Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Pulmonary inflammation characterized by airway eosinophilia and airway hyperresponsiveness (AHR) are hallmarks of asthma (1). Asthma is also known to be accompanied by an increase in allergen-specific and total IgE, with a cytokine profile in bronchoalveolar lavage fluid (BALF) that is representative of a Th2-mediated event (2). Several animal models, including the guinea pig, rat, and nonhuman primates, show symptomatic phenotypes similar to that of atopic asthma upon aerosol antigen challenge (3). Recently, mouse models have been shown to reproduce many features of human asthma (7). Moreover, the availability of specific sensitizing protocols for eliciting a Th2-mediated inflammatory process, specific immunologic modifying reagents, systems for measuring pulmonary function, and the availability of genetically engineered animals make murine models of asthma appealing.

In most models, inhaled antigen in sensitized animals initiates a cascade of proinflammatory events and is known to cause an increase in cell-adhesion-molecule expression on both the endothelial surface of the lung vasculature and the epithelial surfaces of the airways (12, 13). The net result is an inflammatory-cell migration into the lung, characterized by an acute neutrophil influx followed by a chronic airway eosinophilia. Both in vitro and in vivo data suggest that the recruited eosinophils release a host of proinflammatory mediators and cytokines that include interleukin (IL)-5, major basic protein (MBP), eosinophilic cation protein (ECP), and eosinophil peroxidase (14). In this regard, upregulation and function of various adhesion molecules are necessary for the diapedesis and migration of leukocytes from the vasculature and into the airways. Eosinophils and neutrophils express both leukocyte integrin (MAC-1) and lymphocyte function-associated antigen-1 (LFA-1), and may therefore migrate via these ligands through intercellular adhesion molecule-1 (ICAM-1) expressed on the endothelium and pulmonary epithelium (17). The other major pathway, involving very late antigen-4 (VLA-4) and vascular cell adhesion molecule-1 (VCAM-1), has also been described in mediating antigen-induced airway eosinophilia and hyperreactivity in guinea pigs (18) and Brown Norway (BN) rats (19). Expression of VLA-4 is limited to eosinophils and lymphocytes, which may bind VCAM-1 or fibronectin for trafficking to inflammatory sites (12, 20).

The use of genetically altered mice as models of the proinflammatory events in asthma offers a distinct advantage over treatments with monoclonal antibody, which may be complicated by: (1) drug metabolism; (2) inherent sensitization to the antibody, concurrent with the production of neutralizing antibody with chronic treatment; and (3) distribution of the antibody to tissues and targeted cell types. However, limitations in forming conclusions from gene-knockout mice should be considered. As an example, Kumasaka and colleagues have described a role for ICAM-1 in a model of endotoxin-induced lung neutrophilia (21). Antisense oligonucleotides and monoclonal antibodies to ICAM-1 provided inhibition of the lung neutrophilia, whereas the ICAM-1 gene knockout was comparable to the wild type. Given these results, Kumasaka and colleagues have postulated that the upregulation of alternative pathways in the absence of ICAM-1, or additional effects of these agents on endothelium, may account for these disparate results.

Although previous studies have shown the importance of ICAM-1 in the development of airway inflammation and hyperreactivity, the relative contribution of ICAM-1 during sensitization to allergen in these models remains unclear. In our experiments, we sought to further define the role of ICAM-1 in a mouse model of asthma by assessing the integrated effects of antigen recognition, T-cell sensitization, and the subsequent development of airway hyperresponsiveness (AHR) and recruitment of eosinophils into the airways of both wild-type mice and mice genetically deficient for the ICAM-1 glycoprotein. Our data provide direct evidence for an important cellular and functional role of ICAM-1 in the induction of airway inflammation and hyperresponsiveness.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Animals

Male wild-type C57BL/6J (ICAM-1^+/+) and ICAM-1-deficient C57BL/6J-ICAM-1 < tm1BAY > (ICAM-1^/) mice weighing 23 to 25 g were obtained from Jackson Laboratories (Bar Harbor, ME). They were housed at 10 animals per cage, allowed food and water ad libitum, and kept under a 12-h light-dark cycle. All experiments were performed according to the directions of the Animal Care and Use Committee at Boehringer Ingelheim Pharm., Ridgefield, CT, with an approved protocol.

