Sepsis is a high-risk clinical setting often resulting in multi-organ failure and death. Release of chromatin NETS (neutrophil extracellular traps) from neutrophils and the toxic role of highly-positively charged histones in late sepsis have been noted previously. Also, for NET formation to occur, peptidylarginine deiminase 4 activity must be present in the neutrophils, leading to citrullinated (cit) histones formation and loss of a portion of the positive charge. The four histones (H2A, H2B, H3 and H4) alone and as octamers of the four units tightly bind DNA. H3 and H4 histones as well as mixed octameric histones can induce a sepsis-like state in mice. One feature previously noted was that histones could inhibit activated protein C (aPC) production in the presence of thrombomodulin (TM). Since aPC generation is felt to protect against vascular damage, it was felt that this might - in part - account for the deleterious effects of histones in sepsis. We have shown that another highly-positive, small molecule, platelet factor 4 (PF4, CXCL4), which exists as a tetramer and which is stored in high concentrations in platelet alpha granules to be released in large amounts post-platelet activation, binds to the chondroitin sulfate (CS) side-chain of TM (TMCS) and enhanced aPC production along a bell-shaped curve with a peak effect around 25 µg/ml. Non-modified mixed histones had a similar bell-shaped effect on aPC generation and [histones + PF4] are additive on affecting aPC generation via TMCS. We wondered, because of this overlapping biology and the fact that significant levels of free PF4 are available in late sepsis, whether PF4 might affect other histone pathobiological pathways in late sepsis focusing on PF4’s interactions with non-modified and cit-histones. We first asked whether released PF4 might affect the binding of histones to DNA within NETS. We found that PF4 binds to DNA with greater affinity than histones in a competitive binding assay and that this effect was more marked for cit-histone consistent with its decreased positive charge. We then studied PF4 biology in three known targets of histone in sepsis. (1) In aPC generation, we examined cit-histones (either mixed, H3 or H4) relative to non-modified histones in stimulating aPC generation and found that they had a more limited effect on aPC generation with TMCS, but that again, PF4 cooperated in inducing aPC generation along a bell-shaped curve. (2) Histones are known to activate platelets (known to involve the toll-like receptor 4), likely contributing to the observed thrombocytopenia in late sepsis. We affirmed this affect with mixed histones and H4. Cit-mixed histones and cit-H4 also activated platelets in a platelet aggregation system, but much more weakly. PF4 had no effect on platelet activation by non-modified histones, but enhanced platelet activation by both cit-mixed histones and cit-H4. This was especially true for platelet activation studies with cit-H4 which on its own had nearly no affect on platelet activation though in the presence of moderate levels of PF4 (25 µg/ml), cit-H4 activated platelets as well as non-modified histones. (3) Finally, both non-modified and cit-histones activate endothelial cells (EC) by binding to their cell surfaces and likely contribute to the vascular damage of late sepsis. Using a microfluidic system involving controlled photochemical injury of the EC lining we found that PF4 enhanced the observed damage after cit-H4 exposure, but not notably after a comparable H4 exposure so that peak damage (as detected by propidium iodide staining) after cit-H4 approached that seen after H4 alone. In conclusion, NET formation involves citrullination of histones, and these modified histones likely contribute significantly to pathobiology in late sepsis. We now propose that in late sepsis, free histones, especially cit-histones, are mobilized out of NETs by PF4 because the PF4 binds DNA with higher affinity. After the histones and cit-histones are released from DNA, PF4 modifies the biology of these histones, especially the cit-histones enhancing their effects on aPC generation, platelet activation and EC injury. These studies provide additional insights of how histones achieve their pathobiological effects in sepsis. Such new insights may be critical for both understanding and monitoring clinical outcome and may lead to new therapeutic targets in sepsis.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution