Malaria and the Red Cell Membrane

G. Pasvol, B. Clough, J. Carlsson S UMMA R Y. Malarial parasites are primarily parasites of red cells and during infection ingest most of the haemoglobin within these cells, leaving the membrane as the only vestige of the original host cell. The red cell membrane thus plays a key role at all stages of infection with malarial parasites, and is modified in many ways during parasitisation, so that at least functionally it has little resemblance to the membrane from which it was originally derived. The highly specific and ordered process of parasite invasion of red cells is regulated at least in part by the uninfected red cell membrane. The red cell sialoglycoproteins or glycophorins of this membrane have been shown to play an important role in invasion by Pfusmodium fulcipurum, the species of most importance to man because of it’s high morbidity and mortality. Structurally, dynamic changes occur within the membrane during parasitisation, and a number of parasite proteins have been found to be associated within it, but changes on the surface of the infected cell have been more difficult to demonstrate. The membrane of the infected cell is important in the many metabolic processes of the parasite, as well as the critical cell-cell interactions that occur when cells containing mature parasites bind to endothelial cells (cytoadherence), bind to uninfected cells (rosetting), or interact with macrophages and other leucocytes. The recognition molecules on the red cell membrane involved in invasion, cytoadherence and rosetting appear to be quite distinct. Structural and functional changes have also been shown to occur in the membranes of uninfected red cells, both in infected patients, and in the presence of parasites in vitro. Interactions of the parasite P. fafciparum with the red cell membrane hold the key to our understanding of the pathogenesis of severe falciparum infection in man.

Malaria due to Plasmodium falciparum is a disease which can involve almost every organ and tissue in the body even though malarial parasites infect only red cells and occasionally platelets. Once within the red cell, the parasite is almost exclusively confined to the intravascular compartment, and it is only in petechial haemorrhages as found in post-mortem specimens of patients dying with cerebral malaria, B. Clough, J. Carbon, St Marys Hospital Medical School, Unit of Infectious Diseases and Tropical Medicine, Lister Unit, Northwick Park Hospital, Middlesex HA1 3UJ. UK. Correspondence to: Geoffrey Pasvol, Professor of Infectious Diseases and Tropical Medicine. Tel: 08 l-869 283 l/2 Fax: 08 l-869 2836. G. Pasvol,

that the parasite is ever found within tissue. This restriction raises the question of how the parasitised red cell, which remains within the intravascular compartment, can produce such major pathophysiological changes. The parasitised red cell is however in many respects a hybrid, possessing a combination of properties of both red cell and parasite. Recently limited parasite maturation has been achieved in cell-free cultures,’ but it is unlikely that such extracellular parasites could survive in vivo. The red cell membrane remains crucial in the interaction of parasite and red cell, and the subsequent discussion will centre on P. falciparum, being the malarial parasite of greatest importance to man.

184

MALARIA

AND THE RED CELL MEMBRANE

The Red Cell Membrane

The red cell membrane is probably one of the best examples of a biological membrane which has been studied in detail and encompasses a bilipid layer, transmembrane molecules and a membrane skeleton. Despite its relative simplicity, many aspects of this membrane remain unknown, and it possesses a number of unique features.

Other polypeptides such as decay accelerating factor (expressing the Cromer antigens), CD44 (expressing the In”/Inb antigens), CD59, acetylcholinesterase and those expressing the Kell, Lutheran, Duffy, Cartwright and LW antigens are present on red cells, but without any apparent relationship to falciparum malaria. Unlike vivax malaria for example, the Duffy antigens do not play a role in the invasion of red cells by falciparum parasites.

Bilipid Laver

Membrane Skeleton

The bilipid layer of the red cell membrane is similar to that in any other cell. One distinctive feature is that the outer leaflet contains relatively high amounts of the phospholipid, phosphatidylcholine (PC) and sphingomyelin (S), whereas the inner leaflet is relatively rich in phosphatidylserine (PS) and phosphatidylethanolamine (PE). Much of the cholesterol of the membrane is found between these two leaflets.

Below the bilipid layer and providing much of the elastic properties essential for the function of the highly deformable cell is a unique system of ‘scaffolding’ otherwise known as the membrane skeleton. The membrane skeleton is made up largely of spectrin dimers which meet in horizontal interactions (i.e. horizontal to the membrane) at sites rich in actin, protein 4.1 and adducin. and at the other end in contact with other spectrin molecules. Protein 2.1 (ankyrin) links the spectrin molecules in vertical interactions with band 3. Protein phosphorylationdephosphorylation is one of the physiological processes controlling red cell membrane stability. Recent data indicate that phosphorylation is involved in molecular interactions at several levels and thereby contributes to the integrity of general membrane structure (reviewed by Boivin3). The intermolecular relationship between spectrin, ankyrin, protein 4.1 and band 3, and their degree of binding to one another, has been shown to be determined by their relative states of phosphorylation.4-7 The rather impervious barrier resulting from the combination of bilipid layer, transmembrane molecules and membrane skeleton, makes the red cell resistant to invasion by most infectious agents, and in this respect the only known pathogens which bind to and invade human red cells are Plasmodia, Babesia and Bartonella spp. Other organisms such as influenza virus, adenovirus, Mycoplasma, Gonococcus, and other infectious agents will bind to red cells, but these organisms do not invade the red cell and such binding is not thought to be the major mechanism, or the site of pathogenesis of these diseases. Malaria remains by far the most important pathogen of red cells. Malarial parasites interact with the red cell membrane at a number of crucial stages during development; firstly in the invasion of parasites into red cells; secondly, in modifications of the red cell membrane which occur in order to facilitate parasite growth, such as the transference of nutrients and waste products; thirdly, in the binding or cytoadherence of parasitised red cells to endothelial cells in a process called ‘sequestration’, and finally, in the ability of parasitised cells to bind uninfected red cells in a process known as ‘rosetting’. Parasitised red cells also become targets for protective immunity, and in this respect the parasitised red cell membrane provides the interface between the parasitised cell and the

