Differences in erythrocyte membrane proteins and glycoproteins in sickle cell disease

Sickle erythrocyte adherence to endothelium at low shear: role of shear stress in propagation of vaso-occlusion

Under venular flow conditions, sickle cell adherence to endothelium is mediated by cell adhesion molecules and adhesive proteins associated with inflammation, coagulation, and endothelial perturbation. Periodic and reduced blood flow are observed in sickle microcirculation during hematologic steady state, suggesting that blood flow is compromised in sickle microcirculation. We tested the hypothesis that low blood flow enhances adherence by quantifying sickle cell adhesion to endothelium under venular flow (1.0 dyne/cm(2) shear stress) and low flow (0.1 dyne/cm(2) shear stress), with and without addition of adhesion promoting agonists. Under low flow, sickle cell adherence to endothelium increases with contact time in the absence of endothelial activation or adhesive protein addition. In contrast, at venular shear stress, sickle cell adherence only occurs following endothelial activation with TNF-alpha or addition of thrombospondin. Analysis of these data with a mathematical model reveals that at low flow adherence is "transport-controlled," meaning that contact time between sickle cells and endothelium is a more important determinant of adherence than high-affinity receptor-ligand interactions. Low-affinity interactions are sufficient for adhesion at low flow. In contrast, at venular flow (1 dyne/cm(2) shear stress) adherence is "affinity-controlled," meaning that adherence requires induction of specific high-affinity receptor-ligand interactions. These findings demonstrate that in addition to activating factors and adherence proteins, microvascular shear stress is an important determinant of sickle cell adhesion to endothelium. This suggests that in vivo, erythrostasis is an important determinant of adhesion that can act either independently or concurrently with ongoing acute events to induce adhesive interactions and vaso-occlusion.

Antioxidant status and susceptibility of sickle erythrocytes to oxidative and osmotic stress

The purpose of this study was to determine if differences in antioxidant status between the red blood cells (RBCs) of sickle cell anemia (SCA) patients and controls are responsible for the differential responses to oxidative and osmotic stress-induced hemolysis. Susceptibility to hemolysis was examined by incubating oxygenated and deoxygenated RBCs at 37 degrees C with 73 mM 2,2' azobis (2-amidinopropane) HCl (AAPH), a peroxyl radical generator, for up to 3.5 hours.f The ability of RBCs to maintain membrane integrity under osmotic stress was determined over a range of diluted saline-phosphate buffer. Sickled RBCs showed a lesser degree of AAPH-induced hemolysis than control groups and were more resistant to osmotic stress-induced hemolysis. SCA patients had higher levels of RBC vitamin E and RBC lipids, but lower RBC GSH, plasma lipids and plasma carotenes than those of the hospital controls. No significant differences were observed in the levels of retinol, vitamin C, vitamin E, MDA and conjugated dienes in plasma, or the levels of MDA and conjugated dienes in RBCs. The results obtained suggest that the differences in antioxidant status between sickled RBCs and controls do not appear to be responsible for their different susceptibility to oxidative or osmotic stress-induced hemolysis observed.

Sialidase in rabbit blood. Characterization of sialidase purified from rabbit erythrocyte membrane

