Introduction

Abstract Introduction Materials and Methods Results Discussion References

Eosinophils are important components of the host-defense system, in particular the response to parasitic infections (1). Eosinophils also appear to play an important role in asthma and lead to some of the tissue damage observed in this condition (2). Eosinophils are released from the bone marrow in response to cytokines such as interleukin (IL)-5 (1). Activation of eosinophil selectins and integrins and IL-4-mediated upregulation of vascular cell adhesion molecule-1 on endothelial cells (3) leads to their adherence to the endothelium, which is followed by their migration into tissues in response to chemoattractants (1).

A variety of peptides, including released on activation, normal T-cell expressed and secreted (RANTES) (4) and the more specific eotaxin (5) are potent chemoattractants for eosinophils. Lipid mediators, particularly platelet-activating factor (PAF), also have chemotactic effects on these cells (6). Leukotriene B₄ (LTB₄) is a potent chemotactic agent for guinea pig eosinophils (7, 8) but is much less active on human eosinophils (9, 10), although it is a very potent activator of human neutrophils (11). 8S,15S-Dihydroxy-5,9,11,13-eicosatetraenoic acid also has chemotactic effects on guinea pig eosinophils, but appears to act via LTB₄ receptors and is less potent than LTB₄ (7).

We recently identified another class of leukocyte-derived 5-lipoxygenase products that are synthesized by the action of a highly specific dehydrogenase on 5S-hydroxyeicosanoids containing 6-trans double bonds (12). This reaction is strongly stimulated by the protein kinase C (PKC) activator phorbol myristate acetate (13). Oxidation of 5-hydroxy-6,8,11,14-eicosatetraenoic acid (5-HETE), the preferred substrate for this enzyme, to 5-oxo-6,8,11,14-eicosatetraenoic acid (5-oxo-ETE) results in a 100-fold enhancement of biologic activity on neutrophils (14). Although 5-oxo-ETE is considerably less potent than LTB₄ as a neutrophil agonist (14, 15), there is considerable evidence that it acts via a specific 5-oxo-ETE receptor. In contrast, we recently found that 5-oxo-ETE is about 2.5 times more active than PAF as an eosinophil chemoattractant, and nearly 40 times more active than LTB₄ (9). In addition to its direct chemotactic effects on eosinophils, low concentrations (1- 3 nM) of 5-oxo-ETE enhanced the responses of these cells to PAF in a synergistic manner, suggesting that these two mediators act by different mechanisms (9). We have also obtained preliminary evidence that 5-oxo-ETE stimulates the pulmonary infiltration of eosinophils in vivo (16).

The objective of the current investigation was to identify rapid responses induced in eosinophils by 5-oxo-ETE that might be involved in the infiltration of these cells into tissues. We hypothesized that there may be differences in some of the responses induced by 5-oxo-ETE and PAF that could explain their synergistic effect on eosinophil chemotaxis.

	Materials and Methods

Abstract Introduction Materials and Methods Results Discussion References

Materials

5-Oxo-ETE was obtained from Cascade Biochem Ltd. (Reading, UK), whereas LTB₄ was synthesized by incubation of arachidonic acid with porcine leukocytes in the presence of 5,8,11,14-eicosatetraynoic acid (17). LTB₄ prepared in this way was at least 98% pure and displayed an EC₅₀ for calcium mobilization in neutrophils of 0.5 nM. PAF (1-O-hexadecyl-2-acetyl-sn-glycero-3-phosphocholine) was obtained from Sigma Chemical Co. (St. Louis, MO). Lipid mediators were dissolved in Hanks' balanced salt solution (HBSS) containing 0.1% bovine serum albumin (BSA; Sigma Chemical).

