Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Blood and tissue eosinophilia characterize a large variety of diseases, including bronchial asthma, infection with helminthic parasites, atopic allergy, and a number of malignant disorders (1). Release of proinflammatory lipid mediators, cytokines, free oxygen radicals, and highly-charged cationic proteins by activated eosinophils contribute to the onset and maintenance of tissue inflammation (1, 2). Eosinophil accumulation in blood and tissues has been related to a defect in their apoptotic death (3). Accordingly, maneuvers aimed at directly promoting eosinophil apoptosis in target tissues have been shown to facilitate the resolution of eosinophilic inflammation (4, 8, 9).

Glucocorticosteroids are the most effective anti-inflammatory drugs for controlling eosinophil-related diseases (10). This property relates to the inhibition of the synthesis and the effects of cytokines that prolong eosinophil survival, such as interleukin (IL)-3, IL-5, and granulocyte-macrophage colony-stimulating factor (10, 11), and to the induction of eosinophil apoptotic death (12, 13). In keeping with these observations, increased numbers of apoptotic eosinophils in the sputum and the bronchial submucosa, and an augmented sensitivity of peripheral blood eosinophils to apoptotic stimulation, have been observed in steroid-treated individuals with asthma (5, 7, 9).

Decisive events during the apoptotic process involve mitochondrial permeabilization and caspase activation (14). These phenomena may occur independently, or they may act in a tight connection, depending on the pro-apoptotic stimuli and the cell type (14). In the case of Fas activation, for instance, initiator caspases, such as caspase-8, activate downstream effector caspases, particularly caspase-3, by inducing the proteolysis of its inactive zymogen form. Activated caspase-3, in turn, cleave cytoplasmic and nuclear components, and promote DNA degradation and cell dismantling (14).

In another scenario, mitochondria amplify, or even initiate, the caspase cascade during a process characterized by the opening of the mitochondrial permeability transition pore (MPTP) (14, 15). This phenomenon leads to a loss in membrane potential (_m) and the subsequent release of soluble factors into the cytosol, such as cytochrome c and apoptosis-inducing factor (AIF) (14, 15). Cytochrome c, in turn, contributes to caspase activation, whereas AIF induces a caspase-independent DNA degradation (16).

Both caspase activation and mitochondrial dysfunction have been shown to occur during eosinophil apoptotic death (17). However, little is known concerning their relative contribution to the cascade of events leading to DNA fragmentation and to changes in cell morphology, particularly in response to glucocorticosteroid stimulation. The present study was aimed at determining to which extent caspase, in particular caspase-3, activation and mitochondrial membrane permeabilization and MPTP opening represent intracellular targets for the pro-apoptotic effect of dexamethasone in human eosinophils.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Eosinophil Purification

Fifty milliliters of peripheral blood was obtained from hypereosinophilic donors ( 500 eosinophils/µl). Informed written consent was obtained from all patients, and none of them received glucocorticosteroid treatment during at least 4 wk preceding the study. Eosinophils were isolated by immunomagnetic negative selection using a modification of the method of Hansel and coworkers (21). Briefly, venous blood was diluted with phosphate-buffered saline (PBS, pH 7.4) and centrifuged on Lymphocyte Separation Medium (density 1.077 g/ml; Eurobio, Les Ulis, France) to remove mononuclear cells. Erythrocytes were then lysed by incubating the granulocyte pellet with an ice-cold isotonic ammonium-chloride solution. In some experiments, cell suspension was additionally layered onto Percoll (density 1.082 g/ml; Pharmacia Biotech Europe GmbH, Orsay, France) to eliminate the residual mononuclear cells. Mixed granulocytes were then incubated with magnetic microbeads-conjugated anti-CD16 monoclonal antibody (mAb) (Miltenyi Biotec, Paris, France) before depletion of CD16⁺ neutrophils through a magnetic separation column (MACS system; Miltenyi Biotec). The purity of the eluted cell population (> 98% eosinophils) was evaluated after staining of cytospin preparations with Diff-Quik dye (Merz Dade, Baxter Dade AG, Duedingen, Switzerland). Each experiment was performed with eosinophils obtained from a single donor.

