Male Balb/c mice, 5 to 6 weeks old, were obtained from B&K Universal AB (Sollentuna, Sweden). All animals were maintained under conventional animal housing conditions and provided with food and water ad libitum. The experimental protocol was approved by the Animal Ethics Committee in Gothenburg, Sweden.

All mice were sensitized by intraperitoneal (i.p.) injection with 8 μg of ovalbumin (OVA; Sigma-Aldrich Sweden AB, Tyresö, Sweden) adsorbed to 4 mg of aluminum hydroxide (Al(OH)₃; Sigma) in 0.5 ml of phosphate-buffered saline (PBS). A booster dose of the OVA-Al(OH)₃ mixture was given on a second occasion, five days after the first injection. Starting eight days after the second sensitization, the animals were briefly anesthetized using aerosolized Isoflurane (Baxter, Deerfield, Ill), and exposed to 100 μg OVA in 25 μl of PBS by intranasal (i.n.) administration on five consecutive days.

Animals were pretreated i.p. with a single dose (100 μg/animal) of purified hamster anti-mouse IL-9 monoclonal antibody (clone D9302C12; Pharmingen, San Diego, CA, USA) or its isotype control, purified hamster IgG (clone G235-2356; Pharmingen). In parallel groups, animals were treated either with a single dose (50 μg/animal) of a monoclonal antibody to mouse IL-5 (clone TRFK-5; R&D Systems Europe Ltd., Barton Lane, Abingdon, UK) or its isotype control, rat IgG₁ (clone R3-34; Pharmingen) in 0.5 ml of PBS, 30 min. before the OVA exposure. Secondly, the combined treatment of anti-IL-9 and anti-IL-5 antibodies was tested. Animals were treated with both anti-IL-9 (100 μg/animal) and anti-IL-5 (50 μg/animal), either anti-IL-9 (100 μg/animal) and rat IgG₁ (50 μg/animal), or anti-IL-5 (50 μg/animal) and purified hamster IgG (100 μg/animal), or both isotype controls together (100 μg of purified hamster IgG and 50 μg of rat IgG₁). Newly produced inflammatory cells were labelled with a thymidine analogue, 5'-bromo-2'-deoxyuridine (BrdU; Boehringer Mannheim Scandinavia AB, Bromma, Sweden), at a dose of 1 mg in 0.25 ml of PBS i.p. twice (7 h apart) on day one and three of allergen exposure (total dose 4 mg/animal).

Samples were collected 24 h after the last allergen exposure. The animals were anesthetized with an i.p. mixture of xylazine (130 mg/kg) and ketamine (670 mg/kg). When in adequately deep anaesthesia, the chest was opened and samples of blood, bronchoalveolar lavage fluid (BALf) and bone marrow were taken.

Blood was obtained by penetration of the right ventricle of the heart with a needle. BAL was performed through the trachea with a cannula by instillation of 0.25 and 0.2 ml of ice-cold PBS. BALf of each animal was pooled; approximately 0.4 ml of the instilled fluid was consistently recovered. Finally, one femur was excised and cut at the epiphyses. Bone marrow cells were removed by perfusion of the femur with 2 ml of PBS. BALf and bone marrow cell suspension were kept on ice until further processing.

Cytospin of blood was obtained by taking 200 μl of blood and mixing it with 800 μl 2 mM EDTA (Sigma) in PBS. The red blood cells were lysed in 0.1% potassium bicarbonate and 0.83% ammonium chloride for 15 min. at 4°C as described previously 8. The white blood cells were re-suspended in PBS with 0.03% bovine serum albumin (BSA; Sigma). Serum was obtained from the remaining volume of blood by centrifugation at 3000 rpm for 10 min. at 4°C.

