INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

EOSINOPHILS PLAY AN IMPORTANT role at inflammatory sites, taking part in the host defense against parasites, pathophysiology associated with eosinophilia, and allergic diseases such as bronchial asthma and atopic dermatitis (8). Neutrophils also play an important role not only in bacterial infection but also in acute respiratory distress syndrome (19), vasculitis (32), idiopathic pulmonary fibrosis (22), and bronchial asthma (33). Activated eosinophils and neutrophils produce oxygen metabolites, such as superoxide, which act as disinfectants and cause airway injury at the inflamed lesion. Leukocyte superoxide anion production is induced by many different reagents, such as phorbol myristate acetate (PMA), platelet-activating factor, N-formyl-methionyl-leucyl-phenylalanine, tumor necrosis factor-, and granulocyte-macrophage colony-stimulating factor (GM-CSF) (13, 31).

CD54 [intercellular adhesion molecule-1 (ICAM-1)] is known to be expressed on leukocytes, fibroblasts, and endothelial cells in cases of allograft rejection and inflammation. It has been reported that CD54 expression on eosinophils is a hallmark of activation (10, 24) and that CD54 is not expressed on freshly isolated eosinophils. ICAM-1 is the common ligand for the ₂-integrin: lymphocyte function-associated antigen-1 (CD11a/CD18) and membrane attack complex-1 [(Mac-1): CD11b/CD18]. Our laboratory found that CD54 on eosinophils is involved in eosinophil degranulation through ₂-integrin (15). However, the role of CD54 on eosinophils and neutrophils in superoxide production is still unknown. We assessed the interaction of CD54 and CD18 on eosinophil and neutrophil superoxide production. We also evaluated differences in CD54 expression between eosinophils and neutrophils.

Theophylline, which elicits bronchial dilation, is one of the most widely used drugs in bronchial asthma therapy (25); phosphodiesterase (PDE) inhibition is thought to be important in the anti-inflammatory mechanism of theophylline (1, 21). We, therefore, evaluated the effects of theophylline and PDE4 inhibitor on the expression of CD54 and CD18, as well as on superoxide production in eosinophils and neutrophils.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Reagents and antibodies. Reagents employed in this study were human recombinant GM-CSF (rGM-CSF) (generously provided by Kirin Brewery, Tokyo, Japan) and PMA (Calbiochem, La Jolla, CA). Theophylline was purchased from Sigma Chemical (St. Louis, MO). KF-19514 (selective PDE4 inhibitor) was donated by Kyowa Hakkou (Tokyo, Japan). Theophylline was dissolved in Hanks' balanced salt solution with 10 mM HEPES at 10¹ M, and KF-19514 in saline at 10² M. The drugs were dissolved immediately before an experiment and diluted to the specified concentrations with the medium buffer. Purified monoclonal antibodies (MAbs), namely, anti-CD18 (L130; mouse IgG1) and anti-CD54 (LB-2; mouse IgG2b), and irrelevant isotype-matched antibodies (mouse IgG1 and IgG2b), were purchased from Becton Dickinson (San Jose, CA) and Organon Teknica (Durham, NC), respectively. FITC-conjugated MAbs anti-CD54 (84H10: mouse IgG1; Immunotech, Marseille, France), anti-CD18 (Dako patts, Glostrup, Denmark), anti-CD9 (M-L13: mouse IgG1; Pharmingen, San Diego, CA), anti-CD16 (3G8: mouse IgG1; Immunotech), and control mouse IgG1 (Dako patts A/S) were used for flow cytometric analysis.

Eosinophil preparation. Eosinophil isolation was performed with essentially the same method as described by Hansel et al. (11). Briefly, heparinized venous blood was obtained from normal subjects and diluted with PIPES buffer (25 mM PIPES, 50 mM NaCl, 5 mM KCl, 25 mM NaOH, 5.4 mM glucose, pH 7.4) at a ratio of 1:1. Diluted blood was overlaid on isotonic Percoll solution (density, 1.082 g/ml; Sigma Chemical) and centrifuged at 1,000 g for 30 min at 4°C. The supernatant and mononuclear cells at the interface were carefully removed, and erythrocytes in the sediment were lysed with two cycles of chilled hypotonic water lysis. Isolated granulocytes were washed twice with PIPES buffer, with 1% inactivated FCS (Life Technologies, Gaithersburg, MD). The resultant pellet of granulocytes was incubated with anti-CD16-coated immunomagnetic particles (50 µl for 5 × 10⁷ cells; Militenyi Biotec, Bergisch Gladbach, Germany) for 60 min on ice. Magnetically labeled neutrophils were then depleted by passing the granulocytes through a magnetic cell separation column in a strong magnetic field. The purity of eosinophils, as verified with Randolph's stain, and their viability assessed by means of Trypan blue dye exclusion were both >98%. Purified eosinophils were washed twice in PIPES buffer with 1% FCS and suspended in the reaction medium.

Neutrophil preparation. Heparinized venous blood was obtained from normal subjects, diluted with RPMI-1640 at a ratio of 1:1, layered onto a Lymphoprep (density, 1.077g/ml; Nycomed Pharma, Oslo, Norway), and centrifuged at 400 g for 30 min at 4°C. The supernatant and mononuclear cells at the interface were carefully removed, and erythrocytes in the sediment were lysed in chilled distilled water. Isolated granulocytes were washed twice with PIPES buffer with 1% FCS. The neutrophils used for the experiments were of >98% purity with <2% of contaminated eosinophils, and the viability was >99% as determined by Trypan blue dye exclusion.

Flow cytometric analysis of eosinophil and neutrophil CD54 and CD18 expression. Flow cytometric analysis was performed with a previously described method (15). Briefly, purified human eosinophils or neutrophils were suspended in RPMI-1640 supplemented with 5% FCS at 1 × 10⁶/ml. One-milliliter aliquots of cell suspension were incubated with 1 ng/ml of PMA or 10 ng/ml of rGM-CSF in polypropylene tubes (Falcon no. 2096, Becton Dickinson) kept in a slanted position for up to 4 h at 37°C. In some experiments, isolated cells were pretreated with theophylline or KF-19514. After incubation, cell suspensions were washed with PIPES supplemented with 1% FCS. Samples of 5 × 10⁵ cells in 50 µl of cold PIPES were incubated for 30 min on ice with a saturating amount of FITC-conjugated anti-CD18 MAb, anti-CD54 MAb, anti-CD9 MAb, anti-CD16 MAb, or mouse IgG1 MAb as a control. The incubated cells were then washed with PBS containing 1% FCS at 300 g for 10 min, resuspended in 1 ml of PBS containing 1% FCS, and kept at 4°C until flow cytometric analysis. The presence of gated eosinophils or neutrophils was confirmed by means of FITC-conjugated anti-CD9 MAb or anti-CD16 MAb, respectively. The cells (1 × 10⁴) were counted in the gated region and shown as histograms of fluorescence intensity vs. cell number. Mean fluorescence intensity was calculated with the Consort 32 program (Becton Dickinson).

