Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Human allergic asthma can be characterized by high levels of antigen-specific immunoglobulin E (IgE) in serum, airway hyperresponsiveness to bronchoconstrictive stimuli, and infiltration of lymphocytes and eosinophils in the airways (1). There is increasing evidence that the inflammatory response as seen in allergic asthma is regulated by Th2 cells (2). The Th2 cell may therefore be an important target cell for drug therapy in the treatment of allergic asthma. Interferon- (IFN-) produced by Th1 cells is capable of downregulating the proliferation of Th2 cells in vitro (3), suppressing IgE antibody production (4), and stimulating the proliferation of Th1 cells (3). Therefore, treatment with IFN- of patients with allergic asthma might be of therapeutic use. However, more knowledge is needed to establish the in vivo role of IFN- in the sensitization and effector phases of the allergic response.

We have described in BALB/c mice a model with features reminiscent of human allergic asthma. In this model, mice are sensitized with ovalbumin without adjuvant and repeatedly challenged with ovalbumin resulting in antigen-specific IgE in serum, airway hyperresponsiveness, and infiltration of inflammatory cells in the bronchoalveolar lavage fluid (BALF) (5). In the present study, the role of endogenous IFN- was investigated by comparing IFN- gene knockout (GKO) (8) and wild-type mice. To determine the systemic and airway effects of exogenous IFN-, we treated wild-type mice with intraperitoneal or nebulized recombinant murine IFN- during the challenge period. In addition, direct effect of IFN- on the cytokine production by antigen-stimulated lung tissue and lung draining lymph node cells was investigated.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

IFN- Detection in Serum and BALF

In a separate experiment, naive BALB/c mice were treated with a single dose of IFN- intraperitoneally, 5 µg in 0.25 ml saline or nebulized 12 µg per aerosol in 3 ml saline, as described subsequently. At different time points (t = 2, 5, 10, 15, 20, and 45 min, and 1, 2, 4, 6, 8, and 24 h) after IFN- treatment, mice were killed by cervical dislocation and blood samples were obtained via cardiac puncture. Two mice were used per time point. Blood samples were rotated for 10 min at 14,000 rpm and serum samples were kept at 20°C until measured. After IFN- nebulization, the lungs were lavaged with 1 ml saline. IFN- levels in serum and BALF were detected by enzyme-linked immunosorbent assay (ELISA) according to the instructions of the manufacturer (PharMingen, San Diego, CA). The detection limit of the ELISA was 156 pg/ml and the standard curve started with 5,000 pg/ml. The half-life of IFN- in serum after the intraperitoneal injection was calculated.

Major Histocompatibility Complex Class I and II Upregulation

In a separate experiment, naive BALB/c mice (n = 4) were repeatedly (daily for 10 consecutive days) treated with IFN- (intraperitoneal or nebulized) or control as described subsequently. At 24 h after the last treatment the lungs were lavaged as described subsequently, and the spleens and thoracic lymph nodes from the paratracheal and parabronchial regions were removed and put in cold phosphate-buffered saline (PBS). A single-cell suspension was prepared with a 70-µm cell strainer (Falcon, Lelystad, The Netherlands). Erythrocytes were lysed from the spleens by 1-min treatment of the pellet fraction with ice-cold isotonic ammonium chloride solution (155 mM NH₄Cl; 10 mM KHCO₃, 0.1 mM ethylenediaminetetraacetic acid [EDTA], pH 7.4). Total cell numbers were determined. The cells (5 × 10⁵ in 25 µl) were preincubated with PBS containing 5% mouse serum (4°C, 15 min). Thereafter, spleen and lymph node cells were incubated (4°C, 30 min) with the fluorescein isothiocyanate (FITC)-labeled monoclonal antibodies against I-A^d, H-2D^d, or the respective isotype control diluted in PBS with 1% mouse serum (all PharMingen). The BALF cells were incubated with FITC-labeled monoclonal antibodies against I-A^d-, H-2D^d-, and PE-labeled monoclonal antibody against F4/80 or the isotype controls. The cells were washed, resuspended in PBS, and subsequently analyzed by flow cytometry under identical setting for all samples (FACScan, Becton-Dickinson, Mountain View, CA). At least 10,000 cells were measured. The mean fluorescence intensity of the major histocompatibility complex (MHC) classes I and II expressing population was determined and corrected for nonspecific binding of the isotype control.

