INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Airway inflammation is now accepted as a defining characteristic of the asthma phenotype (1). It has long been known that eosinophils are present in increased numbers in the airways and peripheral blood of patients with severe asthma, and it is now clear that even patients with mild intermittent asthma recruit these cells to their airways (4). The mechanisms that are responsible for selective eosinophil recruitment are of interest because eosinophilia is linked to asthma severity (5, 6). Eotaxin is a chemotactic cytokine or chemokine of the CC class that selectively attracts eosinophils by activating CCR3 receptors (7, 8), which are also present on basophils (9) and T lymphocytes (10). Its relevance to asthma has been suggested by animal experiments where abrogation of eotaxin action by ligand neutralization or by manipulations in some (11) but not all strains (12) of genetically altered mice have demonstrated a role for eotaxin in producing allergen-induced airway eosinophilia (13). It is known that the airway epithelium and subepithelium of patients with mild asthma and airway eosinophilia express more eotaxin mRNA and protein than those of normal subjects (16, 17). It is also known that eotaxin expression in endobronchial biopsies correlates directly with the sensitivity of subjects with asthma to contractile agonists (17). On the basis of these observations, we tested the hypothesis that airway cells in humans respond to segmental allergen challenge by expressing eotaxin mRNA and secreting eotaxin protein in direct relation to the mobilization of eosinophils.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subject Characterization

Approval for this study was prospectively granted by our Institutional Review Board, and all subjects gave prior written informed consent. We studied five normal subjects who had never smoked or had a physician's diagnosis of any respiratory disorder including asthma. All of these subjects were free of comorbid disease and responded negatively to all questions on a screening questionnaire for allergic and respiratory disease. All five subjects in this group had normal lung function and experienced a 20% or greater reduction in baseline FEV₁ value in response to 8 mg/ml of inhaled methacholine. We also studied six nonsmoking subjects who met the ATS criteria for an asthma diagnosis (18). All six subjects demonstrated a stable baseline FEV₁ of at least 70% of the predicted normal value, with 15% or greater reversibility to inhaled -agonist or a greater than 20% decline in FEV₁ on exposure to less than 8 mg/ml of nebulized methacholine. All of these subjects used only short-acting -agonists for control of asthma symptoms (one subject used loratadine for nasal symptoms); none of the subjects was taking inhaled steroids, methyl xanthines, or leukotriene modifiers. Each subject with asthma had an abnormal allergen skin prick test defined as a wheal of greater than 5 mm (averaged orthogonal diameters) in response to at least 1 of 12 standard allergens in 50% vol/vol glycerin solution (Eastern 10-tree mix, ragweed mix, cat hair, 3-weed mix, dog epithelium/mixed breeds, timothy grass, D. farinae, D. pteronyssinus, Alternaria tenius, cockroach, Cladosporium herbarum, and Aspergillus mix; Greer Laboratories Inc., Lenoir, NC) and cat hair standardized extract (ALK Laboratories Inc., Milford, CT). Following skin prick testing, quantitative skin testing was performed with aqueous solutions of at least one symptom-associated allergen that provoked a positive skin prick test. The weakest dilution that produced a 3-mm wheal 15 min after administration was determined. At a second visit at which lung function was determined to be stable, with an FEV₁ of at least 70% of the predicted normal value, aqueous-phase allergen solutions of increasing concentration were administered by nebulization according to the method of Cockcroft and coworkers (19), and the provocative concentration required to reduce the FEV₁ by 20% (PC₂₀) was determined. This concentration was used in segmental allergen challenge as detailed below.