Superoxide production assay from eosinophils and neutrophils. Production of superoxide by eosinophils or neutrophils was measured by superoxide dismutase-inhibitable reduction of cytochrome c with a slightly modified technique, as previously described (3). PMA or rGM-CSF was diluted in Hanks' balanced salt solution with 10 mM HEPES for each concentration. Freshly isolated cells were suspended in the same medium and mixed with 100 µM cytochrome c (Sigma Chemical). The cell suspension (100 µl; 10⁶ cells/ml) was dispensed onto 96-well, flat-bottom tissue culture plates (Falcon no. 3072, Becton Dickinson) coated with 2.5% human serum albumin (HSA) (Sigma Chemical) for eosinophil experiments or with 100 µg/ml of fibrinogen (Sigma Chemical) for neutrophil experiments. The reactions were initiated by adding 100 µl of each stimulant; the reaction wells were then measured for absorbance at 550 nm in a microplate autoreader (Tecan Austria Ges, Salzburg, Austria) followed by repeated readings. During these absorbance measurements, the plate was maintained at a temperature of 37°C. For the inhibition experiments, eosinophils or neutrophils were reacted with MAb (final concentration, 10 µg/ml) or each drug at room temperature for 10 min before the addition of stimuli and throughout the entire incubation period. Each reaction was performed in duplicate. Superoxide anion production was calculated with an extinction coefficient of 21.1 × 10³ l · mol¹ · cm¹ for reduced cytochrome c at 550 nm and was expressed as nanomoles of cytochrome c reduced per 1 × 10⁶ cells.

Statistical analysis. Statistical significance of the differences between various treatment groups was assessed with one-way ANOVA; Fisher's paired least significant difference method was employed when significant statistical results were noted. P values of <0.05 were taken as significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Inhibitory effect of anti-CD18 and anti-CD54 on superoxide production. Superoxide production of eosinophils was observed in conjunction with rGM-CSF or PMA stimulation (Fig. 1). Anti-CD18 MAb significantly inhibited superoxide production from eosinophils in response to either stimulus. Anti-CD54 MAb inhibited PMA-stimulated, but not rGM-CSF-stimulated, superoxide production. Neutrophil superoxide production was observed as a result of both rGM-CSF and PMA stimulation (Fig. 2). Anti-CD18 MAb significantly inhibited superoxide production of neutrophils in response to either stimulus, showing similar results to those for eosinophils. Furthermore, anti-CD54 MAb inhibited both rGM-CSF- and PMA-stimulated neutrophil superoxide production. PMA was a more potent and quicker stimulator of superoxide production than rGM-CSF.

View larger version (34K): [in this window] [in a new window]   Fig. 1.   Effects of anti-CD18 (; left) or anti-CD54 monoclonal antibody (MAb; ; right) on eosinophil superoxide production. Eosinophils were preincubated with anti-CD18 MAb or anti-CD54 MAb for 10 min at room temperature and stimulated with recombinant granulocyte-macrophage colon-stimulating factor (rGM-CSF; 10 ng/ml; A) or phorbol myristate acetate (PMA; 1 ng/ml; B) for up to 4 h at 37°C. Production of superoxide by eosinophils was measured by means of superoxide dismutase-inhibitable reduction of cytochrome c, as described in MATERIALS AND METHODS. The results (means ± SE) of 8 experiments with duplicate samples are shown. Significant difference from values obtained with eosinophils preincubated with irrelevant isotype-matched antibodies IgG1 () and IgG2b (), ** P < 0.01.

View larger version (34K): [in this window] [in a new window]   Fig. 2.   Effects of anti-CD18 (; left) or anti-CD54 MAb (; right) on neutrophil superoxide production. Neutrophils were preincubated with anti-CD18 MAb or anti-CD54 MAb for 10 min at room temperature and stimulated with rGM-CSF (10 ng/ml; A) or PMA (1 ng/ml; B) for up to 4 h at 37°C. Production of superoxide by neutrophils was measured by means of superoxide dismutase-inhibitable reduction of cytochrome c, as described in MATERIALS AND METHODS. The results (means ± SE) of 8 experiments with duplicate samples are shown. Significant difference from values obtained with neutrophils preincubated with irrelevant isotype-matched antibodies IgG1 () and IgG2b (),** P < 0.01 and * P < 0.05.