IFN- Effect on Cytokine Production by Antigen-Stimulated T Cells

In a separate experiment, ovalbumin-sensitized and -challenged BALB/c mice were intraperitoneally injected with an overdose of pentobarbitone (0.5 g/kg body weight) 24 h after the last challenge. Lymph node and lung cells were antigen-stimulated as described previously (9). The lungs were lavaged five times through a tracheal cannula with 1-ml aliquots of pyrogen-free saline warmed to 37°C. The lungs were perfused via the right ventricle with 4 ml saline containing 100 U/ml heparin to remove any blood and intravascular leukocytes. The thoracic lymph nodes derived from the paratracheal and parabronchial regions and lungs were removed. The lymph nodes were transferred to cold PBS and gently homogenized on a 70-µm cell strainer to obtain a single-cell suspension. The lungs were minced and digested with 3 ml RPMI 1640 containing 2.4 mg/ml collagenase, 1 mg/ml DNAse, and 50 mg/ml gentamicin for 30 min by 37°C. The cell suspension was resuspended and filtered through a 70-µm cell strainer with 10 ml RPMI containing 20% fetal calf serum (FCS). The lymph node and lung cell suspensions were washed and resuspended in culture medium (RPMI 1640 containing 10% heat inactivated FCS, 1% glutamax I, 50 mg/ml gentamicin, and 50 mM -mercaptoethanol), and total cell number was counted. The cells were brought on concentration in culture medium (2 × 10⁵/well lymph node cells and 8 × 10⁵/well lung cells in 200 µl) and were stimulated with ovalbumin (50 µg/ ml) or medium, in the presence or absence of IFN- (100 to 1,000 pg/ml). The cells (viability > 95%) were plated in round-bottom 96-well plates (Costar, Badhoevedorp, The Netherlands) and cultured for 5 d at 37°C with 5% CO₂ in humidified air. The supernatants were removed and kept at 20°C until cytokine levels were measured by ELISA. The IFN-, interleukin-4 (IL-4), and IL-5 ELISAs were performed according to the instructions of the manufacturer (PharMingen). The detection limits of the ELISAs were 156 pg/ml for IFN-, 15.6 pg/ml for IL-4, and 31.3 pg/ ml for IL-5.

Animals

Specified pathogen-free female IFN- (GKO) (8) and wild-type mice were obtained from the breeding colony at Genentech, South San Francisco, CA. Mice with targeted disruption of the IFN- gene and wild-type mice were backcrossed onto the BALB/c background. The GKO mice were bred from F2 homozygous BALB/c mice that were F5. Once the mice acclimatized to the animal facility at the National Institute for Public Health and Environmental Protection, Bilthoven, The Netherlands, sensitization started at the age of 6 to 14 wk. Specific pathogen-free BALB/c mice were obtained from the breeding colony of the Central Animal Laboratory, Utrecht, The Netherlands. The mice were housed in macrolon cages and provided with food and water ad libitum.

Immunization Protocol

Active sensitization was performed (10) by giving seven intraperitoneal injections of 10 µg ovalbumin (grade V) in 0.5 ml pyrogen-free saline on alternate days (one injection per day). This sensitization procedure has been shown to induce high titers of total IgE antibodies in the serum of BALB/c mice, of which 80% was ovalbumin-specific IgE (10). Three weeks after the last injection, the mice were exposed to eight ovalbumin (2 mg/ml) or eight saline aerosols for 5 min on consecutive days (one aerosol per day). The aerosol was performed in a plexiglas exposure chamber (5 liters) coupled to a Jet nebulizer (Pari IS-2 Jet nebulizer; PARI Respiratory Equipment, Richmond, VA; particle size 2 to 3 µm) driven by compressed air at a flow rate of 6 liters/min. Aerosol was given in groups of maximally five animals.

From each mouse, in vivo airway responsiveness, ovalbumin-specific IgE in serum, and cellular infiltration in the BALF were determined 24 h after the last challenge. All experimental groups consisted of at least seven mice.

Treatment with IFN-

Wild-type mice were divided in six age-matched groups; three groups received saline challenge and three groups received ovalbumin challenge. One group of mice received no IFN- (control group), a second group received intraperitoneal IFN- (intraperitoneal IFN- group, 5 µg in 0.25 ml saline), and a third group received IFN- (nebulized IFN- group, 12 µg per aerosol in 3 ml saline per five mice) by aerosol. Aerosol was given as described previously. Treatment with IFN- started 2 d before the first ovalbumin or saline inhalation and was given once a day prior to ovalbumin or saline inhalation. Murine rIFN- (1 µg corresponds to 10,000 U) was generously provided by Genentech. All IFN- preparations contained less than 0.125 EU/ml endotoxin.

