Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

The eosinophil is recognized as a proinflammatory granulocyte implicated in protection against parasitic infection. Eosinophils also play an important role in various allergic diseases, such as bronchial asthma, allergic rhinitis, and atopic dermatitis (reviewed in 1). Major mediators of eosinophil effector function are a series of granule-derived proteins that are released when infiltrating eosinophils are exposed to appropriate stimuli (1). These proteins are potent cytotoxins for various types of cells in vitro and cause pathophysiologic changes in vivo in animals (1). Numerous studies have shown marked eosinophil infiltration and release of eosinophil-derived granule proteins in tissues from patients with allergic diseases (1). These studies have provoked considerable interest in the intracellular signaling events that regulate eosinophil activation.

Accumulating evidence points to important roles of integrin adhesion molecules, such as ₄₁ (VLA-4, CD49d/ CD29) and _M2 (Mac-1, CD11b/CD18), in eosinophil recruitment into the sites of inflammation in allergic diseases. In addition, recent evidence suggests these integrins, particularly ₂ integrin, are capable of modulating the eosinophil degranulation response stimulated with various physiologic secretagogues. For example, eosinophil degranulation induced by IgG immobilized onto tissue culture plates or Sepharose beads was inhibited by anti-CD18 monoclonal antibodies (mAb), an antibody (Ab) against ₂ integrin (2). Cellular adhesion mediated by _M2 influenced the eosinophils' degranulation response triggered by a lipid mediator, platelet-activating factor (PAF), and by a cytokine, granulocyte-macrophage colony-stimulating factor (GM-CSF) (3). The ability of these secretagogues to induce eosinophil degranulation was minimal when cells were kept in suspension (3). Furthermore, extracellular matrix proteins, such as laminin and fibronectin, which bind to eosinophils in a ₁ integrin-dependent fashion, interfered with ₂ integrin-mediated adhesion and concomitantly inhibited eosinophil degranulation provoked by PAF, the complement fragment C5a, or interleukin (IL)-5 (4).

Recent studies on hematopoietic cells and interstitial cells suggest that, in addition to their passive functions as anchors of cell-cell or cell-matrix interactions, adhesion molecules play active roles in cellular regulation. For example, cellular adhesion leads to activation of intracellular signaling events, such as elevations of intracellular pH and Ca²⁺ levels and activation of protein kinases (reviewed in 5). More recently, localization of the protein tyrosine kinase (PTK) focal adhesion kinase (FAK) to the cytoplasmic face of focal adhesion plaques in fibroblasts (6) suggests that a PTK-dependent signaling pathway is initiated by ligand-dependent integrin stimulation. The presence of tyrosine phosphorylated FAK in platelets (7) and basophilic leukemia cells (8) suggests that this PTK may also be involved in signal output from integrins in these hematopoietic cell types. In addition, proteins other than FAK, such as paxillin (9) and an unidentified 115 kD protein (10), were also tyrosine phosphorylated in adherent neutrophils and lymphocytes, respectively. These observations, together with the strong dependency of eosinophil degranulation on the ₂ integrin, lead us to hypothesize that ₂ integrin may participate actively in the intracellular signaling mechanisms and functional activation of eosinophils. Here, we tested this hypothesis by inhibiting adhesion of IL-5-stimulated eosinophils with anti-CD18 mAb and by directly engaging the cell surface ₂ integrin, _M2, with specific mAb.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Reagents

Mouse antiphosphotyrosine mAb (4G10) was purchased from Upstate Biotechnology (Lake Placid, NY). Mouse monoclonal anti-FAK was purchased from Transduction Laboratories (Lexington, KY). Rabbit polyclonal anti-c-Cbl was purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Peroxidase-conjugated rabbit antimouse IgG Ab was purchased from DAKO (Carpinteria, CA). Recombinant human IL-5 was a kind gift from Dr. S. Narula (Schering-Plough Corporation, Kenilworth, NJ). Stock solutions of IL-5 (100 µg/ml) were dissolved in RPMI containing 0.1% bovine serum albumin (BSA), and aliquots were stored at 20°C. Genistein was obtained from Calbiochem (Los Angeles, CA), and stock solutions of genistein (50 mg/ml) were made in dimethylsulfoxide (DMSO) and stored at 20°C. The solutions were diluted in reaction medium immediately prior to use. Preliminary studies demonstrated that concentrations of DMSO used in the experiments (< 0.1%) did not affect eosinophil viability or stimulus-dependent eosinophil-derived neurotoxin (EDN) release. Mouse mAb used to block cell adhesion were anti-CD11a (clone G25.2, IgG2a [11]), anti-CD11b (clone D12, IgG2a [12]), and anti-CD18 (clone L130, IgG1 [13]) (Becton Dickinson, San Jose, CA). These Ab possess blocking activities for cell adhesion (11). Anti-CD11b (clone bear 1, mouse IgG1 [14]), which does not inhibit cell adhesion, was purchased from Immunotech, Inc. (Westbrook, ME). Purified mouse anti-HLA class I antigen (HLA class I; mouse IgG1) was purchased from Pharmingen (San Diego, CA). Isotype-matched Ig (mouse IgG1 and mouse IgG2a) was purchased from Cappel-Organon Teknika Co. (Durham, NC). Human serum albumin (HSA; Catalog no. A3782) was purchased from Sigma Chemical Co. (St. Louis, MO). HSA was globulin-free and its purity was certified by the manufacturer.