Expression of CD18 and CD54 on eosinophils and neutrophils. To study the contribution of CD18 and CD54 to superoxide production of eosinophils and neutrophils, their expression on eosinophils and neutrophils stimulated by PMA or rGM-CSF was examined. A representative profile of CD18 or CD54 expression in five independent experiments is shown in Figs. 3 and 4. As shown in Fig. 3, freshly isolated eosinophils expressed CD18 but not CD54 on the cells, as reported previously (12), whereas freshly isolated neutrophils expressed both CD18 and CD54. The other four experiments produced similar results. CD18 expression on eosinophils and neutrophils and CD54 expression on neutrophils were increased after 2 h of rGM-CSF stimulation (Fig. 4, A and C). However, rGM-CSF stimulation did not induce CD54 expression on eosinophils (Fig. 4A). CD18 expression on eosinophils and neutrophils and CD54 expression on neutrophils were increased after 30 min of PMA stimulation (Fig. 4, B and D). Newly expressed CD54 on eosinophils was observed after 15 min of PMA stimulation (data not shown), and maximum mean fluorescence intensity was observed 30 min after stimulation (Fig. 4B).

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Profile of CD18 (top) or CD54 (bottom) expression on freshly isolated eosinophils (A) and neutrophils (B). The profile is representative of all 5 experiments.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Profile of CD18 (left) or CD54 (right) expression on stimulated eosinophils (solid trace) incubated with 10 ng/ml rGM-CSF for 2 h (A) or with 1 ng/ml PMA for 30 min (B) or on stimulated neutrophils (solid trace) incubated with 10 ng/ml rGM-CSF for 2 h (C) or with 1 ng/ml PMA for 30 min (D) compared with fresh isolated cells (dotted trace). The profile is representative of all 5 experiments.

Inhibitory effect of theophylline and KF-19514 on superoxide production. Superoxide production of eosinophils and neutrophils reaches a plateau from 30 to 60 min after PMA stimulation or ~120 min after rGM-CSF stimulation so that superoxide production in the presence of theophylline or KF-19514 was examined 30 min after PMA stimulation and 120 min after rGM-CSF stimulation. Superoxide production of eosinophils as a result of PMA or rGM-CSF stimulation was inhibited by theophylline or KF-19514 in a dose-dependent manner (Fig. 5). Superoxide production of neutrophils in response to PMA or rGM-CSF stimulation was also inhibited by theophylline or KF-19514 in a dose-dependent manner (Fig. 6).

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Effect of theophylline or KF-19514 on eosinophil superoxide production. Eosinophils were preincubated with theophylline or KF-19514 for 10 min at room temperature and then stimulated with PMA (1 ng/ml; A) for 30 min or rGM-CSF (10 ng/ml; B) for 120 min at 37°C. The results (means ± SE) of 5 experiments with duplicate samples are shown. Significant difference compared with eosinophils incubated with stimuli (PMA or rGM-CSF) alone, ** P < 0.01 and * P < 0.05.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   Effect of theophylline or KF-19514 on neutrophil superoxide production. Neutrophils were preincubated with theophylline or KF-19514 for 10 min at room temperature and then stimulated with PMA (1 ng/ml; A) for 30 min or rGM-CSF (10 ng/ml; B) for 120 min at 37°C. The results (means ± SE) of 4 experiments with duplicate samples are shown. Significant difference compared with neutrophils incubated with stimuli (PMA or rGM-CSF) alone, ** P < 0.01.

Inhibition by theophylline and KF-19514 of CD18 and CD54 expression. Eosinophils and neutrophils stimulated by PMA demonstrated significantly increased expression of CD18 and CD54 at 30 min compared with the medium alone (Fig. 7, P < 0.01 for all). Neutrophils stimulated by rGM-CSF also demonstrated significantly increased expression of CD18 and CD54 at 120 min (P < 0.01 for both). CD18 and CD54 expression on PMA-stimulated eosinophils and neutrophils was not inhibited by theophylline and KF-19514, but the same expression on rGM-CSF-stimulated neutrophils was partially inhibited.