Segmental Allergen Challenge

Segmental allergen challenge with seasonal allergens was performed at times when these agents were not in season. All subjects were free of respiratory symptoms and had stable lung function for at least 4 wk following inhalational challenge before segmental allergen challenge, for which subjects underwent two serial bronchoscopic procedures 4 h apart. Bronchoscopy was performed under midazolam/fentanyl conscious sedation with a total of < 400 mg of topical lidocaine. The bronchoscope was advanced to airway occlusion in the anterior segment of the left upper lobe, three 50-ml aliquots of warmed sterile saline were serially instilled, and bronchoalveolar lavage (BAL) fluid was recovered into Teflon containers by gentle aspiration, after which endobronchial biopsies were performed. In subjects with asthma the bronchoscope was then moved to the right upper lobe, where 2 ml of allergen diluent was delivered; after 5 min the bronchoscope was moved and advanced into the right middle lobe, where 2 ml of allergen solution was delivered (20). The visible effects of diluent and allergen instillation on the airway mucosal surface and on subsegmental airway caliber were recorded by means of electronic images. Allergen solution was prepared just before the procedure, with 1:10 vol/vol dilution of the allergen concentration that caused a 20% reduction in FEV₁, that is, 10% of the PC₂₀ FEV₁ (20). The bronchoscope was then removed, and focal wheezing was confirmed to be present exclusively over the right middle lobe in all subjects by auscultation. No subject reported subjective symptoms of asthma or had a significant decrement in FEV₁ during the 4-h recovery period before the second bronchoscopy was performed. BAL fluid and endobronchial biopsy specimens were obtained from both the right upper lobe (diluent) and the right middle lobe (antigen) sites.

Cell Processing and Eotaxin ELISA

BAL fluid was strained through a gauze mesh and centrifuged (700 × g, 5 min, 4° C); the supernatant was stored at 80° C. After thawing, samples were assayed for eotaxin protein using a solid-phase enzyme-linked immunosorbent assay (ELISA). The monoclonal primary antibody used in this assay is directed to eotaxin epitopes that are not affected by glycosylation; the assay gives similar results when an alternative monoclonal biotinylated secondary antibody is substituted for the polyclonal reagent that was used in this report. The limit of detection for this assay in BAL fluid is 10 pg/ml, with a coefficent of variation of 15%; the assay was performed as previously described except that the primary antibody (2A12) was used at 150 ng/well (21). Aliquots of BAL cells were pelleted onto glass slides by cytocentrifugation (Shandon Inc., Pittsburgh, PA) and stained with Diff-Quik (Fischer Scientific, Pittsburgh, PA) or subjected to immunohistochemical analysis for eotaxin as described below. The remaining BAL cells were washed and cultured in RPMI-1640 with 10% fetal bovine serum (FBS) for 6 h on 10-cm² culture plates at a concentration of 0.5-1.0 × 10⁶ cells/ml. After a 2-h incubation, the cells were cultured for an additional 4 h in the presence or absence of interleukin-1-beta (IL-1, 10 ng/ml), gamma interferon (IFN-, 100 ng/ml), or lipopolysaccharide (LPS, 10 µg/ml). Dexamethasone (10 µM) was added to selected wells 1.5 h after the initiation of cell culture.

RNA Analysis

Total RNA was isolated from freshly harvested cells by guanidinium-thiocyanate-phenol chloroform extraction (Stratagene Inc., La Jolla, CA). For northern analysis 2.5 to 15 µg of total RNA was subjected to gel electrophoresis on a formaldehyde-2% agarose gel and transferred to a nylon membrane (Schleicher and Schuell, Keene, NH). After UV crosslinking, the membrane was hybridized at 68° C in ExpressHyb Hybridization Solution (Clontech, Palo Alto, CA) with a ³²P-labeled 0.35-kb cDNA probe containing the entire coding region of the human eotaxin gene (22) or a glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA probe (Clontech). The membranes were washed for 10 min at room temperature in 2× SSC-0.05% sodium dodecyl sulfate (SDS) and then for 20 min at 50° C in 0.2× SSC- 0.1% SDS. In each case the eotaxin signal was detected at the site expected for a 0.8-kb transcript. To control for differences in RNA loading, the hybridization signal for eotaxin was normalized to that for GAPDH in each sample.