Sensitization and Antigen Challenge

To achieve an immune response mediated by IgE (10), mice were sensitized with an intraperitoneal injection of 0.5 ml of aluminum hydroxide (Amphogel; Wyeth Ayerst Laboratories, Philadelphia, PA) (2 mg/ml) adsorbed to ovalbumin (OVA) (Calbiochem-Novabiochem Corp., La Jolla, CA) (16 µg/ml) in sterile Ca²⁺- and Mg²⁺-free phosphate-buffered saline (PBS) (Gibco BRL Life Technologies, Grand Island, NY). The mice were subsequently boosted, intraperitoneally, with this same mixture 5 d later.

On Day 12 after the initial sensitization, from 8 to 10 mice were placed in a 24 × 34 × 13-cm polypropylene chamber for antigen exposure; this consisted of two exposures, in a single day for 1 h each and separated by 4 h, to an aerosol generated from 0.5% OVA in sterile PBS. This group was termed OVA/OVA. The aerosol was generated with an ultrasonic nebulizer (Devilbiss Ultra Neb 99; Devilbiss Corp., Somerset, PA); a room air bias flow of 0.5 liters/min was maintained through the chamber. Negative controls were sensitized to OVA and challenged with aerosolized PBS, and were designated as OVA/PBS. Following challenge, the mice were returned to their cages for 2 or 3 d, after which time they were anesthetized with sodium pentobarbital (Nembutal; Abbott Laboratories, N. Chicago, IL), 70 mg/kg by intraperitoneal injection, and were prepared for measurement of bronchial reactivity to methacholine (MCh) or bronchoalveolar lavage (BAL), respectively, as subsequently described in detail.

Measurement of Lung Function

Pulmonary resistance (RL) was determined as previously described by De Sanctis and colleagues (22). Briefly, each mouse was anesthetized and its trachea was cannulated with a 19-gauge tubing adapter. The jugular vein was catheterized for the administration of methacholine (MCh) (as methacholine chloride), and a thoracotomy was performed. Following surgery, each animal was placed in a whole-body plethysmograph and ventilated with a rodent ventilator (Model 683; Harvard Apparatus, Natick, MA); a positive end-expiratory pressure (PEEP) of 3 cm H₂O was provided. RL was calculated by analysis of electrical signals proportional to transpulmonary pressure and lung volume. Changes in lung volume were calculated from the measured changes in plethysmograph pressure, and transpulmonary pressure was measured as the pressure difference between the pressure at the airway opening (Pao) and the pressure within the plethysmograph itself. Dose-response curves for MCh were obtained by administering sequentially increasing doses of MCh (33 µg/kg to 3,300 µg/kg) in volumes based on body weight. The peak RL response occurred in this interval and was used for the evaluation of airway reactivity. The vehicle, PBS, had no effect on RL. From the relationship between the dose of MCh administered and RL, the effective dose that would have resulted in an increase in RL was determined by log-linear interpolation. The dose required to increase the RL to 270% of baseline is termed the ED₂₇₀ RL, and can be interpreted as an index of pulmonary reactivity (23); because the doses of agonist are given in geometrically increasing amounts, it is common to logarithmically transform the data. Numerically low values of ED₂₇₀ RL indicate a high level of sensitivity to the administered agonist and are consistent with an asthma-like phenotype. Dose-response curves for MCh were constructed for each animal, and a mean response curve for each group was constructed from these data. From these curves, data were extracted to further describe the physiologic changes of the airways associated with antigen challenge in mice.

In Vitro Spleen-cell Proliferation

Mice were immunized with OVA and aluminum hydroxide as previoulsy described. On Day 12, the animals were killed and the extent of proliferation of splenic lymphocytes upon in vitro challenge with OVA was determined as detailed elsewhere (24). Briefly, spleen cells were isolated and suspended at a concentration of 10⁷/ml in RPMI 1640 medium (Gibco BRL) supplemented with 10% heat-inactivated fetal calf serum (FCS), 25 mM 4-(2-hydroxyethyl)-1-piperazine-N'-2-ethanesulfonic acid (HEPES) buffer, 2 mM glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin. One-hundred microliters of the cell suspension, containing 10⁶ cells, were added per well to flat-bottom microtiter wells (Corning Glassware, Corning NY). The cells were challenged with 100 µl of medium, or with serial 2-fold- diluted OVA (50 to 1.6 mg/ml) or anti-CD3 (Catalog No. 1452C11; Pharmigen, San Diego, CA) at 0.5 µg/ml as a control. The cells were then incubated for 72 h at 37°C with 5% CO₂ and 95% air, and were pulsed with 0.5 µCi of [³H]thymidine during the final 18 h of culture. The cells were harvested on glass-fiber filters, and the extent of [³H]thymidine uptake was determined with a liquid scintillation counter.

