Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Eosinophils are mature inflammatory cells present in increased numbers in tissue and blood in a variety of allergic conditions (1). Eosinophils and basophils arise from a common progenitor cell (2), the eosinophil/basophil colony-forming unit (Eo/B CFU), which circulates in increased levels in atopic individuals depending on symptomatology (3). Fluctuations in the levels of Eo/B CFU occur on natural allergen exposure (4) and during asthma exacerbation (5, 6).

Growth factors that promote Eo/B CFU proliferation and differentiation include granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin (IL)-3, and IL-5 (7). Mature eosinophils contain messenger RNA (mRNA) transcripts for GM-CSF, IL-3, and IL-5 (8), and eosinophils in bronchoalveolar lavage from asthmatics express GM-CSF and IL-5 mRNA before and after allergen challenge (11). While expression of these cytokines by eosinophils may play a role in acute and chronic inflammatory responses, it could also provide an autocrine source of growth factors for differentiating eosinophils. Indeed, the microenvironment of nasal polyp tissue, which includes abundant eosinophils capable of expressing GM-CSF (12) and IL-5 (13), as well as other cytokines (14), has been shown to promote differentiation of Eo/B CFU (15). However, it is not known whether autocrine, eosinophil-initiated hemopoietic effects can occur either in vitro or in vivo. Previous experiments from our laboratory have demonstrated an increased number of Eo/B CFU in atopic versus nonatopic individuals (3), and an increased number of circulating Eo/B CFU in atopic asthmatics after allergen challenge versus before challenge (6). We hypothesized that, in addition to tissue-derived cytokines from fibroblasts and other structural cells (16), circulating Eo/B progenitors may be promoted to differentiate by cytokines derived from developing eosinophils themselves. In the current study, we have examined the expression of the eosinophilopoietins IL-5 and GM-CSF by differentiating eosinophils in semisolid cultures in vitro. We also compared the expression of cytokines by differentiating eosinophils in nonatopics and atopics, and in atopic asthmatics both before and after allergen challenge.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Studies were approved by the Ethics Committee of McMaster University Health Sciences Center. An initial set of experiments compared the expression of GM-CSF in peripheral blood Eo/B colony cells of atopic and normal individuals (Study 1), while a second set of experiments examined expression of GM-CSF and its message in a time-course study (Study 2). Finally, we compared the expression of GM-CSF and IL-5 in Eo/B colony cells before and after allergen inhalation in atopic asthmatics (Study 3). All subjects gave informed consent to participate in the studies.

Subjects

Studies 1 and 2. A total of 12 unchallenged atopic and two nonatopic subjects were studied. Atopy was confirmed with a positive skin test for one of 21 common allergens. Atopic subjects were studied out of season and thus not exposed to sensitizing allergens for at least 4 wk, since colony formation is depressed during prolonged allergen exposure (4). Subjects demonstrating a negative skin test served as nonatopic controls. Five of the atopics were used to examine time-dependent expression of GM-CSF and GM-CSF mRNA, and seven of the atopics were compared with normal subjects for expression of GM-CSF in Eo/B colony cells.

Study 3. Twenty-five nonsmoking subjects with mild atopic asthma (13 female/12 male) were selected for the allergen-inhalation experiments because of a previously documented allergen-induced early and late bronchoconstrictor response of at least 15% reduction in the forced expiratory volume in 1 s (FEV₁). Subjects were not exposed to sensitizing allergens and did not have asthma exacerbations or respiratory tract infections for at least 4 wk prior to entering the study. All subjects had stable asthma with FEV₁ greater than 70% of predicted normal on all study days before allergen inhalation, and used no regular medication other than infrequent (< twice weekly) inhaled ₂-agonist as required to treat their symptoms. All medications were withheld for at least 8 h before each visit, and subjects were instructed to refrain from rigorous exercise, tea, or coffee in the morning before visits to the laboratory.

Study Design

Each normal and each atopic subject attended the laboratory once. A skin test was performed and blood was obtained by direct venipuncture for assay of Eo/B CFU.

Each atopic asthmatic attended the laboratory on three occasions. Baseline measurements of FEV₁, the provocative concentration of methacholine causing a 20% fall in FEV₁ (methacholine PC₂₀), blood differentials, and CFU were determined the morning before allergen challenge. Allergen challenges were carried out the following morning, and FEV₁ was measured during the following 7 h. Measurements of FEV₁, methacholine PC₂₀, and blood were repeated 24 h after allergen challenge.

