, following epithelium thickening and the reduction of the airways lumen. In parallel, collagen accumulated around airways (not shown), as highlighted by acidic Van Gieson's picrofuschine staining (16), and was not affected by vinblastine, but was reduced by dexamethasone (not shown). Vinblastine only slightly reduced the staining intensity (Figure 4G) or the ratio of mucus cells/ciliated cells in the larger bronchioles (Table 1) after rmIL-13, but the RB6-8C5 anti-granulocyte antibody was effective in reducing the staining intensity induced by ~ 50-60% (ratio 0.5/0.6, not shown). Dexamethasone downregulated mucin accumulation in the lungs: the staining intensity was lowered from 3-4+ to 1.5-2+, in parallel with the number of stained cells (Table 1).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Our results raise questions concerning the relevance of granulocytes and of inflammation for rmIL-13- or Ova- induced BHR and mucus accumulation, the interactions between mucus, BHR, and eosinophilia, and the role of the Th1/Th2 balance. Antigen and one of its potential mediators, IL-13, share analogies, some of which are described (6), whereas marked differences are novel. Such variations are: the predominance of acidic mucins in the secretions after Ova or rmIL-13; their modulation by dexamethasone; the effectiveness of granulocyte depletion against BHR by Ova but not by rmIL-13; and finally, the unique role of IL-13.

Role of Granulocytes in Ova- or rmIL-13-Induced BHR

Vinblastine, which suppresses granulocyte accumulation, abolished Ova-induced BHR (Figure 7A) (2), as did the RB6-8C5 anti-granulocyte antibody. This supports the idea that a granulocyte, possibly the eosinophil (17, 18) which is the most massively recruited granulocyte after Ova challenge, participates in BHR. In contrast to Ova, rmIL-13 alone induced a strong BHR in nonimmunized BP2 mice (Figure 8B), as described for other strains (6, 7), and a moderate (15-20%) BALF eosinophilia and neutrophilia. The effects of rmIL-13 on BHR were unaffected by granulocyte depletion. IL-13 is implicated in the Ova-induced BHR in mice, which is abolished by its neutralization (6).

Granulocyte Depletion and Mucus

Mucus accumulation after rmIL-13 was poorly affected by vinblastine, but was inhibited (50-60%) by 0.2 mg/mouse of RB6-8C5, a dose which suppressed the granulocytes from the lungs. This suggests a supportive, probably not exclusive, role of granulocytes in mediating the effects of rmIL-13 on mucus in mice. After Ova challenge, granulocyte depletion by vinblastine had no or only a slight reducing effect on mucus, as evaluated by histology (10-20%, non significant), and was comparatively more effective when it was induced with the anti-granulocyte antibody (20-40%). This interference remainded slight, as described in a rat nasal model of allergy (19). Indeed, vinblastine may show additional effects on mucus, not addressed here. After Ova, IL-13, if involved, may also be produced by other cells resistant to vinblastine, such as lymphocytes (2) or macrophages (20, 21, 22).

BHR, Eosinophils, and Mucus

Both vinblastine and the RB6-8C5 anti-granulocyte antibody suppressed BHR, eosinophils, and neutrophils in blood, and their recruitment into the lungs and into the BALF upon allergenic challenge (Figure 3). By contrast, mucus accumulation persisted (Figures 4C and 4G), thus dissociating mucus from eosinophilia (2, 23) from neutrophilia and from BHR (Figures 7 and 8B).

Drug Modulation

Steroids are the most effective molecules used in treating asthma (24, 25). Dexamethasone, a standard glucocorticoid, inhibits the synthesis of numerous cytokines, including IL-4 and IL-13 (Figure 2), via transcription factors (25). These cytokines induce MUC5AC expression (4, 26, 27). Dexamethasone also suppresses antigen-induced BHR and eosinophilia in mice (Figure 7A and Refs. 2, 28), probably by acting on genes involved in BHR, on cytokines (such as IL-5, unpublished data), chemokines, and adhesion molecules involved in eosinophilia. We show here that dexamethasone interfered with mucus accumulation, and MUC5AC mRNA_s were reduced (Figure 2A). Because inhibition was incomplete, it may result from insufficient dosage (unlikely, since repeated injections were effective against eosinophilia and BHR) or to a partial resistance to the drug.

Dexamethasone also inhibited rmIL-13-induced BHR. Because it may affect transcription factors involved with early events of gene induction (24, 25), rmIL-13 may thus act indirectly via the activation of other genes or processes responsible for BHR and downregulated by dexamethasone. The latter interfered, though to a lesser extent, with mucus accumulation induced by rmIL-13, also suggesting an indirect effect via the early induction of genes involved in mucus accumulation (3). Other mediators, such as lipoxygenase derivatives, may operate and provide an alternative target for the steroid. 15-lipoxygenase is indeed induced by IL-13 and IL-4 (27, 29) and cysteinyl-leukotrienes may induce eosinophilia and/or mucus accumulation in the airways (4, 27, 29).