PD98059 was purchased from Calbiochem (La Jolla, CA), whereas staurosporine and wortmannin were obtained from Boehringer Mannheim (Laval, PQ, Canada) and Research Biochemicals International (Natick, MA), respectively. Indo-1 and N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)phallacidin (NBD-phallacidin) were obtained from Molecular Probes (Eugene, OR). Fluorescein isothiocyanate (FITC)-labeled mouse antihuman CD11b (Bear1) and the corresponding FITC-labeled isotype immunoglobulin G₁ (IgG₁) control antibody were purchased from Immunotech-Coulter (Burlington, ON, Canada). FITC-labeled mouse antihuman L-selectin (Leu-8) and the corresponding control IgG_2a antibody were purchased from Becton-Dickinson (San Jose, CA). Mouse plasma was obtained from Cedarlane (Hornby, ON, Canada).

Preparation of Eosinophils

Human granulocytes were prepared by treatment of whole blood from healthy volunteers with Dextran T-500 (Pharmacia Biotech Inc., Baie d'Urfé, PC, Canada), followed by centrifugation over Ficoll-Paque (Pharmacia) and removal of remaining red blood cells by hypotonic lysis (18). After washing in phosphate-buffered saline (PBS) containing 137 mM NaCl, 2.7 mM KCl, 1.5 mM KH₂PO₄, and 8.1 mM Na₂HPO₄ at a pH of 7.4, the cells were suspended in PBS at a concentration of 10⁹ cells/ml. Eosinophils were purified from the granulocyte fraction through the use of magnetic beads labeled with a mouse monoclonal antibody to CD16 as described in the literature (19). A suspension of anti-CD16-labeled magnetic microbeads (300 µl; Miltenyi Biotec Inc., Bergisch-Gladbach, Germany) was added to a suspension of granulocytes (4 × 10⁸ cells in 400 µl PBS). After incubation for 30 min at 8°C, PBS (10 ml) was added and the mixture was centrifuged at 200 × g for 10 min. The pellet was resuspended in PBS (2 ml) and the cell suspension was divided into two equal parts; each part was passed through a column containing a steel matrix placed in a permanent magnet (MACS; Miltenyi Biotec). Neutrophils were retained on the column, whereas eosinophils (95 ± 4%; mean ± SD) were obtained in the pass-through fraction. The major contaminating cells were lymphocytes and monocytes. After centrifugation at 200 × g for 10 min, the cells were suspended in HBSS.

Measurement of Surface Expression of CD11b and L-Selectin

Eosinophils (3 × 10⁵ cells in 260 µl) were incubated with agonists (10 µl) in HBSS containing 0.1% BSA for 15 min (CD11b) or 10 min (L-selectin) unless otherwise indicated. The times chosen were those at which the responses to the three agonists were maximal. The time courses for the responses to 5-oxo-ETE, LTB₄, and PAF were quite similar to one another. The incubations were terminated by the addition of Isoton-II medium (2 ml at 0°C; Coulter Corporation, Miami, FL) and centrifugation for 5 min at 400 × g at 4°C. The cells were washed once in ice-cold Isoton-II (1.5 ml) containing 0.1% BSA, centrifuged as described previously, and incubated for 10 min at 4°C with mouse plasma (5 µl). The cells were then incubated for 30 min at 4°C with mouse antihuman FITC-labeled monoclonal antibodies (6 µl; 75 ng) to CD11b or L-selectin, or with the appropriate isotype-matched control FITC-labeled antibody. The eosinophils were washed twice with ice-cold Isoton-II/BSA as described previously and were fixed with formaldehyde (1% wt/vol). The distribution of fluorescence intensities among 10,000 cells was measured for each sample, using an Epics-Profile II instrument (Coulter Corporation). All data were corrected for the value obtained for the corresponding isotype control antibody, which was 0.38 ± 0.02 (IgG_2a) for L-selectin and 0.41 ± 0.01 (IgG₁) for CD11b. The binding of the control antibodies was unaffected by the agonists under investigation.

