INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The objective of this investigation was to assess the predominant events associated with corticosteroid-induced inhibition of eosinophil secretion. Prior investigations have demonstrated that both systemic (1, 2) and topical (3, 4) corticosteroid decreased the number of eosinophils in the airways of humans with asthma. Corticosteroids also are known to cause eosinophil death (5), which is thought to occur predominantly through cellular apoptosis (1, 8). However, prior investigations have provided indirect evidence to suggest that corticosteroids also may impair the activity of cellular phospholipase, which in turn prevents the cleavage of arachidonic acid (AA) from the sn-2 position of phospholipids (9, 10). Previous investigations also have indicated that the translocation of the 85 kDa cytosolic phospholipase A₂ (cPLA₂) to the perinuclear membrane of the eosinophil is an essential step in leukotriene production (11, 12). At this site, cPLA₂ in association with 5-lipoxygenase activating protein converts AA first to 5-hydroxyperoxyeicosatetraenoic acid and then to leukotriene A₄ (LTA₄), the precursor for LTC₄ (13, 14).

In this investigation, we used eosinophils isolated from mildly atopic humans to assess the time course in vitro and events by which corticosteroids may inhibit stimulated secretion of LTC₄, a potent bronchoconstrictor in human asthma (14, 15). The use of mildly atopic donors allowed for collection of greater number of eosinophils and permitted within-donor paradigms for all components of each experimental algorithm. No donor had taken anti-allergic or anti-asthmatic drugs for > 3 mo prior to entry in the study. We have shown previously that the stimulated secretory properties of eosinophils from these mildly atopic donors do not differ from nonatopic control subjects (16). Cellular viability was assessed first by direct cell count as a fraction of the initial number of eosinophils incubated over 48 h. Human recombinant interleukin-5 (rhIL-5) was added in quantity sufficient to maximize cellular survival over this period (17, 18), and the ability of increasing concentrations of fluticasone propionate (FP) to promote decreased eosinophil survival was assessed by trypan blue exclusion and propidium iodide (PI) staining (19). FP was selected because it is an extremely lipophilic corticosteroid that has the greatest affinity for the internal glucocorticoid receptor of any drug in its class (20). Studies also were performed to determine if decreased LTC₄ secretion caused by FP resulted solely from eosinophil death, or if attenuated stimulated secretion of LTC₄ occurred independently of changes in eosinophil number and viability. We found that cellular secretion of LTC₄ decreased progressively in a time- and concentration-related fashion during exposure to FP. However, even after correction for cellular viability, LTC₄ secretion still was substantially reduced by treatment with FP. This decreased secretion was independent of the extent of cellular apoptosis, which was not substantial, but corresponded in concentration-related fashion to impaired translocation of cPLA₂ to the nuclear envelope of the eosinophil (11, 21).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Peripheral Blood Eosinophil Isolation

Human peripheral blood eosinophils (PBE) were purified by a modification of the negative immunomagnetic separation technique (22). Peripheral blood (120 ml) was obtained from mildly atopic donors in accordance with an approved institutional protocol. PBEs were isolated by using an antibody against CD16, a surface receptor found on neutrophils but not eosinophils (22, 23). The final cell preparation routinely consisted of > 98% eosinophils as examined by Wright- Giemsa staining.

Evaluation and Analysis of Cell Survival

Trypan blue exclusion analysis. Cytocentrifuge preparations were made with a Shandon Cytospin Model 3 (Shandon, Pittsburgh, PA) and treated cells were examined after Wright-Giemsa stain using light microscopy. Viability was assessed by dye exclusion of 0.4% trypan blue (dipotassium phosphate-sodium chloride). Eosinophils suspended for 5 min by gentle agitation were diluted with Hanks' balanced salt solution (HBSS) containing 0.1% gelatin, and the number of cells stained by the dye was counted by light microscopy. Viability of eosinophils was calculated as a percentage = 

All experiments were performed using sterile microplate wells to preserve the viability of eosinophils during the pretreatment and activation periods.

PI staining. Treated eosinophils (5 × 10⁵ cells) were stained with hypotonic PI (50 µg/ml PI, 0.1% Triton X-100, and 0.1% sodium citrate in double distilled water) overnight, and samples were analyzed on a Becton Dickinson FACScan (San Jose, CA). To collect all cell fragments of PI single-stained samples (acquisition of 10,000 events), the forward scatter threshold was set on range 0-1,000. The fluorescence analysis refers to all acquired fragments (19).

Terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay. Cell suspensions (5 × 10⁵ eosinophils) from all experimental groups were pelleted by centrifugation at 1,000 × g for 10 min at 4° C. Immediately, cells were fixed in 1% paraformaldehyde in phosphate-buffered saline (PBS) and incubated at room temperature for 30 min. After centrifugation, the fixed cells were permeabilized with 0.74% n-octyl--D-glucopyranoside (OG) in 10 mM PBS for 10 min at room temperature. Immediately after washing with labeling buffer, a mixture of 1:50 terminal deoxytransferase deoxyribonucleoside triphosphate (TdT dNTP), 1:50 TdT, and 1:50 Mn⁺⁺ (Trevigen Apoptotic Cell System, Gaithersburg, MD) was added to the treated cells and allowed to incubate for an additional 60 min at 37° C. The reaction mixture was terminated by addition of stop buffer and followed by 25 µl of streptavidin-fluorescein detection solution. After final incubation, the cells were washed with labeling buffer and stored in PBS until analysis. The apoptotic eosinophils were evaluated by Becton Dickinson FACScan (San Jose, CA).

Fluorescence-activated cell sorter (FACS) analysis. The annexin-1 (lipocortin-1) expression was analyzed by single-color indirect immunofluorescence with 10 µg/ml unlabeled monoclonal antibody (mAb) directed against annexin-1. Eosinophils (5 × 10⁵ cells/intervention) were fixed in 1% paraformaldehyde in PBS and stained with anti- annexin-1 mAb or isotype control for 30 min at 4° C. Primary mAb was detected with phycoerythrin (PE)-conjugated goat anti-mouse Ig as previously described (19). The fluorescence of 5 × 10⁴ cells was measured on a FACScan (Becton Dickinson, Bedford, MA), and histograms were generated as mean channel fluorescence (x-axis) versus relative to cell number (y-axis).

Verification of Cell Activation

LTC₄ secretion by enzyme immunoassay. Stimulated secretion of LTC₄ caused by eosinophil activation with either vehicle control or 10⁶ M formyl-methionine-leucine-phenylalanine plus 5 µg/ml cytochalasin B (FMLP/CB) was measured by enzyme immunoassay using reagents obtained from Cayman Chemical (Ann Arbor, MI). This concentration of FMLP/CB has been shown previously to cause maximal secretion of LTC₄ (24), degranulation of eosinophil protein (16, 24), and induction of metabolic burst activity (25), as well as substantial eosinophil migration into the airway lumen in guinea pig tracheal explants (26, 27). In separate studies, secretion of LTC₄ also was quantitated in cells treated with increasing concentrations of FP at different time intervals (see protocol below). All experimental data were meaned from duplicate runs for each intervention, and mean values were expressed as picogram LTC₄ per million eosinophils (pg/10⁶ cells). The final value of LTC₄ secretion was normalized by dividing the actual LTC₄ value by the fraction of total cell number × the fraction of viable cells obtained after PI staining. Secretion was thus expressed as total secreted LTC₄ /total eosinophil number × fraction viable (PI determination).