Preparation of CD11b-coated Polystyrene Beads

Anti-CD11b mAb coupled to polystyrene microbeads (diameter, 1.91 µm; Polyscience, Inc., Warrington, PA) was used as a stimulus for eosinophil activation. Beads were coated according to the manufacturer's recommendations with minor modifications. Briefly, 0.5 ml of a 2.5% suspension of the beads was pelleted in an Eppendorf microcentrifuge tube. After the beads were washed with 0.1 M borate buffer (pH 8.6), 200 µg of the protein, such as anti-CD11b mAb (clone bear 1), anti-HLA class I mAb, or mouse IgG1, was added and incubated overnight at 4°C with gentle mixing. After centrifugation for 5 min, the supernatants were saved for determination of residual unbound protein. Approximately 60% of added protein was routinely immobilized onto the beads; no differences were found in the binding efficiencies of the specific proteins tested in this study. The Ab-coupled beads were mixed with 0.5 ml of 2.5% HSA in borate buffer, pH 8.6, and incubated for 30 min at 4°C to block nonspecific protein-binding sites. After a single wash in borate buffer, the beads were resuspended in 0.5 ml of storage buffer (phosphate-buffered saline [PBS] containing 10 mg/ml BSA, 0.1% NaN, and 5% glycerol, pH 7.4) and stored at 4°C.

Eosinophil Isolation

Eosinophil purification was performed with minor modifications of a procedure described previously (15). All procedures were carried out at 4°C or on ice to prevent cell activation. Briefly, heparinized blood was obtained from normal volunteers (eosinophil counts < 5% of total leukocytes) and overlayered on an isotonic Percoll solution (density 1.082 g/ml; Sigma Chemical Co.). After centrifugation at 1,000 × g for 30 min, the supernatant and the mononuclear cells at the interface were carefully removed. Erythrocytes in the sediments were lysed by two cycles of hypotonic water lysis. Isolated granulocytes were washed with piperazine-N,N(-bis-(2-ethane sulfonic acid) (PIPES) buffer (25 mM PIPES, 50 mM NaCl, 5 mM KCl, 25 mM NaOH, 5.4 mM glucose, pH 7.4) with 1% defined calf serum (DCS) (HyClone Laboratories, Logan, UT) and mixed with an approximately equal volume of anti-CD16 mAb conjugated with magnetic particles (Miltenyi Biotec Inc., Auburn, CA). After 60 min on ice, CD16-positive cells (neutrophils) were removed by a magnetic cell separation system (Becton Dickinson) (15). The purity of eosinophils counted by Randolph's stain was always more than 98%, and the contaminating cells were neutrophils. Purified eosinophils were washed in reaction medium consisting of RPMI 1640 medium supplemented with 25 mM N-(2-hydroxyethyl)piperazine-N'-2-ethanesulfonic acid (Hepes; pH 7.4) before use. This reaction medium was used for most of the experiments described in this manuscript, unless otherwise indicated.

Eosinophil Degranulation

Eosinophil degranulation was induced with mouse mAb coupled to polystyrene microbeads or with IL-5. Ninety-six-well flat-bottom tissue culture plates (Costar 3596; Cambridge, MA) were blocked with 50 µl of 2.5% HSA dissolved in PBS (pH 7.4) at 37°C for 2 h. After incubation, wells were washed three times with PBS before use. Eosinophils were washed with reaction medium and resuspended in the same medium at 5 × 10⁵ cells/ml. Aliquots (100 µl) of cell suspension were added to the wells, and cellular degranulation was initiated by addition of 25 ng/ ml of IL-5, polystyrene microbeads coated with mAb, or medium alone. Experiments with adhesion-blocking mAb, such as anti-CD18 mAb, were performed by preincubating the cells with the mAb for 15 min at room temperature before incubation with stimuli. After an incubation for 4 h at 37°C in a 5% CO₂ atmosphere, the cells were pelleted by centrifugation; cell-free supernatants were collected from each well and were kept at 20°C until assayed for the release of EDN (see below). All experiments were performed in duplicate.

To quantitate eosinophil degranulation, the quantities of an eosinophil granule protein, EDN, in the sample supernatants were measured by double-Ab competition radioimmunoassay (RIA) using radioiodinated EDN, rabbit anti-EDN Ab, and burro antirabbit IgG, as previously reported (16). Total cellular EDN contents were measured in parallel samples from cells lysed with 0.5% NP-40 detergent. All assays were conducted in duplicate.

Adhesion Assay

The numbers of adherent eosinophils were determined by measuring the contents of EDN in adherent cells by RIA, as previously reported (3). Briefly, after preincubation with blocking mAb, including anti-CD11a (clone G25.2), anti-CD11b (clone D12), and anti-CD18 (clone L130), or isotype-matched mouse Ig, cells were added to 96-well flat-bottom tissue culture plates, and stimulated for 1.5 h with IL-5 (25 ng/ml). After stimulation, the supernatant fluids were collected for assay of EDN and the plates were gently rinsed with warm PBS to remove nonadherent cells. Adherent cells were then lysed with 0.5% NP-40, and the EDN contents in the lysates were measured by RIA. Percent adhesion was calculated as a ratio of EDN content in adherent eosinophils to total available EDN after incubation according to the following equation:

(1)

At the endpoint of this adhesion study (1.5 h after stimulation), degranulation of IL-5-stimulated eosinophils was less than 3% of total granule contents. The results, using this method to determine eosinophil adhesion, showed strong linear correlation (r = 0.987) with those of the adhesion assay using ⁵¹Cr-labeled cells, as previously reported (3).