Transmembrane Molecules

The red cell has relatively few transmembrane proteins, although more are being identified.’ The most commonly occurring transmembrane proteins are band 3 (the major anion transport protein -95 kD) and the glycophorins (GPS) of which there are at least five different types (GPS, A, B, C, D. & E). There are about a million copies of band 3 per cell which express AB(H) blood group activity. Band 3 consists of 911 amino acids, 360 of which at the N-terminal end are located in the cytoplasm and bind to the red cell skeleton via protein 2.1 (ankyrin) and possibly protein 4.1. This cytoplasmic domain also binds peripheral proteins such as protein 4.2, the glyceraldehyde 3-phosphate glycolytic enzymes dehydrogenase, aldolase and phosphofructokinase and also haemoglobin. The C-terminal 550 amino acids comprise the membrane associated domain involved in anion transport with a single, but almost always heterogeneous N-linked sugar at the asparagine at position 642. Band 3 forms dimers and probably oligomers in the uninfected membrane, but anion transport is facilitated by the monomer. Each red cell has about a million copies of GPA and 250000 copies of GPB which together express MN blood group activity. The other GPS number about 50000 or less copies each per red cell. The GPS are highly glycosylated with a large number of sialic acid-containing O-linked tetrasaccharides. Whilst GPA and B are homologous in part, GPC is completely unrelated. The cytoplasmic portion of GPC interacts with protein 4.1 of the red cell skeleton. Antigens of the rhesus (Rh) blood group appear to be located on two related hydrophobic polypeptides of 30000 and 50000 kD, but as yet these have not been implicated in any interaction with malaria. Rh null cells are fully susceptible to invasion by P. falciparum (Pasvol G, unpublished observations).

BLOOD

immunological macrophages, plasma cells.*

system of the host and can neutrophils. lymphocytes and

REVIEWS

185

bind even

The Role of the Red Cell Membrane in Invasion Merozoites released from the hepatic (preerythrocytic) schizonts or from asexual blood-stage schizonts are released into the circulation where they can interact with the red cell membrane of uninfected red cells. Invasion is a highly specific and sequential process in which the merozoite attaches to a susceptible red cell, orientates itself such that its apical end is apposed to the red cell membrane (Fig. l), and then slowly moves into a localised invagination of the red cell (Fig. 2).9 This invagination is continuous with the red cell surface, but unlike the rest of the membrane is depleted of electron-dense intramembranous particles. Recent evidence suggests that this parasitophorous vacuole membrane (PVM) is probably of parasite, rather than red cell origin” possibly originating from the rhoptries, one of the microorganelles of the merozoite.’ l-l3 Once surrounded by the PVM. the parasite comes to lie ‘within’ the red cell cytosol. Identification of the molecules on the red cell surface to which the merozoites bind has been the subject of intensive recent research. The sialic acid rich GPS especially GPA and GPB, play a major role in this interaction. r4-16 Moreover it would seem the O-linked tetrasaccharides, rather than the very much larger N-linked sugars on these molecules, are important. ” Invasion by parasites of uninfected cells appears to correlate with the binding of a conserved parasite molecule, the erythrocyte binding antigen of 175 kD (EBA 175) to these cells. l8 EBA 175 has been cloned and sequenced and demonstrates a translation product of 1435 amino acids.” Antibodies to a

Fig. 1 Invasion of an uninfected red cell (right) merozoite (left). The apical end of the merozoite specialised organelles is apposed to the red cell merozoite is covered by a fuzzy coat. (Courtesy M. Aikawa).

by the with it’s membrane. The of Professor

Fig. 2 lnteriorisation of the merozoite into the red cell. The merozoite appears to move into an invagination of the red cell, the origin of which has not been definitively determined. At the site of transition from red cell surface to invagination, a ‘junctional’ zone consisting of electron dense material is apparent. The invaginated portion of the membrane is devoid of intramembranous particles. That part of the merozoite within the invagination is devoid of the ‘fuzzy coat’. (Courtesy of Professor M. Aikawa).

synthetic peptide of 44 residues in length (EBA peptide 4) inhibits invasion and blocks the binding of the parent molecule to red cells. Moreover EBA peptide 4 is conserved amongst a number of strains of P. falciparum from different parts of the world, making it a possible vaccine candidate. EBA 175 recognises the N-acetyl neuraminic acid linked in an CI 2-3 configuration to galactose on the O-linked tetrasaccharide, but not the other neuraminic acid which is linked in the a 2-6 configuration.” Current evidence does not indicate a role for band 3 in merozoite recognition of red cells, although band 3 may be involved in a subsequent step in the invasion process. Once attachment and orientation have occurred, interiorisation follows, but the molecules involved in this process have not as yet been identified. Phosphorylation is important in the invasion process.21 Invasion is significantly inhibited in the absence of cytoplasmic ATP22,23 and adenosine, an inhibitor of phosphorylation inhibits invasion.24 The membrane skeleton appears to be the major target of phosphorylation/dephosphorylation in the red cell relevant to invasion. Such phosphorylated membrane proteins have been identified in P. falciparum infections, the pattern of which appears to be quite different from normal uninfected cells.2s-27 In addition to phosphorylation of red cell membrane proteins, phosphorylation of parasite proteins have also been described,2ss29 and parasite-derived protein kinases have been identified in P. ,falcipczrum.30~31 One of these MESA (the mature-parasite infected erythrocyte surface antigen) is of particular interest as it binds protein 4.1 and shows similarities to other cytoskeletal and neurofilament proteins including myosin. 28 Certain protein kinase inhibitors have been