Localization of cPLA₂ by immunohistochemistry. Cytosolic PLA₂ was localized by fixing the treated cells with 2% paraformaldehyde in PBS for 20 min; standard cytoslides then were prepared (28). IgG₁ antibody was used as a negative control for all subsequent studies. Slides then were washed with PBS buffer at pH 7.40. Slides containing samples were stained with mAb directed against cPLA₂ and processed further according to the Vectastain ABC instruction kit (VECTOR Laboratories, Burlingame, CA). The color reaction product was allowed to develop for 30 min at room temperature. Slides were immediately viewed by light microscopy to assess localization of cPLA₂.

Activation of isolated human PBE. Stimulated LTC₄ secretion was elicited with FMLP/CB in baseline studies using 10 cell isolations from 10 separate donors (16, 24). PBE (5 × 10⁵ cells) were resuspended in 500 µl complete RPMI buffer containing 10 pg/ml rhIL-5 for 0.5 h, 24 h, and 48 h. RhIL-5 was added to maintain eosinophil viability during the experimental period (17). The cell mixture was pipetted onto the sterile microplate wells and placed in an incubator at 37° C (95% O₂, 5% CO₂). The cell suspension was activated with either buffer control or FMLP/CB for 30 min at 37° C, and the reaction was terminated by centrifugation (1,350 rpm × 5 min). The supernatant was collected and saved at 70° C until analysis for LTC₄ release. Secretion of LTC₄ was expressed as picogram LTC₄ per million eosinophils (pg/10⁶ cells). The pellet was processed further to assess the cell survival using either trypan blue exclusion analysis or PI stain (see the previously stated protocols).

Effect of FP on LTC₄ secretion. To determine the effect of FP on stimulated eosinophil secretion of LTC₄, aliquots of 2.5 × 10⁵ cells from the same 10 subjects used in the aforementioned protocol were pretreated with 10¹⁰ M to 10⁶ M FP for 0.5 h, 24 h, and 48 h. At the end of the incubation time, the treated cells were activated with either buffer or FMLP/CB. The reaction mixture was terminated by centrifugation (Model GS6KR; Beckman, Palo Alto, CA), and the supernatants were collected and stored at 70° C until assayed for stimulated LTC₄ release. The FP-treated pellet (no FMLP/CB) was processed further to assess the cell survival using trypan blue exclusion analysis or PI stain (see above protocols). These data were used to normalize the LTC₄-secretion as a function of the number of viable cells.

Because trypan blue dye and PI staining do not distinguish apoptosis from death resulting from necrosis, identical experiments as above were conducted to quantitate the number of apoptotic eosinophils using TUNEL assay as described previously (n = 5 different eosinophil isolations from 5 different donors). This assay specifies apoptotic eosinophils as defined by morphology, wherein cell shrinkage was evident under light microscopy.

Effect of FP on PLA₂ translocation. The effect of FP on the translocation of cPLA₂ to the nuclear membrane as assessed by immunohistochemistry was examined in 4 separate eosinophil isolations from 4 separate donors (28). Experiments were conducted using eosinophils coincubated with 10¹⁰ M to 10⁶ M FP for 0.5 h, 24 h, and 48 h prior to FMLP/CB activation. Cytoslides were fixed in 2% paraformaldehyde solution and treated with specific antibody that recognizes cPLA₂ in treated eosinophils. These stained cells then were viewed under light microscopy to confirm the intracellular translocation of cPLA₂ to the perinuclear membrane of eosinophils.

Effect of exogenous AA on FP-treated eosinophils. To determine whether FP inhibited normalized LTC₄ synthesis caused by FMLP/ CB specifically by inhibiting PLA₂, stimulated LTC₄ secretion was assessed in cells incubated with 1 µM AA prior to treatment with 10⁶ M FP. After 15 min, the cell mixture was activated with FMLP/ CB and the supernatant was collected and stored at 70° C for analysis of LTC₄. Buffer containing either no FP or no AA was added to eosinophils and used as control for all groups studied.