Laboratory Procedures

Methacholine-inhalation test. Methacholine-inhalation challenge was performed as described by Cockcroft (17). Subjects inhaled normal saline, then doubling concentrations of methacholine phosphate from a Wright nebulizer for 2 min. FEV₁ was measured at 30, 90, 180, and 300 s after each inhalation. Spirometry was measured with a Collins water-sealed spirometer and kymograph. The test was terminated when a fall in FEV₁ of 20% of the baseline value occurred, and the methacholine PC₂₀ was calculated.

Allergen-inhalation test. Allergen challenge was performed as described by O'Byrne and colleagues (18). The allergen producing the largest skin-wheal diameter was diluted in normal saline. The concentration of allergen extract for inhalation was determined from a formula described by Cockcroft and associates (19) using the results from the skin test and the methacholine PC₂₀. The starting concentration of allergen extract for inhalation was two doubling concentrations below that predicted to cause a 20% fall in FEV₁. The same doses of allergen were administered during each treatment period, and the FEV₁ was measured at 10, 20, 30, 40, 50, 60, 90, and 120 min after allergen inhalation, then each hour until 7 h after allergen inhalation. The early bronchoconstrictor response was taken to be the largest fall in FEV₁ within 2 h after allergen inhalation, and the late response was taken to be the largest fall in FEV₁ between 3 and 7 h after allergen inhalation.

Differential blood counts. Blood was collected into heparinized tubes by direct venipuncture, and blood smears were made for differential staining (Diff Quik; American Scientific Products, McGaw Park, IL). Differential cell counts were obtained from the mean of two slides with 300 cells counted per slide. Total leukocyte count was determined using a hemocytometer (Neubauer Chamber; Hausser Scientific, Blue Bell, PA), and cell populations were expressed as the number per milliliter of blood by dividing by the total number of cells counted and multiplying by the total leukocyte count.

Methylcellulose assay. Methylcellulose assays for CFU were performed as previously described (4). Mononuclear cells were separated from whole peripheral blood using Percoll density-gradient (Pharmacia, Uppsala, Sweden), then adherent cells were removed by a 2-h incubation at 37°C in plastic flasks. Nonadherent mononuclear cells (NAMC) were cultured in 0.9% methylcellulose (Sigma Chemical Co., St. Louis, MO) at 1 × 10⁶ per 35 × 10 mm tissue culture dish (Falcon Plastics, Oxnard, CA) in Iscove's modified Dulbecco's medium and 20% fetal calf serum (GIBCO, Burlington, ON, Canada) supplemented with 1% penicillin-streptomycin and 5 × 10⁵ mol/liter of 2-mercaptoethanol.

Immunocytochemical staining. Cells collected from all conditions were washed in 0.5 ml Dulbecco's phosphate-buffered saline (DPBS; GIBCO) and resuspended in DPBS at 0.75-1.0 × 10⁶/ml, and cytospins were prepared on aptex-coated glass slides using 50 µl of cell suspension and a Shandon III cytocentrifuge at 300 rpm for 5 min (Shandon Southern Instruments, Sewickly, PA), fixed for 10 min in periodate-lysine-paraformaldehyde, then 10 min in 15% sucrose; slides were stored at 70°C. Cells were stained with mouse monoclonal anti-human GM-CSF antibody (Genzyme, Cambridge, MA) (atopic versus normal subjects) or mouse monoclonal anti-human IL-5 antibody (R&D Systems, Minneapolis, MN) (allergen inhalation). Antibodies were diluted in 1.0% bovine serum albumin (Sigma) and wash buffer made up of DPBS, 0.01 M N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid buffer and 0.01% saponin (Sigma). Labeling of these antibodies was detected by the alkaline-phosphatase antialkaline-phosphatase method (20). Slides were incubated for 60 min with 75% human AB serum (Sigma) and for 30 min with 25% normal rabbit serum (Sigma) to block nonspecific binding of the first and second antibodies, respectively. Slides were incubated overnight at a concentration of 10 µg/ml anti-GM-CSF and 30 µg/ml anti-IL-5. Mouse IgG₁ (Sigma) was used as a negative control. A positive control for GM-CSF immunostaining consisted of peripheral blood monocytes stimulated with lipopolysaccharide: for IL-5, peripheral blood eosinophils stimulated with calcium ionophore. The percentage of cells immunolocalizing GM-CSF and IL-5 was determined from a count of 400 cells under light microscopy, based on a scale from 0 to 5. All cells demonstrating an intensity of stain > 1 were counted as positive, and those < 1 were counted as negative.