Collection and Processing of Samples

For the collection of bronchoalveolar lavage fluid (BALF), separate groups of mice were anesthetized and placed in the supine position, and the trachea was cannulated with a 19-gauge catheter with a beveled tip. The animals were exsanguinated by cutting the brachial artery; for measurements of IgE, blood samples were collected in Microtainer serum separator tubes (Becton Dickinson, Rutherford, NJ). Once bleeding had ceased, the lungs were lavaged three times with 1 ml of Ca²⁺- and Mg²⁺-free PBS. The initial lavage was instilled and retrieved one time, whereas the second and third lavages were each instilled twice. This procedure allowed for a greater number of lung washes with less diluent. Approximately 0.8 ml of lavage fluid was recovered from each of the lavages, for a pooled total of ~ 2.5 ml.

For the measurement of lavage total cell counts, the samples were centrifuged at 25°C and 1,500 rpm for 10 min. The supernatant was collected and stored at 80°C for measurements of eosinophil peroxidase, total protein, and IL-5. The cell pellet was resuspended in 1 ml cold PBS. Lavage total cell counts were made with a Coulter Counter (Model ZM; Coulter Electronics Corp., Hialeah, FL). Smears of the BAL cells were prepared by placing ~ 50,000 cells into a cytocentrifuge (Cytospin 3; Shandon-Lipshaw, Pittsburgh, PA) at 400 rpm for 10 min. The smears were air dried and stained with Wright-Giemsa stain. Differential counts were based on readings of at least 200 leukocytes.

Measurement of BALF Eosinophil Peroxidase, IL-5, Total Protein, and Serum Total and OVA-specific IgE

Eosinophil peroxidase in BALF was measured colorimetrically according to modified techniques previously described by Strath and colleagues (25). One hundred microliters of sample or of standard (porcine eosinophil peroxidase; ExOxEmis Corp., San Antonio, TX) were pipetted, in duplicate, into the wells of a 96-well plate (Cell Wells; Corning), followed by 100 µl of assay reaction mixture containing 0.05 M Tris buffer (Tris[hydroxymethyl]aminomethane, Trizma; Sigma Chemicals, St. Louis, MO), 0.1 µl 30% H₂O₂ (Fisher Scientific, Pittsburgh, PA), and 0.015% Triton X (Sigma Chemicals), pH = 8.0. The plate was incubated in the dark for 30 min, and the reaction was then terminated with 50 µl of 4 M H₂SO₄ per well, after which the samples and standards were read on a plate reader (Spectramax Model 340; Molecular Devices Corp., Sunnyvale, CA) at 490 nm. Regression analysis was done with the SOFTmax Pro analysis software package. The lavage IL-5 concentration was determined colorimetrically with a specific mouse IL-5 enzyme-linked immunosorbent assay (ELISA) kit (TiterZyme; PerSeptive Diagnostics, Cambridge, MA). The detection limit of the IL-5 assay was > 1 pg/ml, with no cross reactivity with other cytokines.

For the measurement of serum total and OVA-specific IgE, blood was collected in Microtainer serum separator tubes, clotted, and centrifuged at 3,500 rpm for 20 min at 25°C. Serum total IgE and OVA-specific IgE were measured with an ELISA, using previously described protocols (26). Antibody titers for OVA-specific IgE were calculated with mouse serum standards, and were reported as ELISA units of optical density as compared with serum standards, as previously described (26). The lower limit of detection for total IgE was ~ 100 pg/ml. Whole blood was collected into ethylenediamine tetraacetic acid (EDTA)- coated Microtainer tubes and analyzed for total white blood cell (WBC) count and differential count, using a Technicon H1E Blood Analyzer (Bayer Corp., Tarrytown, NY).