Measurement of Actin Polymerization

The F-actin content of eosinophils was measured with NBD-phallicidin in a procedure similar to that described by Howard and Oresajo (20). Phallacidin binds strongly to F-actin, the polymerized form of actin, but not to unpolymerized G-actin (20). Eosinophils (3 × 10⁵ cells in 260 µl) were incubated with agonists for 20 s unless otherwise indicated. The cells were fixed by treatment with formaldehyde (30 µl of a 37% solution) at room temperature for 15 min. F-actin was then stained by incubation with lysophosphatidylcholine (30 µg in 15 µl) and NBD-phallacidin (49 pmol in 6.2 µl; final concentration: 0.15 µM) overnight in the dark at 0°C. The cells were then centrifuged at 700 × g for 5 min and resuspended in Isoton-II (0.5 ml). The fluorescence intensity of the stained eosinophils was quantified by flow cytometry using the Epics-Profile II instrument described previously.

Measurement of Intracellular Calcium Levels

Eosinophils (10⁷ cells/ml) were preincubated for 5 min at 37°C in PBS and then incubated with the acetoxymethyl ester of indo-1 (1 µM) for a further 30 min. The indo-1- loaded cells were then washed twice in PBS and resuspended in the same medium to obtain a final cell concentration of 1.07 × 10⁶ cells/ml. Fluorescence was measured at 37°C with a Photon Technology International (Monmouth Junction, NJ) Deltascan 4000 spectrofluorometer with a temperature-controlled cuvette holder equipped with a magnetic stirrer. The excitation and emission wavelengths were set at 331 and 410 nm, respectively. Before the addition of agonists, CaCl₂ and MgCl₂ in 56 µl PBS were added to an aliquot of the eosinophil suspension (934 µl) to give final concentrations of 1 mM of each, a final concentration of cells of 10⁶/ml, and a final volume of 1 ml. Responses to agonists were measured after stabilization of the baseline fluorescence. F_max was determined by adding digitonin (final concentration, 0.1%), whereas F_min was calculated after determination of autofluorescence as described in the literature (21, 22).

Data Analysis

The statistical significance of differences in EC₅₀ values and maximal responses was determined through one-way analysis of variance, with either the Student-Newman- Keuls test or Dunn's test as a multiple-comparison method. Values are presented as means ± SE. Differences with values of P < 0.05 were considered statistically significant.

	Results

Abstract Introduction Materials and Methods Results Discussion References

We investigated the effects of 5-oxo-ETE on a variety of eosinophil responses, including the surface expression of the adhesion molecules L-selectin and CD11b, actin polymerization, and calcium mobilization.

Time Courses for 5-oxo-ETE-Induced Eosinophil Responses

The effects of 5-oxo-ETE (1 µM) on the eosinophil responses discussed previously are shown in Figure 1. 5-oxo-ETE induced very rapid increases in cytosolic calcium levels and F-actin content, both of which reached maximal levels by 5 s. Intracellular calcium levels declined rapidly after stimulation, reaching near-basal levels by about 30 s. On the other hand, 5-oxo-ETE-induced actin polymerization appeared to be biphasic, with a second maximum at 20 s, followed by a subsequent decline to relatively low levels of F-actin by 5 min. LTB₄-induced actin polymerization followed a similar biphasic time course, except that the magnitude of the response was less than that of 5-oxo-ETE (data not shown). 5-oxo-ETE also induced L-selectin shedding and increased the surface expression of CD11b, but the time courses for these responses were slower than those for calcium mobilization and actin polymerization, reaching maximal levels by about 5 to 6 min (L-selectin) and about 9 min (CD11b; data not shown). The times required to reach half the maximal response were about 30 s for L-selectin and about 70 s for CD11b. The time courses for the responses to LTB₄ and PAF were quite similar to those for 5-oxo-ETE (data not shown).

View larger version (19K): [in this window] [in a new window]   Figure 1.   Time courses for the responses of eosinophils to 5-oxo-ETE. Eosinophils (3 × 105 cells in 260 µl or, in the case of calcium measurements, 106 cells in 1 ml) were incubated for various times with 5-oxo-ETE (1 µM). Changes in cytosolic calcium levels (open circles), F-actin content (closed circles), and the surface expression of CD11b (open squares) and L-selectin (closed triangles) were measured as described in MATERIALS AND METHODS. The values are expressed as percentages of the maximum change observed in cells from the same individual. Because maximal calcium and actin responses were not always observed at the same time in all experiments, the maximal responses shown are slightly less than 100%. The maximal response for surface expression of CD11b was observed at 9 min (not shown in the figure). Values are means ± SEM of determinations on cells from five (calcium, F-actin) or three (CD11b, L-selectin) different individuals.