Airway Responsiveness In Vivo

Airway responsiveness was measured in vivo 24 h after the last aerosol exposure using an air-overflow pressure method (5, 11). With this method, the airway resistance to inflation is measured. Mice were anasthetized by intraperitoneal injection of urethane (2 g/kg), and placed on a heated blanket (30°C). Procedures were continued only after careful assurance of the adequacy of anesthesia by checking the foot reflex. If this reflex was absent, the trachea was cannulated and a small polyethylene catheter was placed in the jugular vein for intravenous administrations. The spontaneous breathing was suppressed by intravenous injection of tubocurarine chloride (3.3 mg/kg). When it stopped, the tracheal cannula was attached to a respiration pump (Sanders Brinie, Enschede, The Netherlands). The inflation volume of the pump was 0.8 ml per beat, of which the mice inhaled approximately 0.1 ml per breath with a rate of 190 beats per minute. A pressure transducer (Validyne, Northridge, CA) was located between the trachea and the respiration pump to measure changes in the airway resistance to inflation. Any increase in airway tone causes a reduction of the amount of air floating into the lungs and subsequently the remainder overflows, resulting in an increase in air-overflow pressure. Pressure signal was recorded breath by breath on a Gould Brush 2400 recorder (Godart, Utrecht, The Netherlands). At time intervals of 4 min, and after the response had returned to baseline level, doubling doses of methacholine ranging from 160 to 5,120 µg/kg were administered intravenously. Concentrations of methacholine were prepared in saline and kept on ice for the duration of the experiment. At the end of the dose-response curve, the maximal response was determined by clamping the tracheal cannula. The increase in air-overflow pressure was measured at its peak and expressed as percent increase of the maximal response.

Murine Ovalbumin-Specific IgE ELISA

After measuring the airway responsiveness, we obtained blood samples from the mice via a cardiac puncture and rotated them for 10 min at 14,000 rpm. Serum was collected and samples were kept at 20°C until ovalbumin-specific and total IgE levels were measured. Ovalbumin-specific IgE in serum was quantitated using ELISA as described previously (6). Briefly, microplates (96 wells; Nunc A/S, Roskilde, Denmark) were coated with rHu Fc RI-IgG (2 µg/ ml) at 4°C for 24 h. The ELISA was performed at room temperature. After blocking with ELISA buffer (containing 50 mM Tris; 2 mM EDTA; 136.9 mM NaCl; 0.05% Tween 20; and 0.5% bovine serum albumin [BSA], pH 7.2) for 1 h, appropriate dilutions of the samples and standard, diluted in ELISA buffer, were added for 2 h. Ovalbumin was diluted to 10 µg/ml in ELISA buffer and incubated for 1 h. After incubation, horseradish peroxidase-conjugated goat antiovalbumin (1:5,000) was added for 1 h followed by application of the substrate 2,2'-azino-bis(3-ethylbenz-thiazoline-6-sulfonic acid) (0.5 mg/ml) to develop coloring. Ten minutes after addition of the substrate, the optical density was measured at 405 nm using a Titertek Multiskan (Flow Labs, Irvine, UK).

Antibody titers of samples were calculated by comparison with an internal ovalbumin-specific IgE standard serum that was serially diluted. The standard was obtained by intraperitoneal immunization of mice with ovalbumin, and arbitrarily assigned a value of 10,000 experimental units per milliter (EU/ml).

Murine Total IgE ELISA

Microplates were coated with 1 µg/ml antimouse IgE at 4°C for 24 h. The rest of the ELISA was performed at room temperature. After the ELISA buffer was blocked for 1 h, appropriate dilutions of the samples and mouse IgE reference serum standard were added for 2 h. After incubation, 1 µg/ml biotin antimouse IgE was added for 1.5 h followed by incubation with 0.33 µg/ml peroxidase conjugated streptavidin for 1 h. The substrate o-phenylenediamine-dihydrochloride (0.4 mg/ml) in PBS containing 0.012% hydrogen peroxide was added. After approximately 15 min the reaction was stopped by adding 4 M H₂SO₄. Subsequently, optical density was measured at 492 nm, using a Titertek Multiskan. Antibody titers of samples were calculated by comparison with an internal total IgE reference serum standard that was serially diluted. Detection limit of the ELISA was 0.78 ng/ml IgE.

BAL

BAL was performed in the same animals that were used for airway responsiveness measurements. In pilot experiments, it was found that combining the two techniques had no effect on the total number of cells derived from lavage or on the appearance of different cell types. Directly after the dose-response curve with methacholine, the animals were lavaged five times through the tracheal cannula with 1-ml aliquots of pyrogen-free saline warmed to 37°C. The lavage was kept on ice until further processing. The BALF cells were washed with cold PBS (400 × g, 4°C, 5 min), and the pellet was resuspended in 150 µl cold PBS. Total numbers of lavage cells were counted with a Bürker-Türk chamber. For differential cell counts, cytospin preparations were made and stained with Diff-Quik. After coding, all cytospin preparations were evaluated by one observer using oil immersion microscopy (magnification: ×1,000). Cells were identified and differentiated into mononuclear cells, neutrophils, and eosinophils by standard morphology. At least 200 cells were counted per cytospin preparation, and the absolute number of each cell type was calculated. To evaluate differences between ovalbumin-challenged mice and saline-challenged control mice for the different treatments, total BALF cell number and the numbers of the various cell types were tested with an analysis of variance. For the very low number of eosinophils in control animals a Poisson distribution was assumed, and for differences between treatment groups a Fisher's exact test was used.