Cell Culture

PBS-washed eosinophils were resuspended in RPMI 1640 (Gibco BRL, Life Technologies SARL, Cergy Pontoise, France), supplemented with antibiotic-antimycotic solution (Gibco BRL), and 10% human autologous serum, and cultured at 37°C with 5% CO₂ in a humidified atmosphere for 0-48 h in the absence or presence of 1 µM dexamethasone-21-phosphate (Sigma, St. Quentin Fallavier, France).

In some experiments, eosinophils were pretreated for 2 h with 30 µM of the permeant and irreversible caspase-3 inhibitor Z-Asp-Glu-Val-Asp-fluromethylketone (Z-DEVD-fmk), or of the broad spectrum caspase inhibitor Z-Val-Ala-Asp-fmk (Z-VAD-fmk) (22) (both from Calbiochem, France Biochem, Meudon, France), or with their vehicle, i.e., a 0.4% solution of dimethyl sulfoxide (DMSO). At this concentration both caspase inhibitors prevented apoptosis induced by Fas crosslinking in human eosinophils (23).

In a separate series of experiments, eosinophils were preincubated for 2 h with 62 µM of bongkrekic acid (Calbiochem), or with 100 µM of decylubiquinone (Alexis, Coger, Paris, France), two MPTP inhibitors (24, 25). These concentrations were selected on the basis of previous studies showing the ability of these compounds to stabilize _m in neutrophils, and to inhibit MPTP opening in isolated mitochondria, respectively (25, 26). Control experiments were conducted in the presence of their respective vehicles, i.e., 0.02 N NH₄OH for bongkrekic acid and 0.33% DMSO for decylubiquinone (both from Sigma).

Apoptosis Determination

Eosinophils (0.15 × 10⁶) were resuspended in a hypotonic solution containing 0.1% (wt/vol) sodium citrate, 0.1% (vol/vol) Triton X-100 and 50 µg/ml propidium iodide (all from Sigma), as previously described (27). A total of 5,000 ungated cells were analyzed with an Epics XL flow cytometer using the Expo 32 software (both from Beckman Coulter, Villepinte, France). The percent of hypodiploid apoptotic nuclei was determined. The forward scatter (FS) and side scatter (SS) patterns were used to evaluate changes in size and granularity, respectively, associated with eosinophil apoptosis (28).

To assess eosinophil apoptotic morphology, i.e., chromatin and cytoplasm condensation, cytospin preparations were performed (0.1 × 10⁶ cells). Eosinophils were then fixed in acetone for 10 min at room temperature and stored at -20°C until staining with Diff-Quik dye. A minimum of 200 total cells was counted in randomly selected fields, and eosinophils showing apoptotic morphology were enumerated.

Determination of _m by Flow Cytometry

Changes in _m were evaluated using 3,3'-dihexyloxacarbocyanine iodide (DiOC₆[3]) (Molecular Probes Inc., Eugene, OR), a lipophilic cationic dye which accumulates into polarized mitochondria (29). Eosinophils (0.3 × 10⁶) were washed in PBS and incubated at 37°C for 30 min in the presence of 40 nM of DiOC₆(3) in RPMI (30, 31). Necrotic cells were excluded by the addition of 2.5 µM of propidium iodide immediately before flow cytometry analysis. To rule out the contribution of other intracellular membranes in DiOC₆(3) staining, control experiments were performed using cells previously incubated for 10 min with 200 µM of the oxidative phosphorylation uncoupling agent carbonyl cyanide m-chlorophenylhydrazone (CCCP; Sigma), which collapses _m (31, 32). A total of 5,000 events were analyzed by flow cytometry (DiOC₆[3] emission in FL-1 and propidium iodide in FL-3). The percent of DiOC₆(3)-high stained cells was determined.

Preparation of Eosinophil Extracts

Eosinophils were incubated for 10 min at 4°C in a lysis buffer containing 10 mM N-(2-hydroxyethyl)piperazine-N'-(2-ethane-sulfonic acid), pH 8.0, 1% nonidet P-40, 150 mM NaCl, 500 mM sucrose, 1 mM Na₂EDTA, 2 mM phenylmethylsulfonyl fluoride, 4% -mercaptoethanol, 10 µg/ml aprotinin, 10 µM leupeptin, and 10 µM pepstatin A (all from Sigma). Cells were subsequently sonicated and centrifuged (4°C, 15 min, 12,000 × g). Protein contents in clarified supernatants were determined using the BioRad protein assay (BioRad, München, Germany), by comparison with an ovalbumin standard curve (ICN Biomedical, Costa Mesa, CA).

Evaluation of Caspase-3-Like Activity by Spectrofluorimetry

Caspase-3-like activity was measured using the fluorimetric CaspACE Assay System (Promega, Charbonnieres, France). Briefly, protein extracts (10 µg) from eosinophil lysates were incubated with 50 µM of acetyl-DEVD-7-amino-4-methyl coumarin (Ac-DEVD-AMC), in the absence or presence of the caspase-3 inhibitor, Ac-DEVD-CHO, according to the manufacturer's instructions. AMC release was monitored on a Fluostar II spectrofluorimeter with filter settings at 355 nm for excitation and 460 nm for emission using the Biolise software (both from BMG LabTechnologies, Champigny-sur-Marne, France). Results were calculated by comparison with a standard curve of AMC and were expressed as pmoles of free AMC.

Evaluation of Caspase-3 Expression and Processing by Western Blot

Equal amounts of proteins (25 µg) were fractionated by 13% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred on polivinylidene difluoride membranes (BioRad) and reacted with a 1:1,000 dilution of rabbit anti-human caspase-3 Ab (BD Biosciences, Le pont de Claix, France), or with the mouse anti--actin mAb (clone AC-74; Sigma) at a 1:4,000 dilution. Immunoblots were then incubated with peroxidase-conjugated donkey anti-rabbit or sheep anti-mouse Abs, both at a 1:4,000 dilution, and developed using the ECL western blotting detection system (all from Amersham, Les Ulis, France). The intensities of the expression of zymogen caspase-3 (p32), of its processed form (p17), and of -actin were quantified using a densitometer (CCD-COHU and the Perfect Image data analysis) (Claravision, Orsay, France) and the Gel Analyst Software (Claravision). Results are expressed as a ratio, defined as the optic density (OD) values of p32 or p17 bands/OD values of the corresponding -actin bands.

Statistical Analysis

Data were analyzed statistically using the StatView SE+Graphics program for Macintosh (Abacus Concepts, Berkeley, CA). If ANOVA was significant, a Student's t test for paired values was used to assess comparability between the means. P values of 0.05 or less were considered significant. The results are expressed as the means ± SEM of the indicated number of experiments.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of Dexamethasone on DNA Fragmentation, Eosinophil Morphology, and Caspase-3 Activation

Culture of peripheral blood eosinophils for 0 to 48 h in minimum culture medium resulted in a time-dependent increase in the percent of hypodiploid DNA-containing nuclei and in a decrease in eosinophil size, as measured by flow cytometry (Figure 1). Incubation with 1 µM dexamethasone significantly amplified spontaneous eosinophil apoptosis, as demonstrated by the time-dependent increase in DNA fragmentation and in the progressive changes in cell size (Figure 1). These observations were confirmed by morphologic analysis (Figure 2). Dexamethasone-treated preparations demonstrated typical characteristics of apoptosis, such as cellular shrinkage and intense chromatin condensation (Figure 2A). In some cases, a complete loss of the nucleus was also noted (Figure 2A). Kinetics studies disclosed a significant increase in the number of apoptotic eosinophils after 48 h of culture in the presence of dexamethasone (Figure 2B).

View larger version (24K): [in this window] [in a new window]   Figure 1.   Dexamethasone induces a time-dependent apoptotic death in human peripheral blood eosinophils. Eosinophils were incubated during 0 to 48 h in minimum culture medium alone (open circles) or supplemented with 1 µM of dexamethasone (Dexa; filled circles). (A) Apoptotic nuclei were quantified by flow cytometry after permeabilization of the cells and staining of DNA with propidium iodide. (B) FS-SS patterns, representative of changes in eosinophil size and granularity, respectively, were evaluated by flow cytometry. Results are the means ± SEM of 16-27 experiments. *P < 0.05, as compared with values obtained at time 0; P < 0.05, as compared with dexamethasone-untreated cells.

View larger version (50K): [in this window] [in a new window]   Figure 2.   Morphologic analysis of eosinophil death. Eosinophils were cultured during 48 h in the presence of medium alone, or supplemented with 1 µM dexamethasone. (A) Apoptotic eosinophils were identified after staining of cytospin preparations with Diff-Quick dye. In contrast to eosinophils showing normal morphology (thin arrow), apoptotic eosinophils are characterized by chromatin and cytoplasm condensation (open arrow). In some cases, apoptotic bodies lacking the nucleus were also observed (closed arrow). Scale bar: 10 µM. (B) Apoptotic eosinophils were enumerated in randomly-selected fields and a minimum of 200 total cells was counted. Open bars, medium alone; filled bars, medium supplemented with dexamethasone. Results are mean percentage of apoptotic eosinophils ± SEM of six experiments. *P < 0.05, as compared with dexamethasone-untreated cells.

To determine whether caspases were activated during these processes, lysates obtained from dexamethasone-treated and -untreated eosinophils were tested for the presence of caspase-3-like activity. Eosinophil apoptosis occurring in the absence of dexamethasone was accompanied by the proteolysis of the fluorogenic substrate Ac-DEVD-AMC at 12 and 24 h (Figure 3A). This activity was markedly augmented by the addition of dexamethasone, particularly at 24 h (Figure 3A). In both cases, caspase-3-like activity returned to basal levels after 48 h of eosinophil stimulation (Figure 3A). Preincubation of protein extracts with the caspase-3 inhibitor, Ac-DEVD-CHO, abrogated Ac-DEVD-AMC cleavage, suggesting that this protease activity was attributable to caspase-3 (data not shown).

View larger version (31K): [in this window] [in a new window]   Figure 3.   Dexamethasone induces caspase-3 activation in human eosinophils. Cells were stimulated with 1 µM of dexamethasone (Dexa), as described in the legend of Figure 1. (A) Whole cell lysates were tested for their ability to cleave the caspase-3 substrate Ac-DEVD-AMC by a fluorimetric assay. Results are expressed as pmoles of free AMC. Open squares, medium alone; filled squares, dexamethasone. (B) Representative Western blot experiment showing caspase-3 expression and proteolysis in protein extracts from dexamethasone-unstimulated and -stimulated eosinophils. Blots were probed with rabbit polyclonal anti-caspase-3, or mouse anti--actin Abs. Positions of molecular weight markers are indicated in kD. Each blot is representative of 3-8 experiments. (C and D) The intensities of the expression of zymogen and cleaved forms of caspase-3 were determined by densitometry and defined as the ratio OD p32 (C) or p17 (D) band/OD -actin band values, respectively. Results are expressed as means ± SEM of three to eight experiments. *P < 0.05, as compared with values obtained at time 0; P < 0.05, as compared with dexamethasone-untreated cells.

Caspase-3 activation during dexamethasone-induced apoptosis was confirmed by analyzing its processing by Western blot. Freshly purified human eosinophils constitutively expressed the inactive form of caspase-3 (p32), but very low levels of its active subunit, p17 (Figure 3B). No changes in the amounts of p32 were observed during spontaneous eosinophil apoptosis, although a moderate raise in p17 expression was detected at 12 and 24 h of culture (Figures 3B-3D). In contrast, dexamethasone stimulation was followed by a marked increase in the levels of active caspase-3 at 24 h, without changes in those of p32 (Figures 3B-3D). A reduction in the amounts of p32 was observed after 48 h of dexamethasone treatment (Figures 3B and 3C). At this time point, a pronounced decrease in -actin expression was also noted (Figure 3B), a phenomenon probably reflecting a nonspecific degradation of cell constituents during late apoptosis.

Effect of Caspase Inhibitors on Dexamethasone-Induced Human Eosinophil Apoptosis

To evaluate the role of caspases during eosinophil apoptosis, cells were incubated for 2 h with 30 µM of either the pan caspase inhibitor, Z-VAD-fmk, or the caspase-3 inhibitor, Z-DEVD-fmk, before dexamethasone stimulation. A moderate and comparable degree of inhibition in spontaneous and dexamethasone-induced eosinophil apoptosis (Figure 4) and in the accompanying changes in cell size (data not shown) was noted at 24 and 48 h in the presence of Z-VAD-fmk and Z-DEVD-fmk.

