Eosinophils as a novel cell source of prostaglandin D2: autocrine role in allergic inflammation

Abstract

PGD(2) is a key mediator of allergic inflammatory diseases that is mainly synthesized by mast cells, which constitutively express high levels of the terminal enzyme involved in PGD(2) synthesis, the hematopoietic PGD synthase (H-PGDS). In this study, we investigated whether eosinophils are also able to synthesize, and therefore, supply biologically active PGD(2). PGD(2) synthesis was evaluated within human blood eosinophils, in vitro differentiated mouse eosinophils, and eosinophils infiltrating inflammatory site of mouse allergic reaction. Biological function of eosinophil-derived PGD(2) was studied by employing inhibitors of synthesis and activity. Constitutive expression of H-PGDS was found within nonstimulated human circulating eosinophils. Acute stimulation of human eosinophils with A23187 (0.1-5 μM) evoked PGD(2) synthesis, which was located at the nuclear envelope and was inhibited by pretreatment with HQL-79 (10 μM), a specific H-PGDS inhibitor. Prestimulation of human eosinophils with arachidonic acid (10 μM) or human eotaxin (6 nM) also enhanced HQL-79-sensitive PGD(2) synthesis, which, by acting on membrane-expressed specific receptors (D prostanoid receptors 1 and 2), displayed an autocrine/paracrine ability to trigger leukotriene C(4) synthesis and lipid body biogenesis, hallmark events of eosinophil activation. In vitro differentiated mouse eosinophils also synthesized paracrine/autocrine active PGD(2) in response to arachidonic acid stimulation. In vivo, at late time point of the allergic reaction, infiltrating eosinophils found at the inflammatory site appeared as an auxiliary PGD(2)-synthesizing cell population. Our findings reveal that eosinophils are indeed able to synthesize and secrete PGD(2), hence representing during allergic inflammation an extra cell source of PGD(2), which functions as an autocrine signal for eosinophil activation.

Figure 1. Human eosinophils are able to synthesize PGD₂ in an H-PGDS-dependent manner

Panel A presents epifluorescence images of cytoplasmic immuno-detection of H-PGDS (green) in non-stimulated human eosinophils. Blue fluorescence shows eosinophil nuclei stained with DAPI. Eosinophils incubated with isotype irrelevant IgG are shown. Panel B shows PGD₂ amounts secreted by human eosinophils stimulated for 15 min with A23187 (0.1 to 5 μM). In C, eosinophils were pre-treated for 30 min with HQL-79 and then stimulated with 5 μM of A23187. Panel D displays confocal images of intracellular immuno-fluorescence for PGD₂ in non-stimulated, A23187 (0.1 μM)-stimulated, and HQL-79-treated A23187-stimulated human eosinophils (as indicated). Arrows indicate PGD₂ immuno-staining at the nuclear envelope of A23187-stimulated human eosinophils. Panel E shows production of PGD₂ and PGE₂ by human eosinophils stimulated for 1 h with AA (10 μM) and then challenged with A23187 (0.1 μM). Eosinophils were pre-treated for 30 min with HQL-79. In F, confocal images of EicosaCell preparations display intracellular immuno-detection of newly formed PGD₂ (green) and of ADRP (red) in human eosinophils stimulated with AA (10 μM). Eosinophils were pre-treated with HQL-79 for 30 min. Overlay images of identical fields are shown in the larger images. In G, constitutive levels of H-PGDS mRNA and its up-regulation in human eosinophils that were stimulated for 1 h with AA (10 μM) were assessed by RT-PCR. Values are expressed as means ± SEM of at least three distinct donors. ⁺ P ≤ 0.05 compared with control. ^* P ≤ 0.05 compared with A23187- or AA-stimulated eosinophils. All the images are representative of three independent experiments with distinct donors. Bar, 5 μm.