View larger version (33K): [in this window] [in a new window]   Fig. 7.   Increased CD18 (A, C, E) or CD54 (B, D, F) expression on eosinophils (A and B) or neutrophils (C-F) after PMA (A-D) or rGM-CSF (E and F) stimulation. The cells were preincubated with theophylline (Theo) or KF-19514 for 10 min at room temperature and then stimulated with PMA (1 ng/ml) for 30 min (A-D) or rGM-CSF (10 ng/ml) for 120 min (E and F). Mean fluorescence intensity (means ± SE) of 4-7 experiments is shown. Significant difference from values obtained with eosinophils or neutrophils incubated with PMA or rGM-CSF alone (no drugs), ** P < 0.01 and * P < 0.05.

Microscopic examination of eosinophils and neutrophils by rGM-CSF or PMA stimulation. Morphological changes of eosinophils and neutrophils stimulated by rGM-CSF or PMA were examined by using an inverted microscope (Olympus, Tokyo, Japan). As shown in Fig. 8A, eosinophils after 2 h of rGM-CSF stimulation were flattened and aggregated. When eosinophils were pretreated with anti-CD18 MAb, many cells remained round and less aggregated, and only a few eosinophils adhered to the plates (Fig. 8B). When eosinophils were pretreated with anti-CD54 MAb, some of them assumed a spindle shape (Fig. 8C); however, the degree of aggregation was similar to that induced by rGM-CSF stimulation alone. As shown in Fig. 8E, eosinophils after 30 min of PMA stimulation showed strong degranulation and marked aggregation. Anti-CD18 and anti-CD54 MAbs attenuated aggregation (Fig. 8, F and G), and only anti-CD18 MAb partially blocked eosinophil adhesion to the HSA-coated plates. When eosinophils were pretreated with 10³ M theophylline, eosinophils stimulated by rGM-CSF were round and refractive with a low-grade adhesion to the plates (Fig. 8D). Theophylline attenuated eosinophil degranulation and aggregation induced by PMA stimulation (Fig. 8H).

View larger version (65K): [in this window] [in a new window]   Fig. 8.   Light photomicrographs of eosinophils adhered to tissue culture plates. Eosinophils were incubated with rGM-CSF (10 ng/ml; A-D) for 2 h or PMA (1 ng/ml, E-H) for 30 min. To determine the role of CD18, CD54, or theophylline in eosinophil adhesion and aggregation, eosinophils were preincubated with 10 µg/ml anti-CD18 MAb (B and F), 10 µg/ml anti-CD54 MAb (C and G), or 103 M theophylline (D and H) for 10 min (magnification ×200).

3

View larger version (57K): [in this window] [in a new window]   Fig. 9.   Light photomicrographs of neutrophils adhered to tissue culture plates. Neutrophils were incubated with rGM-CSF (10 ng/ml; A-D) for 2 h or PMA (1 ng/ml, E-H) for 30 min. To determine the role of CD18, CD54, or theophylline in neutrophil adhesion and aggregation, neutrophils were preincubated with 10 µg/ml anti-CD18 MAb (B and F), 10 µg/ml anti-CD54 MAb (C and G), or 103 M theophylline (D and H) for 10 min (magnification ×200).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

It is thought that cell-to-cell interaction through adhesion molecules on eosinophils and neutrophils plays a role in superoxide production. Human eosinophils and neutrophils constitutively express Mac-1 (CD11b/CD18) on the cell surface. Mac-1 is known to be a ligand of HSA (14) and fibrinogen (18). In our preliminary experiments, eosinophils stimulated with PMA or rGM-CSF showed much greater superoxide production in albumin-coated tissue plates than in fibrinogen-coated tissue plates. On the other hand, neutrophils showed much greater superoxide production in fibrinogen-coated tissue plates than in albumin-coated tissue plates. Such differences in superoxide production between eosinophils and neutrophils may be due to conformational changes of ₂-integrin after PMA or rGM-CSF stimulation.