5-oxo-ETE Is a Potent Stimulator of L-Selectin Shedding by Eosinophils

Surface expression of L-selectin on eosinophils was measured after incubation with agonists for 10 min. 5-oxo-ETE induced the shedding of L-selectin from eosinophils with an EC₅₀ of 5.0 ± 1.1 nM (Figure 2). PAF (EC₅₀: 46 ± 23 nM), LTB₄ (EC₅₀: 90 ± 36 nM), and 5-HETE (EC₅₀: 71 ± 37 nM) also stimulated L-selectin shedding, but were all considerably less potent in this respect than 5-oxo-ETE (P < 0.005). The highest concentration (1 µM) of 5-oxo-ETE stimulated the shedding of 87 ± 3% of the L-selectin expressed on the eosinophil surface (Figure 2).

View larger version (25K): [in this window] [in a new window]   Figure 2.   Effects of 5-oxo-ETE, 5-HETE, PAF, and LTB4 on the shedding of L-selectin from human eosinophils. Eosinophils (3 × 105 cells in 260 µl) were incubated for 10 min with 5-oxo-ETE (closed circles; n = 7), 5-HETE (closed inverted triangles; n = 4), PAF (open triangles; n = 7), or LTB4 (open circles; n = 6). Surface expression of L-selectin was measured by flow cytometry using an FITC-labeled monoclonal antibody as described in MATERIALS AND METHODS. The results are expressed as percentages of the mean fluorescent intensities of the control values (0.72 ± 0.07 arbitrary units) obtained after incubation of eosinophils with diluent in the absence of agonist after correction for the isotype control (0.38 ± 0.02 arbitrary units), and are means ± SE of determinations on cells from the numbers of different individuals indicated previously. The inset shows the relative numbers of cells (ordinate) exhibiting different intensities of fluorescence (abscissa) after treatment with vehicle (dotted line) or 5-oxo-ETE (100 nM; solid line).

5-oxo-ETE Increases the Surface Expression of CD11b on Eosinophils

Surface expression of CD11b on eosinophils was measured after incubation with agonists for 15 min. 5-oxo-ETE was a potent activator of CD11b expression, with an EC₅₀ of 9.5 ± 4.3 nM (Figure 3). However, in contrast to the effects of these agonists on L-selectin shedding, LTB₄ (EC₅₀: 1.3 ± 0.4 nM) was somewhat more potent than both 5-oxo-ETE and PAF (EC₅₀: 4.4 ± 1.0 nM) (P < 0.05). The maximal response to LTB₄ appeared to be slightly weaker than those to 5-oxo-ETE and PAF, but this difference was not statistically significant. It is possible that maximal responses to the latter two agonists were not reached in the current study because we did not use concentrations exceeding 1 µM.

View larger version (24K): [in this window] [in a new window]   Figure 3.   Effects of 5-oxo-ETE, PAF, and LTB4 on the surface expression of CD11b by human eosinophils. Eosinophils (3 × 105 cells in 260 µl) were incubated for 15 min with 5-oxo-ETE (closed circles), PAF (open triangles), or LTB4 (open circles). Surface expression of CD11b was measured by flow cytometry using an FITC-labeled monoclonal antibody as described in MATERIALS AND METHODS. The results are expressed as percent increases in the mean fluorescent intensities above the control value (5.0 ± 0.4 arbitrary units) obtained in the absence of agonist, and are means ± SE of determinations on cells from seven (5-oxo-ETE), four (PAF), or six (LTB4) different individuals. The maximal response to 5-oxo-ETE in these experiments was 6.7 ± 0.6 arbitrary units. The inset shows the relative numbers of cells (ordinate) exhibiting different intensities of fluorescence (abscissa) after treatment with vehicle (dotted line) or 5-oxo-ETE (100 nM; solid line).