Sialidase activities of rabbit blood cells and serum were measured. The leucocyte particulate fraction showed the highest specific activity of sialidase towards mixed gangliosides and sialyllactose, and the cytosolic fraction showed for fetuin. Predominant sialidase activity in the blood was detected in erythrocyte particulate fraction when mixed gangliosides were used as substrate. The sialidase for ganglioside was solubilized from the erythrocyte ghosts by using Triton X-100. The solubilized sialidase was purified 1886-fold by sequential chromatographies on DEAE-cellulose, EAH-Sepharose 4B, Octyl-Sepharose CL-4B, Sephadex G-100, concanavalin-A--Sepharose, N-(p-aminophenyl)oxamic acid-agarose and Heparin-Sepharose CL-6B. The optimum pH of purified sialidase was 4.5 for ganglioside mixture, and this enzyme exhibited M(r) = 48,000 by gel filtration. When the purified sialidase was subjected to SDS/PAGE, a major sialidase-active protein band at M(r) = 54,000 and another fainter inactive protein band with M(r) = 115,000 were observed. The purified enzyme was active towards oligosaccharides, gangliosides, fetuin glycopeptide and 4-methylumbelliferyl alpha-D-N-acetylneuraminic acid except for glycoproteins tested. Fe2+, Fe3+ and dithiothreitol significantly inhibited the enzyme activity, while Triton X-100 activated the enzyme. Inside-out vesicles and unsealed ghosts of rabbit erythrocyte showed the sialidase activity for mixed gangliosides but not for resealed ghosts or intact erythrocytes. These results indicate that the active site of this sialidase is oriented mainly on the inside of the erythrocyte membrane and not on the outside. Treatment of rabbit erythrocyte unsealed ghosts with phosphatidylinositol-specific phospholipase C liberated no sialidase activity toward mixed gangliosides from the ghosts.

Red blood cells from patients with sickle cell disease exhibit an increased adherence to cultured endothelium pretreated with tumour necrosis factor (TNF)

Red blood cells (RBCs) from 24 patients with sickle cell disease were more adherent to cultured endothelium pretreated with the inflammatory cytokine, tumour necrosis factor (TNF) than RBCs from 22 healthy subjects. The enhanced sticking was apparent in RBC preparations from patients who were in crisis (mean 190% increase from controls) and out of crisis (mean 220% increase) and was not related to the number of circulating RBCs, reticulocytes, platelets, leucocytes or haemoglobin levels. When irreversibly sickled RBCs, enriched by centrifugation on density gradients, were added to TNF-treated endothelium they were found to be significantly more adherent (mean 411% increase; P < 0.001) than the unfractionated RBCs from the same patients. There was no difference between the adherent properties of sickle RBCs and normal RBCs for untreated endothelium. Contributing factors to the enhanced adhesion to TNF-treated endothelium may be the low surface change of sickle RBCs, and increased levels of fibrinogen and von Willebrand's factor (vWF) in the patients' plasma. By acting on vascular endothelium to increase its adhesiveness for sickled RBCs, it is concluded that inflammatory cytokines such as TNF may have a prominent role in mediating the events that lead to microvascular occlusions in sickle cell disease.

Increased adherence of oxidant-treated human and bovine erythrocytes to cultured endothelial cells

Bovine erythrocytes, which normally lack phosphatidyl choline in their membranes, when treated with either H2O2 or diamide (1-3 mM), showed a partial appearance of phosphatidyl ethanolamine (PE 40%) and phosphatidyl serine (PS, 30-33%) in the external leaflet of the bilayer and a concomitant increased (four- to five-fold) propensity to adhere to cultured bovine aortic endothelial cells. Similar treatment of normal human erythrocytes caused an alteration in the organization of the phospholipid bilayer and also resulted in their increased adherence to endothelial cells derived either from human umbilical vein or bovine aorta. Treatment of RBCs with H2O2 at low concentration (0.5 mM) resulted in cross-linking of spectrin without significant changes in the orientation of aminophospholipids but the RBCs exhibited 15-20% increase in adherence to endothelial cells. Pretreatment of either human or bovine erythrocytes with antioxidants such as vitamin E (2 mM) prevented both oxidant-induced reorganization of phospholipids in the bilayer and enhancement of adherence to endothelial cells. Introduction of either phosphatidyl serine or phosphatidyl ethanolamine but not phosphatidyl choline into erythrocyte membranes increased their adherence to endothelial cells threefold. Oxidant-treated RBCs exhibited enhanced binding and fluorescence of Merocyanine 540 dye (MC-540), which is sensitive to the packing of lipids in the lipid bilayer. On flow cytometric analysis, 78% of H2O2 (0.5 mM)-treated erythrocytes compared to 30% of untreated RBCs exhibited MC-540 binding and fluorescence, indicating differences in the lipid packing in the outer leaflet of the bilayer. Oxidant-treated erythrocytes adhere preferentially to endothelial cells rather than to bovine aortic smooth muscle cells and skin fibroblasts. It is suggested that the alterations in the erythrocyte membrane surface due to spectrin cross-linking and the organization of the phospholipids concomitant with less ordered packing in the external leaflet of the bilayer, either induced by oxidative manipulation in normal RBC or in pathological erythrocytes, play a role in erythrocyte-endothelial cell interaction.

