Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Eosinophils have a proposed vital role in the pathogenesis of a number of disease states, with tissue and blood eosinophilia being associated with asthma, atopic allergy, and helminthic parasite infections (1- 4). The accumulation and persistence of these granulocytes at sites of inflammation is due to a number of factors, such as selective adhesion and migration (5), as well as increased survival in response to growth factors such as interleukin (IL)-5 (6), IL-3, and granulocyte macrophage colony-stimulating factor (GM-CSF) (7), either from exogenous sources, such as T cells, or as a result of autocrine elaboration by eosinophils themselves (8). These hematopoietins have been proven to abrogate eosinophil apoptosis in vitro (9), and ex vivo using nasal polyposis as a model of tissue inflammation (10, 11). The observation that protection from apoptosis afforded by IL-5 is inhibitable by cycloheximide and actinomycin D suggests that RNA and protein synthesis is required (12).

The release of eosinophil intracellular mediators has been found to be histotoxic, causing tissue lesions and airway inflammation (reviewed in 13). In contrast to necrosis, apoptosis provides a mechanism whereby membrane integrity is maintained throughout cell death, allowing ingestion and clearance of whole cells or discrete vesicles ("apoptotic bodies") by resident phagocytes without toxic mediator release (6, 14). Apoptosis of eosinophils may therefore provide a crucial mechanism for the limitation of inflammation (15).

Eosinophils exhibit the classic apoptosis-associated morphological changes such as internucleosomal nucleic acid cleavage, cytoplasmic condensation (18), and redistribution of membrane phospholipids (19), allowing binding of AnnexinV as an early marker of apoptosis (20, 21). A number of extracellular modulators of eosinophil apoptosis have been elucidated. Monoclonal antibody (mAb)- dependent ligation of the Fas receptor (22, 23) or CD69 (24) and treatment with glucocorticoids (25) have been shown to initiate programmed cell death (PCD). Eosinophil survival is enhanced by adhesion to tissue fibronectin via very late antigen-4 (26, 27) or treatment with lipopolysaccharide (28), leading to elaboration of GM-CSF. However, little is known about the intracellular signaling mechanisms employed in the regulation of eosinophil apoptosis.

The bcl-2 proto-oncogene was first described at chromosomal breakpoint of t(14:18) found in B-cell lymphomas (29, 30), and was determined to be the mammalian homologue of the Caenorhabditis elegans death repressor gene, ced-9 (31, 32). A large family of genes with a sequence homology to bcl-2 has emerged that possesses important divergent functions in the mediation of PCD, consisting of death antagonists (bcl-2, bcl-x_L, bcl-w) or death agonists (bax, bcl-x_S, bak) (33). Alternative transcription of the bcl-x gene results in two distinct mRNAs, the larger transcript, bcl-x_L, encoding a death suppressor, and the second variant, bcl-x_S, smaller due to an internal deletion, encoding a protein capable of antagonizing the effects of Bcl-x_L (34, 35). Bcl-2 homologues form a dynamic equilibrium of homo-heterodimerization via conserved regions and associate with unrelated proteins, such as protein kinase Raf-1 (36, 37), to form a complex life-death rheostat linked to growth factor receptors and kinase signaling cascades. Most have a transmembrane region allowing localization to the mitochondrial membrane and the potential to mediate the release of apoptogenic factors such as cytochrome C from the organelle (38). Peripheral blood lymphocytes express Bcl-2, Bcl-x, and Bax, although neutrophils express only the latter homologue (41, 42). Also, neutrophils transgenic for the expression of bcl-2 have inhibited apoptosis but not engulfment by phagocytes in vitro (43). The literature is currently unclear regarding the expression of Bcl-2 homologues in eosinophils. Druilhe and colleagues (44) have recently shown by immunoblot analysis the constitutive expression of Bax, with Bcl-x_L and Bcl-2 essentially absent from peripheral blood eosinophils (PBE). The lack of constitutive Bcl-2 expression supports the findings of a previous investigation (45). However, these studies report conflicting observations on the potential role of IL-5 in the regulation of Bcl-2 in PBE. In contrast, Yasui and coworkers (46) indicated constitutive expression of Bcl-2, assessed by flow cytometry, which is downregulated by the potential anti-inflammatory agent theophylline.