BAL and bone marrow cell suspension were centrifuged at 1000 rpm for 10 min. at 4°C. Supernatants were aspirated and cells were re-suspend in PBS with 0.03% BSA. The total number of cells in BAL, bone marrow and blood was determined using standard hematological procedures. Cytospins of BALf, bone marrow and blood cells were prepared and stained with May-Grünwald-Giemsa for differential cell counts. Cell differentiation was determined by counting 300–400 cells using a light microscope (Zeiss Axioplan 2; Carl Zeiss, Jena, Germany). The cells were identified by standard morphological criteria, and bone marrow mature and immature eosinophils were determined by nuclear morphology, May-Grünwald-Giemsa staining properties, and cytoplasmic granulation, as previously described 78. Cytospin preparations for immunocytochemistry were air-dried and stored at -80°C until further examination.

Intranuclear BrdU was detected in BAL, bone marrow and blood cytospins using a mouse monoclonal antibody against BrdU. Cytospin preparations taken out of the freezer were fixed overnight in 4% paraformaldehyde, and subsequently washed with tris-buffer saline (TBS) and subjected to digestion in trypsin (No.232-650-8; Sigma) diluted in distillated H₂O at 37°C for 20 min. The slides were then incubated in 4 M HCl for 15 min. to denature the DNA, followed by neutralization in Holmes Borate-Borax buffer (1.24% H₃BO₃ in distillated H₂O, pH 8.5) for 10 min. The samples were treated with peroxidase blocking solution (No. S2023; DAKO) for 1 h. Non-specific binding sites were blocked with 5% normal rabbit serum (No. X0902; DAKO) for 15 min. Subsequently, the slides were incubated with 2.5 μg/ml rat anti-BrdU antibody (clone BU1/75; No. MAS 250p; Harlan-Sera Lab) or isotype control, rat IgG_2a (clone R35-95; Pharmingen) in incubation buffer (0.5% BSA/TBS) for 1 h. After washing in 0.05% Tween/TBS and TBS, the slides were incubated with 1:50 dilution of rabbit F(ab')₂ anti-rat Ig-HRP (No.6130-05, SBA) and 2% normal mouse serum (No.X0910; DAKO) for 1 hour. After further washing, the staining with DAB substrate-chromogen system (No. K3466; DAKO) was developed for 10–15 min. by monitoring in microscope. The slides were counterstained with Mayer's hematoxylin for 30 sec. and eosin for 2 min., dehydrated and mounted in Mountex. All slides were evaluated on light microscope in random fields of view. Cells with any nuclear brown staining together with the pink staining in cytoplasma were counted as BrdU-labelled (BrdU⁺) eosinophils.

Cytospin preparations from blood, BAL and bone marrow were fixed in 2% formaldehyde for 10 min. All incubations were performed at room temperature (RT). After washing in TBS, endogenous biotin was blocked with DAKO Biotin Blocking system. The preparations were incubated with a monoclonal biotinylated rat anti-mouse CD34 antibody (0.5 μg/ml; clone RAM34, BD Biosciences) or its isotype control (biotinylated Rat IgG_2a, BD Biosiences) for 2 h. After further washings, the preparations were incubated with Streptavidin Alkaline-Phosphatase (DAKO) for 45 min. Bound antibodies were visualized with Vector Red Alkaline Phosphatase Substrate kit (Vector Laboratories Inc. Burlingame, CA, USA). The preparations were washed in Tris-HCl buffer (100 mM, pH 8.2–8.5), rinsed in distilled H₂O and counterstained with Mayer's hematoxylin for 30 sec., dehydrated and mounted in Mountex. Three hundred cells were counted in random fields of view.

IL-5 concentration in serum and BALf from anti-IL-9 or its control treated animals was measured using a commercial murine IL-5 enzyme-linked immunosorbent assay (ELISA) kit (R&D Systems, Inc, Abingdon, UK) according to the manufacturer's instruction. The detection limit for IL-5 was 5 pg/ml.