In situ reverse transcription-polymerase chain reaction. Cells were picked from colonies and washed once with diethyl pyrocarbonate-treated saline, then resuspended at a concentration of 1.2 × 10⁶/ml. Droplets of this cell suspension (20 µl) were placed at three sites on in situ polymerase chain reaction (PCR) glass slides coated with an aminoalkylsilane (Perkin Elmer, Mississauga, ON, Canada), then fixed for 20 min in 4% PBS paraformaldehyde and stored at 20°C. Two representative slides were randomly chosen from each of the three time points. These cells were permeabilized with 2 µg/ml proteinase K digestion (Sigma) and treated overnight with DNase (Boehringer Mannheim, Laval, PQ, Canada) at 37°C to digest nuclear material. Reverse transcription (RT) by incubation with the 3' downstream primer and reverse transcriptase (Superscript^TM RNase H; GIBCO) for 3 h at 37°C converted the mRNA to complementary DNA, and this product underwent 35 cycles of PCR (GeneAmp 1000 System; Perkin Elmer) using the optimal annealing temperature for human GM-CSF of 60°C, in buffer including the 5' upstream (ATG TGG CTG CAG AGC CTG CTG C) and 3' downstream (CTG GCT CCC AGC AGT CAA AGG G) primers for human GM-CSF (Applied Biosystems, University of Alberta, Edmonton, AB, Canada), and 115 U/ml TAQ polymerase enzyme (GIBCO) and digoxigenin-11 deoxyuridine triphosphate (dUTP) (Boehringer Mannheim). Slides were incubated with anti-digoxigenin (Boehringer Mannheim) and color-developed with a solution of 4-nitro blue tetrazolium chloride (Boehringer Mannheim) and X-phosphate/5-bromo-4-chloro-3-indolyl-phosphate (Boehringer Mannheim). Slides were counterstained 10 min in 1% carbol chromotrope 2R (Sigma), which is specific for eosinophils. The slides were enumerated by examining 500 cells per slide with conventional light microscopy. Positive and negative controls were processed in parallel with the test site, on the remaining two sites on the slide. The positive control was not treated with DNAse or reverse transcriptase, and the negative control was not treated with reverse transcriptase. To eliminate the possibility of false positivity due to high background on the test site, one slide was treated with an RNase solution (Amersham Life Science, Cleveland, OH); another was treated with an irrelevant primer sequence. Both tests showed a high background that remained high even when mRNA was not amplified. Human 2 microglobulin was used as a system control.

Statistical Analysis

All summary statistics are expressed as mean and SEM, except for methacholine PC₂₀, which is expressed as geometric mean and geometric standard error of the mean (GSEM). Methacholine PC₂₀ was measured by linear interpolation of log dose-response curves resulting in logarithmic values for PC₂₀, which were then subjected to statistical analysis. Student's paired t test was used to compare the allergen-induced changes in Eo/B CFU, blood eosinophils, percentage of Eo/B cells immunolocalizing GM-CSF and IL-5, and methacholine airway responsiveness. Repeated-measures analysis of variance was used to compare atopic and normal GM-CSF expression in colony cells grown under three different conditions, and to examine immunoreactive GM-CSF in Eo/B colony cells over time.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

GM-CSF Expression in Colony Cells Increases Temporally

Eo/B colony cells from atopic subjects grown with rhSCF plus rhIL-3 demonstrated immunoreactivity to anti-GM-CSF and anti-IL-5 (Figures 1A and 1C). Although only cells staining homogeneously with intensity > 1 were counted as positive, many of the cells demonstrated a faint positive signal localized next to the nucleus (Figure 1B). The percentage of Eo/B colony cells with immunoreactive GM-CSF after 14 and 18 d of culture was significantly higher than cells grown for 10 d (P < 0.007), being 8.1 ± 2.2%, 21.3 ± 4.6%, and 18.0 ± 4.0% after 10, 14, and 18 d, respectively. In addition, these cells were demonstrated by in situ RT-PCR to contain mRNA for GM-CSF at these times, being 2.7 ± 0.4%, 3.5 ± 0.2%, and 5.5 ± 0.8% after 10, 14, and 18 d, respectively (Figure 2).