186

MALARIA

AND THE RED CELL MEMBRANE

shown to inhibit invasion at low concentration (Carlsson and Brown, unpublished observations). In Southeast Asian ovalocytosis, in which the cells are resistant to invasion in vitro by P. falciparum and P. knowlesi, a molecular defect of band 3 has been found involving a deletion of nine amino acids (position 400-408) at the interface between the cytoplasmic tail and the first transmembrane domain of the protein.32 It has been reported that ovalocyte band 3 binds to ankyrin with higher affinity than normal resulting in a cell with increased membrane rigidity.33 However, other results have indicated that the interaction between band 3 and ankyrin is norma1,32 and suggest that conformational changes in band 3 at the site of insertion into the membrane could result in altered deformability and that this also results in defective anion transport.34 In order for invasion to proceed beyond the step of attachment, a number of physico-chemical interactions of the membrane need to take place in which the membrane may deform.35 However, whilst the rates of invasion can be correlated with deformability of the red cell membrane induced by a number of red cell antibodies,36 a direct relationship of membrane rigidity or the lateral mobility of proteins such as GP and band 3 to invasion has not been established (Clough, unpublished observations). Recently, parasites which can invade red cells in vitro by pathways less dependent on neuraminic acid have been identified,37 which may reflect the ability of these parasites to switch their requirements for invasion.38 One of these alternative pathways is trypsin-sensitive and may be independent or less dependant on GPA or GPB. However, such parasites have yet to be identified in field isolates. Another recent finding is that heparin with low antithrombin III affinity (i.e. low anticoagulant efficacy) is capable of inhibiting invasion in vitro and provides the basis for possible adjunctive therapy in severe disease,39 although a recent trial of low dose subcutaneous heparin (70 units/Kg three times a day for 5 days) did not alter the course of mild malaria.40

The Membrane of the Parasitised Cell

Transport Functions of the Membrane of the Parasitised Red Cell The membrane of the parasitised red cell differs in many respects from the parent membrane. Nutrients need to be acquired and waste products removed. A number of transport channels have been found to be present in this membrane which are not present in the uninfected red cell (Fig. 3) (reviewed by4r). As early as 6 h after invasion, the membrane of the infected cell gradually becomes more permeable to a number of substrates. Membrane transport for a number of nutrients increases; for amino acids especially glutamine, glutamic acid, cysteine, methionine and isoleucine; for substances such as the hexitols

e.g. mannitol and sorbitol; for phospholipid substrates such as choline, myoinositol and fatty acids, and for nucleosides. In addition, endogenous specific carriers for glucose, nucleosides and tryptophan are activated by parasite maturation. One of the major waste products of the parasite is lactate, and the parasite appears to be incapable of gluconeogenesis. Transport of lactate across the red cell membrane increases 600-fold in infected cells probably via three parallel pathways; firstly as an anion or acid through new permeability pathwayswhich is the predominant route; secondly via the endogenous lactate carrier which can be inhibited by cinnamic acid derivatives4* and thirdly as an anion via the modified anion exchanger (band 3).43 Unfortunately as with many of these new pathways, the biochemical nature of them has yet to be characterised. There is still debate as to whether these permeant substances are taken up by a single or multiple pathways, and if these pathways are of parasite or modified host cell origin. Proteins of parasite origin are undoubtedly deposited in the membrane of the parasitised cell (see below), but none has been ascribed a transport function. Inhibition of the development, rather than function of these pathways by inhibitors of protein synthesis, would suggest that they are the result of parasitederived membrane components rather than red cell origin. Conflicting evidence has arisen as to whether parasites are capable of ingesting macromolecules such as dextrans, protein A and IgG. It has been suggested that these relatively large molecules gain direct access to the aqueous space between the PVM and parasite plasma membrane through a parasitophorous duct (Fig. 3).44 These molecules can then be endocytosed by the parasite. Such a duct could be a relic which persists from the time of merozoite invasion or a membrane-bound pathway which develops later. However, other workers have shown that macrointraduring molecules are not internalised erythrocytic maturation of the parasite, but are possibly internalised during merozoite invasion.45 Lipid Changes in the Infected Cell Membrane During intra-erythrocytic development of the malaria parasite, modifications of the lipid bilayer of the red cell membrane occur. Infected cells contain more PC and phosphoinositol (PI), but less S than uninfected cells. The fatty acid content is modified with large increases in palmitic and oleic acids, and a decrease in arachidonic and docosahexaenoic acid.46 The asymmetric arrangement of the phospholipids with PC and S localised in the outer (exoplasmic) leaflet and phosphatidylserine (PS) and phosphatidylethanolamine (PE) in the inner (cytoplasmic) leaflet is altered after invasion.47*48 An increase of PS and PE occur in the outer leaflet, whereas PC increases in the inner leaflet. However, some studies have argued

BLOOD

RED CELL PATHWAYS Band3 Amino acids

i

I

REVIEWS

187

PARASITE INDUCED PATHWAYS amino acids hexitols nucleosides

choline

Pumps I

ptimalarial

Drug4

Fig. 3 Schematic representation of permeation pathways in the host membrane of malaria-infected cells4’ The major permeation pathways present in normal erythrocytes are depicted on the left side of the scheme, grouped as carriers, pumps and channels (leaks). The parasite-induced or modified pathways (neotransporters) are depicted on the right side, some of which might display pore- or channel-like properties (parallel lines), and others are pathways for simple diffusion either through the modified bilayer of the host cell membrane (shaded rectangles) or parasite-derived or modified proteins (shaded circles). The new pathways can serve as pharmacologic targets for antimalarial drugs (AMD) or as route for selective admission of AMDs into infected cells. A proposed parasitophorous duct linking the parasite with the outside of the red cell is also shown.