Chemicals

Ovalbumin (chicken egg albumin, crude grade V), 2,2'-azino- bis(3-ethylbenz-thiazoline-6-sulfonic acid), tubocurarine chloride, cell-cultured tested BSA (fraction V), and o-phenylenediamine-dihydrochloride were purchased from Sigma Chemical Company (St. Louis, MO); heparin was purchased from Leo Pharmaceuticals (Weesp, The Netherlands); PBS, RPMI, FCS, gentamicin, and glutamax I from Gibco Life Technologies (Merelbeke, Belgium); pentobarbitone (nembutal) was purchased from Sanofi Sante B.V. (Maassluis, The Netherlands); collagenase and DNAse were purchased from Boehringer Mannheim (Mannheim, Germany); mouse serum was purchased from Cedar Lane (Hornby, ON, Canada); saline was purchased from B. Braun Medical B.V. (Oss, The Netherlands), and urethane and methacholine (acetyl--methylcholine) were purchased from Janssen Chimica (Beerse, Belgium). rHu FcRI-IgG was kindly provided by Genentech. Peroxidase-conjugated streptavidin and horseradish peroxidase-conjugated goat antiovalbumin were purchased from Jackson (West Grove, PA). Mouse IgE reference serum was purchased from ICN Biomedicals (Aurora, OH), and antimouse IgE and biotin antimouse IgE were purchased from PharMingen and Diff-Quik from Merz & Dade A.G. (Düdingen, Switzerland).

Data Analysis

Unless stated otherwise, data are expressed as means ± SEM and were evaluated using an analysis of variance followed by a post hoc comparison between groups. A probability value P < 0.05 was considered statistically significant. Statistical analyses were carried out using Systat, version 5.03 (NaG Inc., Oxford, UK).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Serum Levels of IFN- after Nebulized or Intraperitoneal IFN- Treatment

In our first experiment, IFN- levels in serum of BALB/c mice were determined after a single dose of intraperitoneal or nebulized IFN- as used in our subsequent experiments. After mice were treated with nebulized IFN-, no IFN- could be detected in BALF and serum at all time-points measured (data not presented). This indicates that nebulized IFN- in the dose used in our experiments does not reach the serum and might therefore only act locally in the lung tissue but not systemically.

After intraperitoneal IFN- injection, the serum levels of IFN- increased immediately and reached a maximal serum level of 9.1 ng/ml IFN- at 2 h after injection (Figure 1). Thereafter the level gradually declined, until IFN- was undetectable in serum 24 h after injection. The calculated half-life of IFN- in serum is 2.16 h.

View larger version (11K): [in this window] [in a new window]   Figure 1.   Time course of serum levels of IFN- of naive BALB/c mice after treatment with a single intraperitoneal dose of IFN- (5 µg/mouse). Each point represents a single animal.

MHC Class I and II Upregulation by IFN- Treatment

Because IFN- is known to increase MHC class I and II expression on different cell types (12), in the second experiment the biologic effects of repeated IFN- treatment were determined by measuring MHC class I and II expression on splenocytes, thoracic lymph node cells, and alveolar macrophages of BALB/c mice. The mean fluorescence intensity for MHC class I and II staining on splenocytes was significantly (P < 0.05, P < 0.01, respectively) increased in intraperitoneal IFN- treated mice (108 ± 17 and 110 ± 9, respectively) when compared with control mice (60 ± 5 and 74 ± 3, respectively) (Table 1). However, nebulized IFN- treatment did not alter the MHC class I (53 ± 3) and II (59 ± 4) expression on splenocytes when compared with control mice.

MHC class I and class II expression in thoracic lymph node cells was upregulated in intraperitoneal IFN--treated mice (79 ± 17 and 68 ± 14, respectively) when compared with control mice (11 ± 3 and 25 ± 12, respectively) (Table 1). In addition, repeated IFN- nebulization slightly increased the expression of MHC class I and II (19 ± 6 and 38 ± 16, respectively) on thoracic lymph node cells when compared with the expression in control mice.

The expression of MHC class I and II expression on alveolar macrophages tended to be upregulated in nebulized IFN- treated mice (640 ± 66 and 14 ± 5, respectively) when compared with control mice (559 ± 21 and 3 ± 2, respectively) (Table 1).

IFN- Effect on Cytokine Production by Antigen-Stimulated T Cells

In a third experiment, we investigated the direct effect of IFN- on the cytokine production by antigen-stimulated cells. Therefore, cells from lung-draining lymph nodes and lungs from ovalbumin-sensitized and -challenged BALB/c mice were stimulated with ovalbumin or medium for 5 d in the absence or presence of IFN-. In supernatants of ovalbumin-stimulated lymph node cells, levels of IL-4, IL-5, and IFN- could be detected, whereas in supernatants of ovalbumin-stimulated lung cells, levels of IL-5, but not IL-4 and IFN-, could be detected (Figure 2). No cytokine production was detectable after culturing of lymph node and lung cells in medium alone (data not presented). Titration of IFN- in both lung-draining lymph node and lung cell cultures had no influence on the IL-4 and IL-5 production (Figure 2). Therefore, IFN- does not act directly on the cytokine production by ovalbumin-specific lymphocytes.

View larger version (34K): [in this window] [in a new window]   Figure 2.   IL-4 (black bars) and IL-5 (striped bars) levels in supernatants of ovalbumin-stimulated lung-draining lymph node (A) and lung tissue (B) cells in the absence or presence of IFN- (100 to 1,000 pg/ml) after culturing for 5 d. Cells were pooled from ovalbumin-sensitized and -challenged mice and cultured in at least triplicate. Shown are the mean values ± SEM. Shown is a representative experiment of three different independent experiments.

Serum Levels of Ovalbumin-Specific and Total IgE

Ovalbumin-specific and total IgE levels were measured in the serum. The serum levels of ovalbumin-specific IgE in wild-type mice were significantly potentiated (P < 0.05) after repeated ovalbumin aerosols (162 ± 57 EU/ml) when compared with saline-challenged mice (4 ± 4 EU/ml) (Figure 3A). Ovalbumin challenge (59 ± 21 EU/ml) in GKO mice did not significantly potentiate ovalbumin-specific IgE levels when compared with saline-challenged (23 ± 12 EU/ml) mice. However, the level of ovalbumin-specific IgE in ovalbumin-challenged wild-type mice was significantly increased (P < 0.05) compared with ovalbumin-challenged GKO mice.

View larger version (20K): [in this window] [in a new window]   Figure 3.   Levels of (A) ovalbumin-specific IgE (in EU/ml) and (B) total IgE (in ng/ml) in serum as determined by ELISA of ovalbumin-sensitized GKO mice and wild-type mice (WT) treated without () or with IFN- by intraperitoneal injection (ip) or by nebulization (neb) and repeatedly challenged with saline (white bars) or ovalbumin (black bars). Values are expressed as means ± SEM (7 to 10 mice per group). *P < 0.05, **P < 0.01 compared with the corresponding saline-challenged control mice; #P < 0.05 compared with ovalbumin-challenged control-treated mice as determined with analysis of variance followed by post hoc comparison between groups.

In ovalbumin-challenged wild-type mice treated with nebulized IFN- (103 ± 46 EU/ml), an upregulation of ovalbumin-specific IgE was still observed when compared with saline-challenged mice (15 ± 10 EU/ml), although this did not reach the level of significance (P = 0.1).

However, intraperitoneal IFN- treatment of ovalbumin- challenged (48 ± 25 EU/ml) wild-type mice significantly (P < 0.05) decreased the upregulation of ovalbumin-specific IgE in serum, compared with ovalbumin-challenged wild-type mice. No difference in ovalbumin-specific IgE serum levels was observed between saline- (30 ± 15 EU/ ml) and ovalbumin-challenged wild-type mice treated with intraperitoneal IFN-.

In wild-type mice, ovalbumin challenge (451 ± 96 ng/ml) induced a significant (P < 0.01) increase in total serum IgE, compared with saline-challenged mice (171 ± 33 ng/ ml) (Figure 3B). In ovalbumin-challenged GKO mice (300 ± 68 ng/ml), a tendency (P < 0.07) of increased total serum IgE levels was observed compared with saline-challenged GKO mice (107 ± 26 ng/ml). Treatment with intraperitoneal IFN- induced a slight but not significant upregulation of the total IgE levels in the saline-challenged wild-type mice (334 ± 88 ng/ml), whereas the levels of total IgE were unaltered in the ovalbumin-challenged mice (405 ± 134 ng/ml). Nebulized IFN- treatment had no effect on the total serum IgE levels in both saline- and ovalbumin-challenged wild-type mice (183 ± 84 ng/ml and 332 ± 98 ng/ml, respectively).

