Extracellular histones induce erythrocyte fragility and anemia

Histones are highly conserved, cationic proteins that play an integral role in DNA condensation and gene expression.¹  However, multiple studies have indicated that extracellular histones and neutrophil extracellular traps (NETs), the predominant source of extracellular histones, play an important role in many pathologies, namely, sepsis,²  trauma,³  atherosclerosis,⁴  cardiac ischemia reperfusion injury,⁵  renal⁶  and liver injury,⁷  and autoimmune disorders.⁸  It is thought that extracellular histones mediate these pathological effects by being cytotoxic for nucleated cells, such as endothelial and epithelial cells,⁹  and by promoting platelet activation with resultant thrombosis and acute thrombocytopenia.¹⁰ 

In contrast to nucleated cells and platelets, there is limited literature describing the interaction between histones and erythrocytes. An early investigation reported that polycations, such as histones, agglutinate erythrocytes,¹¹  and subsequently it was demonstrated that erythrocytes express phosphatidylserine on their surface after exposure to histones and, in doing so, become procoagulant.¹²  The focus of this study, therefore, was to investigate whether histones bind to erythrocytes and alter their biology and function in vitro and in vivo.

Washed human erythrocytes from healthy donors were resuspended as a 20% (volume-to-volume ratio) suspension in phosphate-buffered saline (PBS)/0.1% bovine serum albumin before exposure to concentrations of calf thymus histones as indicated (Sigma-Aldrich, Sydney, NSW, Australia).

An automated 96-well microplate pipetting system (epMotion 5070; Eppendorf AG, Hamburg, Germany), using TS 50 tips with an aperture radius of 0.25 mm and MicroAmp Optical 96-Well Reaction Plates (Applied Biosystems, Norwalk, CT), was used to mimic the mechanical sheer stress encountered by circulating erythrocytes. Shear exposure was determined by the number of passages (5-40×) and the speed of passage (25-100 mm/s) through the TS 50 tips. The erythrocyte samples were incubated with the histones for 60 minutes at 37°C in 60% tonicity PBS/0.1% bovine serum albumin before shear stress exposure, and erythrocyte fragility was measured by hemoglobin release (optical density, 540 nm) after lysis. In parallel, an artificial human spleen assay was used in vitro to detect, by filtration, the deformability of erythrocytes¹³  after histone exposure.

The Australian National University Animal Experimentation Ethics Committee approved all animal experimental procedures; 5- to 8-week-old female BALB/c mice, bred at the Australian National University, were used in this study.

Mice were injected IV with histones (50 mg/kg) in PBS. Retro-orbital bleeds were performed, spleens were harvested at 5, 10, 20, and 30 minutes postinjection, and single-cell suspensions of splenocytes were prepared.¹⁴  Hematologic analysis of blood samples was subsequently performed with an ADVIA 2120i hematology analyzer (Siemens, Munich, Germany). The splenocyte hemoglobin content was quantified with a hemoglobin assay kit (Sigma-Aldrich).

Histone-induced aggregation of erythrocytes was quantified by flow cytometry based on increases in forward scatter (FSC) and side scatter (SSC) or by increases in erythrocyte autofluorescence (FL-1 channel) using a BD FACSCalibur instrument (BD Biosciences, North Ryde, NSW, Australia). Aggregation of erythrocytes was also monitored by scanning electron microscopy (SEM)¹⁵  and by the erythrocyte sedimentation rate (ESR).¹⁶ 

The initial studies revealed that when human erythrocytes were incubated with histones, substantial erythrocyte aggregation occurred that could be readily detected by flow cytometry based on increases in FSC and SSC and erythrocyte autofluorescence (Figure 1A), a result that was confirmed by SEM (Figure 1B). Subsequent experiments used erythrocyte autofluorescence and ESR to demonstrate that aggregation and increased sedimentation of erythrocytes induced by histones was concentration dependent and reached statistical significance at histone concentrations of ≥100 μg/mL and ≥50 μg/mL, respectively (P < .0001) (Figure 1C).