In this study, to elucidate the regulation of eosinophil apoptosis in vitro, the expression of bcl-2 and its homologues bax, and splice variants bcl-x_S and bcl-x_L, were investigated by reverse transcriptase-polymerase chain reaction (RT-PCR), immunoblot, and flow cytometric analysis. Potential modulation of these apoptosis mediators by survival-enhancing concentrations of IL-5 was also examined. The results indicate that anti-apoptotic bcl-2 is absent from unstimulated eosinophils but is upregulated by IL-5. Bax and Bcl-x were constitutively expressed in freshly isolated PBE. Both bcl-x_L and bcl-x_S splice variants were detectable, but bcl-x_L predominated. Expression of Bax and Bcl-x was not modulated by IL-5.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cytokines and Antibodies

Recombinant human IL-5 was obtained from R&D Systems (Abingdon, UK). Murine human Bcl-2-specific mAb 124, and Bax (N-20) and Bcl-x_S/L (S-18)-specific rabbit antihuman polyclonal antibodies (pAb) were obtained from Autogen Bioclear UK Ltd. (Wilts, UK for Santa Cruz Biotech, Santa Cruz, CA). Human -actin-specific mAb was obtained from Sigma Chemical Co. (Poole, UK). Control mouse and rabbit immunoglobulin Gs, fluorescein isothiocyanate (FITC), and biotin-conjugated antimouse/rabbit secondary antibodies were obtained from Dako Ltd. (High Wycombe, UK).

Cell Lines

Human myeloid cell line HL-60 and T-cell line Jurkat were obtained from European Collection of Cell Cultures (Wilts, UK).

Isolation of PBE and Cell Culture

Heparinized peripheral venous blood was taken from volunteers with a normal or slightly raised eosinophil count. Eosinophils were purified by a two-step method of density gradient centrifugation, followed by negative immunomagnetic selection as described previously (58). Briefly, removal of erythrocytes by dextran sedimentation was followed by slow centrifugation (100 × g, 15 min, room temperature) of leukocyte-rich supernatant, and resuspension in Hanks' balanced salt solution without Ca²⁺ and Mg²⁺ supplemented with 2% heat-inactivated fetal calf serum (FCS), and 0.02 mol/L ethylenediaminetetraacetic acid. Leukocytes were then centrifuged (400 × g, 25 min, room temperature) on a Histopaque 1083 (Sigma Chemical Co.). Mononuclear cells were carefully removed, and erythrocytes contaminating the granulocyte pellet were lysed by hypotonic shock using sterile, ice-cold water. Eosinophils were separated from neutrophils by negative immunomagnetic selection on the basis of CD16-coated immunomagnetic beads (Miltenyi Biotec Inc., Auburn, CA). Postisolation purity and viability, assessed by Kimura stain and exclusion of trypan blue, respectively, were routinely > 99%. Freshly isolated PBE were washed and cultured at 3 × 10⁶ cells ml¹ in RPMI 1640 L-glutamax (GIBCO-BRL, Paisley, UK) supplemented with 10% FCS and antibiotics. HL-60 and Jurkat cell lines were similarly cultured in media that lacked antibiotics. Eosinophils were stimulated by the presence of 100 pg/ml IL-5 where indicated.

Assessment of Eosinophil Apoptosis and Viability

Eosinophil viability was assessed by trypan blue exclusion counting at least 200 cells by light microscopy. Apoptosis was assessed by light microscopy on the basis of cells exhibiting classic apoptotic morphology, such as nuclear and cytoplasmic condensation in the presence of Kimura stain (21).

Immunoprecipitation and Immunoblotting

Cells were washed in cold phosphate-buffered saline (PBS), lysed in ice-cold 1% Triton X-100 isotonic lysis buffer containing freshly added protease inhibitors (100 µg/ml phenylmethyl sulfonyl fluoride, 1 µg/ml aprotinin, 1 µg/ml pepstatin A), and incubated on ice for 30 min. Nuclei and unlysed cellular debris were removed by centrifugation (12,000 × g, 5 min, 4°C). For immunoprecipitation, the lysate was precleared overnight at 4°C with constant agitation, with 50 µl of 1:1 protein A-Sepharose, which was removed by centrifugation at 400 × g for 2 min. Specific antibody or control were added for 60 min and captured by protein A-Sepharose for 60 min at 4°C. Immunoprecipitates were sequentially washed twice in dilution buffer (0.1% Triton X-100; 0.1% bovine hemoglobin; 0.01 M Tris-Cl, pH 8.0; 0.14 M NaCl), followed by a wash in dilution buffer without Triton X-100, followed by a wash in 0.05 M Tris-Cl, pH 6.8.

Equal volume of Laemmli buffer (5% -mercaptoethanol) was added to immunoprecipitate or whole-cell lysate and boiled for 4 min prior to electrophoresis through 12.5% sodium dodecyl sulfate polyacrylamide gels. Proteins were electrotransferred onto Hybond nitrocellulose membrane (Amersham Life Science Ltd., Bucks, UK). Nonspecific binding sites on the membrane were blocked by incubation (1 h, room temperature) with 5% nonfat milk (Marvel) in PBS 0.1% Tween-20 (PBS-T). Filters were incubated with specific primary antibody/negative control, diluted to 0.1 µg/ml in blocking buffer (2 h, room temperature). The membrane was then incubated (1 h, room temperature) with species-specific biotinylated secondary antibody (1:3,000), followed by reaction with horseradish peroxidase-streptavidin (1:1,000) (1 h, room temperature). Each incubation was performed with constant agitation and followed by 1 × 15 min and 2 × 5 min wash with PBS-T. Membranes were then developed using enhanced chemiluminescent system according to the manufacturer's instructions and exposed to Hyperfilm (Amersham).

Intracellular Flow Cytometric Analysis

Cells were washed in PBS prior to fixation and permeabilization using a kit according to the manufacturer's instructions (Bradsure Biologicals, Bucks, UK). During the permeabilization step, cells were incubated with 1 µg/ml primary antibody (25 min, room temperature). Cells were then washed in PBS and incubated with species-specific FITC-conjugated secondary antibody, diluted 1:100 in PBS (25 min, room temperature, in dark). After being washed in PBS, cells were analyzed on FACScan (Becton Dickinson, Oxford, UK). Quantification of the flow cytometry data was performed on the basis of specific median fluorescence, which was calculated using the following equation:

Specific median fluorescence = Median fluorescence of test antibody  Median fluorescence of isotype-matched control.

Analysis of mRNA

Total RNA was extracted from cells using TRIzol according to manufacturer's protocol (GIBCO-BRL). PolyA⁺ RNA was primed with oligo dT (GIBCO-BRL) and reverse transcribed with Superscript RT (GIBCO-BRL) for 1 h at 37°C. The cDNA product was amplified by PCR using primers obtained from Cruachem Ltd. (Glasgow, UK) specific for bcl-2 (5' primer GATGTCCAGCCAG CTGCACCTG; 3' primer-CACAAAGGCATCCCAGCCTCC), bax (5' primer-GGACCCGGTGCCTCAGGA; 3' primer-CAAAGATGGTCACGGTCTGC), bcl-x_S/L (5' primer-TTGGACAATGGACTGGTTGA; 3' primer-GTAGAGTGGATGG TCAGTG), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as a positive control (5' primer-GGGAAGCTCACT GGCATGGCCTTCC; 3' primer-CATGTGGGCCATGAGGTCCACCAC). The bcl-x_S/L primers corresponded to the 5' and 3' untranslated regions and allowed simultaneous amplification of bcl-x_S (0.6 kb) and bcl-x_L (0.8 kb). A 50-µl PCR reaction was set up containing 5 µl cDNA, 10 mM Tris-HCl, 25 pmol each primer, 1.5 mM MgCl₂, 0.2 mM dNTP, and 2.5 U BioTaq DNA polymerase (Bioline, London, UK). Amplification consisted of 30 to 35 cycles of denaturation at 94°C for 20 s, annealing at 60°C (bcl-2) or 58°C (bax, bcl-x_S/L) for 30 s, and extension at 72°C for 1 min. Amplification products were electrophoresed on 1.2% agarose gels and visualized by ethidium bromide staining under ultraviolet light and photographed. Photographs were scanned and imported into Microsoft Power Point 97 (Redmond, WA).

Verification of RT-PCR Products

RT-PCR product was excised from the agarose gel, and the DNA was extracted from the gel using gel extraction columns according to the manufacturer's protocol (Qiagen Ltd., Sussex, UK) and eluted with water. The product was then verified by direct automated sequencing or "nested PCR."

Sequencing reactions were performed using a 30-ng PCR product, with a dRhodamine terminator cycle sequencing kit (Perkin Elmer Applied Biosystems, Bucks, UK) according to the manufacturer's protocol. Automated sequencing of the product was kindly performed by the Protein and Nucleic Acid Chemistry Laboratory, Leicester University, Leicester, UK.

