The results confirm previous findings that CD55 levels on the surface of RBC are lower in children with malarial anaemia , and in addition indicate that CD55 loss begins early in the course of disease and affects RBC of all ages. It is also shown that CD55 and CD59 levels decrease progressively as RBC age. Altered RBC age profiles during P. falciparum infections could therefore modify complement regulatory proteins (CrP) levels, but specifically do not account for the decrease in CrP during early malarial anaemia. The correlation between CD55 and haemoglobin levels in anaemic children suggests that CD55 loss may at least partially mediate the disease.
This study focused on acute episodes of mild to moderate anaemia. Children in the cohort were monitored intensively with routine blood smears, clinical examinations and prompt treatment in case of illness, and therefore chronic malaria cases were unlikely. Significant increase in number of young RBC in anaemic children suggests body's attempt to correct the reduction in cell mass resulting from malaria infection, and also suggests that these cases were acute and not chronic episodes . Earlier studies focused on hospitalized children with severe malarial anaemia and presumably many of these were chronic cases . Although CD55 levels were found to be higher in young RBC, the increase in the fraction of young RBC during malarial anaemia in this cohort did not lead to an overall increase of CD55 levels. RBC of all ages had lower CD55 levels in anaemic versus non-anaemic children, suggesting that an active process may be removing CD55 from RBC in cases of malarial anaemia.
Red blood cells were separated into fractions of different ages using Percoll density gradients method. Density gradient separation is a well validated technique which yields distinct cell populations with progressive shift in the cell age with density, and differ by physical and biochemical properties [15–17]. Decrease in the mean corpuscular volume, expression of GPI-anchored glycoproteins (CD55 and CD59) and increased exposure of phosphatidylserine (PS) on the external leaflet of cell membranes are related to the RBC aging process [10–18]. Indeed, in normal donor blood samples, there was a decrease in MCV, CD55 and CD59 inversely related to RBCs density. These observations, together with the increased exposure of phosphatidylserine (PS) in the more dense RBCs, supports the notion that in density -separated subsets, an enrichment of young, mature, old and very old RBCs subsets was attained.
Apparently, this is the first study to show variations in CD55 and CD59 levels in relation to RBC age during malaria. Earlier study reported a progressive decrease in CD55 and CD59 levels with RBC age in healthy donors [18, 19]. Reductions in CD55 and CD59 molecules during RBC aging in healthy individuals suggest that non-pathological mechanisms exist to mediate CrP loss throughout RBC life. These mechanisms are not fully understood, although several studies suggest that CrP molecules are lost from RBC through vesicle formation and extrusion from the cell surface [11–20]. Additional mechanisms may include proteolytic cleavage of CrP during transport and clearance of immune complexes from the RBC surface in the liver and spleen . However, the loss of RBC surface molecules during malarial anaemia appears to be a selective process. CD55 but not CD59 levels were significantly lower in anaemic donors. It is suggested that an active process separate from physiological cell ageing may be taking place during malaria that preferentially removes specific RBC surface molecules.
Alternatively the same mechanism responsible for physiological loss may be taking place, but at an accelerated rate for CD55. A process resembling transfer reaction of CR1 has been proposed to explain the loss of CD55 on RBC during malaria . During transfer reaction immune complexes are removed from RBC surfaces by phagocytes. In healthy individuals, RBC act as passive shuttles for the transport of complement-coated immune complexes from the circulation to the reticuloendothelial phagocytes in the liver and spleen.
The correlation between CD55 and haemoglobin as observed in this study suggests that CD55 depletion contribute to RBC loss during malarial anaemia. Erythrophagocytosis resulting from C3b deposition on RBC could explain the concomitant decrease of CD55 and haemoglobin. Severe malarial anaemia is associated with elevated levels of circulating immune complexes [22, 23], which could be adsorbed by CD55 and subsequently transferred to macrophages . The loss of CD55 during this process could compromise its regulatory function and allow the deposition of opsonin C3b on RBC [24, 25], leading to increased RBC destruction by the phagocytes in the reticuloendothelial tissues. Complement binding to RBC has been reported to be associated with macrophage activation and reduced haemoglobin in P. falciparum malaria .
Alternatively, direct lysis of RBC by membrane attack complex (MAC) could explain the concomitant decrease of CD55 and haemoglobin, but seems less likely. Children with severe malarial anaemia have been reported to have low levels of CR1 and CD55 . Such deficiencies would likely render the RBC vulnerable to complement attack, especially during complement activation which is common during malaria . However, CD59 did not decrease significantly during anaemia episodes in this or earlier studies , which imply that complement-mediated lysis is not the likely mechanism for RBC loss during malaria. CD59 is a principal regulatory protein of complement attack . Weisner et al  found that despite the activation of all lytic complement factors, no complement-mediated lysis of RBC occurred in the presence of functional intrinsic CD59. Because CD59 levels do not change significantly during malarial anaemia, this probably limits complement-mediated lysis.
The differential loss of CD55 and CD59 may be related to their different roles in protection from complement attack. Activation of complement occurs in a step-wise fashion, and each regulatory protein acts at a different step in the cascade. CD55 acts at the initial enzymatic step to prevent the activation of C3 to C3b by accelerating the dissociation of the C3 convertase C4-2a and C3bBb [28–30]. CD59 prevents formation of polymeric C9 complex at the final step of MAC assembly . At least during acute malarial anaemia, CR1 and CD55 may sufficiently regulate the complement cascade to limit formation of MAC, thereby consuming CD55 and sparing CD59.
Demonstration of higher amounts of C3 activation and degradation fragments bound to the RBCs of the anaemic children and an inverse correlation with CD55 would be another approach to support the hypothesis that CD55 loss supports the causal role in malarial anaemia. Moreover, demonstration of C3b on the CD55 low cells, would argue that sublytic C5b-9 induced the ectocytosis of infected and bystander cells, which resulted in the preferential loss of CD55 over CD59. This phenomenon was not explored in the current study, but earlier studies by other researchers did demonstrate an increase in C3b deposition on low red cell CR1 and CD55 levels in children with severe malarial anaemia [24, 25].