Figure 2. Eotaxin elicits production of biologically active PGD₂

Both human and mouse eosinophils were pre-treated for 30 min with HQL-79 before eotaxin stimulation. In A and B, for analysis of PGD₂ and cysLTs synthesis, human eosinophils were stimulated for 1 h with eotaxin (6 nM) and then challenged with A23187 (0.1 μM). In C, for analysis of lipid body biogenesis, human eosinophils were stimulated for 1 h with eotaxin (6 nM). In D, confocal images of EicosaCell preparations display intracellular immuno-detection of newly formed PGD₂ (green) in human eosinophils stimulated with rh eotaxin (6 nM). Overlay images of immuno-fluorescence and light microscopy of identical fields are shown in the larger images. In E, for analysis of PGD₂ synthesis, mouse eosinophils were stimulated for 1 h with rm eotaxin (6 nM). Panel F shows confocal images of PGD₂ immuno-detection of H-PDS (green) in mouse eosinophils stimulated with rm eotaxin (6 nM). Blue fluorescence shows eosinophil nuclei stained with DAPI. Bar, 5 μm. Results are expressed as means ± SEM for at least three independent experiments with eosinophils from distinct donors. ⁺ P ≤ 0.05 compared with control. ^* P ≤ 0.05 compared with eotaxin-stimulated eosinophils.

Figure 3. Endogenous eosinophil-derived PGD₂ mediates AA- and eotaxin-, but not PAF or PGD₂-induced lipid body biogenesis

For in vitro analysis of lipid body biogenesis, human (A, B, C and D) or mouse (E, F and G) eosinophils were pre-treated for 30 min with HQL-79 and then stimulated for 1 h with AA (10 μM; A and F), MIF (50 ng/mL; B), PAF (1 μM; C and G) or PGD₂ (25 nM; D and E). In vitro results are expressed as means ± SEM for at least three independent experiments with eosinophils from distinct donors. ⁺ P ≤ 0.05 compared with control. ^* P ≤ 0.05 compared with stimulated eosinophils.

Figure 4. Eosinophil-derived PGD₂ controls eosinophil activation via interaction with specific PGD₂ receptors

In A, for analysis of cysLTs production, human eosinophils were pre-treated for 30 min with HQL-79 or Bay-u3405, stimulated for 1 h with AA (10 μM) and then challenged with A23187 (0.1 μM). From B to F, for in vitro analysis of lipid body biogenesis, human eosinophils were pre-treated for 30 min with neutralizing anti-PGD₂ antibody, BWA868c, Bay-u3405 or Cay10471 and then stimulated for 1 h with AA, eotaxin, or PAF, as indicated. The results are expressed as the means ± SEM for at least three independent experiments with eosinophils purified from distinct donors. ⁺ P ≤ 0.05 compared with control. ^* P ≤ 0.05 compared with stimulated eosinophils.

Figure 5. Infiltrating eosinophils are an auxiliary cell source of PGD₂ at sites of allergic inflammation

In A and B, for in vivo analysis of eosinophil influx and PGD₂ production, respectively, sensitized mice were sacrificed 48 h after allergic challenge with ovalbumin. Animals received HQL-79 1 h prior to allergic challenge. In vivo results are expressed as the means ± SEM for at least eight animals. ⁺ P ≤ 0.05 compared with control. ^* P ≤ 0.05 compared with ovalbumin-challenged animals. In C, EicosaCell images show intracellular immuno-fluorescence for PGD₂ pleural leukocytes recovered from non- and HQL-79-treated ovalbumin-challenged mice (as indicated). Overlay images of identical phase contrast fields are shown to facilitate the identification of the immuno-fluorescent cell type. E, M, and MC indicate eosinophils, macrophages, and mast cells, respectively. Arrows show immuno-labeled PGD₂ within a resident mast cell. Bar, 10 μm. In D, epifluorescence image (left panel) shows allergic infiltrating eosinohils as SiglecF⁺PE-labeled (red staining) cells with DAPI-stained polymorphic nuclei (blue staining). Right panel shows EicosaCell analysis of PGD₂ synthesis within SiglecF⁺ pleural eosinophils found in the site of allergic inflammation of HQL-treated and non-treated animals. Histogram is a representative data of 5 animals per group.