Tyrosine Phosphorylation

For analyses of tyrosine-phosphorylated proteins following engagement of _M2, the cells were stimulated with mAb coupled to polystyrene microbeads in Eppendorf microcentrifuge tubes, or by IL-5 in tissue culture plates blocked with HSA. To reduce nonspecific binding of proteins, the microcentrifuge tubes were blocked with 2.5% HSA for 2 h and then washed twice with PBS prior to use in cell stimulation experiments. Aliquots of eosinophils (1 × 10⁶ cells/sample) were suspended in 50 µl of reaction medium, and stimulated with 50 µl of microbeads suspended in reaction medium for the times indicated. After incubation, cells were centrifuged quickly (11,000 × g, 15 s), and cell pellets were lysed with sodium dodecyl sulfate (SDS) sample buffer (62.5 mM Tris-HCl, pH 6.8, 2.3% SDS, 10% glycerol, 5% 2-ME, 0.001% bromphenol blue). Cell lysates were boiled for 15 min before electrophoresis. When the effect of genistein was studied, cells were preincubated for 1 h in the presence or absence of various concentrations of genistein before the addition of microbeads.

To stimulate eosinophils with IL-5 in tissue culture plates, 48-well flat-bottom tissue culture plates (Costar 3548) were blocked with 200 µl per well of 2.5% HSA at 37°C for 2 h, and then washed three times with PBS. After preincubation with 10 µg/ml of blocking mAb (e.g., anti-CD18 mAb) for 15 min at room temperature, eosinophils (2 × 10⁶ cells/sample) were suspended in 100 µl of reaction medium, added to the HSA-coated plates, and stimulated by addition of 50 µl IL-5 (25 ng/ml final concentration) for the times indicated. After the incubation, cells were lysed with 50 µl ice-cold 4× SDS sample buffer to each well and by mixing well with a cell scraper (Becton Dickinson Labware, Lincoln Park, NJ) on ice. Cell lysates were transferred to Eppendorf microcentrifuge tubes and boiled for 15 min.

For immunoblotting, the samples were loaded onto an 8.75% SDS-polyacrylamide gel electrophoresis (PAGE). After electrophoresis, the separated proteins were transferred to an Immobilon P membrane (Millipore Corp., Bedford, MA) in 25 mM Tris, 192 mM glycine for 60 min at 100 V. The membranes were incubated overnight at room temperature in blocking solution (Tris-buffered saline [TBS], pH 7.2, 0.5% Tween 20, and 2% BSA) and incubated for 2 h with 0.1 µg/ml of the antiphosphotyrosine mAb (4G10) in fresh blocking solution without BSA. The membranes were washed three times with TBS and 0.2% Tween 20, and incubated for 1 h with 0.5 µg/ml of peroxidase-conjugated rabbit antimouse IgG Ab in TBS, 0.5% Tween 20, 2% BSA. The membranes were washed four times and immunoreactive proteins were detected with the enhanced chemiluminescence system (Amersham Life Science Inc., Arlington Heights, IL) (11). In some experiments, the films of immunoblot were scanned by ScanJet 4c Scanner (Hewlett Packard, Palo Alto, CA), and analyzed by Macintosh computer using NIH Image software (version 1.61) with macros for densitometric analysis of 1-D gels.

Immunoprecipitation

To identify tyrosine-phosphorylated proteins, cell lysates were immunoprecipitated with specific antibodies and the immunoprecipitated proteins were separated by SDS-PAGE and analyzed by immunoblotting with specific antibodies. Briefly, eosinophils (6-10 × 10⁶ cells/sample) were stimulated in Eppendorf microcentrifuge tubes or 48-well tissue culture plates, as described above. Subsequently, the cells were extracted with lysis buffer containing 20 mM Tris-HCl, pH 7.4, 30 mM Na₄P₂O₇, 50 mM NaF, 40 mM NaCl, 5 mM ethylenediamine tetraacetic acid (EDTA), 1% NP-40, 10 µg/ml leupeptin, 5 µg/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride (PMSF), 2 mM Na₃VO₄, and 0.5% deoxycholic acid for 10 min on ice. The cell lysates were centrifuged at 3,000 × g for 10 min at 4°C. Supernatants were collected and mixed with the indicated amounts of rabbit anti-c-Cbl polyclonal antibody for 1 h at 4°C, followed by the addition of 25 µl of packed protein A-Sepharose beads for an additional 1 h. The samples were then washed four times with lysis buffer and the pellets were used for electrophoresis and immunoblot, as described above.

Inositol Phosphate Production

As described earlier (17), eosinophils (4-6 × 10⁶ cells/ml) were cultured overnight in labeling medium (inositol-free RPMI 1640 medium supplemented with 10% DCS and 10 mM Hepes) containing 40 µCi/ml myo-[2-³H]inositol. The labeled cells were washed and suspended at 2-3 × 10⁶ cells/ml in RPMI 1640 containing 15 mM LiCl, 0.1% HSA and 25 mM Hepes, and 100 µl/sample of cell suspension was used. The procedures for cell stimulation are identical to those described above for tyrosine phosphorylation, except that a 30-min stimulation time was used. When cells were stimulated in microcentrifuge tubes, reactions were terminated with 0.6 ml chloroform/methanol/HCl (1:2:0.02). When cells were stimulated in protein-coated wells, the reactions were terminated with 400 µl of ice-cold methanol and 40 µl of 0.22 M HCl, and the lysates were transferred to polypropylene tubes with the addition of 200 µl of ice-cold chloroform. In either case, after 30 min on ice, water-soluble inositol phosphate (IP) was extracted by addition of 0.2 ml of chloroform and 0.2 ml of 1 M NaCl. The aqueous phases were removed and diluted with 10 ml of aqueous 30 mM sodium formate-2.5 mM sodium tetraborate. Radiolabeled IP was isolated by anion exchange chromatography, as previously described (18). Total IP (inositol mono, bis, and triphosphates) was eluted from the anion exchange columns with 2 ml of 1.2 M ammonium formate- 0.1 M formic acid. ³H-labeled IP were assayed by mixing the column eluates with 20 ml of scintillation cocktail (Ultima Gold; Packard Instrument Company, Downers Grove, IL), followed by liquid scintillation counting.

Statistical Analysis

Statistical significance of the difference between various treatment groups was assessed by paired Student's t-test.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Tyrosine Phosphorylation of Proteins in IL-5-stimulated Adherent Eosinophils

To investigate whether adhesion molecules, specifically ₂ integrin, participate in eosinophil signal transduction, we tested whether cellular adhesion modulates tyrosine phosphorylation of intracellular proteins. As shown in Table 1, IL-5 induced adhesion of eosinophils to the tissue culture plates coated with albumin. This IL-5-dependent eosinophil adhesion was significantly inhibited by anti-CD11b mAb (P < 0.05), but not by anti-CD11a mAb or by isotype-matched controls. Anti-CD18 mAb completely abolished eosinophil adhesion, suggesting that IL-5-stimulated eosinophils adhere to albumin-coated plates through ₂ integrins.

When tyrosine phosphorylation of proteins was examined by immunoblot, several faint bands were observed between 50 and 73 kD in unstimulated, nonadherent eosinophils as shown in Figure 1A (lane 1). When eosinophils were stimulated with IL-5 and allowed to adhere to albumin-coated plates for 30 min (lane 2), several additional tyrosine-phosphorylated proteins, including those at 100, 105, 115, and 125 kD, were detected, consistent with previous observations (19). The intensities of the 50-73-kD proteins were also markedly increased. When eosinophils were stimulated with IL-5 but adhesion was inhibited by pretreatment with anti-CD18 mAb (lane 4), the intensities of the 105- and 115-kD bands decreased. In contrast, the intensity of the 100-kD band was not affected by anti-CD18 mAb pretreatment. Isotype-matched control Ig (mouse IgG1) showed minimal effects on protein tyrosine phosphorylation of IL-5-stimulated eosinophils (lane 3), consistent with the failure of this antibody to inhibit eosinophil adhesion (Table 1).

View larger version (59K): [in this window] [in a new window]   View larger version (42K): [in this window] [in a new window]   Figure 1.   Tyrosine phosphorylation of proteins in IL-5-stimulated, adherent eosinophils and the effects of anti-CD18 mAb on tyrosine phosphorylation of these proteins. Panel A: Antiphosphotyrosine immunoblot. Eosinophils (1 × 106 cells/sample) were preincubated with 10 µg/ml of anti-CD18 mAb (lane 4), control IgG1 (lane 3), or medium alone (lanes 1 and 2) for 30 min on ice. Cells were then transferred to HSA-coated plates and stimulated with (lanes 2, 3, and 4) or without (lane 1) 25 ng/ml IL-5 for 15 min at 37°C. Cells were lysed and tyrosine-phosphorylated proteins were analyzed by immunoblot using antiphosphotyrosine mAb (4G10). From bottom to top, the arrows indicate 100-, 105-, 115-, and 125-kD proteins. The figure shows representative results from four independent experiments with similar results. Panel B: Densitometric analyses of the effects of anti-CD18 mAb on the tyrosine phosphorylation of proteins. The immunoblot films were scanned and the densitometric analysis of bands was performed. The results were normalized to the density of bands from the cells stimulated with IL-5 in the absence of mAb as 100%. The data are shown as mean ± SEM from four independent experiments. Asterisk denotes significant (P < 0.05) differences compared with the cells stimulated with IL-5 in the absence of mAb.

Figure 1B shows the summary of the densitometric analyses of four experiments. When cells were stimulated with IL-5 and adhered to the plates, the densities of all of the bands at 100, 105, 115, and 125 kD increased compared with those in unstimulated eosinophils. The changes in the 115-kD band were most pronounced, resulting in a 12-fold increase in density. When cells were stimulated with IL-5 but cellular adhesion was inhibited by anti-CD18 mAb pretreatment, there was no increase in density compared with the control without IL-5. The density of the 105-kD band was also significantly reduced (P < 0.05, n = 4). In contrast, the densities of the 125- and 100-kD bands were not affected by anti-CD18 mAb. Kinetic studies showed that tyrosine phosphorylation of the 115-kD protein was detected 5 min after stimulation of the cells with IL-5, peaked at 15 min, and decreased slightly at 30 min (data not shown). When cells were stimulated with IL-5 in the presence of anti-CD18 mAb, the phosphorylation of the 115-kD band was virtually abrogated at all time points (data not shown). These findings suggest that tyrosine phosphorylation of several intracellular proteins, especially the 115-kD protein, is strongly dependent on ₂ integrin-mediated adhesion of the cells.

Engagement of _M2 Induces Protein Tyrosine Phosphorylation in Human Eosinophils

The results in Table 1 suggest that _M2 (Mac-1, CD11b/ CD18) is involved in adhesion of IL-5-stimulated eosinophils to albumin-coated plates. Therefore, we investigated whether direct engagement of _M2 induces tyrosine phosphorylation of proteins in eosinophils. First, we studied tyrosine phosphorylation of proteins in eosinophils following ligation of _M2 by anti-CD11b mAb, an antibody against the -chain of _M2. As shown in Figure 2 (lane 1), a number of faint bands, including protein bands at 100, 105, 115, and 125 kD and multiple protein bands between 60 and 68 kD, were detected in unstimulated eosinophils. When eosinophils were stimulated with anti-CD11b mAb (clone bear 1) coupled to polystyrene microbeads for 30 s, the intensities of the bands at 100, 105, and 115 kD were increased. The elevated levels of phosphorylation in these proteins were maintained for at least 30 min. The intensity of the immunoreactive band migrating at 125 kD was elevated slightly after 30 s, and increased even more for stimulation times up to 30 min. In contrast, when eosinophils were challenged with microbeads coated with anti-HLA class I mAb, an isotype-matched control mAb, the 115-kD phosphoprotein did not appear at any time point. Furthermore, the intensities of the protein bands at 100, 105, and 125 kD were increased at 30 s, but the intensities of these bands decreased in a time-dependent manner. The densitometric analyses of four experiments showed that stimulation of eosinophils with anti-CD11b mAb microbeads for 15 min induced a 575 ± 162% increase in the density of the 115-kD band compared with the cells incubated with medium alone for the same period (mean ± SEM, n = 4). The increases in the densities of the 100-, 115-, and 125-kD bands were modest (69 ± 36, 81 ± 39, and 50 ± 42% increases, respectively, mean ± SEM, n = 4). In contrast, microbeads coated with anti-HLA class I mAb did not show a significant effect on the densities of any of these four bands (less than 34% increase, mean of four experiments). Furthermore, microbeads coated with mouse IgG1 showed no increase in the density of the 115-kD band. Thus, as revealed by whole-cell antiphosphotyrosine immunoblot, several eosinophil intracellular proteins were tyrosine phosphorylated after ligation of cell surface _M2. In particular, the tyrosine phosphorylation of the 115-kD protein was pronounced, consistent with the results of IL-5-stimulated, adherent eosinophils.