Effect of FP on the surface membrane expression for annexin-1 (lipocortin-1). Experiments were performed to determine whether inhibition of LTC₄ synthesis resulted specifically from selective inhibition of LTC₄ synthesis or from global inhibition of eosinophil synthetic functions caused by treatment with FP. Membrane expression of annexin-1 thus was measured in eosinophils treated with increasing concentrations of FP at different time intervals as previously specified. Treated eosinophils (5 × 10⁵ cells) were stained with mAb directed against annexin-1, and the surface membrane expression for annexin-1 was analyzed by FACScan as previously outlined.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of FP on Eosinophil Survival and Apoptosis

Incubation of eosinophils at 37° C during the experimental period with rhIL-5 prevented decrease in total cell number. There were no changes from initial cell numbers with any concentration of FP at 24 h and only a minimal decrease from 2.5 × 10⁵ total cells per well to 2.3 × 10⁵ cells per well for cells treated with 10⁶ M FP after 48 h (Figure 1). By contrast, cell death resulted from treatment by FP, despite incubation with rhIL-5 (Figure 2). There was complete survival of eosinophils at 30 min for all concentrations of FP. However, by 24 h, survival of eosinophils incubated with rhIL-5 as measured by trypan blue exclusion decreased from 85 ± 4.5% (no FP) to 69 ± 5.9% for eosinophils treated with rhIL-5 + 10⁶ M FP (p < 0.05) (Figure 2). At 48 h, viability for cells treated with rhIL-5 alone decreased to 73 ± 6.8% and decreased to 41 ± 5.2% after incubation with 10⁸ M FP (p < 0.004) (Figure 2).

View larger version (53K): [in this window] [in a new window]   Figure 1.   Effect of FP on eosinophil number at different time intervals. Eosinophils were treated with increasing concentrations of FP in complete RPMI 1640 buffer + 10 pg/ml rhIL-5, and cell number for all experimental interventions was assessed by light microscopy after Wright-Giemsa stain. The initial cell number (2.5 × 105 eosinophils) remained unchanged over a 24-h period; a minimal reduction in cell number was observed after 48 h with  108 M FP.

View larger version (43K): [in this window] [in a new window]   Figure 2.   Effect of FP on eosinophil viability as assessed by trypan blue exclusion analysis. Eosinophils were treated with increasing concentrations of FP in complete RPMI 1640 + 10 pg/ml rhIL-5 at different time intervals. Viability of cells decreased 24 h after treatment with 106 M FP and decreased further at 48 h with  108 M FP despite coincubation with rhIL-5. (*p < 0.05, **p < 0.004 compared with vehicle control.)

Similar results were obtained for cells stained with PI after treatment with FMLP/CB (Figure 3). At 24 h, eosinophil survival was 58 ± 3.6% for cells treated with rhIL-5 + 10⁶ M FP (p < 0.01 versus time-matched buffer control receiving rhIL-5 alone). At 48 h, eosinophil survival was 37 ± 3.3% for cells treated with rhIL-5 + 10⁸ M FP.

View larger version (42K): [in this window] [in a new window]   Figure 3.   Effect of FP on eosinophil viability as assessed by PI staining. Eosinophils were treated with increasing concentrations of FP in complete RPMI 1640 + 10 pg/ml rhIL-5 at different time intervals. Comparable effect was observed for treated eosinophils as for trypan blue exclusion (Figure 2). This analysis was performed using flow cytometry to quantitate cell death occurring after FP treatment (see METHODS). (* p < 0.01, ** p < 0.02, compared with vehicle control.)

Cell death caused by FP did not result predominantly from apoptosis. At 48 h > 62% of all cells exposed to rhIL-5 and 10⁸ M FP had died as measured by both trypan blue exclusion (Figure 2) or PI (Figure 3). By contrast, at 10⁸ M FP, only 23% of these cells had undergone programmed cell death by apoptosis as measured by TUNEL assay by 48 h (Figure 4).