Actin Polymerization in Eosinophils Is Strongly Enhanced by 5-oxo-ETE

The effects of different concentrations of agonists on actin polymerization were determined after incubation for 20 s, the time at which these responses were maximal. As shown in Figure 4, LTB₄ had an EC₅₀ value (0.6 ± 0.2 nM) lower than those of both PAF (EC₅₀: 6.6 ± 5.2 nM) and 5-oxo-ETE (EC₅₀: 6.3 ± 0.6 nM) (P < 0.02). However, the maximal response to 5-oxo-ETE was more than 2.5 times greater than that to the other two agonists (P < 0.002).

View larger version (26K): [in this window] [in a new window]   Figure 4.   Effects of 5-oxo-ETE, PAF, and LTB4 on actin polymerization in human eosinophils. Eosinophils (3 × 105 cells in 260 µl) were incubated for 20 s with 5-oxo-ETE (closed circles), PAF (open triangles), or LTB4 (open circles). Polymerized actin (i.e., F-actin) was measured by flow cytometry after the cells were fixed with formaldehyde and treated with NBD-phallacidin in the presence of lysophosphatidylcholine as described in MATERIALS AND METHODS. The results are expressed as percent increases in mean fluorescent intensities compared with control values (26 ± 4 arbitrary units) obtained after incubation of eosinophils with diluent in the absence of agonist. The values are means ± SE of determinations on cells from six (5-oxo-ETE), three (PAF), or four (LTB4) different individuals. The maximal response to 5-oxo-ETE was 39 ± 5 arbitrary units. The inset shows the relative numbers of cells (ordinate) exhibiting different intensities of fluorescence (abscissa) after treatment with vehicle (dotted line) or 5-oxo-ETE (100 nM; solid line).

5-oxo-ETE-Induced Actin Polymerization Is Subject to Homologous But Not Heterologous Desensitization

We have previously shown that 5-oxo-ETE-induced calcium mobilization in neutrophils can be prevented by prior treatment with this agonist, but not with LTB₄ (14). To determine whether this was also true for 5-oxo-ETE- induced actin polymerization in eosinophils, these cells were preincubated for 5 min with various concentrations of 5-oxo-ETE, PAF, or LTB₄, followed by treatment for a further 20 s with 5-oxo-ETE (100 nM) and measurement of F-actin (Figure 5). Preincubation of eosinophils with 1 µM 5-oxo-ETE for 5 min nearly completely prevented the response to subsequent addition of this agonist. The IC₅₀ for this effect was 120 nM. In contrast, preincubation of eosinophils with PAF or LTB₄ had no detectable effect on the response to 5-oxo-ETE.

View larger version (23K): [in this window] [in a new window]   Figure 5.   Desensitization of eosinophils to 5-oxo-ETE-induced actin polymerization. Eosinophils (3 × 105 cells in 260 µl) were preincubated for 5 min with either 5-oxo-ETE (closed circles), PAF (open triangles), or LTB4 (open circles) prior to the addition of 5-oxo-ETE (100 nM) and incubation for a further 20 s. F-actin was then measured by flow cytometry as described in the legend to Figure 4. The results are expressed as percentages of the responses to 5-oxo-ETE (100 nM) after preincubation for 5 min with vehicle alone, and are means ± SE of determinations on four (5-oxo-ETE and LTB4) or three (PAF) different individuals. The baseline response in the absence of any agonist was 22 ± 5 arbitrary units, whereas the response to 100 nM 5-oxo-ETE was 33 ± 6 arbitrary units.

Effects of Kinase Inhibitors on 5-oxo-ETE-Induced Actin Polymerization

We tested the effects of various inhibitors on 5-oxo-ETE- induced actin polymerization. Eosinophils were preincubated for 20 min with staurosporine (25 nM), an inhibitor of PKC, PD98059 (50 µM), an inhibitor of mitogen-activated protein (MAP) kinase kinase, or wortmannin (25 nM), an inhibitor of phosphatidylinositol (PI)-3-kinase. These concentrations of staurosporine (13), PD98059 (23), and wortmannin (24, 25) have been reported to inhibit strongly the aforementioned enzymes. 5-oxo-ETE was then added and F-actin levels were measured 20 s later. Although staurosporine (P < 0.05) and wortmannin (P = NS) appeared to have slight inhibitory effects on 5-oxo-ETE-induced actin polymerization, none of the inhibitors tested had pronounced effects on this phenomenon (Figure 6).