that the phospholipid composition and organisation remains unchanged provided that the cells are not energy deprived and protected against ATP and magnesium depletion.49-51 The alterations in the organisation of the membrane phospholipids if they do occur, may account for a number of the membrane abnormalities observed in infected cells including increased membrane fluidity,52,53 and permeability to cations and nutrientss4 Additionally it appears that lipids themselves are capable of crossing the bilipid layer of the infected cell,47.49 as the parasite is incapable of synthesising fatty acids or cholesterol de novo,55 whereas it can synthesise phospholipids using fatty acids from serum or the red cell membrane as substrate.56 Cytoadherence A major factor thought to be responsible for the severe pathology of falciparum malaria is the ability of red cells infected with mature parasites to ‘bind’ or cytoadhere to endothelial cells in the capillaries of key target organs and tissues.57*58 Cytoadherence of infected cells has most recently been shown to occur under conditions of flow in vitro with peak adhesion occurring at 0.1-0.2 Pa. 59 Since adherence occurs within a narrow range of flow, it has been suggested that the flow characteristics themselves might partly explain the tissue distribution of cytoadherence. In one study the density of packing of schizonts could

be ranked in the following descending order; brain, heart, liver, lung, kidney and lastly blood.60 However, the ability of parasitised cells to cytoadhere in vitro using C32 melanoma cells, human umbilical vein endothelial cells (HUVECs) or the macrophage cell line U937 does not appear to correlate with disease severity. If examined by scanning electron microscopy, a number of regular, symmetrically arranged ‘knobs’ appear on the surface of the cell as the immature parasite of P. falciparum develops within these cells (Fig. 4).61 These knobs are thought to be the site at which the parasitised red cell attaches to the endothelial cell in the deep tissues and the site of contact can be visualised by transmission electron microscopy (Fig. 5). However, parasite lines have recently been described that lack knobs and yet are capable of cytoadherence in vitro. 62,63 Moreover knobless parasites, although only very few in number have been isolated from patients with malaria after limited culture in vitro.* Several proteins have been identified with the knobs and include: (i) The histidine rich protein (HRP) which induces an excrescence below the bilipid layer and which is thought to be attached to the cytoskeleton.64 This protein is also known as KAHRP (knob-associated histidine rich protein). (ii) A number of high molecular weight proteins which are thought to protrude from the HRP of

188

MALARIA

AND THE RED CELL MEMBRANE

Fig. 4 Scanning electronmicrograph of a red cell containing a mature malarial parasite. As the malarial parasite matures within the red cell a regular array of excrescences or ‘knobs’ appear on the surface of the infected cell. (Courtesy of Professor M. Aikawa).

Fig. 5 Cytoadherence of an infected red cell (above) to an endothelial cell (below). The site of attachment of the infected cell appears to occur at the site of the knob. (Courtesy of Professor M. Aikawa).

which, the best known is the P. falciparum erythrocyte membrane protein 1 (Pf EMP 1),65 which show variation in size between 220-350 kD. In addition a modification of band 3 may also be involved,66-68 as well as an antigen recognised by the cross-reactive monoclonal antibody 33G2.62 (iii) A knob-associated structural protein such as MESA, also known as Pf EMP2. This protein has been suggested to be associated with, or bound to, protein 4.1 of the red cell membrane skeleton.28*69 Pf EMP 1 undergoes phenotypic antigenic variation, analogous in many ways to the variable coat of the trypanosome which is shed under immune pressure to be replaced by a different variant. Such

Fig. 6 Transmission electronmicrograph of rosetting of uninfected red cells around a parasitised red cell (centre). The rosetting cells lie in close juxtaposition to the infected cell as in cytoadherence but the molecular interactions appear to be distinct. (Courtesy of Professor M. Aikawa).

a mechanism may be responsible at least in part for host immune evasion. Purification of Pf EMP 1 has been elusive, although adult hyperimmune sera can precipitate this antigen which seems to be exclusively located at the site of the knob. Pf EMP 1 is thought to bind to a number of potential ligands on the surface of endothelial cells which at present include: (a) the adhesion molecule CD36 expressed on platelets, monocytes/macrophages, and early red cell precursors,‘I” (b) thrombospondin, a major component of the platelet c1 granule,71 and (c) ICAM 1, the ligand for the leucocyte integrin LFA- 1.72 However, the specificity of binding of parasitised cells to each of these various molecules is thought to be quite different, and it has yet to be established which of these ligands is/are the definitive receptor(s) in the disease in man. Whether Pf EMPl is the molecule which binds all three, has yet to be established. Binding to endothelial cells is not exclusive; parasitised red cells can also bind to autologous white cells which include monocytes, neutrophils, lymphocytes and plasma cells, and appears to correlate with cytoadherence to melanoma or endothelial cells.’ It thus appears that there are few cells in the intravascular compartment to which infected cells cannot bind-and as will be seen below, not even uninfected red cells are spared. Rosetting In an interaction quite distinct from cytoadherence, red cells containing mature stages have recently been found to bind uninfected red cells to their surface.73s74