View larger version (112K): [in this window] [in a new window]   Figure 2.   Tyrosine phosphorylation of proteins in eosinophils stimulated with anti-CD11b mAb coupled to polystyrene microbeads. Eosinophils (1 × 106 cells/sample) were suspended in Eppendorf tubes and stimulated by medium alone or microbeads coated with anti-CD11b mAb (clone bear 1) (lanes 2-6) or with anti-HLA class I mAb (isotype-matched control mAb, lanes 7- 11) for times indicated, at 37°C. After stimulation, supernatants were removed and cells were lysed and analyzed by immunoblot using antiphosphotyrosine mAb (4G10). From bottom to top, the arrows indicate 100-, 105-, 115-, and 125-kD proteins. The figure shows representative results from four independent experiments with similar results.

Second, to examine whether the tyrosine phosphorylation of the proteins requires the activation of tyrosine kinase, we studied the effects of a tyrosine kinase inhibitor, genistein. Eosinophils were pretreated for 1 h with up to 50 µg/ml genistein, and then stimulated with anti-CD11b mAb microbeads for 15 min. Pretreatment with the drug did not affect eosinophil viability, as determined by trypan blue dye exclusion and by the eosinophil degranulation response stimulated by phorbol myristate acetate (PMA) (20). As shown in Figures 3A and 3B, the 100-, 105-, 115-, and 125-kD proteins were tyrosine phosphorylated in eosinophils stimulated for 15 min by anti-CD11b microbeads in the absence of genistein, consistent with the observations in Figure 2. Pretreatment of eosinophils with genistein inhibited tyrosine phosphorylation of the 115-kD protein in a concentration-dependent manner; the 115-kD band was barely detectable with 50 µg/ml genistein. In contrast, the intensities of the bands of the 100- and 125-kD proteins were less affected by genistein treatment, and the intensity of the band of the 105-kD protein was not at all affected by genistein. Densitometric analysis of the gel (Figure 3C) showed that about 5 µg/ml genistein produced the 50% inhibition (IC₅₀) for the tyrosine phosphorylation of the 115-kD protein, similar to IC₅₀ values of genistein as an inhibitor for the Src protein tyrosine kinase (6 to 7 µg/ml) (21). These findings suggest that tyrosine phosphorylation of the 115-kD protein in eosinophils stimulated with anti-CD11b mAb is closely associated with the activation of a genistein-sensitive tyrosine kinase.

View larger version (60K): [in this window] [in a new window]   View larger version (67K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   Figure 3.   Effects of genistein on tyrosine phosphorylation of proteins in eosinophils stimulated by anti-CD11b mAb coupled to polystyrene microbeads. Panel A: Eosinophils (1 × 106 cells/sample) were suspended in Eppendorf tubes and preincubated with (lanes 2-4) or without (lane 1) the indicated concentrations of genistein for 1 h. After incubation, cells were stimulated by microbeads coated with anti-CD11b mAb for 10 min at 37°C. Cell lysates were prepared and protein tyrosine phosphorylation was examined by immunoblot analysis with antiphosphotyrosine mAb (4G10). From bottom to top, arrows to the left of the gel indicate 100-, 105-, 115-, and 125-kD proteins. Panel B shows a high magnification of the 100-, 105-, 115-, and 125-kD bands in panel A. Panel C shows densitometric analyses of these four bands. The results were normalized to the density of bands from the cells without genistein as 100%. The figure is representative of three independent experiments showing similar results.

Identification of Specific Tyrosine-phosphorylated Proteins

To investigate key molecule(s) involved in adhesion-dependent signal transduction of eosinophils, we performed a series of experiments to identify tyrosine-phosphorylated protein(s) in adherent eosinophils. Initially, we attempted to characterize the 115-kD protein; however, to date, we have been unable to do so in spite of several trials of immunoprecipitation and immunoblot (data not shown). The presence of tyrosine-phosphorylated 125-kD FAK in platelets (7) and basophilic leukemia cells (8) suggests that FAK may also be involved in the signal output from integrins in these hematopoietic cell types. Therefore, we next examined whether FAK is tyrosine phosphorylated in adherent eosinophils. Eosinophils in albumin-coated plates were stimulated by IL-5 for 30 min in the presence or absence of anti-CD18 mAb. Cell lysates were immunoprecipitated with anti-FAK mAb, and immunoblotted with antiphosphotyrosine mAb (4G10). As shown in Figure 4, FAK was found to be tyrosine phosphorylated in unstimulated, nonadherent eosinophils (lane 1), and its level of tyrosine phosphorylation did not change in IL-5-stimulated, adherent eosinophils (lane 2). Similarly, the level of tyrosine phosphorylation of FAK was not affected by anti-CD18 mAb treatment of IL-5-stimulated eosinophils (lane 3). These findings suggest that tyrosine phosphorylation of FAK is not modified by cellular adhesion in eosinophils.