View larger version (26K): [in this window] [in a new window]   Figure 6.   Effects of various inhibitors on 5-oxo-ETE-induced actin polymerization in eosinophils. Eosinophils were preincubated at 37°C for 20 min with either vehicle (Control), staurosporine (Ssp: 25 nM), PD98059 (PD; 50 µM), or wortmannin (Wort; 25 nM). 5-oxo-ETE (100 nM; hatched bars) or vehicle (solid bars) was then added and the incubation continued for a further 20 s. The cells were then fixed and F-actin content determined as described in the legend to Figure 4. The values are means ± SE of determinations on eosinophils from five different individuals. The baseline response in the absence of any agonist or inhibitor was 21 ± 4 arbitrary units, whereas the response to 100 nM 5-oxo-ETE was 30 ± 6 arbitrary units; *P < 0.05 compared with the increase in polymerized actin induced by 5-oxo-ETE in control cells without staurosporine.

5-oxo-ETE Induces Calcium Mobilization in Eosinophils

Concentrations of 5-oxo-ETE as low as 1 nM stimulated calcium transients in eosinophils (Figure 7). However, the EC₅₀ (19 ± 2 nM) for this response was somewhat higher than those for the other responses to this agonist investigated in the present study. The concentration-response for LTB₄ (EC₅₀: 71 ± 38 nM) appeared to be quite similar to that for 5-oxo-ETE, although there was considerable variability among the EC₅₀ values for individual experiments. In contrast, PAF (EC₅₀: 4.6 ± 2.0 nM) was both more potent (P < 0.02) and more efficacious (P < 0.01) than 5-oxo-ETE and LTB₄, inducing a maximal response that was two to three times greater than that for the latter two agonists. The calcium response to 5-oxo-ETE was quite reproducible among donors, as can be seen by the small error bars in Figure 7. In contrast, there was a high degree of variability in LTB₄-induced calcium responses among donors. Of the five donors tested, the maximal responses to LTB₄, expressed as percentages of the maximal responses to 5-oxo-ETE by cells from the same donor, were 16, 81, 116, 194, and 322%. There was much less variability in the response to PAF than to LTB₄, since the maximal responses elicited by PAF ranged from 220% to 330% of the maximal response to 5-oxo-ETE for eosinophils from the same individual.

View larger version (17K): [in this window] [in a new window]   Figure 7.   Effects of 5-oxo-ETE, PAF, and LTB4 on calcium mobilization in human eosinophils. Various concentrations of 5-oxo-ETE (closed circles), PAF (open triangles), or LTB4 (open circles) were added to indo-1-loaded eosinophils (10 6 cells/ml), and changes in fluorescence were monitored as described in MATERIALS AND METHODS. The data are expressed as maximal increases in the concentration of intracellular calcium, which were usually observed approximately 5 s after the addition of agonist. The points are the means ± SE of determinations on eosinophils from five different individuals. Baseline cytosolic calcium levels before stimulation were 142 ± 5 nM.

	Discussion

Abstract Introduction Materials and Methods Results Discussion References

We have previously shown that 5-oxo-ETE is a potent eosinophil chemotaxin (9), and this has been confirmed by other laboratories (26, 27). Furthermore, preliminary experiments that we have conducted suggest that 5-oxo-ETE is also active in vivo, inducing eosinophil infiltration into the lungs of Brown Norway rats (16). In the present investigation we examined some of the early responses of eosinophils to 5-oxo-ETE that may be involved in tissue infiltration of these cells. We found that 5-oxo-ETE is a potent stimulator of the surface expression of CD11b and of L-selectin shedding, as well as of calcium mobilization and actin polymerization in human eosinophils.