View larger version (30K): [in this window] [in a new window]   Figure 4.   Effect of caspase inhibitors on dexamethasone-induced human eosinophil apoptosis. Eosinophils were incubated during 2 h with 30 µM of Z-DEVD-fmk (DEVD), or Z-VAD-fmk (VAD), or with their vehicle, i.e., 0.4% DMSO, and next stimulated for 0 to 48 h with 1 µM of dexamethasone (Dexa). Apoptotic nuclei were quantified by flow cytometry after permeabilization of the cells and staining with propidium iodide. Results are the means ± SEM of 15-17 experiments. Open bars, vehicle; shaded bars, DEVD; diagonally striped bars, VAD; solid bars, Dexa+vehicle; horizontally striped bars, Dexa+DEVD; vertically striped bars, Dexa+VAD. *P < 0.05, as compared with dexamethasone-untreated cells; P < 0.05, as compared with dexamethasone-treated cells.

To ascertain that caspase inhibitors efficiently penetrated cells, inhibitor-treated eosinophils were lysed and assayed for caspase-3 activity. Both Z-VAD-fmk and Z-DEVD-fmk, added during eosinophil culture, abolished caspase-3 fluorogenic substrate cleavage (data not shown).

Role of the MPTP in Dexamethasone-Induced Eosinophil Apoptosis

To investigate the involvement of mitochondrial alterations during eosinophil apoptosis, the reduction in _m, which reflects membrane permeabilization (15), was assessed by flow cytometry.

Freshly purified eosinophils contained polarized mitochondria, as demonstrated by the incorporation of the lipophilic dye, DiOC₆(3) (Figure 5A). Treatment with the proton ionophore CCCP abolished _m, confirming that DiOC₆(3) staining reflected its accumulation into polarized mitochondria (Figure 5A). Eosinophil culture in the absence of dexamethasone led to a slight, but not significant, reduction in DiOC₆(3) uptake (Figures 5A and 6A). This phenomenon was dramatically amplified by 24 and 48 h stimulation with dexamethasone (Figures 5A and 6A). In both cases, _m disruption was concomitant with DNA fragmentation (Figures 6A and 6B).

View larger version (42K): [in this window] [in a new window]   Figure 5.   Effect of bongkrekic acid on dexamethasone-induced reduction in eosinophil size, DNA fragmentation, and m dissipation. Eosinophils were incubated with 62 µM bongkrekic acid (BA), or with its vehicle, i.e., 0.02 N NH4OH for 2 h, and next stimulated with 1 µM dexamethasone (Dexa) for 24 h. (A) Cells were incubated with 40 nM DiOC6(3) and stained with 2.5 µM propidium iodide for the detection of changes in m by flow cytometry. Recorded events were gated on propidium iodide negative population to exclude nonviable cells. Negative control experiments were performed using CCCP-pretreated cells. (B) Eosinophils were permeabilized and DNA was stained with propidium iodide (PI) for the quantification of apoptotic nuclei by flow cytometry. (C) FS-SS pattern showing modifications in cell morphology occurring during eosinophil stimulation. These histograms and plots are representative of three to five experiments.

View larger version (32K): [in this window] [in a new window]   Figure 6.   Bongkrekic acid inhibits dexamethasone-induced loss in m and eosinophil apoptosis. Eosinophils were stimulated during 0 to 48 h with 1 µM dexamethasone (Dexa) and treated with 62 µM bongkrekic acid (BA), as described in the legend of Figure 4. (A) the population of DiOC6(3) high-stained cells and (B) the proportion of apoptotic nuclei were quantified by flow cytometry. Results are the means ± SEM of three to five experiments. Open bars, vehicle; thinly striped bars, BA; thickly striped bars, Dexa+BA; filled bars, Dexa + vehicle. *P < 0.05, as compared with values obtained at time 0; P < 0.05, as compared with vehicle-treated cells; P < 0.05, as compared with dexamethasone-treated cells.

The role of the MPTP in _m disruption and DNA fragmentation was then evaluated by treating dexamethasone-unstimulated and -stimulated eosinophils with 62 µM bongkrekic acid, a ligand of the MPTP component, adenine nucleotide translocator (ANT, 24). This compound failed to modify significantly mitochondrial permeabilization associated with spontaneous eosinophil apoptosis, but markedly inhibited that induced by dexamethasone at 24 and 48 h (Figure 6A). In parallel, bongkrekic acid attenuated spontaneous apoptosis and suppressed dexamethasone-induced DNA fragmentation and reduction in eosinophil size (Figures 5 and 6). To confirm these observations, we evaluated the effectiveness of another MPTP inhibitor, decylubiquinone (25), on these parameters. At the concentration of 100 µM, decylubiquinone prevented dexamethasone-induced loss of _m in three out of seven eosinophil preparations (81% and 65% inhibition at 24 and 48 h, respectively, data not shown). This was associated to a reduction in eosinophil apoptosis (99% and 56% inhibition at 24 and 48 h, respectively, data not shown).