View larger version (44K): [in this window] [in a new window]   Figure 4.   Antiphosphotyrosine immunoblot of FAK in IL-5-stimulated, adherent eosinophils. Eosinophils (6 × 106 cells/sample) were preincubated with 10 µg/ml of anti-CD18 mAb (lane 3) or medium alone (lanes 1 and 2) for 30 min on ice. The pretreated cells were transferred to HSA-coated plates and stimulated with (lanes 2 and 3) or without (lane 1) 25 ng/ml IL-5 for 15 min at 37°C. Cells were lysed and lysates were immunoprecipitated with anti-FAK mAb. Tyrosine-phosphorylated proteins were analyzed by immunoblot using antiphosphotyrosine mAb. The figure is representative of three independent experiments showing similar results.

Recent studies suggest that tyrosine phosphorylation of the 120-kD c-cbl protooncogene product, Cbl, may play a critical role in signal transduction mediated by several receptors, including the T-cell antigen receptor (22), the B-cell antigen receptor (23), and receptors for polypeptide growth factors (24). In contrast to the FAK results (discussed earlier), unstimulated, nonadherent eosinophils (Figure 5, lane 1) showed a minimal level of tyrosine phosphorylation of Cbl. Tyrosine phosphorylation of Cbl was increased in IL-5 stimulated, adherent eosinophils (Figure 5, lane 2). Pretreatment of the cells with anti-CD18 mAb prevented cellular adhesion (Table 1) and inhibited tyrosine phosphorylation of Cbl (Figure 5, lane 3), suggesting that tyrosine phosphorylation of Cbl was not due to the direct effect of IL-5 on eosinophils and that cellular adhesion is required for tyrosine phosphorylation of Cbl. Furthermore, incubation of cells with anti-CD11b mAb microbeads for 15 min stimulated tyrosine phosphorylation of Cbl (Figure 6, lane 2). Although the immunoprecipitated amounts of Cbl were the same among the cells treated with medium alone or with microbeads coated with mouse IgG1 or anti-CD11b mAb (Figure 6, lanes 4 to 6), medium- or mouse IgG1-treated eosinophils did not show tyrosine phosphorylation of Cbl (Figure 6, lanes 1 and 3). These findings suggest that Cbl is the substrate for the PTK activated by cellular adhesion or by _M2 engagement.

View larger version (55K): [in this window] [in a new window]   Figure 5.   Tyrosine phosphorylation of c-Cbl in IL-5-stimulated, adherent eosinophils. Eosinophils (6 × 106 cells/sample) were preincubated with 10 µg/ml of anti-CD18 mAb (lanes 3 and 6) or medium alone (lanes 1, 2, 4, and 5) for 30 min on ice. The pretreated cells were transferred to HSA-coated plates and stimulated with (lanes 2, 3, 5, and 6) or without (lanes 1 and 4) 25 ng/ml IL-5 for 15 min at 37°C. Cells were lysed and lysates were immunoprecipitated with anti-Cbl polyclonal Ab. Tyrosine-phosphorylated proteins were analyzed by immunoblot using antiphosphotyrosine mAb (clone 4G10, lanes 1-3). The blot was then stripped and reprobed with polyclonal anti-Cbl Ab (lanes 4-6). The arrow indicates the protein reactive with anti-Cbl Ab. The figure shows representative results from two independent experiments with similar results.

View larger version (55K): [in this window] [in a new window]   Figure 6.   Tyrosine phosphorylation of c-Cbl in eosinophils stimulated by anti-CD11b mAb coupled to polystyrene microbeads. Eosinophils (6 × 106 cells/sample) were suspended in Eppendorf tubes and stimulated for 15 min at 37°C with medium alone (lanes 1 and 4) or with microbeads coated with anti-CD11b mAb (clone bear 1, lanes 2 and 5) or mouse IgG1 (isotype-matched control Ig, lanes 3 and 6) for 15 min at 37°C. Cells were lysed and lysates were immunoprecipitated with polyclonal anti-Cbl Ab. Tyrosine-phosphorylated proteins were analyzed by immunoblot using antiphosphotyrosine mAb (clone 4G10, lanes 1-3). The blot was then stripped and reprobed with polyclonal anti-Cbl (lanes 4-6). The arrow indicates the protein reactive with anti-Cbl Ab. The figure shows representative results from two independent experiments with similar results.

Engagement of _M2 Induces IP Production by Eosinophils

In numerous receptor systems, ligand binding triggers a rapid increase in phosphoinositide hydrolysis induced by PLC activation (25). Earlier reports described phosphoinositide-specific PLC activation by eosinophils incubated with PAF (26) or immobilized Ig (17, 20), and activation of this enzyme was strongly associated with the degranulation response of the cells (17). There has been no report indicating that cytokine receptors belonging to the cytokine/hemopoietin receptor superfamily, such as IL-5 and GM-CSF receptors, are directly or indirectly associated with phosphoinositide-specific PLC. However, we found that when eosinophils were stimulated with IL-5 in albumin-coated wells, they produced increased amounts of IP (Figure 7A). This increased IP production was abolished by preventing cell adhesion with anti-CD18 mAb, but not with control mouse IgG. Thus, activation of PLC in IL-5-stimulated eosinophils is not a direct outcome of IL-5 receptor occupancy, but is likely mediated by ₂ integrin.