BLOOD

Interest in rosetting derives from its potential as a parasite virulence factor. In a study in The Gambia, all isolates of P. fulciparum obtained from children with cerebral malaria were capable of rosetting, whereas many isolates from patients with mild disease did not.75 Moreover, plasma from children with mild disease could more often disrupt rosettes than plasma from patients with cerebral malaria.75.76 A monoclonal antibody to HRP has been shown to specifically a surprising finding, since the disrupt rosettes,” location of this protein is thought to be below the bilipid layer in the red cell cytoskeleton. However, immunoprecipitation using the monoclonal antibody and metabolically labelled parasite material, showed that another parasite derived histidine-rich polypeptide of 28 kD was also recognised by the antibody, and it is possible that this crossreactive antigen could be involved in rosetting. Studies by other authors have not demonstrated a correlation of rosetting and disease severity. Ho and collaborators were unable to find a significant association between rosetting and the biochemical indices of severe malaria, but noticed an inverse relationship with cytoadherence (perhaps implying that if parasites could not rosette they would cytoadhere).‘* There are undoubtedly parasites that are capable of both (i.e. rosetting and cytoadherence).s*79 If rosettes are first disrupted, these same parasites are not only capable of cytoadherence to C32 melanoma cells but also appear to adhere more avidly than their nonof rosetting rosetting counterparts. ” The importance however is emphasised by the finding that the forces that bind rosettes together in vitro exceed those of cytoadherence by at least a factor of five.81 In the rat mesocaecal model, rosetting parasites resulted in greater obstruction to microvascular blood flow than cytoadherent lines incapable of resetting.” Rosetting has characteristics quite distinct from that of cytoadherence. Unlike cytoadherence, rosetting is calcium dependent and is sensitive to heparin.74 There is however wide variation in rosette sensitivity to heparin. Heparin with a low affinity for antithrombin III (i.e. low anticoagulant activity) and high molecular weight was as effective as standard heparin of rosetting in disrupting rosettes. 83 The insensitivity to neuraminidase or trypsin treatment of uninfected cells also distinguishes it from the interactions which occur during invasion. 74 It can be concluded that the recognition molecules on the red cell membrane involved in invasion, cytoadherence and rosetting are all quite distinct.

The Membrane of the Uninfected Infection

Cell in Malarial

In addition to the modifications that occur in the red cell membrane of the infected cell, a number of changes occur in the membranes of uninfected cells and in view of the anaemia that may develop, which is often in excess of the relatively few cells that are

REVIEWS

18’)

infected, it is likely that these uninfected cells are modified in some way. The phospholipid asymmetry in uninfected cells has been reported to be changed in P. knowlesi infection.84 However, at least 30% of animals acutely infected were subsequently shown not to have this phospholipid asymmetry change, although it was a constant feature of chronically infected, and splenectomised uninfected animals,85 implying that the spleen plays an important role in maintaining normal lipid asymmetry, which may be particularly relevant to the insidious development of 86 However. studies on in malarial anaemia in man. vitro cultures of P. falciparum and P. knowllesi in splenectomised monkeys have not confirmed any abnormal lipid membrane redistribution in uninfected cells in the presence of parasitised red ceIls.47.49.87 Antigen-antibody complexes which fix complement and then bind via the C3b receptor to the surface of uninfected cells may result in these cells becoming the targets for phagocytosis. Membrane rigidity of uninfected red cells in the presence of parasites is quite different from normal red cells where exposure of uninfected cells to conditioned media from parasite cultures reduces membrane deformability.88 These mechanisms affecting uninfected cells may be important especially where profound anaemia may develop slowly, and the number of infected cells is small. The uninfected red cell membrane is also involved in rosetting, but the nature of the molecules on the surface of red cells responsible for binding to schizont-infected cells has yet to be established. Cells lacking GPA and GPB (MkMk) are capable of rosetting (Clough B, unpublished observations).

Conclusion The malarial parasite is a unique agent of infection in that it targets red cells in the circulation and thereby affects all organs and tissues of the body. Since the parasite may ingest up to 75% of the haemoglobin contained within the red cell during maturation, the red cell membrane is the major component of the original uninfected cell that remains after infection. The red cell membrane therefore holds the key to the major alterations initiated by the parasite which might be crucial in the interaction of the parasite and host. Improved knowledge of these interactions should provide a better understanding of the pathophysiology and immunology of malaria, and may provide novel means by which specific agents can be targetted at some of these membrane functions and lead to the destruction of these parasites or to modification of disease severity.

Acknowledgements We would like to thank Dina Shah and Barbara Bixby for help in preparing the script and the Weilcome Trust for their support. We also thank Professor H. Ginsburg, Drs R. N. Davidson and J. Brown, and MS A. Atilola for their critical comments.