Effect of Caspase Inhibitors on Dexamethasone-Induced Eosinophil _m Disruption

We then determined whether caspase activation played a role during dexamethasone-induced _m dissipation. Eosinophil treatment with the caspase inhibitors, Z-VAD-fmk and Z-DEVD-fmk, failed to modify the decrease in DiOC₆(3) uptake observed after 24 h and 48 h of culture in the absence or in the presence of dexamethasone (Figure 7).

View larger version (24K): [in this window] [in a new window]   Figure 7.   Caspase inhibitors fail to modify dexamethasone-induced loss in m. Eosinophils were incubated during 2 h with 30 µM of Z-DEVD-fmk (DEVD), or Z-VAD-fmk (VAD), or with their vehicle, i.e., 0.4% DMSO, and next stimulated for 0 to 48 h with 1 µM of dexamethasone (Dexa). The percent of DiOC6(3)-labeled viable cells was determined by flow cytometry. Results are the means ± SEM of three to four experiments. Open bars, vehicle; shaded bars, DEVD; diagonally striped bars, VAD; filled bars, Dexa+vehicle; horizontally striped bars, Dexa+DEVD; vertically striped bars, Dexa+VAD. *P < 0.05, as compared with vehicle-treated cells.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study demonstrated that eosinophil apoptosis induced by dexamethasone was associated with caspase-3 activation, as ascertained by the processing of caspase-3 and the concomitant generation of a DEVDase activity in cell lysates. In our hands, caspase-3-like activity became detectable 12 h after dexamethasone addition, reached a maximum at 24 h, and decreased thereafter. These findings are in agreement with the kinetics of dexamethasone-induced caspase-3 proteolysis recently reported in human eosinophils (17) and with the absence or the presence of very low amounts of activated caspase-3 in eosinophil lysates at early (2 h), or late (48 h) time points after dexamethasone stimulation, respectively (18, 19).

Although essential in several cell types for the development of the typical features of apoptosis, such as chromatin condensation and DNA fragmentation (14), caspase-3 is not necessarily activated during glucocorticosteroid-induced apoptosis. For example, pre-B leukemia cells undergo apoptosis upon dexamethasone treatment by a mechanism involving caspase-6, and not caspase-3, activation (33).

To evaluate the role of caspase, and particularly caspase-3, activation during dexamethasone-induced apoptosis, eosinophils were treated with the broad caspase and the caspase-3 inhibitors, Z-VAD-fmk and Z-DEVD-fmk, respectively. Both compounds slightly decreased the rate of dexamethasone-induced apoptosis (~ 18% and 4% inhibition at 24 h, respectively). This effect was comparable to that observed against spontaneous eosinophil apoptosis (~ 25% and 7% inhibition at 24 h, respectively). Because under these conditions Z-VAD-fmk and Z-DEVD-fmk abolished caspase-3-like activity generation in cell lysates, we concluded that caspase activation does not play a critical role in dexamethasone-induced apoptosis in human eosinophils. These results are in agreement with the lack of effect of caspase inhibitors in preventing dexamethasone-induced apoptosis in mouse thymocytes (34). In other studies, however, Z-VAD-fmk was able to protect murine thymocytes, as well as human monocytes and pre-B leukemia cells, from dexamethasone-induced apoptosis (33, 35). Altogether, these findings contrast with our observations and suggest that, depending on the cell type, different intracellular pathways may be activated during glucocorticosteroid-induced apoptosis.

Recent reports have demonstrated that dexamethasone-induced eosinophil apoptosis was reduced by concentrations of Z-VAD-fmk equal or above 50 µM, with a maximal potency observed with 200 µM of this antagonist (18, 19). Because the effect of Z-VAD-fmk was not tested against spontaneous eosinophil apoptosis (19), no firm conclusion can be raised from these observations concerning the ability of caspase inhibitors to specifically antagonize the pro-apoptotic effect of glucocorticosteroids. Furthermore, at high concentrations (50-100 µM), Z-VAD-fmk was shown to prevent the activation of calpains, another family of cystein proteases with apoptotic properties (38). This suggests that the potential of Z-VAD-fmk to attenuate dexamethasone-induced eosinophil apoptosis previously reported (18, 19) may be unrelated to caspase inhibition.