Immunohistochemical Staining

Immunocytochemical staining was performed on serial 3-µm sections as previously reported (23). Due to substantial variability among the subjects regarding the relative immunostaining of airway biopsy tissue and background, staining conditions were optimized individually for each case by techniques similar to those previously reported (24). The technique that resulted in the best balance between immunostaining and background staining was used for all samples from a given subject. Two different murine monoclonal antibodies to eotaxin (LS59 2G6 and LS42 6H9; LeukoSite Inc., Cambridge, MA) were tested on the biopsy samples, and both clones resulted in the same immunostaining patterns. For the bronchoalveolar lavage samples, cells were fixed in 4% paraformaldehyde for 10 min and then treated with 0.3% triton for 5 min. Nonspecific immunoglobulin binding was blocked with 10% normal horse serum. The purified murine monoclonal antibody diluted 1:200 in phosphate-buffered saline (PBS)/2% bovine serum albumin (BSA) was applied to the samples, which were then incubated at 4° C overnight. The slides were then incubated in the secondary antibody (biotinylated horse antibody to goat immunoglobulin G [IgG]; Vector Laboratories, Burlingame, CA) diluted in 5% powdered milk made with PBS at 4° C for 2 h. Endogenous peroxidase activity was quenched with methanol with 1% hydrogen peroxide. ABC standard (Vector Laboratories) was applied to the samples, which were then incubated at room temperature for 1 h (25). Biotinylated tyramide (NEN Life Science Products, Boston, MA) was applied to the samples for 6.5 min at room temperature, followed by strepavidin-horseradish peroxidase (NEN Life Science Products) for 30 min at room temperature. Immunopositivity was localized with the chromagen diaminobenzidine (0.025%) in PBS and 0.1% hydrogen peroxide. As a negative control, an irrelevant IgG (MOPC, Sigma) was substituted for the primary antibody. Immunostaining for paraffin-embedded biopsy samples used a technique similar to that used for the BAL samples. However, for the cases with very high background staining, the primary antibody was prepared with the DAKO Animal Reasearch Kit (DAKO Corporation, Carpinteria, CA), acording to the manufacturer's instructions for a biotinylated primary antibody. The remainder of the immunostaining technique was as described above, except that no secondary antibody was used for the biotinylated primary antibodies. Immunostaining for well-characterized markers for cell types, including a rabbit polyclonal antibody to keratin to identify epithelial cells (used at 1:300, DAKO Corp.), a murine monoclonal antibody to EG2 to identify activated eosinophils (used at 1:200; Pharmacia), and a murine monoclonal antibody to CD68 to identify macrophages (clone EBM11, used at 1:100; DAKO Corp.), was performed as stated above for the BAL samples.

Analyses of Immunostaining

The BAL samples were examined under high power; 200 cells were counted, and the percentage of immunopositive cells was calculated. The cell type was determined by the histologic appearance, including the nuclear morphometry. For example, epithelial cells were identified as ciliated mononuclear cells. The immunostaining on the paraffin-embedded biopsies was analyzed by morphometry. Each biopsy was digitized with an Optronics digital camera (Optronics, Goleta, CA) interfaced with Scion Image version 1.3 software (Scion Corporation, Frederick, MD); the tissue area was obtained with Optimas image analysis software (Media Cybernetics, Silver Springs, MD). The total number of immunopositive cells in each biopsy were counted, and a density of immunopositive cells was calculated for each sample. Cell types were determined by serial immunostaining with well-characterized markers, which included keratin for epithelial cells, EG2 for eosinophils, and CD68 for macrophages, and by examination of nuclear morphology.

Statistical Analysis

Values for eotaxin levels were tested for normalcy and analyzed by t test, paired t test, or analysis of variance as appropriate. Data are presented as the mean and standard error of the mean (SEM); a p value of less than 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Characteristics

As expected, on the basis of our inclusion and exclusion criteria, the subjects with asthma had significantly lower values of FEV₁ (p = 0.013), were more sensitive to the contractile effects of methacholine (p < 0.005), and demonstrated more positive dermal reactions by skin prick test (p < 0.005) than the normal subjects (Table 1).

Bronchoscopic Findings

The instillation of allergen was associated with visible alterations in the airway mucosa in all subjects; these changes were not observed in the segments into which diluent had been instilled. Allergen instillation caused an almost immediate increase in the prominence of the vascular pattern (Figure 1A), which was followed in minutes by diffuse erythema localized to the airways distal to the position of the bronchoscope (Figure 1B) and then by visible airway narrowing (Figure 1C) associated with a reduction of 25 to 40% of the area available for air flow. The focal nature of the airway narrowing was supported by the presence of focal wheezing exclusively over the right middle lobe and the absence of subjective symptoms of asthma.

View larger version (34K): [in this window] [in a new window]   Figure 1.   Airway images before and after bronchoscopic delivery of allergen solution in two representative subjects.

