Neutralizing the pathological effects of extracellular histones with small polyanions

Abstract

Extracellular histones in neutrophil extracellular traps (NETs) or in chromatin from injured tissues are highly pathological, particularly when liberated by DNases. We report the development of small polyanions (SPAs) (~0.9-1.4 kDa) that interact electrostatically with histones, neutralizing their pathological effects. In vitro, SPAs inhibited the cytotoxic, platelet-activating and erythrocyte-damaging effects of histones, mechanistic studies revealing that SPAs block disruption of lipid-bilayers by histones. In vivo, SPAs significantly inhibited sepsis, deep-vein thrombosis, and cardiac and tissue-flap models of ischemia-reperfusion injury (IRI), but appeared to differ in their capacity to neutralize NET-bound versus free histones. Analysis of sera from sepsis and cardiac IRI patients supported these differential findings. Further investigations revealed this effect was likely due to the ability of certain SPAs to displace histones from NETs, thus destabilising the structure. Finally, based on our work, a non-toxic SPA that inhibits both NET-bound and free histone mediated pathologies was identified for clinical development.

Fig. 1. Polyanions inhibit histone-mediated endothelial cell cytotoxicity and RBC fragility with minimal structural requirements for the activity identified.

a Human endothelial cells (HMEC-1) (left panel) or red blood cells (RBCs) (right panel) were incubated (1 h, 37 °C) with histones (400 μg ml⁻¹) in the presence of different polyanion concentrations and IC₅₀ values (mean of three separate determinations ± s.e.m.) for HMEC-1 cytotoxicity or RBC fragility determined. Heparan sulfate (lo) and (hi) represent low and highly sulfated heparan sulfate. b Chemical structures of compounds selected for future study. c Sulfated disaccharide mCBS has minimal structural requirements to approximate inhibitory effect of heparin on histone-mediated cytotoxicity and RBC fragility, the monosaccharides, methyl β-glucoside per-O-sulfate and glucose per-O-sulfate having little or no activity. d Effect of the level of sulfation of mCB (di-, tri-, tetra-, and penta-O-sulfated) on its ability to be an effective inhibitor of histone-mediated cytotoxicity and RBC fragility compared to fully sulfated (hepta-O-sulfated) mCBS. Data are presented as mean ± s.e.m. (n = 3) and analyzed by two-way ANOVA with Tukey’s correction for multiple comparisons. Source data are provided as a Source Data File.

Fig. 2. Histones promote RBC and platelet aggregation, reduce RBC deformability, and induce platelet degranulation, processes that are inhibited by SPAs.

a Human RBCs were incubated for 1 h at 37 °C alone (control) or in the presence of histones (400 μg ml⁻¹) with or without mCBS (200 μg ml⁻¹). Percentage of RBC aggregation measured by flow cytometry based on forward (FSC) and side (SSC) scatter parameters and an appropriate gating strategy to discriminate aggregated from normal (non-aggregated) RBC. b Scanning electron micrographs depicting the level of RBC aggregation at low and high magnification following the three treatments (n = 1) depicted in a and each representative of three fields per magnification. c Concentration-dependent inhibition of histone-mediated RBC aggregation by mCBS and MTS, in this case RBC aggregation being calculated by flow cytometry as fold change in RBC auto-fluorescence relative to RBC in the absence of histones. The fold change values also provide an estimate of the number of RBC in each aggregate. Data were analyzed by two-way ANOVA with Sidak’s correction for multiple comparisons (n = 3). d Retention of RBC in an artificial spleen that measures RBC deformability. RBCs were incubated with increasing concentrations of histones for 1 h and, at the highest concentration used (400 μg ml⁻¹), also incubated with either mCBS or MTS (100 and 200 μg ml⁻¹), prior to passage through the artificial spleen. Data are presented as mean ± s.e.m. (n = 2–3) and analyzed by one-way ANOVA with Dunnett’s multiple comparisons tests. e Histone-induced aggregation of isolated platelets. f SPA inhibition of histone (HIS) (150 μg ml⁻¹) induced platelet aggregation. Data were analyzed by two-way ANOVA with Dunnett’s correction for multiple comparisons (n = 6). g Histone-induced degranulation of platelets in whole blood, as measured by ATP release. Dotted line ATP release from thrombin-activated platelets (n = 1). h SPA inhibition of histone-induced platelet degranulation. Data in g, h are representative of one of two experiments. Source data are provided as a Source Data File.