190

MALARIA

AND THE RED CELL

MEMBRANE

-

References 1. Trager W, Williams J 1992 Extracellular (axenic) development in vitro of the erythrocytic cycle of Pfasmadium Fulciparum. Proceedings of the National Academy of Science USA 89: 5351-5355 2. Anstee D J 1990 Blood group-active surface molecules of the human red blood cell. VOX Sanguinis 58: I-20 3. Boivin P 1988 Role of the phosphorylation of red blood cell membrane proteins. Biochemical Journal 256: 689-695 4. Lu P-W, Soong C-J, Tao M 1985 Phosphorylation of ankyrin decreases its affinity for spectrin tetramer. The Journal of Biological Chemistry 260: 14958814964 5. Eder P S, Soong C-J, Tao M 1986 Phosphorylation reduces the affinity of protein 4.1 for spectrin. Biochemistry 25: 1764-1770 6. Soong C-J, Lu P-W. Tao M 1987 Analysis of band 3 cytoplasmic domain phosphorylation and association with ankyrin. Archives of Biochemistry and Biophysics 254: 509-517 7. Subrahmanyam G, Bertics P J, Anderson R A 1991 Phosphorylation of protein 4.1 on tyrosine-418 modulates its function in vitro. Proceedings of the National Academy of Science USA 88: 5222-5226 W, Afzelius B A, Helmby H et al 1992 8. Ruangjirachuporn Ultrastructural analysis of fresh Plasmodium fulciparuminfected erythrocytes and their cytoadherence to human leukocytes. American Journal of Tropical Medicine and Hygiene 46: 511-519 W C, Shiroishi T 1975 9. Dvorak J A, Miller L H, Whitehouse Invasion of erythrocytes by malaria merozoites. Science 187: 748-749 10. Dhtzewski A, Mitchell G, Fryer P et al 1992 Origins of the parasitophorous vacuole membrane of the malarial parasite, Plasmodium falciparum, in human red blood cells. Journal of Cell Science 102: 527-532 11. Bannister L H, Mitchell G H, Butcher G A, Dennis E D 1986 Lamellar membranes associated with rhoptries in erythrocytic merozoites of Plasmodium knowlesi: a clue to the mechanism of invasion. Parasitology 92: 291-303 J P 1986 Rhoptry 12. Stewart M J, Schulman S. Vanderberg secretion of membranous whorls by Plasmodium falciparum merozoites. American Journal of Tropical Medicine and Hygiene 44: 37-44 13. Mikkelsen R B, Kamber M, Wadwa K S, Lin P-S, SchmidtUllrich R 1988 The role of lipids in Plasmodium fulciparum invasion of erythrocytes: A coordinated biochemical and microscopic analysis. Proceedings of the National Academy of Science USA 85: 5956-5960 14. Pasvol G, Wainscoat J S, Weatherall D J 1982 Erythrocytes deficient in glycophorin resist invasion by the mararial parasite Plasmodium falciparum. Nature 297: 64-66 15. Pasvol G, Jungery M, Weatherall D J et al 1982 Glycophorin as a possible receptor for Plasmodium falciparum. The Lancet 947-950 16. Pasvol G 1984 Receptors on red cells for Plasmodium fufciparum and their interaction with merozoites. Phiiosophical Transactions of the Royal Society of London 3@7: 189-200 17. Pasvol G, Hodson C, Tanner M J A. Newbold C I 1987 The relative roles of N- and O-linked carbohydrate in the invasion of human red cells by merozoites of Plasmodium fulciparum. In: Chang K-P, Snary D Host-Parasite (eds) Cellular and Molecular Interactions in Protozoa1 Infections. Springer-Verlag, Berlin, pp. 245-254 18. Camus D, Hadley T J 1985 A Plasmodium falciparum antigen that binds to host erythrocytes and merozoites. Science 230: 553-556 19. Sim B K L, Orlandi P A, Haynes J D et al 1990 Primary structure of the 175K Plasmodium falciparum erythrocyte binding antigen and identification of a peptide which elicits antibodies that inhibit malaria merozoite invasion. The Journal of Cell Biology 111: 1877-1884 20. Orlandi P A, Klotz F W. Haynes J D 1992 A malaria invasion receptor, the 175-kilodalton erythrocyte binding antigen of Plasmodium fulciparum recognizes the terminal NeuSAc(a2-3)Galsequences of glycophorin A. The Journal of Cell Biology 116: 901-909 K, Dluzewski A, Wilson R J M, Gratzer W B 21. Rangachari

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

1986 Control of malarial invasion by phosphorylation of the host cell membrane cytoskeleton. Nature 324: 364-365 Olson J A. Kilejian A 1982 Involvement of spectrin and ATP in infection of resealed erythrocyte ghosts by the human malaria parasite, Plasmodium falcipnrum. The Journal of Cell Biology 95: 757-762 Dluzewski A R, Rangachari K, Wilson R J M, Gratzer W B 1983 A cytoplasmic requirement of red cells for invasion by malarial parasites. Molecular and Biochemical Parasitology 9: 145-160 Dluzewski A R, Rangachari K, Gratzer W B, Wilson R J M 1983 Inhibition of malarial invasion of red cells by chemical and immunochemical linking of spectrim molecules. British Journal of Haematology 55: 629-637 Jones G L, Edmundson H M 1990 Protein phosphorylation during the asexual life cycle of the human malarial parasite Plasmodium falciparum. Biochimica et Biophysics Acta 1053: 1188124 Murray M C, Perkins M E 1989 Phosphorylation of erythrocyte membrane and cytoskeleton proteins in cells infected with Plasmodium falciparum. Molecular and Biochemical Parasitology 34: 229-236 Wasserman M, Vernot J P, Mendoza P M 1990 Role of calcium and erythrocyte cytoskeleton phosphorylation in the invasion of Plasmodium falciparum. Parasitology Research 76: 681-688 Coppel R L 1992 Repeat structures in a Plasmodium fulciparum protein (MESA) that binds human erythrocyte protein 4.1. Molecular and Biochemical Parasitology 50: 335-348 Foley M, Murray L J, Anders R F 1990 The ring-infected erythrocyte surface antigen protein of Plasmodium falciparum is phosphorylated upon association with the host cell membrane. Molecular and Biochemical Parasitology 38: 69-76 Wiser M F, Eaton J W, Sheppard J R 1983 A plasmodium protein kinase that is developmentally regulated, stimulated by spermine, and inhibited by quercetin. Journal of Cellular Biochemistry 21: 3055314 Read L K, Mikkelsen R B 1990 Cyclic AMP- and Ca2+dependent protein kinases in Plasmodium faleiparum. Experimental Parasitology 71: 39-48 Schofield A E, Tanner M J A, Pinder J C et al 1992 Basis of unique red cell membrane properties in hereditary ovalocytosis. Journal of Molecular Biology 223: 949-958 Liu S-C, Zhai S, Palek J et al 1990 Molecular defect of the band 3 protein in southeast asian ovalocytosis. New England Journal of Medicine 22: 1530- 1538 Schofield A E, Reardon D M, Tanner M J A 1992 Defective anion transport activity of the abnormal Band-3 in hereditary ovalocytic red blood cells. Nature 355: 836-838 Dvorak J A, Miller L H, Whitehouse W C, Shiroishi T 1975 Invasion of erythrocytes by malaria merozoites. Science 187: 148-749 Pasvol G, Chasis J A. Mohandas N et al 1989 Inhibition of malarial parasite invasion by monoclonal antibodies against glycophorin A correlates with reduction in red cell membrane deformability. Blood 74: 1836-1843 Hadley T A, Klotz F W, Pasvol G et al 1987 Falciparum malaria parasites invade erythrocytes that lack glycophorin A and B (MkMk). The Journal of Clinical Investigation 80: 1190-1193 Dolan S A. Miller L H, Wellems T E 1990 Evidence of a switching mechanism in the invasion of erythrocytes by Plasmodiumfalciparum. The Journal of Clinical Investigation 86: 618-624 Kulane A, Ekre H-P, Perlmann P et al 1992 Effect of different fractions of heparin on Plasmodium falciparum merozoite invasion of red blood cells in vitro. American Journal of Tropical Medicine and Hygiene 46: 589-594 Hemmer C, Kern P, Holst F, Nawroth P, Dietrich M 1991 Neither heparin nor acetylsalicylic acid influence the clinical course in human Plasmodium falcparum malaria: a prospective randomized study. American Journal of Tropical Medicine and Hygiene 45: 608-612 Cabantchik Z I 1989 Altered membrane transport of malaria-infected erythrocytes: A possible pharmacologic target. Blood 74: 146441471 Kanaani J, Ginsburg H 1992 Effects of cinnamic acid derivatives on in vitro growth of Plasmodium falciparum and