View larger version (15K): [in this window] [in a new window]   Figure 7.   IP production by eosinophils. Panel A: IP production by IL-5-stimulated, adherent eosinophils and the effect of anti-CD18 mAb on the IP production. [3H]inositol-labeled eosinophils were preincubated for 15 min with 10 µg/ml of anti-CD18 mAb or IgG1 (10 µg/ml) or with medium alone for 30 min on ice. Cells were then transferred onto HSA-coated plates and stimulated with or without 25 ng/ml IL-5 for 30 min at 37°C. After incubation, total [3H]IP was isolated and measured, as described in MATERIALS AND METHODS. The values are expressed as mean ± SEM from four independent experiments. Significant differences from the values obtained with cells treated without IL-5 are shown, *P < 0.05. Panel B: IP production by eosinophils stimulated by anti-CD11b mAb. [3H]inositol-labeled eosinophils were suspended in microcentrifuge tubes and stimulated with microbeads coated with HSA, anti-CD11b mAb (clone bear 1), mouse IgG1 (isotype-matched control Ig), or anti-HLA class I mAb (isotype-matched control mAb) for 30 min at 37°C. After incubation, cells were lysed and total [3H]IP was measured, as described in MATERIALS AND METHODS. The values are expressed as mean ± SEM from four independent experiments. Significant difference from the values obtained with cells treated without beads is shown, *P < 0.05.

We next examined whether direct engagement of _M2 by anti-CD11b mAb activates PLC in eosinophils. Suspended cells were stimulated for 30 min with anti-CD11b mAb (clone bear 1), anti-HLA class I mAb (a control mAb), or mouse IgG1 (a control Ig) coupled to polystyrene microbeads. As shown in Figure 7B, incubation with anti-CD11b mAb microbeads significantly enhanced IP production of eosinophils (P < 0.05), whereas beads coated with anti-HLA class I or mouse IgG1 did not affect eosinophils. Pretreatment of eosinophils with 2 µg/ml genistein inhibited the increased IP production in cells stimulated with anti-CD11b mAb microbeads (91 ± 19% of inhibition, mean ± SEM, n = 3), suggesting that activation of a protein tyrosine kinase is required for activation of PLC following ligation of _M2. This finding, as well as the observations with IL-5-mediated adherent eosinophils, suggests that the engagement of _M2 leads to activation of phosphoinositide-specific PLC in eosinophils.

Engagement of _M2 Induces Degranulation of Eosinophils

Finally, we investigated whether cellular activation signals in eosinophils, provoked by engagement of _M2, is connected to effector functions in the cells. As shown in Figure 8, stimulation of suspended eosinophils with anti-CD11b mAb coupled to polystyrene microbeads induced significant cellular degranulation (P < 0.05). Control Ab and Ig, including anti-HLA class I mAb and mouse IgG1, as well as uncoated microbeads (HSA beads), did not induce degranulation of eosinophils. This finding suggests that direct engagement of _M2 with anti-CD11b mAb microbeads causes eosinophil degranulation without any other exogenous stimuli.

View larger version (15K): [in this window] [in a new window]   Figure 8.   Degranulation of eosinophils stimulated by anti-CD11b mAb. Eosinophils were suspended in microcentrifuge tubes and stimulated by microbeads coated with HSA, anti-CD11b mAb (clone bear 1), mouse IgG1 (isotype-matched control Ig), or anti-HLA class I mAb (isotype-matched control mAb) for 4 h at 37°C. After incubation, concentrations of EDN in the cell-free supernatants were analyzed as an indication of eosinophil degranulation. The values are expressed as mean ± SEM from four independent experiments. Significant difference from the values obtained with cells treated without beads is shown, *P < 0.05. Total cellular EDN content was 1904 ± 358 ng/106 cells (mean ± SEM, n = 4).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Previously, several reports showed eosinophil adhesion, especially that mediated by a ₂ integrin, is a crucial step for effector functions of eosinophils (3, 4, 27). However, the molecular basis for this ₂ integrin-dependent activation of eosinophils has not been explored. In the present study, we used two approaches to examine the intracellular activation events triggered by engagement and clustering of ₂ integrin in eosinophils. First, we stimulated eosinophils with IL-5 and allowed them to adhere to albumin-coated tissue culture plates in a ₂ integrin-dependent manner (Figures 1, 5, and 7A). Second, we crosslinked a ₂ integrin, _M2, on suspended eosinophils by allowing the cells to interact with anti-CD11b mAb coupled to polystyrene microbeads (Figures 2, 6, and 7B). Results obtained from these two different approaches were essentially the same. Stimulation of eosinophils with IL-5 and subsequent cellular adhesion to albumin-coated plates resulted in tyrosine phosphorylation of a number of proteins, including a 115-kD protein and Cbl, and increased phosphoinositide hydrolysis. Prevention of cellular adhesion by anti-CD18 mAb inhibited these cellular events. Similarly, direct ligation of _M2 by anti-CD11b mAb resulted in rapid and prolonged tyrosine phosphorylation of several proteins, most notably a 115-kD protein and Cbl, together with increased phosphoinositide hydrolysis. Furthermore, engagement of _M2 induced the degranulation response of eosinophils (Figure 8). These findings suggest that cellular adhesion through ₂ integrins and engagement of _M2 on eosinophils trigger cellular activation, including activation of PTK and PLC, and cause eosinophil degranulation. Therefore, ₂ integrin is likely to participate actively in intracellular signaling and functions of eosinophils.

