Red blood cells infected with Plasmodium falciparum (IRBCs) undergo changes primarily in their membrane composition that contribute to malaria pathogenesis. However, all manifestations (eg, anemia) cannot be accounted for by IRBCs alone. Uninfected erythrocytes (URBCs) may play a role, but they have been underresearched. We wanted to document changes in the erythrocyte membrane that could contribute to URBC reduced life span and malaria-associated anemia. Human erythrocytes were cultured with P falciparum and washed at the trophozoite stage. IRBCs and URBCs were separated on Percoll density gradient, thus obtaining erythrocyte fractions of different densities/ages. IRBC-and URBCpurified membranes were analyzed and compared with control normal erythrocytes (NRBCs) of the same age, from the same donor, kept in the same conditions. P falciparum accelerated aging of both IRBCs and URBCs, causing a significant shift in the cell population toward the denser (old) fraction. Protein, phospholipid, and cholesterol content were reduced in IRBCs and young URBCs. Young and medium uninfected fractions had higher levels of lipid peroxidation and phospholipid saturation (because of the loss of polyunsaturated fatty acids, PUFAs) and lower phosphatidylserine. In IRBCs, thiobarbituric reactive substances (TBARSs) were higher, and PUFAs and phosphatidylserine lower than in NRBCs and URBCs. In comparison, trophozoite membranes had lower phospholipid (particularly sphingomyelin and phosphatidylserine) and cholesterol content and a higher degree of saturation. Parasiteinduced peroxidative damage might account for these modifications. In summary, we demonstrated that membrane damage leading to accelerated senescence of both infected and uninfected erythrocytes will likely contribute to malaria anemia. (Blood. 2003;102:705-711)

In falciparum malaria, both infected and uninfected red cells have structural and functional alterations. To investigate the mechanisms of these modifications, we studied the effects of two Plasmodium falciparum haem products (haematin and malaria pigment in the synthetic form beta-haematin) on isolated human red blood cells (RBCs) and purified RBC ghosts. A dose- and time-dependent incorporation of haematin into RBC ghosts and intact cells was observed, which was in proportion to the extent of haematin- induced haemolysis. RBCs pre-incubated with haematin were more sensitive to haemolysis induced by hypotonic shock, low pH, H2O2 or haematin itself. Haemolysis was not related to membrane lipid peroxidation and only partially to oxidation of protein sulphydryl groups and it could not be prevented by scavengers of lipid peroxidation or hydroperoxide groups. N-acetylcysteine partly protected the oxidation of SH groups and significantly reduced haemolysis. In contrast, beta-haematin was neither haemolytic nor oxidative towards protein sulphydryl groups. Beta-haematin did destabilise the RBC membrane, but to a lesser extent than haematin, inducing increased susceptibility to lysis caused by hypotonic medium, H2O2 or haematin. This study suggests that the destabilising effect of haematin and, to a much less extent, beta-haematin on the RBC membrane does not result from oxidative damage of membrane lipids but from direct binding or incorporation which may affect the reciprocal interactions between the membrane and cytoskeleton proteins. These changes could contribute to the reduced red cell deformability associated with severe malaria.