Histology and Immunohistochemistry

For conventional histology of the lung, separate groups of animals were sensitized and challenged under the same protocol as previously described. The trachea was cannulated in each mouse. For light microscopy, the lungs were gravity-inflated in situ to 25 cm H₂O with 10% buffered formalin (Fisher Scientific). The whole lung was embedded in paraffin, sectioned tangentially at a thickness of 5 µ, and stained with hematoxylin and eosin (H&E). The inflammatory infiltrate and ultrastructure of the lungs in each group of animials were evaluated in a blind fashion by an independent histopathologist. For immunohistochemical staining, the lung tissues were inflated with 50% ornithine carbamyltransferase (OCT)/PBS in situ, embedded in OCT (Tissue-Tek, Miles, IN), snap-frozen in liquid nitrogen, and stored at 70°C. Cryostat sections were fixed in cold acetone for 10 min and air dried. Nonspecific binding of antibody was blocked with Power Block containing casein (BioGenex, San Ramon, CA). The slides were incubated overnight with each primary antibody at room temperature, using the Sequenza Immunostaining Center (Shandon-Lipshaw). The negative control was developed with tissues incubated with rat IgG (BioGenex). The tissue was quenched with 0.3% H₂O₂ in methanol, and immunohistochemical labeling was achieved with the avidin-biotin-peroxidase method (BioGenex). Peroxidase activity was detected with 3-amino-9-ethylcarbazole (AEC) and counterstaining with hematoxylin.

Monoclonal Antibodies for Immunohistochemistry

The monoclonal antibodies used for immunohistochemistry were derived from hybridoma cell lines. YN1/1.7 (rat antimouse ICAM-1 [a gift from Dr. Fumio Takei]) and MK 2.7 (rat antimouse VCAM-1 [American Type Culture Collection, Rockville, MD]) were grown according to standard cell-culture technique, and were subsequently purified. Each antibody was tested for endotoxin with the Limulus amebocyte lysate test, and was then purified so that endotoxin levels were below 5 EU/ml.

Statistical Analysis

The results for each experiment are represented as mean ± SEM. The changes in cellular and biochemical components of the BALF were analyzed with Tukey's multiple comparison test (MCT). The serum IgE and ED₂₇₀ comparisons were done with Dunnett's t test. MCh dose- response curves were analyzed through a univariate analysis, comparing each dose administered among the groups.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Development of Airway Eosinophilia and Inflammation in ICAM-1^+/+ Mice

Evidence of lung inflammation in the mouse model used in the study was demonstrated by significant changes in both the cellular and biochemical lavage parameters in ICAM-1^+/+ OVA/OVA-challenged mice as compared with ICAM-1^+/+ OVA/PBS group controls. OVA challenge resulted in a significant increase in the total cell count recovered in the lavage from the ICAM-1^+/+ OVA/OVA as compared with the ICAM-1^+/+ OVA/PBS group (279 ± 66 × 10⁴ cells/ml versus 47 ± 7 × 10⁴ cells/ml; P  0.05) (Figure 1). This increase in total cells was primarily due to the influx of eosinophils, as represented by the percent eosinophil composition in the lavage (66 ± 8% versus 0; P  0.05) (Figure 1). Additionally, evidence of airway inflammation consistent with a Th2 response in this model was shown by the presence of IL-5 in the lavage fluid; lavage-fluid IL-5 levels were significantly greater in the OVA/ OVA than in the OVA/PBS mice (113 ± 44 pg/ml versus 39 ± 4 pg/ml [P < 0.05], respectively) (Figure 2). The levels of eosinophil peroxidase were significantly higher in the BALF of OVA/OVA than in that of OVA/PBS mice (109 ± 12 pg/ml versus 28 ± 3 pg/ml; P  0.05) (Figure 2). Furthermore, we observed changes in lavage protein consistent with an altered lung permeability in the OVA/ OVA as compared with the OVA/PBS mice (496 ± 74 mg/ dl versus 172 ± 35 mg/dl, P  0.05) (Figure 3).

View larger version (18K): [in this window] [in a new window]   Figure 1.   BAL total cells (solid bars) and percent eosinophils (open bars) in wild-type (+/+) and ICAM-1 deficient (/) mice. Total cells and percent eosinophils of the (+/+) OVA/ OVA mice were significantly increased (*P  0.05) compared with all other groups by Tukey's multiple comparison test. Total leukocytes and percent eosinophils in the ICAM-1-deficient mice were not significantly different from those in the PBS-challenged groups. Values are mean ± SEM.