BAL Cell Recovery

Total cells as well as macrophages, lymphocytes, neutrophils, and eosinophils increased markedly in allergen-challenged compared with unchallenged or diluent-exposed segments (p < 0.05, Table 2). Recovery of total cells and eosinophils was greater in allergen-challenged segments than in control segments in every subject with asthma studied, but we observed significant variability between subjects with respect to eosinophil mobilization after segmental allergen challenge. Eosinophils constituted 2-3% of the cells recovered before allergen challenge and 22% of the cells recovered after segmental allergen challenge (p < 0.05 by paired t test and by ANOVA based on ranks).

BAL Cell RNA Analysis

BAL cell eotaxin RNA expression was undetectable in all five normal subjects studied and in our subjects with asthma in the absence of allergen challenge. Eotaxin RNA expression was not increased in BAL cells treated with IL-1, IFN-, or LPS alone but was significantly increased after allergen challenge in the five subjects for whom Northern analysis was successful (p < 0.05, Figure 2B). Allergen-stimulated eotaxin RNA expression was significantly diminished after treatment with 10 µM of dexamethasone. The majority of cells cultured under these conditions had monocytic morphology.

View larger version (22K): [in this window] [in a new window]   Figure 2.   (A) Northern analysis detecting eotaxin mRNA expression in BAL cells from lung segments of subjects with atopic asthma before allergen challenge (prechallenge), after diluent exposure (sham), or following allergen challenge (antigen) in the presence or absence (US) of 10 ng/ml IL-1, 100 ng/ml IFN-, or 10 µM dexamethasone (Dex). ( B ) Eotaxin mRNA expression was greater in cells from antigen-challenged segments than from either control segments or antigen-challenged segments treated with dexamethasone (n = 5, p < 0.05 by ANOVA on ranks).

Eotaxin Protein Quantification

Eotaxin protein was detectable in the BAL fluid of all subjects. In the absence of allergen challenge, BAL eotaxin levels were significantly higher in subjects with asthma than in normal subjects (25 ± 3 versus 15 ± 2 pg/ml, p < 0.05, Figure 3A). Segmental allergen challenge was associated with an increase in BAL eotaxin in every patient with asthma studied, whereas diluent infusion did not increase eotaxin levels. BAL eotaxin levels increased from 22 ± 4 to 53 ± 10 pg/ml (p = 0.013 by paired t test, Figure 3B). Eotaxin recovery increased by a factor of 2 or greater in four subjects and increased only modestly in the other two. Increased BAL levels of eotaxin were associated with increased recovery of eosinophils. We detected a strong direct correlation between the increase in BAL eotaxin concentration and the fold increase in BAL eosinophils in subjects with asthma (r² = 0.88, p = 0.0036, Figure 4).

View larger version (16K): [in this window] [in a new window]   Figure 3.   BAL eotaxin levels were significantly higher in subjects with atopic asthma (n = 6) than in nonatopic normal subjects (n = 5, p < 0.05 by t test, A). BAL eotaxin levels at 4 h were greater in allergen-challenged segments than in diluent-exposed segments (n = 6, p < 0.05 by paired t test, B).

View larger version (20K): [in this window] [in a new window]   Figure 4.   Allergen-induced increases in BAL eotaxin levels were strongly associated with increased recovery of eosinophils in BAL fluid after allergen challenge (n = 6, r2 = 0.88, p = 0.004).