Quantification of IL-9 was measured in BALf, serum and fresh bone marrow supernatant (the femur was perfused twice with 1 ml of PBS) from OVA-sensitized and exposed for 5 days to OVA or PBS (control) animals by ELISA method. ELISA plate (Maxisorp F96, NUNC) was incubated with 100 μl of purified rat anti-mouse IL-9 antibody (No.551218, BD) at a concentration of 1 μg/ml in coating buffer (0.1 M NaHCO₃, pH 9.5) at 4°C for 18 h. All incubations were performed in dark. After washings with PBS containing 0.05% Tween 20, non-specific bindings were blocked with assay diluent (10% fetal calf serum, FCS, in PBS) for 1 h at RT. After further washings, bone marrow supernatant, BALf and serum samples (100 μl of each) were incubated in duplicate at 4°C for 18 h. A standard curve was generated by using serial dilutions of recombinant mouse IL-9 (No.551867, BD). After washing, 100 μl of biotinylated anti-mouse IL-9 antibody (No. 554473, BD) at a concentration 0.5 μg/ml in the assay diluent was added and incubated at RT for 1 h. After further washings, the plates were incubated with 1:6000 dilution of Extr-Avidin HRP (No. E2886, Sigma) for 30 min. After another washing, 100 μl of TMB substrate solution was added and incubated for 30 min. The reaction was terminated by adding 50 μl of stop solution (1.0 M H₂SO₄) and assessed with an ELISA reader (type-349; Labsystems, Stockholm, Sweden) at 450 nm. The detection limit for IL-9 was 125 pg/ml.

Bone marrow cells were harvested from OVA-sensitized and exposed for six days animals, after 4 h of the last exposure. Cells were cultured in a 48-well plate in RPMI 1640 culture medium, complemented with 10% FCS, 1% penicillin-streptomycin, 1% sodium pyruvate and 2 mM L-glutamine (all obtained from Sigma) in a concentration of 1 × 10⁶ cells/500 μl of medium at 37°C in an atmosphere containing 5% CO₂. Cells were cultured only with plain medium (for negative control), either stimulated with 100 μg/ml of OVA or stimulated with phorbol myristate acetate (PMA) and calcium ionophore (end conc. 2 ng/ml and 1 μg/ml respectively, both obtained from Sigma). After 2 h, Brefeldin A solution was added in a final concentration of 10 μg/ml, and incubated for additional 6 h.

After incubation, bone marrow cells were harvested, washed and double-stained (CD4 surface/ IL-9 intracellular), using a standard saponin protocol 26 with some modifications. Unspecific binding was blocked with 2% mouse sera (Dako) for 15 min. The cells were thereafter incubated with a phycoerythrin-conjugated anti-CD4 (clone H129.19, Pharmingen) and its isotype-matched control for 30 min. at 4°C. Surface immunostained cells were fixed in 4% paraformaldehyde at room temperature for 10 min., followed two washings in 1% FCS/PBS. After resuspention in 2 ml of SAP Buffer (0.1% saponin and 0.05 NaN₃, w/v in HBSS, Sigma), the cells were incubated with a biotin-conjugated anti-mouse IL-9 monoclonal antibody (clone D9302C12, Pharmingen) at RT for 40 min., then with streptavidin-FITC for 20 min. Finally, the cells were washed with 1% FCS/PBS, resuspended in the same Buffer, and analyzed using a FACScan flow cytometer (Becton Dickinson, Mountain View, CA). Ten thousand cells were computed in a list mode and analyzed using the CellQuest Software (Becton Dickinson).

Results are presented as mean values ± SEMs. The Mann-Whitney U-test was employed for comparison of data between groups. Value of p < 0.05 was considered as statistically significant.

Total cell numbers in BAL were not different between anti-IL-9 and its control IgG treated groups (Figure 1a). There was no significant difference in eosinophil, neutrophil, and lymphocyte numbers in BAL. Animals treated with anti-IL-9 had significantly higher number of macrophages in BAL (Figure 1a) and significantly lower percentage of eosinophils in comparison to its control treated animals (63.4 ± 6.7 vs 80.1 ± 1.4 % of total cells in BAL, p = 0.024). Anti-IL-5, in comparison to control IgG₁ treated mice, significantly decreased the total cell number in BAL, mainly due to decrease in eosinophil number (Figure 1b).