BLOOD on the permeability of the membrane of malaria-infected erythrocytes. Antimicrobial Agents and Chemotherapy

36:

1102-l 108 of lactate in 43. Kanaani J, Ginsburg H 1991 Transport Plasmodium falcipurum-infected human erythrocytes. Journal of Cellular Physiology 149: 4699476 44. Pouvelle B, Spiegel R. Hsiao L et al 1991 Direct Access to Serum Macromolecules by Intraerythrocytic Malaria Parasites. Nature 353: 73-75 45. Haldar K. Uyetake L 1992 The movement of fluorescent endocytic tracers in Plasmodium falciparum infected erythrocytes. Molecular and Biochemical Parasitology 50: 16lll78 46. Hsiao L L. Howard R J, Aikawa M, Taraschi T F 1991 Modification of host cell membrane lipid composition by the intra-erythrocytic human malaria parasite Plusmodium fhlciparunr. Biochemical Journal 274: 121-132 47. Schwartz R S, Olson J A. Raventos-Suarez C et al 1987 Altered plasma membrane phospholipid organization in Plasmodium~fkdc@nwnr-infected human erythrocytes. Blood 69: 4Oll407 48. Joshi P. Gupta C M 1988 Abnormal membrane phospholipid organization in Plasmodium falcipartrm-infected human erythrocytes. British Journal for Haematology 68: 255-259 49 Van der Schaft P H, Beaumelle B, Vial H et al 1987 Phospholipid organization in monkey erythrocytes upon Plasmodium knotc~le.Ginfection. Biochimica et Biophysics Acta 901: I-14 50 Mall G N. Vial H J. Bevers E M et al 1989 Phospholipid asymmetry in the plasma membrane of malaria infected erythrocytes. Biochemical Cell Biology 68: 579-585 K, Dluzewski A R, Wilson R J M, 51 Field S. Rangachari Gratzer W B 1992 Effect of intraerythrocytic magnesium ions on invasion by Pla.vnzodium,falciparum. Parasitology 105: 15-19 52 Sherman I W, Greenan J 1984 Altered red cell membrane fluidity during schizogonic development of malarial parasites (Plusmodirtm ,fh/riparzm~ and P. lophurae). Transaction of the Royal Society of Tropical Medicine and Hygiene 78: 641-644 53 Taraschi T F. Parashar A. Hooks M, Rubin H 1986 Perturbation of red cell membrane structure during intracellular maturation of Plasmodium ,falcipurum. Science 232: 102-104 54 Ginsburg H, Kutner S. Krugliak M, Cabantchik Z I 1985 Characterization of permeation pathways appearing in the host membrane of Plusmodiunz~falcipurum infected red blood cells. Molecular and Biochemical Parasitology 14: 3133322 55 Holz G G J 1977 Lipids and the malarial parasite. Bulletin of the World Health Organization 55: 237-248 56 Vial H J, Thuet M J. Philippot J R 1982 Phospholipid biosynthesis in synchronous Plasmodium falcipurtrm cultures. Journal of Protozoology 29: 2588263 57 Luse S A. Miller L H 1971 Plasmodium ,falcipurum malaria. Ultrastructure of parasitized erythrocytes in cardiac vessels, American Journal of Tropical Medicine and Hygiene 20: 6555660 58 Udeinya I J. Schmidt J A. Aikawa M, Miller L H. Green I 198 I Falciparum malaria-infected erythrocytes specifically bind to cultured human endothelial cells. Science 213: 555 -557 59 Wick T M, Louis V 1991 Cytoadherence of Plasmodium falriparum-infected erythrocytes to human umbilical vein and human dermal microvascular endothelial cells under shear conditions. American Journal of Tropical Medicine and Hygiene 45: 5788586 60. MacPherson G, Warrell M J. White N J, Looareesuwan S, Warrell D A 1985 Human cerebral malaria. A quantitative ultrastructural analysis of parasitised erythrocyte sequestration. American Journal of Pathology 119: 385-401 61 Allred D, Gruenberg J, Sherman I 1986 Dynamic rearrangements of erythrocyte membrane internal architecture induced by infection with Plasmodium fulcipurum. Journal of Cell Sciences 81: I-16 62. Udomsangpetch R. Aikawa M. Berzins K, Wahlgren M. Perlmann P 1989 Cytoadherence of knobless Plasmodium ,faLcipurunz-infected erythrocytes and its inhibition by a human monoclonal antibody. Nature 338: 763-765 63. Biggs B A. Culvenor J G. Ng J S, Kemp D J, Brown G V

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

15.

76.

77.

78.

79.

80.

81.

REVIEWS

191

1989 Plasmodium falcipurum: Cytoadherence of a knobless clone. Experimental Parasitology 69: 189-197 Kilejian A 1979 Characterization of a protein correlated with the production of knob-like protrusions on membranes of erythrocytes infected with Plasmodium.fulcipurum. Proceedings of the National Academy of Science USA 76: 4650-4653 Leech J H. Barnwell J W, Miller L H. Howard R J 1984 Identification of a strain-specific malarial antigen exposed on the surface of PIusmodium,fulciparum-infected erythrocytes. Journal of Experimental Medicine 159: I5677 1575 Winograd E. Greenan J R T. Sherman I W 1987 Expression of senescent antigen on erythrocytes infected with knobby variant of the human malaria parasite Plaw7odium fulcipurum. Proceedings of the national Academy of Science USA 84: 1931-1935 Winograd E. Sherman I W 1989 Characterization of a modified red cell membrane protein expressed on erythrocytes infected with the human malaria parasite Plusmodium fulcipurum: Possible role as a cytoadherent mediating protein. The Journal of Cell Biology 108: 23-30 Winograd E, Sherman I W 1989 Naturally occuring antiband 3 autoantibodies recognize a high molecular weight protein on the surface of Plasmodium fulcipurum infected erythrocytes. Biochemical and Biophysical Research Communication 160: I3577 1363 Lustigman S. Anders R F. Brown G V, Coppel R L 1990 The mature parasite-infected erythrocyte surface antigen (MESA) of Plasmodium,f&iparum associates with the erythrocyte membrane skeletal protein, band 4.1. Molecular and Biochemical Parasitology 38: 261-270 Barnwell J W, Ockenhouse C F. Knowles D M 1985 Monoclonal antibody OKM5 inhibits the in vitro binding of Plasmodium ftilcipurum-infected erythrocytes to monocytes, endothelial. and C32 melanoma cells. Journal of Immunology 135: 3494-3497 Roberts D D, Sherwood J A. Spitalnik S L et al 1985 Thrombospondin binds falciparum malaria parasitized erythrocytes and may mediate cytoadherence. Nature 318: 64-66 Berendt A R, Simmons D L, Tansey J, Newbold C I. Marsh K 1989 Intercellular adhesion molecule-l is an endothelial cell adhesion receptor for P/u.vmodium /&iparum. Nature 341: 57-59 Handunnetti S M, David P H. Perera K L R L, Mendis K N 1989 Uninfected erythrocytes form ‘Rosettes’ around Plusmodium fulcipurum infected erythrocytes. American Journal of Tropical Medicine and Hygiene 40: I 15-l IX Udomsangpetch R, Wahlin B, Carlson J et al 1989 Plusmodium,fulcipurum-infected erythrocytes form spontaneous eythrocyte rosettes. Journal of Experimental Medicine 169: 1835-1840 Carlson J, Helmby H. Hill A V S et al 1990 Human cerebral malaria: association with erythrocyte rosetting and lack of anti-rosetting antibodies. The Lancet 336: 1457-1460 Treutieer C-J. Hedlund I. Helmbv H et al 1992 Rosette formation in Plasmodium /hlciparkm isolates and anti-rosette activity of sera from Gambians with cerebral or uncomplicated malaria. American Journal of Tropical Medicine and Hygiene 46: 503-510 Carlson J. Holmquist G. Taylor D W. Perlmann P, Wahlgren M 1990 Antibodies to a histidine-rich protein (PfHRPI) disrupt spontaneously formed Plusmodium fulcipurum erythrocyte rosettes. Proceedings of the National Academy of Science USA 87: 35 I I-251 5 Ho M, Davis T M E. Silamut K, Bunnag D. White N J I99 I Rosette formation of Plasmodium j&ipuruminfected erythrocytes from patients with acute malaria. Infection and Immunity 59: 2 135-2139 Hasler T. Handunnetti S M, Aguiar J C et al 1990 In vitro rosetting. cytoadherence, and microagglutination properties of Plasmodium fulcipurum-Infected erythrocytes from Gambian and Tanzanian patients. Blood 76: 184551852 Handunnetti S M. Hasler T H, Howard R J 1992 Plasmodiunz,fulciparum-infected erythrocytes do not adhere well to C32 melanoma cells or CD36 unless rosettes with uninfected erythrocytes are first disrupted. Infection and Immunity 60: 928-932 Nash G. Cooke B. Carlson J, Wahlgren M 1992 Rheological properties of rosettes formed by red cells parasitised by

192

MALARIA

AND

THE RED CELL

MEMBRANE

Plasmodium falciparum. British Journal of Haematology (In Press) 82. Kaul D, Roth E, Nagel R, Howard R, Handunetti S 1991 Rosetting of Plasmodium fulciparum infected red blood cells with uninfected red blood cells enhances microvascular obstruction under flow conditions. Blood 78: 812-819 83. Carlson J, Ekre H-P, Helmby H et al 1992 Disruption of Plasmodium faleiparum erythrocyte rosettes by standard heparin and heparin devoid of anticoagulant activity. American Journal of Tropical Medicine and Hygiene 46: 595-602 84 Gupta C M, Alam A, Mathur P N, Dutta G P 1982 A new look at nonparasitized red cells of malaria-infected monkeys. Nature 299: 259-261

85. Joshi P, Alam A, Chandra R, Puri S K, Gupta C M 1986 Possible basis for membrane changes in nonparasitized erythrocytes of malaria-infected animals. Biochimica et Biophysics Acta 862: 220-222 86. Phillips R, Pasvol G 1992 Anaemia of Plasmodium falciparum malaria. Balliere’s Clinical Haematology 5: 315-330 in malaria. 87. Gupta C M 1988 Red cell membrane alterations Indian Journal of Biochemistry and Biophysics 25: 20-24 88. Naumann K. Lloyd Jones G, Saul A, Smith R 1991 A Plasmodium falciparum exo-antigen alters erythrocyte membrane deformability. FEBS Letters 292: 95-97