Immunohistochemical Analysis

Similar to our analysis of eotaxin RNA expression and protein recovery, significantly more cells demonstrated immunohistochemical evidence of the presence of eotaxin after segmental allergen challenge. A greater percentage of BAL cells from allergen-challenged segments demonstrated eotaxin immunopositivity than cells from diluent- or unexposed control segments (77 [41-100]%, A, versus 22 [13-30]%, D, or 17 [13- 28]%, U, median [interquartile range], p < 0.05). The majority of immunopositive cells were identified as macrophages and eosinophils. Segmental allergen challenge was associated with a significant increase in the percentage of BAL macrophages (50 [27-100]%, A versus 18 [12-24]%, D, or 14 [5-23]%, U, p < 0.05) and eosinophils (5 [3-30]%, A, versus 0.4 [0.2-0.7]%, D, or 0.3 [0.2-0.4]%, U, p < 0.05, Figure 5C) that were immunopositive for eotaxin. To confirm these findings and better identify eotaxin-producing cell types in the airway, we performed a morphometric analysis of endobronchial biopsy tissue taken 4 h after segmental allergen challenge. The density of eotaxin-immunopositive cells was significantly greater in the airways of allergen-challenged segments than in normal airways, asthmatic airways before allergen challenge (preantigen), and airways from (sham) segments exposed only to allergen diluent (p < 0.01, n = 5, Figure 6). The difference in density of immunopositive cells between the preantigen and normal groups did not reach statistical significance (p = 0.067). The distribution of immunopositivity among the eotaxin-expressing cell types was also affected by allergen challenge. In normal subjects 41 ± 14% of the immunopositive cells were epithelial cells and 59 ± 15% were macrophages. In subjects with asthma before antigen and after exposure to the antigen diluent, 23 ± 17% of immunopositive cells were epithelial cells, 10 ± 10% were eosinophils, and 67 ± 16% were macrophages. In allergen-challenged segments 72 ± 7% of the immunopositive cells were epithelial cells, 2 ± 1% were eosinophils, and 25 ± 7% were macrophages. Cell type was ascertained by staining serial sections with keratin for identification of epithelial cells, EG2 for eosinophils, and CD68 for macrophages, and by assessment of nuclear morphology. These findings demonstrate that eotaxin expression is increased in airway epithelial cells (Figure 6).

View larger version (70K): [in this window] [in a new window]   Figure 5.   Eotaxin in BAL cells from subjects with asthma in the absence (A) or presence (B) of segmental allergen challenge (original magnification: ×100). Eotaxin was detected in significantly more macrophages and eosinophils from allergen-challenged segments than from diluent-exposed segments or from a distinct segment before allergen exposure (n = 6, *p < 0.05 by ANOVA on ranks C ).

View larger version (86K): [in this window] [in a new window]   Figure 6.   Immunohistochemical detection of eotaxin in endobronchial biopsy tissue from representative antigen-challenged (original magnification: ×20, A) and (original magnification: ×40, B) and control segments (original magnification: ×20, C ). Eotaxin detection in endobronchial biopsy tissue of a nonatopic normal subject (original magnification: ×20, D). Density of immunopositive cells in tissue from subjects with asthma before challenge and after diluent or allergen challenge and from nonatopic normal subjects (right).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study demonstrates that eotaxin is produced and released by pulmonary resident cells after allergen challenge. The immunolocalization of eotaxin to the epithelial surface of the lung not only is consistent with the established sites of constitutive eotaxin expression (16, 17) but also extends these observations by demonstrating an increase in the expression of both eotaxin protein and mRNA at these same loci after allergic stimulation. This finding is relevant because segmental allergen challenge is associated with BAL eosinophilia and the recruitment of eosinophils into the airway epithelium. The hypothesis that eotaxin recruits eosinophils to the asthmatic lung is supported by the demonstration that segmental allergen challenge elicits eotaxin production at the sites to which eosinophils are recruited. Our finding that eotaxin is produced and released by lung tissue 4 h after segmental allergen challenge is consistent with findings in animal models (11, 13), and verifies these findings in humans with asthma at a time when eosinophils are first being recruited to the lung. In addition, our findings are in agreement with those reported after inhalational allergen challenge in allergic asthmatics (26).

After allergen challenge, the concentrations of eotaxin in epithelial lining fluid, as estimated from BAL levels, are sufficient to induce chemotaxis in human eosinophils (27, 28). Mediator levels in airway epithelial lining fluid are expected to be higher than levels in BAL fluid, which is a mixture of lining fluid and saline. Although the admixture of lining fluid and saline makes absolute levels difficult to estimate, differences in BAL mediator levels are relevant, and for mediators that are not equally released from all microenvironmental loci, they underestimate the magnitude of local concentration differences. BAL eotaxin levels approximately doubled to 50 pg/ml after segmental allergen challenge. BAL fluid is a dilution of airway lining fluid of approximately 1:100 to 1:200 (29, 30) or more (31); a finding of 50 pg/ml thus reflects a concentration of approximately 1 to 10 ng/ml in airway lining fluid. These are the threshold concentrations for chemotaxis of human eosinophils in vitro (27, 28), thus supporting the concept that allergen exposure can change eotaxin levels from those below the threshold to those sufficient to induce efficient eosinophil chemotaxis. Furthermore, we demonstrated that BAL eotaxin levels were strongly and directly correlated (r² = 0.88) with the recruitment of eosinophils to the lung after segmental allergen challenge. An association of this strength could have occurred if a large portion of BAL eotaxin was from BAL eosinophils, but we demonstrated that the majority of eotaxin immunoreactivity was present in alveolar macrophages rather than eosinophils (Figure 5). This association between increases in BAL eotaxin and the mobilization of eosinophils would also occur if eotaxin were important for attracting eosinophils after allergic exposure or if its release was associated with other mechanisms, such as other chemokines or phospholipid chemoattractants, that cause eosinophil recruitment after allergen challenge. The timing, tissue location, and recovered quantities of eotaxin all support a role for this chemokine in recruiting eosinophils to the airways of subjects with allergic asthma or define eotaxin as a biomarker for this process.

