Mark D. Scott

Thalassaemic erythrocytes : cellular suicide arising from iron and glutathione-dependent oxidation reactions ?

Summary. Both β‐thalassaemic red blood cells and normal red blood cells (RBC) artificially loaded with unpaired α‐haemoglobin chains exhibit increased amounts of membrane‐bound haem and iron. In the model β‐thalassaemic RBC the amount of free haem and iron was as much as 20 times that which could have been contributed by the entrapped α‐haemoglobin chains alone. This excess haem iron arises from destabilization of haemoglobin via reactions between ferric iron (Fe3+), initially contributed by the unpaired a chains, and cytoplasmic constituents, primarily reduced glutathione (GSH). Indeed, in the presence of Fe3+ (100μM) addition of even small amounts of GSH (0‐5 mM) to dilute RBC haemolysates (0‐15 mg haemoglobin/dl) greatly accelerated methaemoglobin formation. In contrast, lysates from GSH‐depleted RBC demonstrated a significantly reduced rate of iron‐mediated haemoglobin oxidation which was reversible by addition of GSH. The initiation, and subsequent propagation, of Fe3+‐mediated haemoglobin oxidation was significantly inhibited by iron chelators. Finally, Fe3+‐driven haemoglobin oxidation was synergized by low amounts of H2O2, an oxidant spontaneously generated in thalassaemic RBC. To summarize, the release of small amounts of free iron from unpaired α‐haemoglobin chains in the β‐thalassaemic RBC can initiate self‐amplifying redox reactions which simultaneously deplete cellular reducing potential (e.g. GSH), oxidize additional haemoglobin, and accelerate the red cell destruction.

Camouflaged blood cells: Low-technology bioengineering for transfusion medicine?

The small number of studies done on the covalent modification of RBC with PEG, or PEG-derivatives, suggests that the immunocamouflage of intact cells significantly reduces the antigenicity and immunogenicity of the foreign cell. Importantly, this protective immunologic effect can be accomplished without adversely affecting the structure, function, or viability of the modified cell (e.g., RBCs and lymphocytes). As a consequence, PEG-RBC may have significant practical value in the treatment of the chronically transfused patient as a prophylactic measure against allosensitization. The PEG-RBC also may be useful in treating the already allosensitized individual. As shown, preexisting antibodies do not effectively recognize nor bind to the modified donor cells. A finding of further interest to transfusion medicine is that pegylation of contaminating lymphocytes within RBC products may prove efficacious in preventing graft-versus-host disease in the immunocompromised patient. However, the main emphasis of our research continues to be the immunocamouflage of RBC for use in chronic transfusion therapy of the SCD and thalassemic patient.

Phospholipid composition and organization in model β-thalassemic erythrocytes

The membrane phospholipid organization in human red blood cells (RBC) is rigidly maintained by a complex system of enzymes. However, several elements of this system are sensitive to oxidative damage. An important component in the destruction of β‐thalassemic RBC is the generation of reactive oxygen species and the release of redox‐active iron by the unpaired α‐hemoglobin chains. Consequently, we hypothesized that the presence of this oxidative stress to the RBC membrane could lead to alterations in membrane lipid organization. Model β thalassemic RBC, prepared by the introduction of excess α‐globin in the cell, have previously been shown to exhibit structural and functional changes almost identical to those observed in β‐thalassemic cells. After 24 hr at 37°C, the model β thalassemic cells exhibited a significant loss of deformability, as measured by ektacytometric analysis, indicative of extensive membrane damage. However, a normal steady‐state distribution of endogenous phospholipids was found, as evidenced by the accessibility of membrane phospholipids to hydrolysis by phospholipases. Similarly, the kinetics of transbilayer movement of spin‐labeled phosphatidylserine (PS) and phosphatidylethanolamine (PE) in all samples was in the normal range and was not affected by the presence of excess α‐globin chains. In contrast, a faster rate of spin‐labeled phosphatidylcholine (PC) transbilayer movement was observed in these cells. While control RBC exhibited a complete loss of their initial (2 mol%) lysophosphatidylcholine (LPC) levels following 24 hr of incubation at 37°C, 1.5 mol% LPC was still present in model β‐thalassemic cells, suggesting an altered phospholipid molecular species turnover, possibly as a result of an increased repair of oxidatively damaged phospholipids.

Superoxide is not the proximate cause of paraquat toxicity

Paraquat toxicity is thought to occur through the production of superoxide O2(.-) and it has been argued that this oxygen radical species is, itself, an important mediator of the toxicity of this drug. If so, a direct relationship should exist between the steady-state amounts of O2(.-) produced and the lethal effects of paraquat. We have therefore examined O2(.-) mediated chemiluminescence and paraquat sensitivity in bacteria with widely varying superoxide dismutase (SOD) activities. As expected, bacteria with high SOD activity exhibit minimal (lucigenin enhanced) chemiluminescence in the presence of paraquat whereas SOD-deficient bacteria show >90-fold higher chemiluminescence compared to parental strains. Nonetheless, high SOD bacteria are more readily killed by paraquat whereas SOD-deficient organisms show no increased susceptibility to this agent. This further supports our earlier conclusions that hypertrophied SOD activity is inadequate defense against paraquat and that O2(.-) is probably not the proximate toxin by which paraquat mediates cellular injury.

Pharmacologic interception of heme: a potential therapeutic strategy for the treatment of β thalassemia?

Pro-oxidant effects of hemoglobin-derived heme and iron contribute to the progressive damage observed in β thalassemic and sickle (HbS) red blood cells. Agents that prevent heme/iron release and inhibit their redox activity might diminish such injury. Consequently, the inhibitory effects of chloroquine (CQ), a heme-binding antimalarial drug, and a novel dichloroquine compound (CQ-D2) on iron release and lipid peroxidation were investigated. In contrast to normal hemoglobin, significant amounts of iron were released from both purified hemin and α-hemoglobin chains during incubations with exogenous reduced glutathione (GSH) and/or H2O2. Addition of either CQ or CQ-D2 effectively inhibited GSH- and GSH/H2O2-mediated iron release from hemin (P<0.001). During prolonged incubations (6 h), both CQ and CQ-D2 significantly decreased the release of heme-free iron from both purified hemoglobin and α-hemoglobin chains. Interestingly, CQ and CQ-D2 differentially affected the redox availability of the heme-bound iron. The CQ: heme complex significantly enhanced membrane lipid peroxidation whereas CQ-D2 dramatically (P<0.001) inhibited heme-dependent peroxidation to almost baseline levels. In summary, CQ-derivatives which render heme redox inert and prevent the release of free iron from heme might be beneficial in the treatment of certain hemoglobinopathies and, perhaps, other pathologies promoted by delocalized heme/iron.