Fig. 3. Histone-mediated cytotoxicity for cells does not require cell surface heparan sulfate, histones directly disrupting lipid bilayers and inducing a cellular Ca²⁺ flux that can be blocked by SPAs.

a Effect of removal of cell surface heparan sulfate by bacterial heparinases 1, 2, and 3 or human platelet heparanase on the sensitivity of HMEC-1 to histone cytotoxicity. b Sensitivity of wild-type CHO-K1 and GAG-deficient pgsA-745 CHO-K1 cells to histone cytotoxicity. Data are presented as mean ± s.e.m. (n = 3) and analyzed by two-way ANOVA with Sidak’s multiple comparisons test. c Lifetime of artificial lipid bilayers exposed to histones (HIS) (1 μΜ) alone (n = 47) or in the presence of the SPAs CBS (n = 52) or MTS (n = 40) (10 μM). Control bilayers (n = 125) contained the RγR1 ion channel protein. Data are presented as mean ± s.e.m. and analyzed by non-parametric Kruskal–Wallis test. d Upper panel: Representative flow cytometry plots, using Ca²⁺-sensitive dye Indo-1, showing Ca²⁺ fluxing HMEC-1 1 min following histone addition (100 μg ml⁻¹). Lower panel: Time course of effect of CBS and MTS (100 μg ml⁻¹) on histone-induced Ca²⁺ flux by HMEC-1. Data are presented as mean ± s.e.m. (n = 2) from one of two experiments. Source data are provided as a Source Data File.

Fig. 4. In vivo SPAs inhibit histone-induced tissue injury, thrombocytopenia, anemia, and DVT.

a Mice injected i.p. with SPA and heparin (Hep) doses (as indicated) 10 min prior to i.v. injection of histones (50 mg kg⁻¹), had blood collected 4 h post histones for assessment of liver (alanine aminotransferase, ALT), kidney (creatinine), and general tissue (lactate dehydrogenase, LDH) damage. Data pooled from 15 separate experiments, with n = 2–50 mice per treatment. b Mice (n = 5/group), treated as above but receiving one SPA dosage (100 mg kg⁻¹), had their blood and spleens collected 10 min post histones for assessment of circulating platelets and RBC and splenic hemoglobin (Hb). c Impact of CBS and MTS on a mouse model of histone-induced DVT (n = 7–10 mice per group). All data, except LDH values, are presented as mean ± s.e.m. and analyzed by one-way ANOVA with Dunnett’s correction for multiple comparisons. LDH values at or above the detection limit of 3325 U l⁻¹ are reported in the LDH data of panel a. For some of these samples, sufficient material was available for dilution to obtain accurate LDH levels. The LDH data are presented as median with 95% confidence intervals, instead of the mean that may be biased by the values at the detection limit. The data above the detection limit was set to 3325 U l⁻¹ prior to performing a non-parametric Kruskal–Wallis test with Dunn’s correction for multiple comparisons. Source data are provided as a Source Data File.

Fig. 5. In vivo SPAs inhibit ongoing tissue injury induced by histones.

a Mice (n = 3/group) were initially injected i.v. with histones (50 mg kg⁻¹), and 2 h later, when ongoing tissue injury was evident, they were injected i.p. with mCBS or MTS (100 mg kg⁻¹), with plasma collected 30 min later and analyzed for content of ALT, LDH, creatinine, and Hb. b Mice were treated as in a but 5 min before sacrifice they were injected i.p. with propidium iodide (PI), and liver, lung, and kidneys were collected and examined by confocal fluorescence microscopy for dead (fluorescent red) cells, with representative fields from the various treatments being shown; scale bar 250 µm. c Same animals were quantified for the number of dead (PI+) cells per field in the different treatment groups (n = 4 mice per group), with at least four fields quantified/organ. Data are presented as mean ± s.e.m. and analyzed by one-way ANOVA with Dunnett’s correction for multiple comparisons. Source data are provided as a Source Data File.