Nested PCR was also performed on the purified RT- PCR products. The primary PCR product was used as a template for a second reaction, involving a third specific primer sequence internal to the initial primer pair. The reduction in size of the product could then be verified on an agarose gel. The reaction conditions were as described previously, using only 1 µl of purified PCR product as a template. Nested primers obtained from Cruachem Ltd. (Glasgow, UK), specific for bcl-2 (3' primer-GACGCTCTCCACACACATGACC), bax (3' primer-AGGAAG TCCAATGTC CAG), bcl-x_S (3' primer-CCACAAAAGTATCCTGTT CAAAGC), and bcl-x_L (3' primer-CCAAGCTGCGATCCGACTCAC).

Statistical Analysis

Where applicable, values are expressed as mean ± SEM. The mean data from IL-5-treated cells were compared with the untreated cells by paired Student's t test. A P value of < 0.05 was regarded as significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Detection of Apoptosis-Associated Proteins

Bcl-2 was not detectable in unstimulated eosinophils (Figure 1a) compared with HL-60 cell line as a positive control (Figure 1b), as assessed by intracellular flow cytometry. Good constitutive expression of Bax specific median fluorescence ([SMF], 26.3 ± 0.76) and Bcl-x (SMF, 11.3 ± 1.42) in freshly isolated PBE was observed (Figures 1c and 1d, respectively). It was not possible to distinguish between Bax / and the small and large isoforms of Bcl-x by this method, because the antibodies used targeted regions common to both.

View larger version (24K): [in this window] [in a new window]   Figure 1.   Expression of apoptosis-associated proteins in freshly isolated peripheral blood eosinophils (PBE) assessed by flow cytometry. PBE ( panels a, c, and d) or HL-60 cell line (panel b) were fixed, permeabilized and stained with anti-Bcl-2 (panels a and b), anti-Bax ( panel c), and anti-Bcl-x ( panel d) according to MATERIALS AND METHODS. Test antibody (hollow histogram); irrelevant control (solid histogram). Representative of n = 4.

Consistent with the flow cytometry results, immunoblot analysis detected the 26-kD Bcl-2 in Jurkat and HL-60, but the protein was absent from freshly isolated eosinophils (Figure 2a). Bax has been shown to be expressed as at least two alternative transcripts, bax- and bax-, encoding proteins of 21 kD and 24 kD, respectively (45). Immunoblot analysis of whole-cell lysates revealed that eosinophils express only Bax, with no detectable expression of Bax (Figure 2b). The immunoblot positive control for Bax was a purified Bax fusion protein. Under stringent reducing conditions, a 37-kD band was consistently observed in eosinophil whole-cell lysates probed with anti-Bax pAb. Bcl-x could not be detected by immunoprecipitation, immunoblotting of eosinophil, or HL-60 whole-cell lysates. However, a Bcl-x fusion protein-positive control (Autogen Bioclear UK Ltd.) was readily detectable (data not shown).

View larger version (38K): [in this window] [in a new window]   Figure 2.   Immunoblot analysis of Bcl-2 and Bax expression in whole-cell lysates under reducing conditions. (a) Bcl-2: Freshly isolated PBEs and cell lines Jurkat and HL-60 as positive controls. (b) Bax: Freshly isolated PBE and 45 kD fusion protein as positive control. Whole-cell lysate was probed with irrelevant antibody as negative control.

Analysis of mRNA Expression

RT-PCR indicated that freshly isolated eosinophils do not express message for bcl-2 (Figure 3a) compared with HL-60 as a positive control for bcl-2 at predicted 255 bp, and GAPDH amplification as an internal PCR control. The bax (495 bp) and bcl-x_L (0.8 kb) mRNA were readily detectable in unstimulated PBE (Figures 3b and 3c). Coamplification of bcl-x_S (0.6 kb) mRNA was also consistently detectable, but to a lesser degree relative to the longer splice variant (Figure 3c). The PCR products were verified by direct sequencing and/or nested PCR (data not shown).

View larger version (34K): [in this window] [in a new window]   Figure 3.   Expression of bcl-2, bax, bcl-xL, and bcl-xS mRNA by RT-PCR. Cell line HL-60 as bcl-2-positive control; GAPDH as internal RT-PCR control. Representative of n = 3.

Effect of IL-5 on Eosinophil Survival and Bcl-2 Protein Expression

Culturing eosinophils in the presence of IL-5 caused a significant reduction in apoptosis, assessed by visual examination of morphology, and increased viability, assessed by trypan blue exclusion, compared with medium control (Figure 4A). Occurrence of apoptotic morphology in the presence of 100 pg/ml IL-5 was consistently < 10% after 24 h, compared with 40 to 50% in the absence of IL-5.

View larger version (27K): [in this window] [in a new window]   Figure 4.   (A) Effect of IL-5 on eosinophil apoptosis assessed by morphology under light microscopy after Kimura stain, and survival assessed by trypan blue exclusion. Mean ± SEM, n = 3. *P < 0.01; **P < 0.05. (B) Effect of IL-5 on the expression of antiapoptosis protein Bcl-2. Cells were permeabilized and stained with anti-Bcl-2 and -actin as positive control, according to MATERIALS AND METHODS, at time points during culture. Data show SMF ± SEM, n = 3. *P < 0.025 IL-5 compared with control.

Stimulation of PBE with IL-5 for 24 h induced a limited, but significant, expression of Bcl-2, with no change in expression in eosinophils cultured in the absence of the cytokine as detected by flow cytometry (Figure 4B). This upregulation of Bcl-2 was not as a result of a nonspecific change in protein levels as shown by a lack of significant variation of the internal control, -actin (Figure 4B). Bcl-2 in IL-5-treated eosinophils was undetectable by immunoblot analysis (data not shown).

Incubation of eosinophils with IL-5 for 24 h was found to have no significant effect on the levels of Bax and bcl-x_S/L expression as determined by flow cytometry (Figure 5).

View larger version (23K): [in this window] [in a new window]   Figure 5.   Effect of IL-5 on the expression of apoptosis-associated proteins in PBE assessed by flow cytometry. Freshly isolated PBE (0 h) or PBE cultured in the presence (IL-5) or absence (MED) of 100 pg/ml IL-5 (24 h) were permeabilized and stained as indicated. Test antibody (hollow histogram); irrelevant control (solid histogram). Representative of n = 3.

Effect of IL-5 on the Expression of bcl-2 mRNA

Amplification of bcl-2 mRNA (250 bp) was detectable in eosinophils stimulated with IL-5 for 24 h, but not unstimulated or cells cultured in medium alone (Figure 6). The signal intensity of PCR product relative to internal GAPDH control and HL-60 bcl-2 product indicate a low expression of message in stimulated eosinophils, consistent with the levels of detectable protein (Figure 5).

View larger version (34K): [in this window] [in a new window]   Figure 6.   Effect of IL-5 on the expression of bcl-2 mRNA in PBE. Total RNA was isolated from freshly isolated PBE, eosinophils cultured with IL-5 for 24 h, HL-60 cell line as bcl-2 positive control, and bcl-2 amplified by RT-PCR. GAPDH as internal RT-PCR control. Results representative of n = 3.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

A number of extracellular modulators of eosinophil apoptosis have been elucidated, but little is known of the intracellular mechanisms employed in the regulation of eosinophil programmed cell death. In this study, we examined the expression of proteins previously shown to have significant roles in the mediation of apoptosis, the death antagonists Bcl-2 and Bcl-x_L, and the death agonists Bax and Bcl-x_S. We observed that the PBE from healthy donors constitutively express relatively high levels of Bax determined by intracellular flow cytometry, immunoblotting, and RT-PCR. The bax gene has been shown to be expressed as numerous tissue-specific splice variants, which predict a membrane protein (, 21 kD) or cytosolic proteins (, 24 kD; , 5 kD) (47). Immunoblot analysis of eosinophil whole-cell lysates revealed the sole expression of a 21-kD protein, corresponding to the  isoform of Bax. Interestingly, immunoblot analysis with anti-Bax pAb consistently detected a 37-kD band in eosinophil whole-cell lysates. The nature of this protein is as yet undetermined, but it is unlikely to be a result of the homo-/heterodimerization of Bax observed in the regulation of apoptosis (36, 47) because of the stringent reducing conditions employed. It is possible that this band represents the newly described Bax  splice variant (48), although further investigation is required to confirm this.

Bcl-x protein was also readily detectable by flow cytometry. Alternative splicing of the bcl-x gene generates two distinct splice variants, a long bcl-x_L, and a short bcl-x_S. Variant bcl-x_S is created by a 63-amino acid deletion in the bcl-x_L open-reading frame (33). Variants bcl-x_L and bcl-x_S encode a death antagonist (28 kD) and a death agonist (17 kD), respectively. RT-PCR revealed that bcl-x_L was the predominating isoform, although bcl-x_S mRNA was also detectable. In this study, despite Bcl-x protein and mRNA being detectable by flow cytometry and RT-PCR, respectively, immunoblot analysis consistently failed to reveal the expression of either Bcl-x isoform in the whole-cell lysates of eosinophils and HL-60 cell line. The reason for this is unclear, as a fusion protein-positive control was detectable, possibly suggesting a lack of solubilization of the protein during cell lysis. Considerable effort to solubilize Bcl-x, including pellet homogenization, sonication, freeze fracture, and multiple detergent lysis proved unsuccessful. Attempts to immunoprecipitate the protein prior to immunoblotting also failed to detect Bcl-x specifically. Lack of solubilization may be due to compartmentalization within the cell, because complex interaction and membrane localization are a feature of Bcl-x (49, 50).

Druilhe and colleagues (44) reported a profile of apoptosis-associated proteins in PBE by immunoblot analysis, similarly failing to detect expression of Bcl-x isoforms with this method alone. However, intracellular flow cytometry and analysis of mRNA with respect to Bcl-x were not performed in this study. PBE were observed not to constitutively express Bcl-2 protein or mRNA, supporting the findings of a recent report (45). Immunoblot analysis has previously shown that peripheral blood neutrophils express Bax, with Bcl-2 and Bcl-x essentially absent (41). The lack of major death suppressor Bcl-2 may reflect the relatively short in vivo lifespan of these granulocytes, and the in vitro persistence of eosinophils relative to neutrophils (6) may potentially be afforded by the expression of Bcl-x_L in the former cell type.

We have shown that IL-5 increases eosinophil survival by inhibiting apoptosis, thereby supporting previous studies (6, 12). It has been reported that a requisite of IL-5- dependent abrogation of eosinophil apoptosis is the new synthesis of mRNA and protein (12). It has also been shown recently that Bcl-2 and Bcl-x_L can act as substrates for caspase-3, activated during apoptosis (52, 53). Therefore, we investigated the potential modulation of Bcl-2 members during spontaneous apoptosis and in the presence of survival-enhancing concentrations of IL-5. Spontaneous apoptosis was observed when culturing eosinophils in the absence of cytokines for 24 h (46.8 ± 7.4%), but no variation in the expression of Bcl-2, Bax, or Bcl-x was detected at the level of gene or protein. Culturing PBE in the presence of 100 pg/ml IL-5 for 24 h maintained cell survival (10.7 ± 2.6), and was found to induce detectable bcl-2 mRNA, but at a low level relative to GAPDH internal control and HL-60 bcl-2-positive control. Flow cytometry revealed a small but significant increase in Bcl-2 expression, without significant variation in -actin control, consistent with the observed increase in message. However, the small increase in Bcl-2 protein was undetectable by immunoblotting. There was no observable modulation in the expression of Bax and Bcl-x. This upregulation of Bcl-2 by IL-5 supported the findings of Ochiai and coworkers (45). However, a conflicting report (44) suggested that despite an increase in antiapoptosis protein Mcl-1 in interferon (IFN)--treated umbilical-cord-blood eosinophils, no modulation of Bcl-2 in either IL-5- or IFN--treated PBE was detectable, although immunoblot analysis was the only technique employed.

The susceptibility of a cell to death signals is determined by the ratio of death agonist:death antagonist, and their subsequent interaction via homo- and heterodimerization. Overexpression of Bax, leading to predomination of Bax homodimers, encourages apoptosis (47), whereas prevalence of Bcl-2 affords protection for the cell (53). The modest IL-5-dependent Bcl-2 increase, coupled with the relatively high constitutive expression of Bax in eosinophils, would suggest that upregulation of Bcl-2 is unlikely to be solely responsible for the protective effects of IL-5. This may implicate a role for other death antagonists in the regulation of eosinophil apoptosis, such as Bcl-x_L, determined to be the predominating Bcl-x isoform in freshly isolated eosinophils, or Mcl-1.

In conclusion, we have determined that freshly isolated PBE constitutively express predominantly Bax and Bcl-x_L, and to a lesser degree Bcl-x_S, but Bcl-2 is absent. This profile is distinct from that of peripheral blood neutrophils, which lack Bcl-x by immunoblot analysis. IL-5 induces a modest increase in the expression of death antagonist Bcl-2. The functions of the more recently described Bcl-2 members, such as Bad, Bcl-w, and Bag-1 (54), and the pivotal role of the caspases in eosinophil apoptosis are the subject of further investigation.