To further explore the consequences of histone binding to erythrocytes, a novel shear flow model was developed wherein erythrocytes incubated with histones in a 60% hypotonic solution were subjected to increasing exposure (passage number) (Figure 1D) or flow rates (Figure 1E), and the extent of lysis was then determined by measuring hemoglobin in the supernatant. Similar experiments performed under isotonic conditions gave comparable but less marked fragility results (data not shown). These experiments showed that histones increase erythrocyte fragility in a concentration-dependent manner, with both the exposure level and flow rate significantly affecting the extent of histone-mediated lysis (Figure 1D-E). The experimental flow conditions used in this assay were within the normal physiological range, because the shear rates, which involved a 0.25-mm radius tube, were similar to those encountered in the vasculature,¹⁷  and histone levels in trauma patients have been reported to be as high as 250 μg/mL.^18,19  Furthermore, histone-induced fragility is not restricted to erythrocytes; a histone level of 400 µg/mL causes an ∼50% loss of leukocytes under shear (data not shown).

Erythrocytes are structured for flexibility that enables them to deform and pass through small capillaries, and thus we investigated whether histone binding to erythrocytes would compromise their deformability. Using an artificial spleen model previously used in malaria-infected erythrocyte deformability studies,¹³  we examined erythrocyte retention within the artificial spleen after exposure to increasing concentrations of histones (Figure 1F). The number of erythrocytes that were able to pass through the artificial spleen significantly diminished as the histone concentration increased. This change was not due to erythrocyte lysis, because this model does not apply a significant level of shear force and, consequently, the free hemoglobin content of eluates did not increase when histones were present (data not shown). Furthermore, erythrocytes exposed to histone concentrations insufficient for induction of aggregation (ie, 12.5 and 25 μg/mL) were still retained by the artificial spleen, implying that retention is not merely due to aggregation, but involves additional factors, such as erythrocyte deformability.

Because in vitro conditions cannot fully replicate the complexities inherent in vivo, we injected mice IV with 50 mg/kg of histones to determine the impact on erythrocytes 5 to 30 minutes post–histone injection. Circulating erythrocyte numbers and hemoglobin content were significantly reduced at 10, 20, and 30 minutes post–histone injection (Figure 2A-B). Additionally, profound leukopenia and thrombocytopenia also occurred in these mice (Figure 2D-E). Histone-induced thrombocytopenia has been reported,¹⁰  however, histone-induced leukopenia has not previously been identified. Consistent with splenic removal of damaged erythrocytes, splenic erythrocyte content, based on hemoglobin levels, was seen to significantly increase at all time points studied, concomitant with decreasing circulating erythrocyte numbers (Figure 2C).

Collectively, these data provide compelling evidence that histones cause significant erythrocyte damage by inducing erythrocyte aggregation, increasing erythrocyte lysis under shear stress, and reducing erythrocyte deformability. In vivo, these processes resulted in anemia and accelerated erythrocyte removal by the spleen. Histones present in NETs may also contribute to these processes, because erythrocytes can associate with NETs.²⁰  Furthermore, lysis of erythrocytes with the release of heme may further exacerbate the situation, because heme has been demonstrated to induce NETosis.²¹  Nevertheless, the murine erythropoietic system rapidly replenished erythrocytes after a single bolus injection of histones, with levels returning to the normal range within 1 hour (data not shown). In the acutely ill patient, however, when comorbidities are present, when the production of histones may be continuous over a period of days, or when histone-binding plasma proteins, such as albumin, become depleted,²²  histone-associated anemia is more likely to develop and require correction. The results presented in this study, therefore, provide evidence for several mechanisms of histone-induced erythrocyte loss in histone-associated pathologies that may help to explain the anemia observed with acute illness in the absence of obvious causes.²³ 

The authors thank Anna Bezos and Anna Browne for help with laboratory work and blood donation. The authors also thank Ross Stephens for reviewing the manuscript.

This work was supported in part by Sirtex Technology, Pty., Ltd.

Contribution: F.K., C.H.O., and L.A.C. performed the experiments, with F.K. and C.H.O. making the greatest contribution; C.H.O., F.K., L.A.C., and C.R.P. wrote the paper; and all authors developed the experimental techniques and revised and approved the final version of the manuscript.

Conflict-of-interest disclosure: C.R.P. received research funding from Sirtex Technology Pty., Ltd., for this project and holds stocks in the start-up companies Lipotek and Beta Therapeutics and in the publicly traded companies Biotron and TBG Diagnostics. The remaining authors declare no competing financial interests.

The current affiliation for P.M.L. is Laboratory of Malaria Immunology, Immunology Frontier Research Center, Osaka University, Osaka, Japan.

Correspondence: Christopher R. Parish, Cancer and Vascular Biology Group, Australian Cancer Research Foundation Department of Cancer Biology and Therapeutics, John Curtin School of Medical Research, Australian National University, Canberra, ACT 2601, Australia; e-mail: christopher.parish@anu.edu.au.