View larger version (124K): [in this window] [in a new window]   Figure 1.   Immunoreactivity for GM-CSF (A, B) and IL-5 (C) with isotype control (D) in Eo/B colony cells. Bar = 5.5 µm (A, C, D); bar = 2.8 µm (B).

View larger version (91K): [in this window] [in a new window]   Figure 2.   GM-CSF mRNA expression in Eo/B colony cells. Positive control (A), negative control (B), and test positive (C) for GM-CSF mRNA. Bar = 2.8 µm.

Expression of GM-CSF in Colony Cells Is Greater in Atopic versus Normal Individuals

The number of Eo/B CFU grown with rhGM-CSF in atopics was 36.5 ± 21.5 per 10⁶ NAMC versus 25.5 ± 11.5 per 10⁶ NAMC in normal individuals. When the percentage of GM-CSF-immunoreactive Eo/B colony cells grown under three different conditions was compared, they were significantly increased in atopic subjects versus normal controls: under conditions of stimulation with rhGM-CSF, 53.4 ± 14.9% versus 14.5 ± 1.4%; rhIL-3, 73.2 ± 10.4% versus 24.0 ± 7.5%; and rhIL-5, 56.0 ± 11.3% versus 6.0 ± 1.2%, respectively (P = 0.04) (Figure 3). There was no significant difference in the percentage of GM-CSF-positive colony cells between the three growth conditions (P = 0.26). Cultures from two atopic subjects were excluded from analysis because of technical problems with the assay.

View larger version (9K): [in this window] [in a new window]   Figure 3.   Immunoreactive GM-CSF in Eo/B colony cells from atopics and normal individuals. A comparison of the percentage of GM-CSF-positive Eo/B colony cells after 14 d culture, stimulated with GM-CSF, IL-3, and IL-5 in atopics versus normal controls (P = 0.04).

Allergen Inhalation

Twenty-five subjects completed the allergen-inhalation challenge and developed a dual airway response. The maximal percentage fall in FEV₁ during the early response was 32.4 ± 2.5% and during the late response was a 22.3 ± 2.7% maximal fall from pre-allergen baseline FEV₁ (Figure 4). The subjects also developed allergen-induced airway hyperresponsiveness 24 h following allergen inhalation, with a significant reduction in the methacholine PC₂₀ from 1.60 mg/ ml (1.31 GSEM) to 0.62 mg/ml (1.34 GSEM) (P < 0.001).

View larger version (15K): [in this window] [in a new window]   Figure 4.   Allergen-induced airway responses. Percent change in FEV1 (mean and SEM) in 25 atopic asthmatics after inhalation of diluent (open circles), and after inhalation of allergen (closed circles).

The number of peripheral blood eosinophils increased significantly 24 h after allergen inhalation, from pre-allergen values of 38.0 ± 3.8 × 10⁴/ml to 57.3 ± 5.6 × 10⁴/ml (P = 0.005) (Figure 5). Individual allergen-induced changes in blood eosinophils ranged from a decrease of 28.5% to an increase of 614.6% from pre-allergen baseline, with a mean increase of 50.8%. The number of in vitro Eo/B CFU responsive to GM-CSF stimulation was also significantly elevated 24 h after allergen inhalation, from 10.2 ± 1.5 per 10⁶ NAMC to 13.9 ± 1.8 per 10⁶ NAMC (P = 0.02) (Figure 5). Individual changes in blood Eo/B CFU ranged from a 58.3% decrease to an increase of 433.3%, with an average increase of 36.3%. There was, however, no correlation between the individual changes in blood eosinophils and the individual changes in blood Eo/B CFU (r = 0.07, P = 0.76).

View larger version (10K): [in this window] [in a new window]   Figure 5.   Peripheral blood eosinophils and Eo/B CFU before and after allergen-inhalation challenge. Circulating eosinophils (left panel ) and Eo/B CFU (right panel ) before allergen (open bars) and 24 h after allergen (solid bars). There was a significant increase in the number of eosinophils (P = 0.0001) and Eo/B CFU (P = 0.02) after allergen inhalation.