View larger version (17K): [in this window] [in a new window]   Figure 4.   Effect of FP on eosinophil viability as assessed by TUNEL assay. Eosinophils were treated as for trypan blue and PI staining, and apoptotic cells were determined by TUNEL assay. Survival of eosinophils was sustained at early incubation time with FP but decreased substantially at 48 h in concentration-related fashion. (*p < 0.01, **p < 0.001 versus vehicle control without FP.)

Effect on Stimulated LTC₄ secretion

Substantial inhibition of stimulated LTC₄ secretion occurred after 48 h and was related in both FP-treated and control cells to cell death (Figure 5). When normalized for cell survival as measured by PI (see METHODS: VERIFICATION OF CELL ACTIVATION), inhibition of FMLP/CB-induced secretion of LTC₄ was unchanged at 24 h (Figure 6). However, at 48 h, LTC₄ synthesis in activated eosinophils treated with  10⁸ M FP was substantial (Figure 6). FMLP/CB caused LTC₄ secretion of 1,498 ± 180 pg/10⁶ cells in eosinophils treated with rhIL-5 alone at 48 h versus 762 ± 113 pg/10⁶ cells for eosinophils treated with 10⁸ M FP for the same time (Figure 6).

View larger version (17K): [in this window] [in a new window]   Figure 5.   Effect of FP on stimulated, non-normalized secretion of LTC4. Stimulated secretion is greater for both untreated (*p < 0.002) and FP-treated eosinophils at 24 h versus 48 h because of the larger number of viable cells at 24 h. Significant difference in the normalized LTC4 secretion of viable cells occurs at 48 h (see Figures 3 and 6).

View larger version (15K): [in this window] [in a new window]   Figure 6.   Effect of FP on normalized stimulated secretion of LTC4. FP-treated eosinophils were activated with 106 M formyl-met-leu-phe + 5 µg/ml cytochalasin B (FMLP/CB), and LTC4 secretion (Figure 5) was normalized for cell survival as measured by PI staining (see Figure 3). Comparable basal stimulated secretion of LTC4 was quantified at 24 h and 48 h in cells treated with complete RPMI 1640 buffer + 10 pg/ml rhIL-5. FP did not inhibit the LTC4 secretion at 24 h, but normalized secretion in viable eosinophils decreased substantially for cells treated with  108 M FP at 48 h. *p < 0.003, **p < 0.007 compared with FMLP/CB, no FP.

Reversal of FP-induced Inhibition of LTC₄ Synthesis by Exogenous AA

Exposure of cells to buffer alone had no effect on the normalized synthesis of LTC₄ (19.6 ± 3.10 pg/10⁶ cells; Figure 7). Addition of exogenous AA alone, also had no effect on normalized LTC₄ synthesis (33.1 ± 10.6 pg/10⁶ cells; p = NS versus buffer alone). FMLP/CB alone (no FP, no AA) caused LTC₄ secretion of 804 ± 194 pg/10⁶ cells; treatment with 10⁸ M FP caused a decrease in stimulated LTC₄ synthesis to 242 ± 60.4 pg/10⁶ eosinophils (p < 0.001). Addition of AA after activation with FMLP/CB elicited a significant reversal of inhibited LTC₄ synthesis by FP (663 ± 197 pg/10⁶ cells after AA; p < 0.01 versus FP + no AA; p = NS versus FMLP without FP) (Figure 7).

View larger version (19K): [in this window] [in a new window]   Figure 7.   Effect of exogenous AA on inhibition of LTC4 synthesis by FP at 48 h. Data are normalized for cell viability. A significant reversal of stimulated release of LTC4 was demonstrated for FP-treated eosinophils after addition of 1 µM AA indicating that the results are mediated at the level of the PLA2 pathway.

Effect of Treatment with FP on Intracellular Perinuclear Translocation of cPLA₂

Immunohistochemical staining revealed definite translocation of cPLA₂ to the perinuclear membrane after activation with FMLP/CB (Figure 8). Inhibition of this translocation was observed for cells treated with increasing concentrations of FP. All cells stained for cPLA₂ first were shown to be viable eosinophils. Hence, FP causes inhibition of cPLA₂ translocation prior to cell death or apoptosis, and this corresponded to inhibited secretion of LTC₄ as measured in viable cells caused by the same concentrations of FP.