Decreased sialic acid content of erythrocytes in patients with aplastic anaemia

The sialic acid content of erythrocytes from patients with aplastic anaemia was determined and compared with those in patients with several haematological disorders and healthy individuals. The sialic acid was released enzymatically with Vibrio cholerae sialidase and quantitated by the thiobarbituric acid method (Aminoff, 1961). The sialic acid content of normal erythrocytes was 538 +/- 31 nmol/ml of packed erythrocytes. That of erythrocytes from patients with aplastic anaemia was 480 +/- 35 nmol/ml of packed erythrocytes, which was significantly lower than normal (P less than 0.01). In contrast, erythrocytes from patients with myeloproliferative disorders showed significantly (P less than 0.05) higher sialic acid contents (564 +/- 45 nmol/ml of packed erythrocytes). These results suggest that some membrane changes occur in erythrocytes in patients with these disorders.

The phosphoproteins of the sickle erythrocyte membrane

The uptake of 32P from exogenous 32Pi into membrane proteins of sickle erythrocytes has been analyzed. The phosphorylation of spectrin is normal in sickle cells. There is, however, a substantial increase in 32P in the sialoproteins of the membrane, which can be demonstrated after fractionation or selective proteolysis. Normal and sickle erythrocytes were separated on Stractan gradients and average cell age was determined using the remaining pyruvate kinase activity as a marker. The altered phosphorylation of sickle cells was not seen in young normal cells, suggesting that it was not related to cell age. The altered phosphorylation was also not correlated with the level of reticulocytes in these fractions. This result is further evidence for abnormalities in the sialoproteins of sickle erythrocytes and is the first demonstration of altered sialoprotein phosphorylation in the red cell.

Erythrophagocytosis in vivo in sickle cell anemia

Recent observations that the sickle RBC are excessively susceptible to phagocytosis by macrophages in vitro prompted me to look for evidence of in vivo erythrophagocytosis (Ep) in patients with sickle cell anemia (SS). Freshly prepared smears of unmanipulated blood of 27 patients with SS in steady state were examined for Ep by a 500-cell differential white blood cell (WBC) count performed in duplicate. Ten of 27 (37%) SS patients showed Ep (1-6/1,000 WBC or 1-10/100 monocytes). By contrast, no Ep was found in similarly prepared blood smears of 25 normal adult controls and nine splenectomized subjects. The mean hemotocrit value of the Ep(+) SS patients was significantly lower than that of the Ep(-) patients (21.0 +/- 1.7% vs 24.0 +/- 2.7% p less than 0.01). Considering the rarity of spontaneous Ep in unmanipulated blood from normal subjects and the relative insensitivity of the method used, the finding of Ep in over one third of SS patients indicates a significant membrane injury of the sickle RBC and serves to validate the in vitro observations. The possible role of the "senescence" mechanism in the induction of Ep is discussed.

Qualitative and quantitative analysis of erythrocyte surface membrane sialyl residues using affinity cytochemistry with special reference to diabetic patients

Erythrocyte surface membrane sialyl residues were investigated by means of affinity cytochemistry using the avidin-biotin complex technique. Mild oxidation with the periodate (MO)-biotin hydrazide (BHZ)-ferritin avidin conjugate (FAv) sequence revealed numerous ferritin particles on erythrocytes from healthy donors. The ferritin particles attached on the perpendicularly sectioned membrane were seen at an average distance of 10 to 12 nm from the outer dense leaflet of the cell membrane. Pretreatment with neuraminidase followed by the MO-BHZ-FAv sequence almost eliminated erythrocyte ferritin labeling. Erythrocytes from diabetic patients showed less dense ferritin labeling compared with those from healthy donors. Quantiative analysis of sialyl residues demonstrated a marked reduction in ferritin labeling of erythrocytes from diabetic patients which was significantly less (p less than 0.01) than that of erythrocytes from healthy donors. This observation supports previous biochemical data demonstrating lower levels of surface membrane negative charge and sialyl residues on erythrocytes from patients with diabetes mellitus.