Ligation of a number of integrins, such as ₄₁ and _IIb/ ₃, results in enhanced phosphorylation of a variety proteins in the 100-130-kD range (30, 31). A 125-kD phosphotyrosyl protein has been identified as the protein tyrosine kinase, FAK, in fibroblasts (6), platelets (7), and basophilic leukemia cells (8). In this study, the antiphosphotyrosine immunoblot of detergent-soluble proteins from anti-CD11b-stimulated eosinophils showed multiple protein bands in the 100-125-kD range (Figure 2). However, FAK was constitutively tyrosine phosphorylated in eosinophils and the levels of tyrosine phosphorylation of this protein did not change after cellular adhesion (Figure 4). In contrast, subsequent studies with Cbl-specific antibodies showed that eosinophils express 120-kD Cbl, and that this Cbl became tyrosine phosphorylated in eosinophils after engagement of _M2 (Figures 5 and 6). Recent studies revealed that Cbl undergoes tyrosine phosphorylation in lymphoid cells in response to both antigen-receptor and cytokine-receptor stimulation (22, 32). Our study extends the range of stimuli that provoke tyrosine phosphorylation of Cbl and is the first to report that integrin adhesion molecules, another important class of immunoregulatory receptors, are also associated with Cbl tyrosine phosphorylation. Our findings are also compatible with those made by others who demonstrated that, in p210^BCR/ABL transformed leukemic cell lines, Cbl formed a complex with a focal adhesion protein, suggesting a possible role for Cbl in adhesion- dependent functions of the cells (33). Tyrosine-phosphorylated Cbl has the ability to bind to the SH2 and/or SH3 domain-containing proteins, such as Src-family protein tyrosine kinase p53/56^lyn (34), ZAP-70 (35), the adapter protein Grb2 (36), phosphatidylinositol 3-kinase (23), and PLC- (22). Therefore, Cbl may play an important pivotal role in the nucleation of multienzyme signaling complexes following integrin receptor engagement in hematopoietic cells. Obviously, more work must be done to examine the role of this protein in effector functions of eosinophils.

PLC-mediated hydrolysis of PIP₂ generates two second messengers, IP₃ and diacylglycerol, which, in turn, trigger intracellular calcium mobilization and PKC activation, respectively. Previous studies on endothelial cells and lymphocytes showed production of IP and elevation of intracellular calcium concentration after adhesion to collagen or cross-linking of _L2, respectively (37, 38). In this study, we extended those observations to granulocytes and showed that engagement of eosinophil _M2 by anti-CD11b mAb enhances production of IP (Figure 7B). The role of ₂ integrins for PLC activation in eosinophils was further supported by an increase in IP production by IL-5-stimulated adherent eosinophils, but not by IL-5-stimulated nonadherent eosinophils (Figure 7A). What, then, is the mechanism of _M2-mediated activation of PLC by eosinophils? In T lymphocytes, the PLC-1 isoform is known to be activated by tyrosine phosphorylation (39), and crosslinking of _L2 results in the tyrosine phosphorylation of PLC-1 (38). We showed previously that PLC activity in eosinophils is regulated by protein tyrosine phosphorylation (20). Therefore, it is highly possible that ligation of _M2 activates PLC via tyrosine phosphorylation of proteins in eosinophils. This speculation is compatible with our finding that _M2-mediated IP production was inhibited by a tyrosine kinase inhibitor, genistein. Although the protein(s) which interact with PLC and stimulate PLC activity in eosinophils has not been identified, phosphorylated Cbl is one of the candidates because it can physically associate with the PLC-1 isomer (22).

A clarification must be made regarding the anti-CD11b mAb (clone bear 1) coupled to microbeads used to study the engagement of the _M2 on eosinophils. In preliminary studies, we tested the effects of adhesion-blocking anti-CD11b mAb (clone D12) in place of nonadhesion-blocking anti-CD11b mAb (clone bear 1). We also examined soluble anti-CD11b mAb (clone bear 1) in place of the same antibody immobilized onto microbeads. The results of these experiments showed only weak and temporal tyrosine phosphorylation of the 115-kD protein, no increase in phosphoinositide turnover, and no cellular degranulation in eosinophils (data not shown). Although the reason(s) for the weakness of adhesion-blocking anti-CD11b mAb and soluble mAb in activating intracellular signals of eosinophils is unknown, a review of the literature provides some insights. Meijne and associates showed by immune fluorescence that _L2 redistributes to the ICAM-1-adherent surface and concentrates in the lamellipodia of a spreading T-cell line (40). By immunoelectron microscopy, they found _L2 localized in microclusters of approximately 10 gold particles, suggesting that clustering or "multimerization" of ₂ integrin is required for strong cell adhesion. The importance of "multimerization" for cell adhesion and triggering of "outside-in" signaling has also been suggested by others (reviewed in 41). Therefore, we can speculate that adhesion-blocking anti-CD11b mAb (clone D12) and soluble mAb induced "dimerization" of adjacent _M2 molecules but did not induce "multimerization" of _M2 molecules in our preliminary studies. This may be the reason why these preparations of mAb were weaker agonists than anti-CD11b mAb (clone bear 1) immobilized onto microbeads. Perhaps both the engagement and also the clustering of integrins at the sites of focal adhesion are important for adhesion-dependent signal transduction of eosinophils.

In summary, this study suggests that, with no other exogenous stimuli, direct engagement and clustering of _M2 by anti-CD11b mAb activates tyrosine kinases and stimulates PLC in eosinophils and induces the degranulation response of the cells. This _M2-mediated activation of eosinophils may be particularly important in the pathophysiology of eosinophilic diseases and allergic inflammation, in which various types of cells show increased expression of ligands for _M2, as typified by ICAM-1 expression by bronchial epithelial cells in bronchial asthma (42). Indeed, _M2 is implicated in the degranulation of eosinophils and the development of bronchial hyperactivity in cynomolgus monkeys challenged with allergen as a model of human bronchial asthma (43). Therefore, although the analyses of detailed molecular mechanisms leading to activation of eosinophils following the engagement of _M2 are still underway, identification of several key molecules in the pathway would provide insight into the transmembrane signaling mediated by integrin adhesion receptors and possibly open new avenues for the treatment of patients with eosinophilia and allergic diseases.