We detected less eotaxin immunoreactivity in unconcentrated lavage fluid of normal subjects than we did in that from subjects with well-controlled mild intermittent asthma. This finding is consistent with published reports that demonstrate more eotaxin immunoreactivity and RNA in the epithelium of subjects with asthma than of normal subjects (16, 17) and is similar to reports of BAL fluid that was concentrated 10-fold (16). This difference in BAL eotaxin levels may be relevant to airway biology because patients with mild intermittent asthma have demonstrable airway eosinophilia even when they are asymptomatic (4). The presence of detectable quantities of eotaxin in the airways of normal subjects without detectable airway or BAL eosinophilia implies either that eotaxin is present in normal subjects in an inactive form, that eotaxin has physiological functions distinct from its effects on eosinophils, or that the concentrations of eotaxin released are below those required for biological activity. Antibody-based detection systems can detect substances other than the substance of interest, and it is possible that our methods allowed us to detect a substance other than eotaxin. This possibility seems unlikely for several reasons. First, the same quantity and pattern of expression are detected with two distinct monoclonal antibodies (27) that are not active against a screening battery of potential ligands (21, 23). Second, immunoreactivity was lost after antibody preabsorption with eotaxin ligand. Third, the sites of immunoreactivity noted in the absence (16), and presence, of allergen challenge of immunoreactive eotaxin correlate with the loci of eotaxin mRNA expression.

We measured BAL eotaxin levels 4 h after allergen challenge because chemotactic ligand levels should precede cell recruitment, which has occurred by 24 h following segmental allergen challenge (20, 32, 33), and because 4 h is the time of peak eotaxin expression in animal models of allergen challenge (14, 34). Our finding that the intracellular presence of eotaxin is strikingly increased after segmental allergen challenge has important mechanistic implications: it demonstrates that, in humans, allergic exposure is associated with the de novo synthesis of protein and implies that augmented BAL eotaxin levels are not the exclusive result of enhanced release. Indeed, we demonstrated that segmental allergen challenge is associated with enhanced expression of mRNA in isolated BAL cells. Our findings confirm in the human system observations in guinea pig (35) and mouse models of asthma (15) and demonstrate that enhanced pulmonary expression of eotaxin mRNA is associated with increased eotaxin protein production after segmental allergen challenge. Whereas allergen challenge increased eotaxin mRNA expression in BAL cells, stimulation with single-agent cytokines did not, and dexamethasone was able to suppress eotaxin production ex vivo, thereby indicating a potential mechanism of action for glucocorticoids in asthma.

We identified significantly greater density of eotaxin immunopositive cells in endobronchial biopsy tissue from antigen-challenged segments than in biopsies from unchallenged control segments in these same subjects or in biopsies from normal subjects. This finding not only confirms our findings in BAL fluid and cells but allows us to specify the locus of antigen-induced eotaxin expression within the airway. We found that eotaxin expression was increased in airway epithelial cells, a finding that supports the concept that eotaxin produced by the airway epithelium creates a gradient that attracts eosinophils from blood to the airways after allergen challenge.

We have demonstrated that segmental allergen challenge induces the rapid epithelial production of quantities of airway and alveolar eotaxin sufficient to account for the recruitment of eosinophils to these sites. Our data strongly support the hypothesis that eotaxin participates in processes that recruit eosinophils into the airways after allergen exposure in sensitive subjects with asthma or is a biomarker tightly associated with events that are responsible for eosinophil recruitment hours after allergen challenge.