It is reported that cell adhesion through CD11b/18 is a crucial step for the activation, signaling, and effector function of eosinophils stimulated by IgG (16) and that CD18 is involved in eosinophil degranulation induced by rGM-CSF or platelet-activating factor (14). Furthermore, adhesion-dependent (i.e., adhesion between extracellular matrix proteins and expression of CD11b/CD18 integrins) respiratory bursts of neutrophils are induced by tyrosine phosphorylation (7). Schnitzler and colleagues (30) reported that the lymphocyte function-associated antigen-1 ligand interaction in neutrophils is a stimulatory signal for phagocytotic activation and induces a strong oxidative burst. In our study, anti-CD18 antibody inhibited eosinophil or neutrophil superoxide production by blocking the cell adhesion to the tissue culture plates, suggesting that intracellular signaling through ₂-integrin plays an important role in superoxide production.

It is well known that CD54 is an adhesion molecule functioning as a ligand for Mac-1. CD54 is not expressed on freshly isolated eosinophils, but CD54 expression is observed on eosinophils in peripheral blood and bronchoalveolar lavage fluid in various diseases (24, 36), suggesting that CD54 is newly expressed on activated eosinophils. CD54 expression on neutrophils is significantly increased after separation by means of density gradients (17). Our study confirmed that CD54 was expressed on freshly isolated neutrophils and that CD54 was further strongly expressed on stimulated neutrophils. Anti-CD54 MAb inhibited superoxide production of neutrophils in rGM-CSF or PMA stimulation. Anti-CD54 MAb also inhibited superoxide production of eosinophils in PMA stimulation, but not in rGM-CSF stimulation, because CD54 was newly expressed on eosinophils within 15 min after PMA stimulation, but not until 2 h after rGM-CSF stimulation. On the other hand, our laboratory previously demonstrated that eosinophil degranulation incubated with rGM-CSF for 4 h was inhibited by anti-CD54 (15). When leukocytes were stimulated by rGM-CSF, eosinophil degranulation gradually increased over 4 h (13), whereas eosinophil and neutrophil superoxide production reached a plateau ~2 h after stimulation. We, therefore, hypothesize that superoxide production is related only to early expression of CD54 on leukocytes after stimulation.

Our laboratory has previously reported that rGM-CSF significantly upregulated CD54 expression on eosinophils incubated for 2 h (15); however, in the present study, rGM-CSF did not. One possible explanation for this discrepancy is the differences in FCS concentration in the culture medium: our laboratory used 2% FCS in the previous experiments vs. 5% FCS in this study. For fluorescence-activated cell sorting examination, 5% FCS in culture medium is more suitable than 2% FCS to make leukocytes stabilize and avoid aggregation and adhesion to the polypropylene tube. Stabilization by FCS may make eosinophils less sensitive to stimuli such as rGM-CSF. Furthermore, we observed that CD54 on eosinophils by rGM-CSF stimulation was not expressed within 1-h culture at 2% FCS and was expressed after 2 h. Moreover, CD54 on eosinophils by rGM-CSF stimulation was clearly expressed from 18 to 24 h, even at 10% FCS in our preliminary experiments. Thus CD54 expression on eosinophils is suggested to be related to FCS concentration and time dependence.

The mechanism of CD54 upregulation in response to rGM-CSF in neutrophils has been reported as follows: 1) rGM-CSF phosphorylates extracellular signal-regulated kinase strongly and p38 mitogen-activated protein kinase weakly (34); 2) the CD54 promoter contains several activator protein-1 (AP-1) binding sites that may be important for CD54 expression; and 3) AP-1 is composed of either Jun homodimers or Fos/Jun heterodimers, and both the extracellular signal-regulated kinase and p38 mitogen-activated protein kinase pathways regulate AP-1 activity both by increasing the expression of Jun and Fos and by phosphorylation of newly synthesized AP-1 complex (28). Thus rGM-CSF stimulation is known to lead to CD54 expression.