Abnormalities in membrane phospholipid organization in sickled erythrocytes

In contrast to the wealth of information concerning membrane phospholipid asymmetry in normal human erythrocytes, very little is known about membrane phospholipid organization in pathologic erythrocytes. Since the spectrin-actin lattice, which has been suggested to play an important role in stabilizing membrane phospholipid asymmetry, is abnormal in sickled erythrocytes, we determined the effects of sickling on membrane phospholipid organization. We used two enzymatic probes: been venom phospholipase A2 and Staphylococcus aureus sphingomyelinase C, which do not penetrate the membrane and react only with phospholipids located in the outer leaflet of the bilayer. Our results suggest that the distribution of glycerophospholipids within the membrane of sickled cells is different from that in nonsickled cells. Compared with the normal erythrocyte, the outer membrane leaflet of the deoxygenated, reversibly sickled cells (RSC) and irreversibly sickled cells (ISC) was enriched in phosphatidyl ethanolamine in addition to containing phosphatidyl serine. These changes were compensated for by a decrease in phosphatidyl choline in that layer. The distribution of sphingomyelin over the two halves of the bilayer was unaffected by sickling. In contrast to ICS, where the organization of phospholipids was abnormal under both oxy and deoxy conditions, reoxygenation of RSC almost completely restored the organization of membrane phospholipids to normal. These results indicate that the process of sickling induces an abnormality in the organization of membrane phospholipids to normal. These results indicate that the process of sickling induces an abnormality in the organization of membrane lipids in RSC which become permanent in ISC.

Sialic acid content of erythrocytes in normal individuals and patients with certain hematologic disorders

The sialic acid content of erythrocytes from healthy individuals of different blood types and of patients with known hematological disorders has been determined. The sialic acid was completely released enzymatically with sialidase and quantitated by the thiobarbituric acid method. The sialic acid content of erythrocytes was constant irrespective of ABO blood type, or anticoagulant used; viz, 0.85-0.92 mumoles/ml of packed erythrocytes or 46-53 X 10(6) sialyl residues per cell. Deviations from these normal values were obtained with erythrocytes from patients with a variety of hematological disorders. Patients with the following disorders have significantly (P less than 0.01) lower sialic acid values compared to erythrocytes from healthy individuals (given in the order of decreasing sialic acid content): sickle cell anemia, acute lymphocytic leukemia, chronic lymphocytic leukemia, acute myelomonocytic leukemia, non-Hodgkin lymphocytic lymphoma, chronic granulocytic leukemia, acute myelocytic leukemia, leukemia, and Hodgkin disease.

The erythrocyte membranes in beta-thalassemia. Lower sialic acid levels in glycophorin

The sialic acids content of glycophorin of thalassemic erythrocyte membranes is about 25% lower than in glycophorin of normal erythrocyte membranes. Glycophorin extracted from old thalassemic erythrocytes separated by density centrifugation, has about half the sialic acids content found in glycophorin extracted from young thalassemic erythrocytes. Possible sialidase activty was sought in the plasma and erythrocyte membranes of thalassemic erythrocytes. No increased sialidase activity was detected in the plasma of the patients as compared to that of normal donors. Thus, other sites for sialidase activity, or other possibilities have to be explored to account for the increased sialic acid hydrolysis of glycophorin of the thalassemic erythrocytes.

Organizational differences in the membrane proteins of normal and irreversibly sickled erythrocytes

Using two-dimensional gels, no unique membrane proteins were detected in irreversibly sickled cells. Membranes from irreversibly sickled cells were shown to cross-link much more readily with dithiobis(succinimidyl propionate) than normal erythrocyte membranes. Increased binding of band 4.5 protein and increased intra-chain disulfides were also demonstrated. These changes may correlate to enhanced cellular rigidity.

Abnormal adherence of sickle erythrocytes to cultured vascular endothelium: possible mechanism for microvascular occlusion in sickle cell disease

The abnormal shape and poor deformability of the sickled erythrocyte (RBC) have generally been held responsible for the microvascular occlusions of sickle cell disease. However, there is no correlation between the clinical severity of this disease and the presence of sickled RBC. In searching for additional factors that might contribute to the pathophysiology of sickle cell disease, we have investigated the possibility that sickle RBC might be less than normally repulsive of the vascular endothelium. After RBC suspensions are allowed to settle onto plates of cultured human endothelial cells, normal RBC are completely removed by as few as six washes. In contrast, sickle RBC remain adherent despite multiple washes. On subconfluent culture plates, normal RBC are distributed randomly, whereas sickle RBC cluster around endothelial cells. Sickle RBC adherence is not enhanced by deoxygenation but does increase with increasing RBC density. The enzymatic removal of membrane sialic acid greatly diminishes the adherence of sickle RBC to endothelial cells, suggesting that sialic acid participates in this abnormal cell-cell interaction. Although net negative charge appears normal, sickle RBC mainfest an abnormal clumping of negative surface charge as demonstrated by localization of cationized ferritin. These abnormalities are reproduced in normal RBC loaded with nonechinocytogenic amounts of calcium. We conclude that sickle RBC adhere to vascular endothelial cells in vitro, perhaps caused by a calcium-induced aberration of membrane topography. This adherence may be a pathogenetic factor in the microvascular occlusions characteristic of sickle cell disease.

(Ca2++Mg2+)-ATPase activity of sickle cell membranes: decreased activation by red blood cell cytoplasmic activator

Human red blood cells (RBCs) contain a cytoplasmic protein that activates membrane-bound (Ca2+ + Mg2+)-ATPase and the transport of Ca2+. The (Ca2+ + Mg2+)-ATPase of sickle cells showed a less than normal response to this activator. This was true whether the activator was obtained from normal or sickle cells. Activator present in sickle cell hemolysates fully activated the (Ca2+ + Mg2+)-ATPase of normal RBC membranes. These results demonstrate that membranes of sickle cells are defective in their response to the activator. Neither the apparent affinity for calcium nor the apparent affinity for activator was different comparing the (Ca2+ + Mg2+)-ATPase of sickle and normal membranes. Young, mature, and irreversibly sickled cells were separated by density gradient centrifugation, and membranes were prepared from each of these cell populations. No significant differences in ATPase activities were found based on cell age (density). The (Ca2+ + Mg2+)-ATPase of all populations of sickle cells showed a decreased response to the activator. Thus, it appears unlikely that the decreased response of the (Ca2+ + Mg2+)-ATPase of sickle cells is due to membrane damage caused by repeated sickling during the life-span of the cell. Reduced activation of (Ca2+ + Mg2+)-ATPase by the cytoplasmic activator may account for calcium accumulation in sickle cells.

Spin-label studies of membrane-associated denatured hemoglobin in normal and sickle cells

A maleimide spin label (N-(1-oxyl-2,2,5,5-tetramethylpyrrolidinyl)-maleimide) was reacted with oxyhemoglobin-free cell stromata of normal and sickle cells. The EPR spectrum of spin-labeled red cell membranes showed that the spin labels are attached to at least two different binding sites. There was a major signal, A, which characterized a strongly immobilized environment and a minor signal, B, which characterized a weakly immobilized environment. Quantitative EPR measurements using equal amounts of Hb AA and Hb SS red blood cells demonstrated that Hb SS red cell membranes had an approximately four times higher EPR signal intensity than Hb AA red cell membranes ((7.98 +/- 1.14 . 10(5) and (2.2 +/- 1.2) . 10(5) spin labels/cell, respectively). Moreover, the ratio of signal intensities A and B are different in these cells. Comparative spectrophotometric studies of membrane-associated denatured hemoglobins of Hb AA and Hb SS red cell membranes suggested that the EPR signal A is derived from spin labels attached to membrane-associated denatured hemoglobin, while signal B is mainly from spin labels attached to membranes. The combination of EPR spectrum of Hb AA membranes pretreated with N-ethylmaleimide and that of spin-labeled precipitated hemoglobin further strengthened this conclusion.