View larger version (20K): [in this window] [in a new window]   Figure 2.   Results of analysis of BAL for IL-5 (solid bars) and eosinophil peroxidase (open bars) in wild-type (+/+) and ICAM-1-deficient mice (/). Values are mean ± SEM, with * indicating significance at P  0.05 by Tukey's multiple comparison test. Wild-type (+/+) OVA/OVA mice had a significant increase in both IL-5 and eosinophil peroxidase after antigen challenge, which was not observed in the ICAM-1 (/) mice.

View larger version (22K): [in this window] [in a new window]   Figure 3.   Total protein analysis of BALF. Values are mean ± SEM, with * indicating significance at P  0.05 by Tukey's multiple comparison test of wild-type (+/+) versus ICAM-1 deficient (/) mice. BALF total protein was increased in the wild-type (+/+) mice as compared with all other groups.

Inhibition of Airway Eosinophilia and Inflammation In ICAM-1-deficient (ICAM-1^/) Mice

Mice genetically deficient in ICAM-1 showed a significant reduction in all of the cellular and biochemical factors measured in the BALF. Values for lavage total cells and percent eosinophils in the ICAM-1^/ OVA/OVA mice (37 ± 6 × 10⁴ cells/ml and 5 ± 3% eosinophils) were significantly (P  0.05) smaller than those found in the ICAM-1^+/+ OVA/OVA mice (279 ± 66 × 10⁴ cells/ml and 66 ± 8% eosinophils), and not statistically different from the values for the OVA/PBS groups (Figure 1). Neither the percent neutrophils nor the percent lymphocytes were statistically different in the OVA/OVA^+/+ and OVA/ OVA^/ phenotypes. Analysis of IL-5 and eosinophil peroxidase activity in BALF also demonstrated a significant difference between the ICAM-1^/ mice and the ICAM-1^+/+ controls (36 ± 4 pg/ml versus 113 ± 44 pg/ml IL-5, and 80 ± 5 pg/ml versus 109 ± 12 pg/ml eosinophil peroxidase) (Figure 2). BALF total protein was also attenuated (Figure 3) in the ICAM-1^/ OVA/OVA as compared with the ICAM-1^+/+ OVA/OVA mice (267 ± 44 mg/dl versus 496 ± 74 mg/dl, respectively, P  0.05).

Peripheral Blood Leukocyte Counts in OVA/OVA ICAM-1^+/+ and ICAM-1^/ Mice

Circulating WBC counts for OVA/OVA ICAM-1^/ (n = 5) and OVA/OVA ICAM-1^+/+ (n = 5) mice were evaluated 3 d after antigen challenge. No differences between the ICAM-1^/ and ICAM-1^+/+ mice were observed in the total leukocyte count (8.5 ± 0.8 × 10⁶/ml versus 6.7 ± 0.8 × 10⁶/ml), percent eosinophils (6.5 ± 1.5 versus 5.0 ± 2.4), percent neutrophils (13.0 ± 3.0 versus 18.2 ± 4.7), or percent monocytes (71.9 ± 3.0 versus 69.2 ± 3.3), respectively.

Induction of OVA-specific Immunity in ICAM-1^/ and ICAM-1^+/+ Mice

Active sensitization induced a significant (P < 0.05) increase in both total and OVA-specific IgE in both the ICAM-1^+/+ and ICAM-1^/ animals as compared with unsensitized animals both prior to and 3 d after OVA challenge (Table 1). There were no significant differences observed in the wild type (+/+) versus the ICAM-1-deficient (/) mice for serum total or OVA-specific IgE either before or after OVA challenge.

To determine whether the reduction in airway inflammation observed in OVA-challenged ICAM-1^/ mice was associated with inefficient priming of OVA-specific T cells, the extent of in vitro splenocyte proliferation following OVA challenge in ICAM-1^/ and ICAM-1^+/+ animals was evaluated. Splenocytes from ICAM-1^/ mice proliferated poorly in response to in vitro OVA challenge when analyzed after 3 d of culture (Figure 4). In contrast, strong in vitro proliferation in response to OVA was observed with splenocytes isolated from OVA-immunized, ICAM-1-sufficient mice.