There are considerable differences in the relative abilities of 5-oxo-ETE, PAF, and LTB₄ to elicit different eosinophil responses. 5-oxo-ETE is the most effective chemoattractant, with LTB₄ exhibiting only very weak activity and PAF having an intermediate effect (9). In agreement with this, we found that 5-oxo-ETE is more active than PAF and LTB₄ in stimulating actin polymerization and the shedding of L-selectin. In contrast, PAF is considerably more effective than either 5-oxo-ETE or LTB₄ in stimulating calcium mobilization in eosinophils. Although it has only very weak chemotactic effects on human eosinophils, LTB₄ is more potent than either 5-oxo-ETE or PAF in stimulating the surface expression of CD11b on these cells. All three of these lipid mediators can be released from leukocytes by the same mechanism. Cytosolic phospholipase A₂ can hydrolyze 1-O-alkyl-2-arachidonoyl-glycerophosphocholine to give both free arachidonic acid and lyso-PAF (28). Arachidonic acid can then be converted via 5-lipoxygenase to 5-oxo-ETE (12) and LTB₄ (29), whereas lyso-PAF can be converted to PAF by 1-alkyl glycerophosphocholine acetyltransferase (30). It is therefore conceivable that these three mediators could be released at the same time and then work in concert to induce the accumulation of eosinophils at inflammatory sites through their combined effects on L-selectin-mediated tethering and rolling, tight adhesion, and cell movement.

The maximal increase in surface expression of CD11b in response to 5-oxo-ETE was about 40% above control in eosinophils, compared with about 140% in neutrophils (15). It is possible that this difference was due to the longer procedure required for the purification of eosinophils, which consisted of Dextran sedimentation, centrifugation over Ficoll-Paque, hypotonic lysis, and immunomagnetic separation. The procedure previously used for neutrophils consisted only of Dextran sedimentation (15). Thus, the eosinophils may have been in a more active state than the neutrophils. However, the maximal increase in actin polymerization in response to 5-oxo-ETE was similar (between 50 and 60% above control) in eosinophils and neutrophils.

It was recently reported that 5-oxo-ETE mobilizes calcium (26) and stimulates actin polymerization in eosinophils, but the dose-response relationship for 5-oxo-ETE was not compared with those of PAF and LTB₄ (31). The investigators who made this finding also found that 5-oxo-ETE induced oxygen radical formation in these cells (31). Degranulation of cytochalasin B-treated eosinophils in response to 5-oxo-ETE was observed by O'Flaherty's group (27), although this has not been found in all studies (26). However, 5-oxo-ETE was not as potent as LTB₄ in inducing this response (27).

The relative abilities of 5-oxo-ETE and LTB₄ to stimulate eosinophil responses are quite different from their abilities to stimulate the corresponding responses in neutrophils. Although the maximal responses to the two agonists are similar, LTB₄ is 30 to 100 times more potent than 5-oxo-ETE in inducing a variety of responses in neutrophils, including chemotaxis (14, 26, 27), actin polymerization (15), surface expression of CD11b (15), shedding of L-selectin (W. S. Powell, S. Gravel, and F. Halwani, unpublished work), and adherence (15). LTB₄ is about 10 times more potent than 5-oxo-ETE in stimulating calcium mobilization in these cells (14). The basis for the different responses of eosinophils and neutrophils to these two eicosanoids is not clear. It would appear that receptors for both agents are present on both cell types, but the relative numbers of receptors may differ. It is also possible that there are differences in the natures of the G proteins that are coupled to these receptors in the two types of cells.

Changes in cytoskeletal structure due to polymerization and depolymerization of actin are very important for the regulation of cell movement and hence chemotaxis (32). Actin polymerization is generally considered to be a calcium-independent process (32). In agreement with this, C5a and PAF-induced shape changes in human eosinophils, which are dependent on actin polymerization, cannot be blocked by calcium depletion with Quin-2 (33). On the other hand, it has recently been reported that 1,2-bis(2-aminophenoxy)ethane-N-N-N'-N'-tetraacetic acid can block actin polymerization in response to PAF, RANTES, and C5a in eosinophils, despite its lack of this effect in neutrophils (34). The reason for this discrepancy is not clear. In the present study we found that 5-oxo-ETE was more active than PAF in stimulating actin polymerization, despite its lower potency in inducing calcium mobilization. Thus it would appear that these two responses can be at least partly dissociated in some cases. The chemoattractant properties of 5-oxo-ETE on eosinophils could be at least partly related to its potent effects on actin polymerization, which could explain why it is a more effective chemoattractant for these cells than is either PAF or LTB₄.