Airway Responsiveness

Airway responsiveness in vivo to methacholine was measured 24 h after the last aerosol. For all mice a complete dose-response curve to methacholine ranging from 160 to 5,120 µg/kg methacholine was made. In wild-type mice challenged with ovalbumin, a significant increase in air-overflow pressure was observed when compared with saline-challenged wild-type mice at doses of methacholine ranging from 640 to 5,120 µg/kg (Figure 4A). Figure 4B shows the increase in air-overflow pressure to 640 µg/kg methacholine of the different experimental groups. This dose is representative for the hyperresponsiveness to methacholine. In GKO mice, ovalbumin challenge did not induce airway hyperresponsiveness as compared with the saline-challenged mice (84.7 ± 1.8% and 74.8 ± 4.3%, respectively). In addition, high airway responses of saline-challenged GKO mice were observed in the complete dose-response curve to methacholine. The airway responsiveness in saline-challenged GKO mice, however, was significantly (P < 0.01) increased, compared with saline-challenged wild-type mice (74.8 ± 4.3% versus 49.4 ± 3.8%, respectively).

View larger version (31K): [in this window] [in a new window]   Figure 4.   (A) Increase in air-overflow pressure to methacholine (160 to 5,120 µg/kg) in ovalbumin-sensitized wild-type mice 24 h after the last ovalbumin (black bars) or saline (striped bars) challenge. (B) The increase in air-overflow pressure to 640 µg/kg methacholine in ovalbumin-sensitized wild-type (WT) or GKO mice treated without () or with IFN- by intraperitoneal injection (ip) or by nebulization (neb), and repeatedly challenged with saline (striped bars) or ovalbumin (black bars). Results are expressed as means ± SEM (7 to 10 mice per group). *P < 0.05 compared with the corresponding saline-challenged control mice; #P < 0.05, ##P < 0.01 compared with the control mice of the same challenge and strain; ¥¥P < 0.01 compared with the control saline-challenged wild-type mice and determined with analysis of variance followed by post hoc comparison between groups.

In wild-type mice, airway responsiveness to 640 µg/kg methacholine was comparable between the groups of saline-challenged mice. Intraperitoneal and nebulized IFN- treatment significantly (P < 0.01, P < 0.05, respectively) inhibited the airway hyperresponsiveness observed in ovalbumin-challenged wild-type mice. After intraperitoneal and nebulized IFN- treatment, no differences were observed in air-overflow pressure between saline- (43.6 ± 7.1% and 40.5 ± 4.8%, respectively) and ovalbumin-challenged (47.0 ± 4.6% and 48.3 ± 5.6%, respectively) mice.

BAL

BAL was performed 24 h after the last challenge. In all experimental groups no significant differences were observed in numbers of neutrophils present in the BALF between ovalbumin- and saline-challenged mice (data not presented). No eosinophils were observed in the BALF of saline-challenged wild-type or GKO mice. The numbers of eosinophils in the BALF of all ovalbumin-challenged mice were significantly increased when compared with the corresponding saline-challenged controls (Figure 5). The eosinophilic infiltration in the ovalbumin-challenged GKO mice (3.4 ± 1.7 × 10⁵ eosinophils/BALF) was less pronounced when compared with ovalbumin-challenged wild-type mice (14.5 ± 7.7 × 10⁵ eosinophils/BALF).

View larger version (30K): [in this window] [in a new window]   Figure 5.   Total number of mononuclear cells (striped bars) and eosinophils (black bars) observed in BALF of ovalbumin-sensitized wild-type (WT) or GKO mice treated without () or with IFN- by intraperitoneal injection (ip) or by nebulization (neb), and repeatedly challenged with saline (sal) or ovalbumin (ova). Values are expressed as means ± SEM (7 to 10 mice per group). *P < 0.05, **P < 0.01 compared with the corresponding saline-challenged control mice; #P < 0.05, ##P < 0.01 compared with the control mice of the same challenge and strain and determined with analysis of variance followed by post hoc comparison between groups.

Intraperitoneal IFN- treatment of ovalbumin-challenged mice caused a significant decrease (P < 0.05) of eosinophils (2.4 ± 1.1 × 10⁵ eosinophils/BALF) compared with the ovalbumin-challenged wild-type mice (14.5 ± 7.7 × 10⁵ eosinophils/BALF). Nebulized IFN- treatment, however, did not alter the number of eosinophils in the lavage (12.8 ± 4.4 × 10⁵ eosinophils/BALF) when compared with ovalbumin-challenged wild-type mice.