View larger version (11K): [in this window] [in a new window]   Figure 4.   Spleen-cell proliferation assay with serial twofold dilutions of OVA (50 to 1.6 mg/ml) 12 d after sensitization. ICAM-1/ (solid circles) and ICAM-1+/+ (solid squares) mice were sensitized to OVA but not challenged. Data are represented as the mean ± SEM for counts per minute of incorporated thymidine. CD3 stimulation for the / and +/+ mice was 30K cpm and 70K cpm, respectively. Spleen cells from +/+ mice proliferated strongly in response to OVA, whereas spleen cells from / mice did not.

Development of AHR in ICAM-1^+/+ Mice

Previous studies have indicated a significant change in airway reactivity to MCh in mice following antigen challenge. In the model used in the present study, antigen challenge resulted in a significant increase in airway reactivity in the wild-type mice treated with OVA/OVA as compared with mice treated with OVA/PBS. When dose-response curves for the groups were compared, we found significant (P < 0.05) differences in the peak response to MCh at the 100, 330, 1,000, and 3,300 µg/kg doses. OVA-challenged animals also achieved a higher plateau in the MCh dose- response curve (Figure 5). Log ED₂₇₀ RL comparisons between the two groups could not be made, since several of the ICAM-1^+/+ mice in the OVA/PBS group did not achieve these increases from baseline RL.

View larger version (15K): [in this window] [in a new window]   Figure 5.   Peak increase in lung resistance in response to increasing intravenous doses of MCh. Values are mean ± SEM (n = 6 in each group). The groups were as follows: OVA/OVA ICAM/ mice (solid squares), OVA/OVA ICAM+/+ mice (solid circles), OVA/PBS ICAM/ mice (solid upright triangles), and OVA/PBS ICAM+/+ mice (solid inverted traingles). Significance (*P  0.05) was determined by a univariate analysis, comparing the response in the OVA/OVA ICAM-1+/+ group and the OVA/ PBS ICAM-1+/+ and OVA/OVA ICAM-1/ groups at each dose of MCh.

Inhibition of AHR in ICAM-1^/ Mice

To determine the role of ICAM-1 in the induction of AHR, we generated MCh dose-response curves 2 d after antigen challenge in both ICAM-1^/ and ICAM-1^+/+ mice. The ICAM-1^/ OVA/OVA mice did not show a significant change in airway reactivity as compared with either ICAM-1^+/+ OVA/PBS or ICAM-1^/ OVA/PBS mice; however, there was a significant increase in their log ED₂₇₀ RL when compared with that of ICAM^+/+ OVA/ OVA mice (2.65 ± 0.08 versus 2.36 ± 0.07, respectively). Additionally, there was a significant increase in the maximal response to MCh at doses of 100, 330, 1,000 and 3,300 µg/kg in the ICAM-1^+/+ OVA/OVA over that of the ICAM-1^/ OVA/OVA mice.

Histologic and Immunohistochemical Evaluation

Examination of the lung 72 h after antigen challenge showed pathologic differences between the ICAM-1-deficient and wild-type mice. These were characterized by a large perivascular eosinophil infiltration, airway eosinophilia, and septal-wall and airway-lumen edema in the sensitized ICAM-1^+/+ mice challenged with OVA. In contrast, the lungs of sensitized ICAM-1^/ mice challenged with OVA showed a markedly attenuated inflammatory infiltrate into the perivascular regions and minimal margination to the alveolar spaces (Figure 6). Neither ICAM^+/+ OVA/PBS nor ICAM-1^/ OVA/PBS mice showed any significant histopathologic changes.

View larger version (148K): [in this window] [in a new window]   Figure 6.   Histologic features of the lungs of ICAM-1/ and ICAM+/+ mice 3 d after aerosol antigen challenge. (a and b) Sections from a OVA/OVA ICAM-1+/+ mouse displaying perivascular and airway leukocyte infiltrate ( ), characterized largely by the presence of eosinophils (magnifications: ×40 and ×100, respectively). (c and d) Sections from a OVA/OVA ICAM-1+/+ mouse, showing a noticeably reduced airway and perivascular cell infiltrate (magnifications: ×40 and ×100, respectively). Lung sections from ICAM-1+/+ OVA/PBS and ICAM-1/ OVA/PBS mice appeared normal.

