Figure 1.
Activation of Eosinophils by NK cells.
Autologous NK cells and eosinophils were cultured at the ratio NK cells: eosinophils (NK: Eos ratio) 0∶1 (i.e. eosinophils alone), 1∶1, 5∶1 and 10∶1 in the presence of interleukin-5 for 3 hours and 12 hours. (A) Gating strategy: live eosinophils were identified firstly on the basis of their size difference (FSC) and granularity (SSC) and secondly as Annexin V− 7AAD− cells. CD63 expression (black histogram) is shown on a representative experiment for the ratio NK cells: eosinophils (NK: Eos ratio) 0∶1 (i.e. eosinophils alone) and 10∶1. Isotypic control antibody is represented in dashed grey. Eosinophils were stained for CD63 (B), CD62L (C) and CD69 (D) expression. Results are shown for each donor and mean is indicated with the black line. One-way Anova (Friedman) tests were performed, followed by Dunn’s post tests. *p<0.05; ***p<0.001.
Figure 2.
NK cells induce eosinophil release of ECP and EDN.
Levels of eosinophil cationic protein (ECP) (A) and eosinophil derived neurotoxin (EDN) (B) were measured in co-culture supernatants after 3 and 12 hours for the NK:Eos ratios 1∶0 (i.e. NK cells alone), 0∶1 (i.e. eosinophils alone); 1∶1; 5∶1; 10∶1. Results are expressed as molecule quantity in ng/ml for 2×105 cells for each donor and mean is indicated with the black line. One-way Anova (Friedman) tests were performed, followed by Dunn’s post tests. *p<0.05; **p<0.01; ***p<0.001.
Figure 3.
NK cells induce eosinophil apoptosis.
After purification, NK cells and eosinophils were cultured in the presence of IL-5 for 3 and 12 h at different NK:Eos ratios: 0∶1 (i.e. eosinophils alone); 1∶1; 5∶1; 10∶1. Eosinophils were identified on the basis of their size difference (FSC) and granularity (SSC). Apoptosis and death of eosinophils was evaluated by Annexin/PI labeling. Dead eosinophils were defined as AnnexinV+ cells. Results are expressed as mean percentage of AnnexinV+ cells amongt the eosinophil population ± SEM. n = 16 (3 hours) or 18 (12 hours). Gaussian distributions were verified using D’Agostino and Pearson test, and One-way Anova tests were performed, followed by Bonferroni’s post tests. *p<0.05; **p<0.01; ***p<0.001.
Figure 4.
Eosinophil degranulation and apoptosis are both abrogated after PFA-induced fixation of NK cells and separation of cell types by transwell.
After purification, NK cells and eosinophils were cultured in the presence of IL-5 for 3 hours at the NK:Eos ratios: 0∶1 (i.e. eosinophils alone) and 10∶1. (A) Percentage of CD63 positive eosinophils cultured with NK cells at the 0∶1 (i.e. eosinophils alone) or 10∶1 NK:Eos ratios in the same culture well (direct contact) or in Transwell (separated with 3 μm filter). (B) Percentage of CD63 positive eosinophils cultured alone, with live NK cells or with PFA-fixed NK cells for the 10∶1 NK:Eos ratio. (C) Percentage of AnnexinV+ eosinophils cultured with NK cells at the 0∶1 (i.e. eosinophils alone) or 10∶1 NK:Eos ratios in the same culture well (direct contact) or in Transwell (separated with 3 μm filter). (D) Percentage of AnnexinV+ eosinophils cultured alone, with live NK cells or with PFA-fixed NK cells for the 10∶1 NK:Eos ratio. Results are shown for each donor and mean is indicated with the black line. One-way Anova (Friedman) tests were performed, followed by Dunn’s post tests. *p<0.05; **p<0.01; ***p<0.001.
Figure 5.
MAPK and PI3K pathways are involved in eosinophil degranulation.
AnnexinV− eosinophils (A) and NK cells (B), cultured alone or together at the 10∶1 NK:Eos ratio, were stained with anti-ERK or anti-phosphorylated ERK (pERK) antibodies. Eosinophils and NK cells were identified on the basis of their size difference (FSC) and granularity (SSC). Live eosinophils were defined as AnnexinV− cells. Gray full lines and black dotted lines represent control antibody staining for each cell type alone or in co-culture, respectively. Gray plain histograms and black full lines represent staining with antibody of interest for each cell type alone or in co-culture, respectively. One representative experiment out of three is shown. (C) Eosinophils were pre-incubated with DMSO as control or with inhibitors of signaling pathways for 5 min at 37°C: PD 98059 25 μM an inhibitor of ERK pathway, SB 203580 1 μM an inhibitor of P38 MAPK, SP 600125 20 μM an inhibitor of JNK pathway and LY 294002 10 μM an inhibitor of PI3K pathway for the NK:Eos ratio 10∶1. Results show percentage of CD63+ eosinophils for each donor and mean is indicated with the black line. One-way Anova (Friedman) tests were performed, followed by Dunn’s post tests. *p<0.05, **p<0.01.
Figure 6.
NK cell-induced eosinophil apoptosis involves mitochondrial reactive oxygen species (ROS).
(A) After purification, NK cells and eosinophils were cultured in the presence of IL-5 for 12 hours at different NK:Eos ratios: 0∶1 (i.e. eosinophils alone); 1∶1; 5∶1; 10∶1. Eosinophils were identified on the basis of their size difference (FSC) and granularity (SSC). ROS staining was achieved using dihydroethidium (HE). Results are expressed as mean percentage of HE+ cells amongt the eosinophil population ± SEM. n = 12. Gaussian distributions were verified using D’Agostino and Pearson test, and One-way Anova test was performed, followed by Bonferroni’s post test. *p<0.05; **p<0.01. (B) Eosinophils were incubated with inhibitors of mitochondrial electron transport: rotenone and antimycin for 30 min at 37°C before 12 h co-culture with NK cells at the NK:Eos ratio of 10∶1. Results for each donor are shown and mean is indicated with the black line. Wilcoxon tests were performed. *p<0.05.