Although calcium mobilization does not appear to be sufficient to induce granulocyte migration, it does appear to contribute to this phenomenon. The chemotactic response of neutrophils to a formyl peptide (formyl-norleucyl-leucylphenylalanine) was partly blocked by calcium depletion (35). Neutrophils migrating on fibronectin- or vitronectin-coated surfaces require intracellular calcium transients to enable the cells to detach from the substrate, thus permitting their further movement (36). Chelation of intracellular calcium has also been reported to block chemotaxis in eosinophils (34). The potent effects of PAF on calcium mobilization in eosinophils could therefore be important for its chemotactic effect on eosinophils. The fact that 5-oxo-ETE is more active than PAF in stimulating actin polymerization, whereas the reverse is true for calcium mobilization, could explain the synergistic effect that we previously observed for the effects of these two mediators on eosinophil migration (9).

The effects of 5-oxo-ETE on CD11b and L-selectin expression suggest that this substance is involved in the interaction of eosinophils with endothelial cells, and their transmigration into tissues. ₂-integrins, including CD11b, have been shown to be important for the interaction of eosinophils with vascular endothelial cells and transendothelial migration (37). L-selectin promotes eosinophil rolling over endothelial cells (38) and may be particularly important for the initial tethering of leukocytes to these cells (39). Monoclonal antibodies to L-selectin have been shown to prevent lipopolysaccharide-induced pleural accumulation of eosinophils in mice (40), further supporting a role for this molecule in eosinophil infiltration. That 5-oxo-ETE is a potent stimulator of L-selectin shedding suggests that it induces a conformational change in this adhesion molecule, resulting in exposure of an extracellular site on L-selectin to a protease on the cell surface (41, 42). Although the question was not addressed in the present study, it is conceivable that 5-oxo-ETE can also activate L-selectin, resulting in enhanced interaction with endothelial cell ligands, as has been shown for other chemoattractants in other cell types (43, 44). In agreement with our findings on the effects of 5-oxo-ETE on eosinophil adhesion molecules, we have preliminary evidence that administration of this substance directly into the airways of Brown Norway rats results in marked pulmonary eosinophilia (16).

We and others have previously shown that 5-oxo-ETE activates granulocytes by a pertussis toxin-sensitive mechanism (45), suggesting that it interacts with a G protein-coupled receptor. Although its intracellular signaling mechanism is not well understood, 5-oxo-ETE has been reported to stimulate MAP kinase in neutrophils (47) and eosinophils (27), and there is evidence that it stimulates PI-3-kinase in neutrophils (46). We investigated the effects of various kinase inhibitors on 5-oxo-ETE-induced actin polymerization. Actin polymerization can be stimulated via several intracellular signaling pathways, including activation of PKC (48, 49) and PI-3-kinase (49). However, we were unable to prevent 5-oxo-ETE-induced actin polymerization with either the PKC inhibitor staurosporine or the PI-3-kinase inhibitor wortmannin. This is consistent with the findings of others that chemoattractant-induced actin polymerization is not blocked by inhibitors of PKC (50) or PI-3-kinase (51) in neutrophils. It has also been shown that C5a-induced shape changes in eosinophils, which are dependent on actin polymerization, are not blocked by staurosporine (33).

In conclusion, 5-oxo-ETE is the most effective of the lipid mediators we have investigated in inducing eosinophil chemotaxis (9), actin polymerization, and shedding of L-selectin. This eicosanoid may be an important mediator of eosinophil infiltration into tissues such as the lung, and could thus play an important proinflammatory role in asthma.