Immunohistochemistry revealed staining for ICAM-1 in the lungs at both 24 and 72 h after OVA challenge in all of the groups tested. This feature of the ICAM-1^/ transgenic mouse is consistent with the findings by King and colleagues of novel isoforms of ICAM-1 generated in this mouse (27). Staining for VCAM-1 was detected in both of the OVA/OVA-treated groups 1 d after antigen challenge. More interesting was that the lungs from the ICAM-1^/ mice treated with OVA/OVA stained positively for VCAM-1, whereas there was no apparent increased staining for VCAM-1 in the ICAM-1^+/+ OVA/ OVA mice or either of the PBS-challenged groups at 3 d after antigen challenge (Figure 7).

View larger version (148K): [in this window] [in a new window]   Figure 7.   Immunohistochemical staining for VCAM-1 in the lungs of ICAM-1/ and ICAM+/+ mice 3 d after aerosol antigen challenge. (a) Lung venule from ICAM/ OVA/OVA mouse 3 d after challenge, with negative staining for VCAM-1 on the endothelial-cell surfaces of the vessel. (b) Lung venule from ICAM+/+ OVA/OVA mouse 3 d after challenge, showing increased staining for VCAM-1 on the endothelial-cell surfaces.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The findings in our study demonstrate an important role for the common isoform of ICAM-1 in the pathogenesis of AHR and inflammation in a murine model of allergic asthma. In murine models, airway eosinophilia and inflammation occur as soon as 24 h after antigen challenge, and may be resolved by Day 14 (10). We have shown that the response in C57BL/6J mice is accompanied by a significant increase in serum IgE, and by increases in BALF eosinophil peroxidase activity, IL-5, and total protein. In addition, we found a significant increase in airway responsiveness to intravenous MCh in the sensitized and challenged ICAM-1^+/+ mice as compared with the ICAM-1^/ group. This asthma-like response is specific to allergen challenge, because mice challenged with PBS developed no pathobiologic changes.

Following antigen challenge, sensitized wild-type mice exhibited significantly greater bronchial reactivity to intravenous MCh than did the ICAM-1-deficient mice. These changes included changes in the log MCh ED₂₇₀ RL. Reporting only this value would suggest a shift in the entire dose-response curve for the ICAM-1^/ group, but this was not entirely the case. In fact, the response of all animals was similar in magnitude at the lower doses of MCh, and significant differences between the groups were not apparent until RL increased by more than 200% from its baseline value. The most remarkable changes associated with antigen challenge in the wild-type mice were the significantly greater responses at the higher end of the dose- response curve. The sensitized and PBS-challenged animals and the sensitized and OVA-challenged ICAM-1 knockout mice achieved a plateau, whereas the wild-type antigen-challenged mice responded in a manner consistent with severe bronchoconstriction. This analysis of the individual- and combined-group dose-response curves provided valuable information supporting an important role for ICAM-1 in the pathogenesis of airway dysfunction. The increase in slope and change in the plateau response are important components of the human asthma phenotype (28), and indicate that the inflammatory response that we induced recapitulates critical features of human asthma.

We also examined the role of ICAM-1 in the induction of humoral and cellular immune responses to OVA. We measured serum total and OVA-specific IgE levels in order to determine whether ICAM-1 disruption prevented the T-cell-dependent sensitization of B cells. Analysis indicated no difference between the and ICAM-1^+/+ and ICAM-1^/ mice in their ability to synthesize IgE, suggesting that the sensitization and associated antigen-processing events were independent of ICAM-1. In addition, we observed that after immunization with OVA, spleen cells from ICAM-1^/ mice proliferated poorly in response to OVA challenge when compared with spleen cells from ICAM-1-sufficient mice (Figure 4). Thus, it appears that one contributing factor to the lack of airway inflammation in ICAM-1^/ mice may be inefficient priming of OVA-specific T-cells. Whether a more vigorous immunization protocol will sufficiently override this deficit remains to be determined. It is interesting to note that our observations made in this model of allergic AHR differ from the observed effects of ICAM-1 gene disruption on the induction and development of immunoinflammatory disease in a model of collagen-induced arthritis. Bullard and colleagues (29) showed that arthritis-prone DBA/1 mice homozygous for the ICAM-1 mutation mounted normal cellular and humoral immune responses to type II collagen, but did not develop arthritis. It is possible that the genetic background of the mouse strain used in the aforementioned study contributed to the experimental results. Alternatively, the data of Bullard and colleagues and our current findings suggest that the target organ and specific immune process may also dictate the relative dependency on ICAM-1 of the induction and perpetuation of disease.