Immunocytochemical stainings of BAL cells for BrdU and CD34 are illustrated in Figure 2. Anti-IL-9 did not significantly change newly produced (BrdU⁺) eosinophil numbers in BAL (Figure 3a), but decreased the percentage of BrdU^- eosinophils in BAL (11.5 ± 1.4 vs 16.6 ± 1.5 % of total cells in BAL, p = 0.024). The number of BrdU⁺ BAL eosinophils was extensively reduced by anti-IL-5 (Figure 3b).

Anti-IL-9 did not affect the increase in BAL CD34⁺ granulocytes induced by OVA-exposure, but slightly (p = 0.048) decreased the number of BAL CD34^- granulocytes (Figure 4a). Numbers of BAL CD34⁺ and CD34^- granulocytes were significantly reduced with anti-IL-5 treatment (Figure 4b).

As compared with the control antibody, treatment with anti-IL-9 significantly decreased the number of blood neutrophils, but without effect on other cell types or total blood cell numbers (Figure 5a). Anti-IL-5 treatment significantly decreased blood eosinophil number (Figure 5b).

Immunocytochemical analysis of intranuclear BrdU staining in blood cells showed significant decrease in BrdU⁺ eosinophils from animals treated with anti-IL-9, compared to its control treated animals (Figure 6a). Also anti-IL-9 significantly reduced the number of BrdU⁺ blood neutrophils (0.7 ± 0.1 vs 1.3 ± 0.2 × 10⁶/ml blood, p < 0.05). Treatment with anti-IL-5 decreased BrdU⁺ and BrdU^- blood eosinophil numbers (Figure 6b), without any effect on BrdU⁺ blood neutrophils.

Anti-IL-9 treatment and anti-IL-5 treatment did not significantly affect allergen induced increase in total blood CD34⁺ cell numbers (data not shown).

Anti-IL-9 treatment reduced bone marrow eosinophilia with significant effect on mature eosinophils (Figure 7a). Anti-IL-5 treatment decreased mature and immature bone marrow eosinophils (Figure 7b).

Anti-IL-9 also decreased BrdU⁺ eosinophil number in bone marrow, but did not affect BrdU^- eosinophils (Figure 8a). Anti-IL-5 treatment reduced both BrdU⁺ and BrdU^- eosinophils in bone marrow (Figure 8b).

Anti-IL-9 treatment did not significantly change bone marrow CD34⁺ cell numbers, as compared with control treated animals (p = 0.08). Anti-IL-5 significantly reduced CD34⁺ cell numbers in bone marrow, as compared with its isotype control treatment (26.7 ± 2.7 vs 16.5 ± 2.7 % of total cells, p = 0.03).

Anti-IL-9 treatment, as compared to its isotype control treatment, did not significantly change IL-5 levels in serum (8.7 ± 2.3 vs 30.5 ± 16.7 pg/ml respectively, p = 0.24) and BALf (20.8 ± 6.2 vs 54.6 ± 22.2 pg/ml respectively, p = 0.16).

The concentration of IL-9, measured by ELISA, in bone marrow supernatant from allergen-exposed animals was 369 ± 247 pg/ml, but was not detectable in serum and BALf from the same animals, as well as in PBS-exposed animals. IL-5 in bone marrow supernatant from the same animals was not detectable by ELISA.

Flow cytometry, using intracellular staining of IL-9, revealed that bone marrow cells from sensitized and OVA-exposed for six days animals after in vitro stimulation with OVA and PMA together with calcium ionophore (IC), had increased expression of IL-9 in comparison with baseline (1.04 and 1.84 times fold respectively). This expression was more pronounced in the bone marrow CD4⁺ cells (Figure 9). There was no substantial difference between the percent of bone marrow cells expressing CD4 at baseline (plain medium), or after stimulation with OVA, or PMA+IC (9.1%, 8.5% and 6.2% respectively).

The effects of combined treatment with anti-IL-5 and anti-IL-9 are described in Table I. No additive effect of anti-IL-9 was observed above the effect of anti-IL-5.