Colony cells grown in vitro under conditions of SCF and IL-3 stimulation contained immunoreactive GM-CSF and IL-5. The percent of colony cells expressing GM-CSF increased significantly from baseline values of 15.4 ± 1.3% before allergen to 29.9 ± 2.9% at 24 h after allergen inhalation (P < 0.001). There was a trend for IL-5 expression by colony cells to increase after allergen from 3.9 ± 0.8% to 5.7 ± 0.8%; however, this was not statistically significant (P = 0.12) (Figure 6).

View larger version (13K): [in this window] [in a new window]   Figure 6.   Immunoreactivity for GM-CSF and IL-5 in Eo/B colony cells before and after allergen-inhalation challenge. The percentage of GM-CSF-positive (left panel ) and IL-5-positive (right panel ) Eo/B colony cells before and 24 h after allergen. There was a significant increase in the percentage of Eo/B colony cells expressing GM-CSF (P = 0.0009), but not IL-5 (P = 0.12).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Eosinophils and their progenitors are present at elevated levels in atopic individuals, and can be further induced in atopic asthmatics by allergen-inhalation challenge. However, very little is understood about the mechanisms underlying this eosinophilia. There is evidence supporting the role of the tissue microenvironment in the accumulation of inflammatory cells at the site of inflammation (15). This is likely to be one mechanism contributing to elevation of eosinophils and their progenitors in the circulation of atopic individuals. This study examined the role of the constitutive expression of growth factors in differentiating eosinophils, which may represent a separate, autocrine mechanism enhancing the process of eosinophil differentiation shown to be upregulated in atopic individuals.

We have demonstrated by combined in situ RT-PCR and cytochemistry (particularly, carbol chromotrope 2R+ cells, i.e., eosinophilic) that Eo/B colony cells grown in culture express mRNA for GM-CSF. Carbol chromotrope is known to stain the granules of eosinophils that are cationic in nature (1). We have also demonstrated by immunocytochemistry that a significantly higher percentage of maturing, but not yet fully mature, progeny of eosinophil progenitors (Eo/B CFU) from the blood of atopic individuals express GM-CSF, when compared with controls. For technical reasons, the baseline colony growth and expression of GM-CSF in Eo/B colony cells was higher in these experiments (Study 1) than in the subsequent experiments (Studies 2 and 3). The latter demonstrated increased expression of GM-CSF in cells that have undergone further differentiation and a higher percentage of GM-CSF-positive colony cells 24 h after allergen inhalation in atopic asthmatics when compared with cells grown in culture before the allergen challenge. The higher GM-CSF expression observed in atopics and allergen-challenged asthmatics suggests that the atopic state per se, and/or allergen provocation, can directly or indirectly induce changes in Eo/B CFU or other nonadherent mononuclear cells, which confer an upregulated cytokine profile upon Eo/B CFU progeny in vivo. One implication of this activation state is that it may initiate another autocrine mechanism of Eo/B differentiation, further increasing the number of progenitors committed to this lineage. Similarly, atopy and/or allergen provocation may provide a "priming" effect for these cells such that they mature more quickly and develop the cytokine profile of a more mature cell earlier than nonatopic/unchallenged individuals.

Immunocytochemistry is a sensitive method to demonstrate the presence of intracellular protein, but it does not distinguish between protein that has been synthesized by the cell and extracellular protein that may be receptorbound or bound to the interior of the cell. It is possible that GM-CSF may have been detected because of the binding of exogenous cytokine to receptors on maturing colony cells. We did, however, find differences in colony-cell immunostaining for GM-CSF even when other cytokines (SCF and IL-3) were used to stimulate the culture. This suggested that exogenous cytokine does not account for the immunostaining. However, in these cultures, which also contained GM colonies, it was possible that non-Eo/B cells generated GM-CSF, which then bound to GM-CSF receptors on the Eo/B colony cells. The staining pattern of GM-CSF and IL-5 expression, however, was typical of that observed for intracellular protein, being homogeneous throughout the cell rather than in a halo distribution typical of cells that have protein bound on their surface. In addition, gene transcription of GM-CSF by Eo/B colony cells was demonstrated with the presence of specific mRNA, confirming that these cells were capable of synthesizing this cytokine. Furthermore, the localized immunoreactivity adjacent to the nucleus may support the view that synthesis and storage of GM-CSF can be associated with intracellular organelles of the cell.