Several groups have previously reported a significant role for the ICAM-1 cell-adhesion pathway in animal models of asthma (5, 17, 34). These studies utilized monoclonal antibodies against either ICAM-1 or the ligands LFA-1/MAC-1, and showed efficacy in limiting both airway eosinophilia and AHR. The present study supports the earlier findings but extends them in a more comprehensive model. Mice homozygous for a disrupted ICAM-1 gene do not develop the antigen-induced and eosinophil-mediated airway inflammation that has been recognized as one of the hallmarks of asthma. We demonstrated inhibition of airway eosinophilia and of lavage IL-5 and eosinophil peroxidase production in the lungs of ICAM-1^/ mice. IL-5 is an important cytokine in human asthma, is produced by Th2 cells, and is known to stimulate eosinophil maturation, activation, and survival (30). Elevated levels of eosinophil peroxidase in the lavage fluid of the ICAM-1^+/+ mice, which were attenuated in the ICAM-1^/ mice, most likely reflected the changes in activation state of the eosinophils in the two groups. The increased presence of these mediators in the lavage fluid of the ICAM-1^+/+ but not that of the ICAM-1^/ group is consistent with the diminished accumulation of eosinophils in the airways of OVA-challenged mice. Furthermore, the increased lavage-fluid protein in the wild-type ICAM^+/+ but not the mutant ICAM^/ mice may indicate the development of endothelial-cell injury.

The cell-adhesion pathways for the migration of inflammatory leukocytes in the sensitized and OVA-challenged ICAM-1^/ and ICAM-1^+/+ mice were examined with immunohistochemistry. Staining for ICAM-1 was positive in the lungs of both the ICAM-1^/ and ICAM-1^+/+ mice at 3 d after challenge. This feature of the ICAM-1^/ transgenic mouse is consistent with the findings of King and coworkers of the presence of novel isoforms of ICAM-1 generated in this mouse (27). In their study, alternatively spliced forms of ICAM-1, with an incomplete extracellular domain structure relative to the common isoform of ICAM-1, were detected with both reverse transcription- polymerase chain reaction (RT-PCR) and immunohistochemistry in both wild-type and ICAM-1^/ mutant mice. Despite the ability of these novel isoforms of ICAM-1 to bind LFA-1 and antibodies to ICAM-1, there is no evidence of a role of these ligands in the pathogenesis of inflammatory or immune responses in vivo. In fact, evidence provided by ICAM-1^/ mouse models of arthritis (29) and of systemic autoimmune diseases characterized by glomerulonephritis and vasculitis (33) and resembling lupus erythematosus in humans implicates the common ICAM-1 isoform as essential to the pathogenesis of disease, and extends our findings.

Staining for VCAM-1 revealed increased VCAM-1 in the ICAM-1^/ OVA/OVA mice but not in the ICAM-1^+/+ OVA/OVA or PBS-treated animals 3 d after antigen challenge. It is interesting that despite the persistence of VCAM-1 expression in ICAM-1^/ mice, eosinophils were not recovered in the BALF at 3 d after airway antigen challenge. Both OVA-challenged groups of mice in our study exhibited positive staining for VCAM-1 at 1 d after OVA challenge. These data suggest the persistent availability of VCAM-1-mediated cell-adhesion pathways in the absence of ICAM-1 in the ICAM-1^/ OVA/OVA animals as compared with the ICAM-1^+/+ OVA/OVA mice, in which staining for VCAM-1 was seen only immediately after challenge. Inflammatory cells may predominantly utilize ICAM-1 in the latter stages after antigen challenge, and absence of this adhesion molecule prevents eosinophil migration in the lung.

In conclusion, our data show that ICAM-1 is critical to the development of antigen-induced AHR, inflammation, and T-cell sensitization to allergen in mice, and provide further support for a role of ICAM-1 in human asthma. These data also support the general concept that pharmacologic inhibition or antagonism of cell-adhesion pathways may alter the progression and pathophysiology not only of asthma, but also of other inflammatory diseases including arthritis, inflammatory bowel disease, and autoimmune disease.