Circulating Iron in Patients with Sickle Cell Disease Mediates the Release of Neutrophil Extracellular Traps

Abstract

Introduction: Neutrophils promote chronic inflammation and release neutrophil extracellular traps (NETs) that can drive inflammatory responses. Inflammation influences progression of sickle cell disease (SCD), and a role for NETs has been suggested in the onset of vaso-occlusive crisis (VOC). We aimed to identify factors in the circulation of these patients that provoke NET release, with a focus on triggers associated with hemolysis.

Methods: Paired serum and plasma samples during VOC and steady state of 18 SCD patients (HbSS/HbSβ⁰-thal and HbSC/HbSβ⁺-thal) were collected. Cell-free heme, hemopexin, and labile plasma iron have been measured in the plasma samples of the SCD patients. NETs formation by human neutrophils from healthy donors induced by serum of SCD patients was studied using confocal microscopy and staining for extracellular DNA using Sytox, followed by quantification of surface coverage using ImageJ.

Results: Eighteen patients paired samples obtained during VOC and steady state were available (11 HbSS/HbSβ⁰-thal and 7 HbSC/HbSβ⁺-thal). We observed high levels of systemic heme and iron, concomitant with low levels of the heme-scavenger hemopexin in sera of patients with SCD, both during VOC and in steady state. In our in vitro experiments, neutrophils released NETs when exposed to sera from SCD patients. The release of NETs was associated with high levels of circulating iron in these sera. Although hemin triggered NET formation in vitro, addition of hemopexin to scavenge heme did not suppress NET release in SCD sera. By contrast, the iron scavengers deferoxamine and apotransferrin attenuated NET formation in a significant proportion of SCD sera.

Discussion: Our results suggest that redox-active iron in the circulation of non-transfusion-dependent SCD patients activates neutrophils to release NETs, and hence, exerts a direct pro-inflammatory effect. Thus, we propose that chelation of iron requires further investigation as a therapeutic strategy in SCD.

Keywords: Hemolysis; Iron; Neutrophil extracellular traps; Neutrophils; Sickle cell disease.

Fig. 1

Neutrophils release more NETs when exposed to sera of patients obtained during VOC compared to the steady state. a Serum levels of Hpx in healthy control (HD; n = 20) and SCD patients in the steady state and VOC (n = 17/group). One-way ANOVA with Tukey's multiple comparisons post hoc test was used. ****p < 0.0001. b Levels of serum-free heme in HD (n = 20) versus patients with SCD in the steady state and VOC (n = 17 in each group). A Kruskal-Wallis test with Dunn's post hoc testing was used. *p < 0.05. n.s. indicates not significant. c Levels of LPI in HD (n = 19) versus patients with SCD in the steady state and VOC (n = 19 for each group). We used a Kruskal-Wallis test with Dunn's post hoc testing for statistical analysis. *p < 0.05, **p < 0.01. n.s. indicates not significant. d Levels of IgM in serum of HD (n = 20) and SCD patients in the steady state and VOC (n = 17/group). e Immunostaining for citrullinated histone H3 (CitH3) after exposure to SCD serum. Neutrophils from a HD were exposed to serum from a SCD patient in VOC for 180 min. Then, extracellular DNA was stained with Sytox Green (green). Immunostaining was performed on NETs induced by 3 different SCD sera. Representative images are shown. Scale bar, 50 μm. Original magnifications ×40. f Quantification of extracellular DNA release in response to sera from SCD patients. A Wilcoxon matched-pairs signed rank test was used for statistical analysis. Incubations of neutrophils with paired SCD sera (n = 18) were performed with neutrophils from 3 different HD for each subject group (steady state and crisis). **p = 0.0087.

Fig. 2

Plasma-purified hemopexin does not prevent the release of NETs in sera of patients with SCD. a Quantification of NET release in response to sera from SCD patients in the presence of plasma-derived Hpx. The densities of extracellular NET-DNA over the image area (i.e., the number of Sytox Green+ pixels divided by the total number of pixels × 100) were determined for paired sera from patients with SCD in the presence or absence of 50-μM plasma-purified Hpx. A Wilcoxon matched-pairs signed rank test was used to compare NET release in response to SCD sera in the presence or absence of Hpx. Incubations of neutrophils with HD (n = 6) or SCD sera (n = 18) were performed with neutrophils from 3 different HD for each subject group. n.s. indicates not significant. b, c Correlations between levels of circulating heme and NET release in sera from SCD patients (n = 17) in the steady state (b) and crisis (c). A Spearman test was used to calculate correlation coefficients. p = 0.6873 and **p = 0.0011 in the steady state and crisis, respectively.

Fig. 3

Iron triggers NET formation, and iron-mediated NET release is blocked by chelation with deferoxamine. a Neutrophils isolated from HD were incubated with medium alone (ctrl) or challenged with 50-μM hemin or protoporphyrin IX (PPIX) for 180 minutes. Release of NETs (green in these images) was detected by fluorescence imaging with confocal microscopy using a mixture of 2 DNA-labeling dyes, one cell impermeable (Sytox Green, green) and the other cell permeable (Hoechst 33,342, blue). Depicted are merged images of green and blue fluorescence. All images are representative of 2 independent experiments using neutrophils from different HD. Scale bars, 50 μm. b NET formation was quantified after exposure to hemin or PPIX. The densities of extracellular NET-DNA over the image area (i.e., the number of Sytox Green+ pixels divided by the total number of pixels × 100) were determined after the challenge with hemin or PPIX and depicted as mean NET density ± SD in 2 separate experiments. c In 2 independent experiments, neutrophils from a HD were exposed to 50-μM FeNTA or hemin in the presence or absence of equimolar amounts of deferoxamine (DFO). After 180 minutes, NETs (green in these images) were visualized with confocal fluorescence microscopy as in panel a. Depicted are representative images in which Sytox Green (green) and Hoechst 33,342 (blue) fluorescence are merged. Scale bars, 50 μm. d NET release was quantified as in panel b and depicted as mean NET density ± SD (n = 2). Original magnifications ×20 for panels a, c.

Fig. 4

Iron chelation abrogates NET release in sera from SCD patients. a Neutrophils from a HD were exposed to serum from a nonautologous HD or patients with SCD during VOC for 180 minutes in the presence or absence of deferoxamine (DFO, 50 μM) or apotransferrin (apoTf, 50 μM). Release of NETs (green in these images) was visualized with confocal fluorescence microscopy using 2 DNA-labeling dyes, one cell impermeable (Sytox Green), and the other cell permeable (Hoechst 33342). Depicted are merged images of Sytox Green (green) and Hoechst 33342 (blue) fluorescence. All images are representative of experiments performed with sera from 11 different patients. Scale bars, 50 μm. Original magnifications. ×20. b Quantification of NET release in response to sera from SCD patients in the presence of iron chelators. The densities of extracellular NET-DNA over the image area (i.e., the number of Sytox Green⁺ pixels divided by the total number of pixels × 100) were determined for sera from patients with SCD during VOC in the presence or absence of 50-μM DFO or apoTf. Incubations of neutrophils with SCD sera (n = 11) were performed with neutrophils from 3 different HD for each subject group. n.d. indicates not determined.