In all groups of saline-challenged wild-type mice, the numbers of mononuclear cells in the BALF were comparable. In GKO mice, similar numbers of mononuclear cells were observed in saline- (4.1 ± 0.3 × 10⁵ mononuclear cells/BALF) and ovalbumin-challenged (4.6 ± 0.3 × 10⁵ mononuclear cells/BALF) mice (Figure 5). In wild-type mice, ovalbumin challenge significantly (P < 0.01) increased the number of mononuclear cells (5.4 ± 1.6 × 10⁵ mononuclear cells/BALF) when compared with saline-challenged mice (2.2 ± 0.3 × 10⁵ mononuclear cells/ BALF). Nebulized IFN- treatment did not alter this increased mononuclear cell infiltration after ovalbumin aerosol (4.9 ± 1.1 × 10⁵ mononuclear cells/BALF). However, intraperitoneal IFN- treatment significantly (P < 0.01) abolished the increased mononuclear cell infiltration after ovalbumin challenge (3.1 ± 0.4 × 10⁵ mononuclear cells/BALF).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In this study we observed that sensitized wild-type mice exhibit upregulated total and ovalbumin-specific IgE in serum, airway hyperresponsiveness, and infiltration of eosinophils and mononuclear cells in the BALF after ovalbumin challenge. In contrast, in GKO mice only a reduced eosinophilic infiltration was observed after ovalbumin challenge. Furthermore, in wild-type mice, parenteral IFN- treatment downregulated ovalbumin-specific IgE levels in serum, airway hyperresponsiveness, and cellular infiltration in the BALF whereas aerosolized IFN- treatment only suppressed airway hyperresponsiveness. This modulatory role of exogenous IFN- does not appear to be mediated via a direct effect on the cytokine production by Th2 cells.

The IgE production (15) and eosinophilic infiltration (16) in allergic asthma are thought to be regulated by the Th2 cell- (2) derived cytokines IL-4 and IL-5 (3). IFN- inhibits the proliferation of Th2 cells (3) and thus reduces the influence of Th2 cytokines. We have shown in the present study that IFN- in vitro is not able to downregulate the IL-4 and IL-5 production by antigen-stimulated lymphocytes, suggesting that the effects of exogenous IFN- administration are not mediated via a direct effect on IL-5 production by antigen-specific T cells in vivo.

It has been shown that IFN- can downregulate IgE production in vitro (4). In the present study, ovalbumin challenge significantly increased the levels of total and ovalbumin-specific IgE in serum when compared with saline-challenged wild-type mice. In contrast, despite the lack of IFN- in GKO mice, no significant upregulation of total and ovalbumin-specific IgE levels in serum were observed after ovalbumin challenge. Interestingly, it has been described that GKO mice have reduced MHC class II expression on macrophages (8). Accordingly, it can be assumed that antigen presentation is less efficient in these mice, resulting in a less efficient allergic response. Antigen challenge in GKO mice might, as a consequence, induce less pronounced allergic symptoms such as antigen-specific IgE, eosinophilic infiltration, and airway hyperresponsiveness. Therefore, it seems likely that endogenous IFN- is necessary for optimal IgE production also during a secondary response. However, in a study using mice lacking the IFN- receptor (17), total IgE levels in serum were increased compared with wild-type mice. An explanation for this discrepancy might be found in the differences in experimental protocol used; that is, sensitization occurred with an adjuvant, inducing a much stronger immune response, and the mice were challenged only once.

Intraperitoneal IFN- treatment during the challenge period completely inhibited the ovalbumin-induced increase of ovalbumin-specific IgE in wild-type mice. This is in line with in vitro studies showing that IFN- reduces the level of IL-4-induced germline  -transcript in B lymphocytes by more than 80% (18). Because IFN- treatment of the wild-type mice started prior to the challenge period, when IgE levels are low, this could downregulate the germline  -transcript, resulting in lower levels of IgE. IL-4, which is essential for IgE production, can be produced by Th2 cells (3) and mast cells (19). In the present study, we have shown that in vitro the levels of IL-4 produced by antigen-stimulated Th2 cells are unaltered by IFN-. However, it remains possible that IFN- acts via inhibition of T-cell proliferation or mast-cell proliferation and development (20, 21), which may lead to decreased IL-4 secretion and lower IgE levels in serum.

In the present study, serum levels of ovalbumin-specific and total IgE were unaltered after nebulized IFN- treatment. After IFN- nebulization, we were unable to detect IFN- in serum and BALF, and this induced only a slight upregulation of MHC class I and II expression on thoracic lymph node cells and alveolar macrophages but not on splenocytes. These data suggest that the effect of nebulized IFN- was restricted to the lungs, whereas parenteral IFN- also had systemic effects because it upregulated MHC class I and II expression on splenocytes and thoracic lymph nodes. Because parenteral IFN- treatment inhibited ovalbumin-specific IgE upregulation whereas nebulized IFN- treatment did not, it can be suggested that the secondary antigen-specific IgE production does not occur in the lung tissue of mice. In agreement herewith, in previous experiments we have not been able to detect ovalbumin-specific IgE in BALF of BALB/c mice (E. M. Hessel, unpublished observations). It has been suggested recently that IgE synthesis takes place in the germinal centers of the lung-draining lymph nodes (22). It seems unlikely that the dose of nebulized IFN- was not sufficient to affect the production of IgE, because the airway hyperresponsiveness was completely abolished after nebulized IFN- treatment. In contrast with the present study, Lack and colleagues (23, 24) showed that nebulized, but not parenteral, IFN- treatment could downregulate ovalbumin-specific IgE in serum. Because the sensitization was done by aerosol, it is likely that a local response is induced, leading to IgE production in lung tissue.