Mature eosinophils have the ability to synthesize IL-5 (8), IL-3 (9), and GM-CSF (10). This study demonstrates that immature and nascent eosinophils synthesize at least one of these growth factors. Increased constitutive expression of cytokines in maturing Eo/B CFU from atopic individuals may represent an autocrine mechanism for enhanced proliferation, differentiation, and activation of eosinophils in allergic responses. In support of this, there is also evidence for GM-CSF as an autocrine differentiating factor in an eosinophilic leukemia cell line, EoL-1 cells (21). GM-CSF is not only an important growth factor for eosinophils, but it also plays a role in cell viability by prolonging eosinophil survival in vitro (22). IL-5 is involved in terminal differentiation, stimulates function, and prolongs survival of the eosinophil (23, 24), whereas IL-3 prolongs survival and enhances functional properties of the eosinophil (25). IL-5 and GM-CSF are upregulated in diseases associated with blood and tissue eosinophilia (26), supporting the hypothesis that cytokine inhibition or delay of apoptosis could be a mechanism for the development of blood and tissue eosinophilia in diseases such as asthma (27, 28). Tyrosine phosphorylation of cytokine receptors regulates the activation and inhibition of apoptosis in human eosinophils (29). Activation of IL-5 and GM-CSF receptors by common tyrosine kinases through the  receptor subunit may be essential for anti-apoptotic effects of IL-5 and GM-CSF (30). In our experiments, we did not compare expression of IL-5 in colony cells between atopic and normal individuals, but expression of IL-5 in eosinophil colony cells of atopic asthmatics was increased in some individuals after allergen challenge. The level of expression of IL-5 compared with GM-CSF in the colony cells from atopic asthmatics was considerably lower. This may reflect the relative immaturity of the colony cells, since IL-5 has shown to be expressed only by a limited number of fully differentiated cells (31).

Although immediate progeny (i.e., colony cells) of Eo/B progenitors derived from atopic individuals constitutively express higher levels of GM-CSF, the hemopoietic inductive microenvironment (HIM) itself may play a significant role in the development of eosinophilia during allergic inflammatory reactions. There are several mechanisms by which bone-marrow stroma can provide signals to control the process of hemopoiesis (32). Signals may be directed by cell-cell interactions, secreted soluble bioactive factors, and cell-matrix interactions. Atopic individuals may differ from normals in the set of signals provided for progenitors in the bone marrow. Thus, eosinophilic inflammation in atopic asthmatics may represent the effects of a cascade of cytokines, including growth and differentiation factors derived from resident inflammatory cells (33); airway structural cells such as fibroblasts and endothelial and epithelial cells (15, 34); and an "activated" HIM. Furthermore, subcutaneous injection of GM-CSF has been associated with increases in the numbers of circulating, colony-forming progenitors in human peripheral blood (38). This suggests that elevations of circulating hemopoietic cytokines in chronic tissue inflammation can potentially mobilize specific hemopoietic progenitor cells. Indeed, specific responsiveness to cytokines of progenitor cells can be induced in atopic individuals and allergic asthmatics through a process involving IL-5 receptor modulation on bone marrow and blood progenitors (39, 40). We are currently exploring the mechanism for this. Similarly, allergen challenge in airway hyperresponsive dogs elicits serum hemopoietic activity that can upregulate bone-marrow myeloid progenitors (41).

Previous work from this laboratory has demonstrated the importance of the microenvironment on the development of progenitor cells. The current study adds another dimension to the regulation of progenitor differentiation, and proliferation, suggesting that in atopic individuals, developing eosinophils may be altered by signals released after allergen inhalation to ultimately produce more hemopoietic cytokines. These cytokines in "activated" progenitors of eosinophils could contribute to the mobilization, differentiation, and activation of both maturing and mature eosinophils. Further experiments are required to determine how much of this cytokine expression by maturing eosinophils is secondary to signals from inflamed tissue and how much it reflects a primary (constitutive) upregulation of programs for cytokine expression in developing eosinophils in atopic and normal individuals.