The involvement of mitochondria during Fas ligation, irradiation, or dexamethasone-induced apoptosis has been reported in a number of inflammatory and immune cells, including neutrophils, thymocytes, and splenocytes (14, 15, 26, 39). We thus investigated the involvement of mitochondrial alterations during dexamethasone-induced eosinophil apoptosis. Stimulation with dexamethasone was followed by mitochondrial permeabilization, as demonstrated by the a marked decrease in _m, particularly at 24 and 48 h. These findings are consistent with those reported in murine lymphocytes (30, 31) and, recently, in human eosinophils (20). Because activated caspases have been shown to promote mitochondrial membrane disturbances (14, 15), we evaluated the effect of Z-VAD-fmk and Z-DEVD-fmk on this process. Dexamethasone-induced loss in _m was unaffected by both caspase inhibitors, suggesting that caspase activation is not a critical event upstream mitochondrial disruption in eosinophils. These data parallel similar observations performed on human thymocytes (34), whereas they contrast with the ability of Z-VAD-fmk to prevent dexamethasone-induced _m disruption in mouse thymocytes and human pre-B leukemia cells (33, 35). Together, these findings argue for the existence of caspase-dependent and -independent pathways involved in dexamethasone-induced mitochondrial permeabilization, probably in relation with the cell type.

Opening of the MPTP is one of the caspase-independent physiologic events leading to mitochondrial permeabilization (15). The MPTP is a complex formed by mitochondrial membrane proteins including the apoptogenic constituant, ANT (15, 42). Bongkrekic acid, a specific ANT ligand, and decylubiquinone are potent inhibitors of MPTP-mediated mitochondrial permeabilization (25, 39, 43). We thus used these compounds to evaluate the role of MPTP opening in dexamethasone-induced loss in _m and apoptosis in eosinophils. We show here that bongkrekic acid suppressed dexamethasone-induced _m dissipation and apoptosis. A similar effectiveness was observed for decylubiquinone, which was, however, restricted to approximately half of the eosinophil preparations. This may result from a defect in decylubiquinone incorporation by intact eosinophils, because all data currently available on the efficacy of this compound have been obtained on isolated mitochondria (25). Nevertheless, these observations indicate that MPTP opening regulates mitochondrial depolarization and the subsequent apoptosis induced by glucocorticosteroids in eosinophils.

We have recently demonstrated the lack of involvement of MPTP during Fas-mediated loss in _m and apoptosis in eosinophils (23). These results, and those presently reported, indicate that, depending on the pro-apoptotic stimulus, opening of the MPTP may account, or not, for mitochondrial permeabilization in human eosinophils.

The mechanisms underlying dexamethasone-induced mitochondrial depolarization in eosinophils remain to be elucidated. An interesting hypothesis involves the participation of Bcl-2 family proteins, some of which directly regulate MPTP function (14, 15, 42). It is worthy of note that human eosinophils consitutively express Bax (44, 45), a pro-apoptotic molecule able to induce mitochondrial membrane depolarization by cooperating with ANT (42). Translocation of Bax to the mitochondria has been observed during dexamethasone-induced apoptosis in murine thymocytes, and was recently proposed as a crucial caspase-independent step in the execution of spontaneous and staurosporine-induced eosinophil apoptosis (41, 46). Whether Bax, or other Bcl-2 family proteins, are also targets for the pro-apoptotic effect of glucocorticosteroids in human eosinophils remains to be established.

Apart from its contribution to the activation of the caspase cascade, mitochondrial permeabilization may initiate a caspase-independent apoptotic pathway, in particular through the release of AIF (14). Interestingly, this caspase-independent DNA fragmentation-inducing factor is released from mitochondria of T cell hybridomas during dexamethasone-induced apoptosis (16). Whether AIF is expressed by eosinophils and is a target for glucocorticosteroids is a matter for further investigations.

In conclusion, this study demonstrates that, despite their activation upon dexamethasone stimulation, caspases are neither required for mitochondrial depolarization, nor for eosinophil apoptosis. In contrast, mitochondrial depolarization via MPTP opening mediates the pro-apoptotic effect of glucocorticosteroids in these cells.