Mucus Secretion

We developed a quantitative procedure for testing the effects of drugs on mucus accumulation and secretion, using methacholine as a secretagogue. This allows dissociation of mucus content in the lung and the alveolar compartments, which appear to work in a balanced way. Methacholine or pilocarpine (not shown) were indeed very effective in inducing mucus secretion from lungs (Figures 4B and 4F) (3), but residual neutral mucins persisted. This results either from stimuli of insufficient intensity (which seems unlikely because larger doses of the secretagogue failed to increase secretion [not shown]), or from the presence of other induced but nonsecreted MUC gene products in the respiratory tract (3). Acidic mucins were abundantly secreted (Figures 5 and 4J), as well as the MUC5AC protein (Figure 6), whose sequence contains a potential signal peptide sequence and cutting sites for proteases. Moreover, its predicted amino-acid composition (without the glycosylations) suggests a dominance of the acidic residues Asp and Glu, over the basic residues Lys and Arg, using DNA strider for sequence analysis as previously reported (33). The MUC5AC gene is probably involved in the accumulation of the acidic mucins observed (Figure 5) (3), and their products are secreted into the BALF, as shown by ELISA (Figure 6).

Influence of the Th1/Th2 Balance on BHR, Eosinophilia, and Mucus Accumulation

Because exogenous rmIL-13 induces BHR, eosinophilia, MUC5AC, and mucus accumulation on immunized and on nonimmunized mice, in the absence of IL-4 and of IFN- (Figures 2B and 2C), it is clear that Th2 polarization, particularly IL-4 acitvity, is not required. This observation extends previous reports showing that IL-4 is not necessary for BHR or mucus accumulation; thus, the transfer of Ova-specific T lymphocytes from IL-4-deficient mice induces mucus accumulation in recipient mice (26). Furthermore, inflammation-independent BHR after Ova challenge also occurs in the absence of IL-4 in T cell-deficient mice (34). Moreover, Cohn and colleagues (35) demonstrated that Th1 cells, which do not produce IL-4, can induce mucus accumulation, as do Th2 cells (which produce IL-4), and that the limiting cytokine is not IL-4 but rather IFN-. The latter downregulates eosinophilia and mucus accumulation in the airways, and Th1 cells (which produce IFN-) can induce mucin accumulation only in the absence of IFN- signaling. Finally, because IL-13 is induced for a longer time than IL-4 after Ova (Figures 2A and 2Cb) (as lung IL-13 R-2 mRNA_s, not shown), it may induce progressive accumulation of MUC5AC (Figures 2A, 2B, and 2Cd).

IL-13 thus acts in vivo, independently from IL-4 and from immunization, to induce BHR and mucus, even though IL-4/IL-13 share partially common pathways (IL-13 R-2 [23, 36] and Stat-6 [37]).

Immunization and Cell Recruitment

Neither mucus accumulation nor eosinophil or neutrophil recruitment induced by rmIL-13 require immunization, in BP2, in A/J (6), or in IL-4-deficient mice (18). Antigen- independent eosinophil recruitment might involve accumulation of eotaxin (22, 38, 39, 40), a recognized eosinophil chemoattractant, or of KC, which recruits a mixture of eosinophils and neutrophils (4, 38). Associated with chemokines such as MIP1-, MCP_s,and RANTES (our unpublished results), those factors may explain the increased eosinophils and neutrophils that accumulate in the lungs of immunized BP2 mice (39).

In conclusion, our results extend the concept that, contrary to Ova, which requires immunization, rmIL-13 alone induces BHR, a moderate eosinophilia, neutrophilia, goblet cell hyperplasia, and metaplasia, with mucus accumulation in the airways of BP2 mice, as occurs in other strains (6, 7). These events are downregulated by dexamethasone. As we show, because neither prior immunization, concomitant inflammation, nor IL-4 were required for rmIL-13-induced BHR or mucus accumulation, we propose that BHR and inflammation on one side, mucus accumulation and inflammation on the other, can be dissociated. Accordingly, rmIL-13 has a broader spectrum of activity than antigen, being independent of acquired immunity. Dexamethasone inhibited BHR and eosinophilia by Ova and rmIL-13, and it reduced the expression of MUC5AC involved in mucus metaplasia. Mucus secretion and bronchial obstruction were also inhibited. In contrast, after Ova, granulocyte depletion by vinblastine or by the anti-granulocyte antibody RB6-8C5 abolished BHR, mucus being poorly affected. Granulocyte depletion failed to modify BHR by rmIL-13, but the anti-granulocyte antibody decreased mucus accumulation and bronchial obstruction by rmIL-13, suggesting some involvement of granulocytes in IL-13-mediated mucus accumulation in BP2 mice. Finally, because BHR and mucus induced by exogenous rmIL-13 were surprisingly affected by dexamethasone, the effects of IL-13 may be indirect, supporting the need for further studies on the mode of action of IL-13, and the attractive hypothesis that it acts via secondary mediators.