Calcium-induced erythrocyte membrane changes. The role of adsorption of cytosol proteins and proteases

Changes in the membranes of human red cells similar to those of certain hemolytic anemias were produced by calcium in three model systems and found to result from membrane adsorption of cytosol proteins and from proteolysis. Proteins of the cytosol adsorbed to human erythrocyte membranes in the presence of calcium and extractable by EDTA were compared to those of the total cytosol by polyacrylamide gel electrophoresis and by isoelectric focusing. Catalase (EC 1.11.1.6) and band 8 were adsorbed to the membranes from the supernatant cytosol with calcium. Band 8 was a normal constitutent of the cytosol, apparently a single chain of molecular weight 24,000 with a pI of 5.35. Other calcium-induced membrane changes could be demonstrated to be due to cytosol protease(s) adsorbed to the membrane in the presence of calcium and extractable with EDTA. When membranes were incubated with the proteases and calcium the decrease in bands 1,2,3 and 4.1 and the appearance of multiple low molecular weight peptides typical of calcium-induced membrane effects resulted.

Erythrocyte membrane lipid reorganization during the sickling process

In order to study possible alterations in membrane lipids during sickling, we have measured the difference in susceptibility to lipid peroxidation, binding of trinitrobenzenesulfonic acid (TNBS) to aminophospholipids, and fatty acid uptake in cells containing sickle haemoglobin under aerobic and anaerobic conditions. We have also examined TNBS binding in irreversibly sickled cells in an attempt to evaluate the permanent effects of any such alterations. We found that when erythrocytes were sickled by deoxygenation, the susceptibility to lipid peroxidation and binding of TNBS to aminophospholipids was markedly increased, while normal control cells showed no change. These effects appeared to be specific for the sickled state rather than a nonspecific consequence of cell age or the concentration of sickle haemoglobin within the cell. In contrast, fatty acid incorporation into membrane phospholipids, representing potential lipid renewal, was decreased in the sickled state. Cell fractions enriched in irreversibly sickled cells showed increased TNBS labelling in air and only modest rises with anoxia. Taken together, these data imply a rearrangement of membrane lipids during the sickling process and suggest a permanent reorganization of membrane lipids in the irreversibly sickled cell.

Membrane alterations in irreversibly sickled cells: hemoglobin--membrane interaction

Irreversibly sickled cells (ISCs) are sickle erythrocytes which retain bipolar elongated shapes despite reoxygenation and owe their biophysical abnormalities to acquired membrane alterations. Freeze-etched membranes both of ISCs produced in vitro and ISCs isolated in vivo reveal microbodies fixed to the internal (PS) surface which obscure spectrin filaments. Intramembranous particles (IMPs) on the intramembrane (PF) surface aggregate over regions of subsurface microbodies. Electron microscopy of diaminobenzidine-treated of ISC ghosts show the microbodies to contain hemoglobin and/or hemoglobin derivatives. Scanning electron microscopy and freeze-etching demonstrate that membrane--hemoglobin S interaction in ISCs enhances the membrane loss by microspherulation. Membrane-bound hemoglobin is five times greater in in vivo ISCs than non-ISCs, and increases during ISC production, parallelling depletion of adenosine triphosphate. Polyacrylamide gel electrophoresis of ISC membranes shows the presence of high-molecular-weight heteropolymers in the pre--band 1 region, a decrease in band 4.1 and an increase in bands 7, 8, and globin. The role of cross-linked membrane protein polymers in the generation of ISCs is discussed and is synthesized in terms of a unified concept for the determinants of the genesis of ISCs.