View larger version (37K): [in this window] [in a new window]   Figure 8.   Effect of FP on stimulated translocation of cPLA2 to perinuclear membrane in human eosinophils. Eosinophils were treated as described in METHODS and cytoslides were stained with mAb directed against cPLA2. (A) Immunohistochemical staining identified cPLA2 translocation to the nuclear membrane for cells activated with FMLP/CB. (B) Pretreatment with 108 M FP, a concentration having a significant effect on stimulated LTC4 release, inhibited the intracellular cPLA2 translocation to perinuclear membrane. (C ) IgG1 was used as negative control in all studied groups (n = 4 eosinophil donors).

Annexin-1 Expression after FP Treatment

Despite progressive inhibition of LTC₄ synthetic capacity caused by FP, FP caused very substantial time- and concentration-related synthesis and membrane expression of annexin-1 on eosinophils. Mean fluorescence intensity (MFI) for annexin-1 on the eosinophil surface was 101 ± 25 after 30 min treatment with 10⁶ M FP. However, eosinophils exposed to FP for 24 h had annexin-1 expression of 310 ± 90 MFI; at 48 h, MFI was > 600 for concentrations of FP  10⁸ M (p < 0.001 for both 24 h and 48 h versus 30 min) (Figure 9).

View larger version (28K): [in this window] [in a new window]   Figure 9.   Surface expression of annexin-1 (lipocortin-1) in FP-treated eosinophils. Cells were treated with either vehicle control or increasing concentrations of FP for 30 min to 48 h. The surface expression of annexin-1 was measured by FACScan using specific mAb directed against annexin-1 and its isotype-matched control IgG1. Despite inhibition of LTC4 synthesis (Figure 6) and cPLA2 translocation (Figure 8), annexin-1 membrane expression increased with both time and concentration of FP (*p < 0.001 versus 30 min).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This investigation was undertaken to assess the sequence of events by which a highly lipid-soluble corticosteroid with high affinity for the intracellular glucocorticoid receptor inhibited the production of LTC₄. This eicosanoid is the most bioactive of all leukotrienes in human airways (15) and in prior investigations has been shown to account for contraction of human airway smooth muscle in vitro caused by exogenously activated human eosinophils (29).

Although prior investigations have indicated that corticosteroids are capable of causing eosinophil apoptosis, the relationship between programmed cell death and precise quantitative inhibition of leukotriene production has not been established previously (7, 17, 30). In these studies, we used rhIL-5 to prolong eosinophil survival in vitro, as has been done in prior investigations (17, 18), and we first examined the time course and events associated with FP-induced inhibition of eosinophil survival for cells treated with rhIL-5. RhIL-5 is secreted in large regional concentrations in the conducting airways of human asthmatics and is known to prolong eosinophil survival in vitro and in vivo (31).

Our data indicate that total cell number remained relatively constant over 48 h regardless of treatment (Figure 1). However, cellular viability, as assessed by trypan blue dye exclusion (Figure 2) or PI staining (Figure 3) decreased minimally at 24 h for eosinophils treated with 10⁶ M FP and substantially at 48 h for concentrations of FP  10⁸ M (Figure 3). Total non-normalized eosinophil secretion of LTC₄ caused by exogenous stimulation with FMLP/CB consequently did not decrease in cells treated with any concentration of FP at 24 h, but decreased substantially at 48 h (Figure 5). When normalized for eosinophil survival as measured by PI staining, comparable basal secretion of LTC₄ was observed at 24 h and 48 h for eosinophils not treated with FP (Figure 6); however, significant inhibition of normalized LTC₄ secretion occurred at 48 h for concentrations  10⁸ M FP (Figure 6). Thus, when normalized for cell viability by PI staining, reduction in LTC₄ synthesis caused by FP was > 60% compared with eosinophils treated with rhIL-5 + buffer alone.

