In vitro activation of coagulation by human neutrophil DNA and histone proteins but not neutrophil extracellular traps

Abstract

NETosis is a physiologic process in which neutrophils release their nuclear material in the form of neutrophil extracellular traps (NETs). NETs have been reported to directly promote thrombosis in animal models. Although the effects of purified NET components including DNA, histone proteins, and neutrophil enzymes on coagulation have been characterized, the mechanism by which intact NETs promote thrombosis is largely unknown. In this study, human neutrophils were stimulated to produce NETs in platelet-free plasma (PFP) or in buffer using phorbol myristate actetate or calcium ionophore. DNA and histone proteins were also separately purified from normal human neutrophils and used to reconstitute chromatin using a salt-gradient dialysis method. Neutrophil stimulation resulted in robust NET release. In recalcified PFP, purified DNA triggered contact-dependent thrombin generation (TG) and amplified TG initiated by low concentrations of tissue factor. Similarly, in a buffer milieu, DNA initiated the contact pathway and amplified thrombin-dependent factor XI activation. Recombinant human histones H3 and H4 triggered TG in recalcified human plasma in a platelet-dependent manner. In contrast, neither intact NETs, reconstituted chromatin, individual nucleosome particles, nor octameric core histones reproduced any of these procoagulant effects. We conclude that unlike DNA or individual histone proteins, human intact NETs do not directly initiate or amplify coagulation in vitro. This difference is likely explained by the complex histone-histone and histone-DNA interactions within the nucleosome unit and higher-order supercoiled chromatin leading to neutralization of the negative charges on polyanionic DNA and modification of the binding properties of individual histone proteins.

Figure 1.

Fluorescence and immunofluorescence microscopy of stimulated human neutrophils for NETs formation. Stimulated normal human neutrophils were seeded on poly-l-lysine–coated slides and incubated at 37°C + 5% CO₂ for 3 hours. After incubation, the preparations were fixed using 4% paraformaldehyde and stained using 2.5 μM Sytox Green, a DNA staining dye. Unstimulated neutrophils did not form NETs (A) and were intact after treatment with DNAase (20 μM) (B). Robust NET release after stimulation with 600 nM PMA (C) or 5 μM A23187 (E) as shown by the bright and scattered Sytox Green signal pattern. Digestion of PMA-stimulated (D) and A23187-stimulated (F) NETs after treatment with DNAase (20 μg/mL), indicating the extracellular location of NETs. (G) Dual immunofluorescence staining of NETs by Sytox Green (2.5 μM) for DNA and citrillunated histone H3 (red). A23187, calcium ionophore.

Figure 2.

hnDNA and individual human histone H3 and H4 trigger coagulation in plasma. TG in recalcified normal PFP (A) and normal PRP (B) containing hnDNA. TG in recalcified FXII-deficient (FXII-Def), FXI-deficient (FXI-Def), or FVII-deficient (FVII-Def) PFP in the presence or absence of 30 µg/mL of hnDNA (C). No TG was observed in any individual deficient plasma after recalcification in the absence of DNA, represented by a single flat curve (PFP no DNA, panel C). Quantification of FXIa-AT after activation of the contact system by hnDNA in the synthetic contact system activation assay as described in “Methods” (D). TG in recalcified PRP containing recombinant human histone H3 (rH3) (E) or recombinant human histone H4 (rH4) (F). Effect of citrullination of histones H3 (G) and H4 (H) on TG in recalcified PRP. All the figures are representative of at least 3 independent experiments.

Figure 3.

Human intact NETs do not trigger coagulation in plasma or in a purified contact system. (A) TG in recalcified PFP containing unstimulated neutrophils (Unstim PMNs) or NETs prepared by stimulation of normal human neutrophils with PMA (600 nM), ionophore (A23187; 5 µM), or LPS (5 µg/mL) (B). TG was performed in recalcified plasma containing LPS-induced NETs in the presence of platelets (200 000/µL). TG was performed in PFP (C) and PRP (D) unstimulated neutrophils, washed NETs, or DNA purified from the same amount of washed NETs. NETs were washed to remove PMA or ionophore before being added to plasma (see “Methods”). Quantification of FXIa-AT in the SCSA in the presence of NETs (E). All figures are representative of at least 3 independent experiments. LPS, lipopolysaccharide.

Figure 4.

Purified hnDNA enhances low-concentration TF–initiated TG by amplifying thrombin-dependent FXI activation. Amplification effect of hnDNA on TG triggered by 1 pM TF in PFP (A) or in FXII-deficient plasma (Def-XII; 25 μg/mL hnDNA) (B) or in FXI-deficient plasma (Def-XI; 25 μg/mL hnDNA) (C). (D) Comparison of peak TG (mean + standard deviation) in the presence or absence of 25 μg/mL hnDNA in FXII- or FXI-deficient plasma. (E) Amplification effect of 50 μg/mL hnDNA on FXIa generation following 10 nM α-thrombin (IIa)–dependent activation of 30 nM FXI in the presence of physiologic concentration of HMWK in buffer. (F) Absence of an amplification effect of intact NETs on 1 pM TF-initiated TG in PFP. A23187, ionophore (5 μM); PMN, human neutrophils. PMA (600 nM). Panels A-C and F are representative of independent experiments performed at least 3 times. Mean values for 3 independent experiments in panels C and D.

Figure 5.

Individual human histones H3 and H4, but not octameric core histones, trigger TG in PRP. TG performed in recalcified PRP containing individual human histone proteins was compared with that of recombinant core histone octamers (A) or purified calf thymus histones (B). CThist, mixture of purified calf thymus histones; rOctamer, octameric core histone reconstituted with recombinant human histone proteins. Panels A and B are representative of 3 independent experiments.

Figure 6.

Purified hnDNA, but not mononucleosome particles, trigger TG in plasma. (A) TG generation performed in recalcified PFP containing hnDNA or recombinant mononucleosomes. (B) TG generation performed in recalcified PRP containing hnDNA or recombinant mononucleosomes. rmNucl, mononucleosomes consisting of reconstituted recombinant human histone proteins and DNA. The unit of rmNucl is expressed as the ratio of final concentrations of histones to DNA. Panels A and B are representative of 3 independent experiments.

Figure 7.

Effect of chromatin reconstitution on purified DNA-dependent activation of coagulation. Chromatin was reconstituted using DNA and histones purified from normal human neutrophils. (A) Detection of nucleosomes in samples following reconstitution; positive and negative controls are provided with the ELISA kit. (B) TG performed in recalcified PFP in the presence or absence of 30 μg/mL kaolin (positive control), 50 μg/mL dialyzed hnDNA, dialyzed histones, or reconstituted chromatin. (C) Effect of dialyzed hnDNA, histones, and reconstituted chromatin on FXIa generation in the presence of physiologic concentrations of FXII, FXI, and HMWK in buffer. (D) Effect of dialyzed hnDNA, histones, and reconstituted chromatin on 10 nM α-thrombin-dependent activation of FXI (30 nM) in buffer in the presence of physiologic concentrations of HMWK. hnDNA, purified hnDNA. Panel B is representative of 3 independent experiments. Mean ± standard deviation of 3 independent experiments in panels A, C, and D.