Our experiments show that ovalbumin challenge in wild-type mice induced in vivo airway hyperresponsiveness compared with ovalbumin challenge in saline-challenged mice. However, no differences in airway responses were observed in GKO mice after repeated saline and ovalbumin challenges. Interestingly, the airway responses of saline-challenged GKO mice were significantly higher than saline-challenged wild-type mice, and intraperitioneal and nebulized IFN- treatment of saline-challenged GKO mice could downregulate these increased responses (C. L. Hofstra, unpublished observations). These data indicate a role for endogenous IFN- in airway responsiveness. It is known that IFN- can induce nitric oxide production (25), which causes dilatation of the airway smooth muscle in patients with allergic asthma and reverses bronchoconstriction (26, 27). In GKO mice, a defect in nitric oxide production by macrophages has been observed (8) that could explain the high airway responsiveness in GKO mice compared with wild-type mice.

The airway hyperresponsiveness in ovalbumin-challenged wild-type mice could be completely abolished by intraperitoneal or nebulized IFN- treatment. In the literature, the role of IFN- in airway hyperresponsiveness remains unclear. As in the present study, diminished airway hyperresponsiveness has been reported after intraperitoneal IFN- (28) or nebulized IFN- (24) treatment. It seems that IFN- has a dual role in airway hyperresponsiveness, because it has also been shown that airway hyperresponsiveness could not be induced in ovalbumin-challenged BALB/c mice treated with antibodies to IFN- prior to the challenge period (5). Blocking endogenous IFN- may alter the normal allergic response as also demonstrated in GKO mice. The effect of IFN- might be mediated via stimulation of macrophage-derived nitric oxide, which can cause, depending on the local concentration, contraction (29) or relaxation (26, 27) of the airway smooth muscles. Thus, IFN- may mediate both proinflammatory and anti-inflammatory activities, depending on the local concentration and timing within the immune response.

In this paper, we showed that ovalbumin inhalations resulted in eosinophilic and mononuclear cell infiltration in the BALF in wild-type mice. Repeated ovalbumin aerosol in GKO mice increased the eosinophilic, but not mononuclear cell, infiltration, although this was less pronounced compared with wild-type mice. It has been shown that IFN- is capable of upregulating the expression of ICAM-1 on endothelial cells (30, 31), which is important for the migration of eosinophils into lung tissues (32). This offers an explanation for the reduced infiltration of eosinophils in the airways of GKO mice. Because IFN- is capable of inducing survival of eosinophils (33), the lack of IFN- in GKO mice might decrease the survival of eosinophils, presumably leading to fewer airway eosinophils.

In the present study intraperitoneal IFN- treatment inhibited the eosinophil and mononuclear cell infiltration that was unaltered by nebulized IFN- treatment in ovalbumin challenged wild-type mice. These data suggest that the eosinophilic infiltration is regulated not at the level of the lungs but systemically, or that the dose of nebulized IFN- used was not sufficient. A likely target could be lung-draining lymph nodes, which are reached by the systemic circulation. Although the airway hyperresponsiveness was inhibited by treatment with nebulized IFN-, it still remains possible that more IFN- is needed to suppress eosinophilia. These data confirm the suggestion that airway hyperresponsiveness and eosinophilic infiltration are dissociated (5). It seems that nitric oxide plays a role in the eosinophilic infiltration, because this infiltration can be inhibited by nitric oxide synthase inhibitors (34). In the literature it has been shown that IFN- treatment inhibits (28, 35), whereas treatment with antibodies to IFN- (5, 38) do not influence, the recruitment of eosinophils into the lungs after antigenic challenge. In addition, in studies using GKO mice (17, 39) no altered eosinophilic infiltration in comparison with wild-type mice was observed. IL-5, produced by Th2 cells among others (3), is a potent chemoattractant for eosinophils (16), and in mice treated with antibodies to IL-5 the eosinophilic infiltration was completely blocked (5, 40, 41). The downregulatory effect of IFN- on Th2 cell proliferation, resulting in less IL-5 production, might therefore result in less eosinophilic infiltration in the BALF. The decreased number of mononuclear cells in the BALF of intraperitoneal IFN--treated wild-type mice when compared with the ovalbumin-challenged control wild-type mice supports this idea. However, we showed that IFN- is not able to decrease directly the IL-5 production by lung lymphocytes.

The present study shows that parenteral IFN- administration in particular can downregulate the allergic response, but at the same time, endogenous IFN- plays a role in an optimal response as well. More studies are needed to elucidate the precise proinflammatory and anti-inflammatory roles of IFN- in the disease development of allergic asthma.