As a ligand of ₂-integrin, CD54 may act as a glue between leukocytes themselves. In fact, microscopic examination revealed that leukocyte aggregation was observed after rGM-CSF stimulation, as also previously reported (15), and this aggregation was also observed after PMA stimulation. Anti-CD54 MAb reduced eosinophil aggregation stimulated by PMA, but not by rGM-CSF, possibly because CD54 on eosinophils was not expressed by rGM-CSF stimulation. In addition, anti-CD54 MAb reduced neutrophil aggregation stimulated by either rGM-CSF or PMA. These findings suggest that CD54 on neutrophils and eosinophils may be at least partly involved in superoxide production because of cell-to-cell interaction through ₂-integrin (CD18) and CD54. However, anti-CD18 or anti-CD54 could not completely inhibit leukocyte aggregation. Eosinophils and neutrophils are also known to express L-selectin and its ligand P-selectin glycoprotein ligand-1 (PSGL-1). It has been reported that L-selectin and PSGL-1 expressed on neutrophils support a collisional cell-to-cell interaction that represents the first step in neutrophil aggregation (9) and that cross-linking of L-selectin and Mac-1 initiates changes in intracellular calcium and superoxide production in neutrophils (5). Therefore, we think that other surface antigens, such as L-selectin and PSGL-1, may also be involved in eosinophil or neutrophil aggregation.

The ability of CD54 to function as a signaling molecule has been demonstrated in previous studies. It has been reported that CD54 cross-linking leads to tyrosine phosphorylation of cytoskeleton proteins (6), that CD54 can associate via its cytoplasmic domain with -actinin, an actin-binding cytoskeletal portion (4), and that the small GTP-binding protein Rho is a key mediator of actin cytoskeletal remodeling induced by intracellular signals (27). CD54 is also implicated in a cascade of signaling events including Rho; this signaling pathway is probably involved in the subsequent transmigration of leukocytes (29). Thus CD54 may participate in signal transduction through outside-in signaling events, resulting in superoxide production.

We also studied the effect of two drugs on superoxide production and CD18/CD54 expression. Theophylline is known to increase intracellular cAMP and protein kinase A, resulting in inhibition of leukocyte function such as chemotaxis (35), degranulation (21), and superoxide production (20). As shown in Figs. 5 and 6, theophylline, as well as KF-19514, was found to inhibit rGM-CSF- or PMA-induced superoxide production of eosinophils and neutrophils. However, Schultz (31) reported that rolipram, a PDE4 inhibitor, inhibited neutrophil superoxide production resulting from GM-CSF stimulation but not from PMA stimulation, whereas our data indicate that PMA-induced superoxide production was significantly inhibited by theophylline or KF-19514. This discrepancy may be due to the difference in exposure time used for the measurement of superoxide production.

As for surface antigens, the increase in the expression of CD18 and CD54 on PMA-stimulated eosinophils and neutrophils was not inhibited by theophylline or KF-19514 but was on rGM-CSF-stimulated neutrophils. It has been reported that PDE inhibitors do not inhibit tumor necrosis factor-induced ICAM-1 expression on human umbilical vein endothelial cells (26) or on human lung microvascular endothelial cells (2). Our data suggest that PDE4 inhibitors may decrease leukocyte superoxide production partly via inhibition of CD18 and CD54 interaction, although these drugs mainly inhibit signal transduction through protein kinase A activity. Because of the inhibitory effect on cell-to-cell interaction, which causes superoxide production and increased expression of adhesion molecules at inflammatory sites, drugs such as theophylline and KF-19514 may be potentially useful for the treatment of acute respiratory distress syndrome and pulmonary fibrosis because of their anti-inflammatory effects, similar to their effect in the treatment of bronchial asthma (23). Further experiments are, therefore, needed to expand the potential of these drugs for the treatment of inflammatory lung diseases.

In conclusion, eosinophil interaction or neutrophil interaction through CD54 and CD18 adhesion is involved in superoxide production. The inhibition of superoxide production by theophylline is suggested to be, at least partly, due to the inhibition of CD54 and CD18 expression in addition to signal transduction.