We further sought to determine if the normalized reduction of stimulated LTC₄ synthesis resulted from programmed cell death (apoptosis) or from inhibition of endogenous synthesis of LTC₄ caused directly by FP. For these studies, we used TUNEL assay, which distinguishes programmed cell death of apoptosis from general loss of cellular viability (e.g., necrosis). In contrast to the > 60% reduction of LTC₄ synthesis at 48 h for eosinophils treated with > 10¹⁰ M FP, apoptosis as measured specifically by TUNEL assay (Figure 4) was only 10% greater than in buffer-exposed cells. Hence, apoptosis per cell accounted for < 16% of the normalized reduction in LTC₄ synthesis caused by FP.

In a final series of studies, we examined by immunohistochemistry, the ability of FP to inhibit the translocation of cPLA₂ from the cytosol to the perinuclear envelope of the cell. Whereas complete translocation of cPLA₂ occurred in stimulated eosinophils treated with rhIL-5 alone, there was substantial inhibition of cPLA₂ translocation to nuclear membrane in cells from the same donor treated with rhIL-5 + FP (Figure 8). These data indicate that treatment with FP prevents translocation of cPLA₂ to the nuclear membrane during exogenous activation, a step that has been associated previously with the initiation of leukotriene synthesis (12, 32, 33). Specificity of cPLA₂ blockade also is suggested by the substantial restoration of LTC₄ synthetic capacity in eosinophils simultaneously treated with FP after receiving exogenous AA (Figure 9). Bypass of the substantial blockade (Figure 8) caused by FP with AA suggests selective blockade of PLA₂ and demonstrates intact downstream cellular function in the synthesis of LTC₄. We further demonstrated that surface expression of annexin-1, which was minimally present on the surface of freshly isolated eosinophils, increased in time- and concentration-related fashion. Thus, FP caused progressive synthesis of annexin-1 simultaneously with progressive inhibition of cPLA₂ translocation and subsequent stimulated LTC₄ synthesis.

It is important to consider the limitations of our findings. Studies were conducted in vitro under conditions substantially different from the environment in which migrating eosinophils enter human airways in conditions such as asthma. Hence, it is not possible to relate these in vitro concentrations, which are uniformly distributed to all cells, to the variable concentrations of FP that are delivered by metered dose inhaler (MDI) or powder devices. Concentrations for these studies were based upon estimations determined from preclinical studies by scientists at Glaxo Wellcome, Inc. Because the stimulus that causes LTC₄ secretion from eosinophils has not been demonstrated under either physiological or pathophysiological conditions, we used exogenous activation of LTC₄ production by FMLP/CB to model this yet undetermined activation process. This method for exogenous activation of eosinophils has been described extensively in prior publications (16, 24). Finally, although we demonstrated that apoptosis is not the primary mechanism by which eosinophil inhibition of LTC₄ initially occurs, we did not establish definitely the mechanism of inhibition. Although we observed that FP inhibits translocation of cPLA₂ to the nuclear envelope in viable cells, it remains to be elucidated that this is the definitive mechanism of inhibition of leukotriene synthesis in eosinophils caused by corticosteroid therapy.

We conclude that in vitro treatment of isolated human eosinophils with FP in the presence of rhIL-5 causes decreased eosinophil survival and initial decrease in synthesis of LTC₄ by a mechanism not related to cellular apoptosis. While cell death caused by FP eventually would serve to eliminate all leukotriene synthesis by eosinophils, our data suggest that initial inhibition of synthesis corresponds to inhibition of nuclear translocation of cPLA₂, a fundamental first step in the synthetic process. Preservation of LTC₄ synthetic capacity after addition of AA and progressive synthesis of annexin-1 during continuous treatment with FP demonstrates a specific inhibitory effect on IL-5 exposed eosinophils prior to induction of cell death.