Fig. 6. SPAs inhibit a range of pathologies involving extracellular histones.

a Survival of rats (n = 8 per group) subjected to cecal ligation and puncture (CLP) and receiving saline (Control), CBS, or MTS. P values obtained with survival analysis log-rank (Mantel–Cox) test. b Kidney and liver damage in CLP rats, as measured by ALT and creatinine blood levels (n = 4–8/group, blood could not be collected or was clotted from some rats). c Effect of CBS and MTS (n = 6–12/group) on cardiac IRI in rats, with the area at risk (AAR) in left ventricle (LV), microvascular obstruction (MVO), and infarct area being measured. d Effect of mCBS and MTS on a skin flap model of IRI in rats (n = 3–5/group), with representative photos shown (skin flap reduced to 25% normal size). Data are presented as mean ± s.e.m. and analyzed by one-way ANOVA with Dunnett’s correction for multiple comparisons. Source data are provided as a Source Data File.

Fig. 7. Sera from sepsis and acute myocardial infarction patients inhibit endothelial cell proliferation, an effect neutralized by DNase I, anti-histone antibodies, and SPAs.

a Proliferation of HMEC-1, as measured by ³H-thymidine incorporation, to detect the anti-proliferative activity of histones, which is histone concentration dependent (left panel, n = 3), but is totally neutralized by mCBS and MTS (right panel, n = 3). b Correlation (Spearman’s r value) of APACHE II scores with anti-proliferative effect of sepsis patients sera on HMEC-1 (n = 20 patients). c Correlation of APACHE II scores with extracellular DNA content of sepsis patient’s sera, with serum from patient 5 (red circle) having greatest anti-proliferative activity (b) and DNA content (c). d Effect of DNase I or pAbs against histone 3 (αH3 10 μg ml⁻¹) and histone 4 (αH4 10 μg ml⁻¹) (n = 4/treatment) on anti-proliferative effect of serum from sepsis patient 5. e Ability of SPAs CBS and MTS to neutralize the anti-proliferative effect of sepsis patient sera (SS) (n = 10 patients). f Extracellular DNA content of occlusion site sera from STEMI patients (n = 12) and stratified according to low (<35 µmol l⁻¹, n = 4) and high (>35 µmol l⁻¹, n = 8) troponin I (Trop) levels. Control sera (n = 3) from the peripheral blood of healthy volunteers. g Ability of SPAs CBS and MTS to neutralize the HMEC-1 anti-proliferative effect of sera from high troponin I STEMI patients (CS) (n = 7 patients). Data are presented as mean ± s.e.m. and analyzed by one-way ANOVA with Dunnett’s correction for multiple comparisons. Source data are provided as a Source Data File.

Fig. 8. Polyanion-induced changes in chromatin and NETs.

a Enhanced uptake of the fluorescent DNA-specific dye, Sytox Green, by chicken chromatin (CC) following exposure to the polyanions MTS and heparin, but not mCBS at concentrations ranging from 0.9 to 500 µg ml⁻¹ (n = 3/treatment). Data suggest that MTS and heparin can displace histones from CC over a wide concentration range, whereas mCBS cannot. b CC exposed to 500 µg ml⁻¹ of mCBS, MTS, and heparin, pelleted by centrifugation, and supernatants and pellets collected and run on SDS-PAGE. Representative data from one gel (137% normal size), confirming that MTS and heparin can displace histones from CC, whereas mCBS cannot. c Histogram including all the SDS-PAGE data and showing that MTS and heparin displace histones from CC, an effect that is significant (n = 4/treatment). d Sytox Green uptake by human NETs following incubation with mCBS, MTS, and heparin as in b above (n = 3/treatment). Data similar to that obtained with CC incubated with MTS and heparin. However, unlike CC, mCBS reduced Sytox Green uptake by NETs by ~50%, suggestive of NET stabilization. Solid circles are individual data points and are presented as mean ± s.e.m. and analyzed by two-way ANOVA with Tukey’s correction for multiple comparisons. Source data are provided as a Source Data File.