Widespread adhesion of erythrocytes in the vasculature would be incompatible with life. However, in illnesses such as sickle cell disease, malaria and diabetes, adhesive RBCs have been documented and are likely contributors to disease severity, inflammation and coagulation activation. End-stage renal disease is characterized by profound, global inflammation and relative thrombophilia. While we, and others, have reported the extensive exposure of RBC phosphatidylserine in patients on hemodialysis, the role of RBC adhesion in inflammation has never been examined in this condition. We performed an analysis of RBCs from hemodialysis patients to characterize their ability to adhere to other cells, elucidate the potential mechanisms of adhesion, and to relate RBC adhesion to the inflammatory state of hemodialysis.
This study was conducted in a cohort of 20 African American patients receiving in-center hemodialysis at 2 separate dialysis clinics as part of a study of sickle cell trait in hemodialysis. There were 9 of 20 patients with documented sickle cell trait. Blood samples were drawn immediately pre-dialysis and prior to administration of heparin. Plasma samples were processed to minimize platelet activation, aliquoted, and frozen for batch analysis. Plasma factors (RANTES, CRP & CD40L) were analyzed via ELISA. Adhesion events in whole blood were detected via two color flow cytometry. Cause of renal failure, mode of venous access and relevant clinical data were obtained from patient charts. All adhesion assays with washed blood cells were conducted under static conditions using microvascular endothelial cells. All studies were compared to those results of healthy control patients (n =11). Spearman's regression analysis was performed to analyze the correlation between continuous variables. Mann-Whitney U and Kruskal Wallis tests were used to compare continuous variables between groups. A p<0.05 was considered significant.
We found that RBCs from patients on hemodialysis were significantly more adhesive than those from healthy controls. In whole blood, we detected marked RBC adhesion to T-cells (median % of T-cells bound to RBCs: 67.9 [hemodialysis] vs. 10.55 [healthy control]) and platelets (median % of platelets bound to RBCs: 33.0 vs. 1.1%). We also noted significant RBC adhesion to neutrophils (median % of neutrophils bound to RBCs: 11.0% vs. 0.1%). Incubation of healthy RBCs with plasma from hemodialysis patients, but not healthy control plasma, was sufficient to induce RBC adhesion to cultured endothelial cells (median RBCs/mm2: 8.0 vs. 0.5) and to T-cells (median % T-cells bound to RBCs: 30.0 vs. 5.0). Plasma from hemodialysis patients also induced phosphatidylserine exposure on healthy RBCs. Phosphatidylserine exposure on the RBC appeared to mediate RBC adhesion to endothelial cells as annexin V significantly reduced RBC adherence (8.0 vs 3.8 RBC/mm2). The extent to which RBCs in hemodialysis patients adhered to T-cells directly correlated with both plasma RANTES (rspearman = 0.65, 95% CI: 0.247–0.864) and CD40L levels (rspearman= 0.60, 95% CI: 0.1645 – 0.847), but not plasma CRP levels. There was also no significant difference in adhesion of RBCs due to cause of renal failure (diabetes, hypertension, or glomerulonephritis) or presence of sickle cell trait.
We describe for the first time a novel adhesion of RBCs in patients receiving hemodialysis and how this adhesion may relate to inflammation in these patients. Our findings also suggest that factors in uremic plasma are sufficient to induce phosphatidylserine exposure on RBCs. This exposure, in addition to providing a site for thrombin generation, also serves as an adhesive moiety on RBCs for endothelial cells. These data may describe an unrecognized etiology of inflammation in hemodialysis and end-stage renal disease.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.