In response to infection and inflammation, neutrophils release NETs, histone-decorated nuclear DNA that ensnares bacteria, but also damages host tissues and promotes thrombosis. Platelet Factor 4 (PF4, CXCL4) is a CXC chemokine stored in high concentrations in platelet alpha-granules and released during platelet activation. Tetrameric PF4 has a very high affinity for heparans and other polyanionic molecules, including DNA. At the proper molar ratio, PF4 can form high molecular weight complexes with heparin and other heparans, and these complexes are antigenic targets for pathogenic HIT antibodies. In light of this information, we chose to investigate whether PF4 can interact with NETs and whether these complexes contribute to the prothrombotic nature of HIT. Using an ELISA assay, we confirm that PF4 binds to DNA to form HIT-like complexes just as it does with heparin following a similar bell-shaped curve of HIT antigenicity. Using immunofluorescence studies and confocal microscopy, we found that exogenous PF4 adheres readily to NET DNA. We then investigated PF4-NET interactions under intravascular flow conditions by using neutrophils isolated from healthy human donors to create NET-coated microfluidic channels through which we infused recombinant human PF4. We found that PF4 selectively adhered to extracellular NET fibers but did not bind to the surface of intact neutrophils. We next noted that PF4 infusion led to a change in NET morphology with compaction of the extracellular DNA to approximately 30% of the original area (p<0.001, N=10 per arm). KKO, a monoclonal anti-PF4-heparin antibody, bound readily to NETs following PF4 incubation, indicating that PF4-NET complexes are antigenic. Of note, KKO binding did not induce additional NET compaction. We then observed that while NETs were highly susceptible to endonuclease digestion prior to PF4 incubation, PF4-NET complexes developed resistance to endonuclease digestion. PF4-NET complex incubation with either KKO or HIT IgG isolated from patient samples further enhanced resistance to endonuclease digestion by >2-fold (p<0.001, N=2 studies, 5 NETS per study), while incubation with a polyclonal anti-PF4 antibody did not have this effect. We also show that neutrophils bind readily to endothelial cells when whole blood was flowed through microfluidic channels lined with TNFα-injured endothelium. The number of adherent neutrophils increased markedly when KKO was added to the whole blood, but no significant changes were observed when an isotype control antibody was included. These findings suggest that neutrophils may contribute to thrombosis in HIT through three sequential steps: (1) HIT-antibody activated neutrophils selectively bind to injured endothelium, increasing the numbers of localized neutrophils at sites of thrombus. (2) Subsequently released NETs are bound by PF4 and HIT antibodies, generating immunogenic complexes that are likely prothrombotic, and finally, (3) PF4 and HIT antibody binding induces resistance to endonuclease digestion leading to a prolongation of NET half-life and an increased opportunity to contribute to clot formation. We believe these data support a set of mechanisms by which neutrophils can contribute to the observed prothrombotic nature of HIT, including a novel NET stabilization process. Further in vivo mouse models will now be pursued to confirm these findings.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution