Lipids and the malarial parasite*

Bulletin of the World Health Organization, 55 (2-3): 237-248 (1977) Lipids and the malarial parasite* GEORGE G. HOLZ, JR 1 Merozoite endocytosis initiates Plasmodium development in a vacuole bounded by an erythrocyte-derived membrane, whose asymmetrical distribution of lipids and proteins is reversed in its orientation with respect to the parasite plasma membrane. Reorientation may accompany the proliferation of the membrane associated with the parasite's growth and phagocytic and pinocytic feeding. Increases in the membrane surface area of the parasite, and in some cases of the erythrocyte, parallel parasite growth and segmentation. Augmentation of all the membrane systems of the infected erythrocyte causes the lipid content to rise rapidly, but the parasite lipid composition differs from that of the erythrocyte in many respects: it is higher in diacyl phosphatidylethanolamine, phosphatidylinositol, polyglycerol phosphatides, diacylglycerols, unesterified fatty acids, triacylglycerols, and hexadecanoic and octadecenoic fatty acids and lower in sphingomyelin, phosphatidylserine, alkoxy phosphatidylethanolamine, cholesterol, and polyunsaturated fatty acids. Active lipid metabolism accompanies the membrane proliferation associated with feeding, growth, and reproduction. Plasmodium is incapable of de novo biosynthesis offatty acids and cholesterol; however, it can fabricate its glycerides and phosphoglycerides with host- supplied fatty acids, nitrogenous bases, alcohols, A TP, and coenzyme A, and can generate the glyceryl moiety during glycolysis. Cholesterol is obtained from the host but nothing is known of sphingolipid origins. Lipid metabolism of the parasite may be associated with alterations in the amounts of octadecenoic fatty acids and cholesterol in the erythrocyte plasma membrane, which in turn are responsible for changes in permeability and fragility. LIPIDS AND ERYTHROCYTE MEMBRANE BIOLOGY The merozoite of the malarial parasite enters the erythrocyte by endocytosis. At the point of attach- ment of the merozoite, invagination and expansion of the erythrocyte plasma membrane occur, ac- companied by swelling of the erythrocyte. Deforma- tion of the surface is seen within 5-10 s of attach- ment and continues for 10-15 min after completion of endocytosis. Continuity of the erythrocyte surface plasma membrane with the expanding membrane of the invagination is evident, and the merozoite comes to occupy the space formed. There follows a contraction of the mouth of the invagination and a resealing of the erythrocyte plasma membrane to isolate the parasite in a parasitophorous vacuole within the host-cell cytoplasm (1-7). This sequence of events is similar to that induced by primaquine (8). * This work was supported by grant AI 05802 from the National Institute of Allergy and Infectious Diseases, US Public Health Service. 1 Professor and Chairman, Department of Microbiology, State University of New York, Upstate Medical Center, Syracuse, NY 13210, USA. The rapidity and local nature of the invagination suggest membrane expansion by the reordering of interactions among extant membrane proteins and lipids. By the time endocytosis is completed, how- ever, the surface area of the vacuole membrane surrounding a merozoite 1.5 um long and 1.0 ,um in diameter in a newly-infected 8-,tm erythrocyte is approximately 3 % of the total area of the erythro- cyte plasma membrane. The parasitophorous vacuole membrane, there- fore, appears to originate from the plasma membrane of the erythrocyte and to be reversed in its orienta- tion, with its original outer (environmental) surface facing the parasite and its original inner (cyto- plasmic) surface facing the exterior of the erythro- cyte, although still in apposition to the cytoplasm. Profound structural and functional consequences of this reversal of orientation must arise from the fact that the erythrocyte plasma membrane is asymmetric with respect to the transverse distribution of its lipids and proteins (9, 10). Its glycolipids and choline-containing lipids (phosphatidylcholine and sphingomyelin) are concentrated in the outer 3602 237 G. G. HOLZ, JR leaflet of the lipid bilayer of the membrane and phosphatidylethanolamine, phosphatidylserine, and phosphatidylinositol (when present) in the inner leaflet. Cholesterol is found mainly in the outer leaflet among the choline-containing lipids, and has a fluidity-controlling role. The outer leaflet is less fluid than the inner leaflet; this property is associated with the high proportion of saturated fatty acids acylating its phospholipids and with the nature of the polar head group, choline. In contrast, the phospho- lipids of the highly fluid inner leaflet contain large amounts of polyunsaturated fatty acids. Glyco- proteins of the membrane lipids have their carbo- hydrate antigenic determinants and receptors facing the environment. Membrane-associated enzymes, some of them vectorial, are variously positioned in the membrane. As the malarial parasite lies in the parasitophorous vacuole, its cytosol and organelles are separated from the blood plasma environment of the erythro- cyte by three membranes, two with normal asym- metry with respect to the parasite (the plasma mem- branes of the erythrocyte and parasite) and a third with reversed asymmetry lying between them (the parasitophorous vacuole membrane). The per- meability properties of such a three-membrane system are unknown but obviously unusual; for example, vectorial enzymes in the newly-formed parasitophorous vacuole membrane would have incorrect orientations for translocation of molecules from plasma to parasite, and the passive barrier properties of the lipid bilayer would be compromised. Logically, one would assume that such structural and functional aberrations are quickly changed, since isotopically-labelled molecules of considerable variety enter parasites within erythrocytes and parasites freed from erythrocytes by immune haemolysis or saponin lysis but still enclosed in the parasitophorous vacuole membrane (11-16). Also, parasites can be grown in vitro within or without their host erythrocyte (16-18). Sterling et al. (19) have noted that the space between the parasitophorous vacuole membrane and the plasma membrane of the young trophozoite diminishes as the parasite develops and the mem- branes become closely associated. Langreth (20) has observed that in P. lophurae-infected erythrocytes Na+-K+ adenosinetriphosphatase (EC 3.6.1.3) ac- tivity is localized on the inside of the erythrocyte plasma membrane but is not concentrated on the cytoplasmic side of the parasitophorous vacuole membrane, as would be expected if that membrane were derived from the erythrocyte plasma membrane but reversed in orientation. In fact, the majority of the adenosinetriphosphatase activity is found on the side of the membrane facing the plasma membrane of the parasite. It was suggested that, although the vacuole membrane might initially be derived from the erythrocyte plasma membrane, it is modified and controlled by the parasite. In support of this hypothesis, Trager (21) cited an observation that rabbit anti-duck erythrocyte antiserum fails to agglutinate P. lophurae freed from erythrocytes by immune haemolysis but still encompassed by the parasitophorous vacuole membrane. Also, Sherman (personal communication, 1976) has found that P. lophurae (also freed by immune haemolysis) does not bind colloidal iron or cationic ferritin, thus demonstrating that there is no strong negative charge on the side of the vacuole membrane facing the cytoplasm, such as would be associated with erythro- cyte glycoproteins. The observations of Trager and Sherman could also be interpreted as evidence that the highly charged and immunogenic oligosaccha- rides of the erythrocyte surface glycoproteins, in their inverted position on the surface of the vacuole membrane facing the parasite, were masked from the agglutinating antibodies and from the Fe membrane markers. There is little precedent for direct alteration in the asymmetry of membranes. Phosphatidylcholine can migrate slowly from one leaflet of the bilayer to the other (flip-flop) (22), but phosphatidylethanolamine and cholesterol flip-flop very slowly or not at all (23, 24). Evidence for flip-flop of proteins, including vectorial enzymes, is lacking. By the time the devel- opment and segmentation of the parasite is com- plete, however, the surface area of the vacuole membrane has approached that of the erythrocyte plasma membrane, and much of the membrane has been ingested and recycled during the phagocytic and pinocytic feeding of the parasite on erythrocyte cytoplasm (25). Such membrane formation would provide opportunities for intercalation of new, parasite-derived membrane domains into the vacuole membrane so that its asymmetry would ultimately become oriented to match that of the erythrocyte and parasite plasma membranes. Since the mature mammalian erythrocyte is incapable of protein synthesis or of de novo lipid synthesis (26), the formation of new vacuole membrane (and plasma membrane in the case of erythrocytes that are enlarged during infection) must be supported by the use of plasma lipids, catalyzed by erythrocyte 238 LIPIDS AND THE MALARIAL PARASITE membrane-associated enzyme systems, or by protein and lipid biosynthesis by the parasite, or by com- binations of these sources. Further strong support for an indispensable functional association of the parasitophorous vacuole membrane and the parasite comes from observations that the membrane is necessary for successful extracellular cultivation of the parasite (5). Trophozoite growth and merozoite formation within the vacuole proceed in a relatively normal manner in culture and upon the completion of segmentation the vacuole membrane ruptures and the merozoites are liberated. The merozoites, however, do not grow and segment to yield a second generation of parasites. LIPIDS OF ERYTHROCYTES AND PARASITES Growth of the malarial parasite in the erythrocyte, accompanied as it is in some cases by an increase in the area of the erythrocyte plasma membrane, and in all cases by an increase in the area of the para- sitophorous vacuole membrane and by proliferation of the membrane systems of the parasite, results in a marked rise in the lipid content of the infected erythrocyte, particularly in the phospholipid fraction (P. knowlesi, 27-29; P. berghei, 30, 31; P. galli- naceum, 32, 33; P. lophurae, 34). Notable increases occur in phosphatidylethanolamine and phos- phatidylcholine and in those phospholipids par- ticularly associated with the malarial parasite, i.e., phosphatidylinositol and polyglycerol phosphatides (30, 31, 34-38). The cholesterol content of the infected erythrocyte also increases but, since the parasite con- tains proportionally more phospholipid and less cho- lesterol than the host cell, and the host cell loses cholesterol (33), the cholesterol/phospholipid ratio falls (28-32, 34, 35). Other neutral lipids found only in traces in normal erythrocytes are elevated in Table 1. Proportions of the lipid classes and major lipid types of normal erythrocytes and of their malarial parasites Total lipids (%) a Lipids Duck Plasmodium Monkey Plasmodium Rat Plasmodium erythrocyte lophurae erythrocyte knowlesi erythrocyte berghei(34) (34) b (14, 35) (14, 35) c (31) d (31) e Neutral lipids 21 12 33 25 24 24 cholesterol 20 8 27 10 24 f 24 diacylglycerols trace 3 - 2 unesterified fatty acids trace 1 1 6 - trace triacylglycerols trace < 1 3 4 - trace cholesterol esters trace trace 2 3 - trace Glycosphingolipids 5 2 7 10 - - Phospholipids 74 86 60 65 76 9 76 Y less polar lipidsh 1 4 - - 3 4 phosphatidylethanolamine 20 36 19 27 10 18 phosphatidylserine < 1 < 1 7 1 7 9 phosphatidylinositol 1 4 2 5 1 3 phosphatidylcholine 40 40 22 29 35 29 sphingomyelin 11 2 9 2 13 7 lysolipids 1 < 1 2 1 7 6 a- = not observed or not reported. b Freed by immune haemolysis; parasitaemia oo 90%, schizonts c- 80%, reticulocytosis c 7%. c Freed by saponin lysis; parasitaemia c. 40 %, schizonts c 70 %. d Phenylhydrazine-treated; reticulocytosis ca 45 %. e Infected erythrocyte, parasitaemia c 60 % (asynchronous), reticulocytosis cO 45 %. f Cholesterol majority of neutral lipids (31). Properly washed normal erythrocytes contain little neutral lipid other than cholesterol (42). g Contained glycosphingolipids (?). h Phosphatidic acid and polyglycerol phosphatides. 239 G. G. HOLZ, JR Table 2. Fatty acids of the total lipids of normal erythrocytes and of their malarial parasites Total fatty acids (%) Fatty acid Duck Plasmodium MonkeY Plasmodium Rat Plasmodium crytesoa lophurae a erythro- knowlesi (46) erythro- bergheicytes lophurae a cytes (46) cytes (48) b (44, 45) c 14:0 1 < 1 1 < 1 - 2 16:0 24 26 22 34 24 42 16:1 1 2 - - <1 4 18:0 10 16 15 9 17 15 18:1 18 33 18 36 8 21 18:2 21 12 15 15 11 7 18:3 1 1 1 1 - 2 20:2 1 < 1 < 1 1 - - 20:3 1 1 2 < 1 - - 20:4 10 3 17 2 31 5 20:5 1 1 2 < 1 - - 22:5 2 1 2 < 1 2 - 22:6 7 3 2 <1 - - a Holz et al., unpublished observations, 1976. b Normal, mature erythrocytes. c Infected erythrocytes; parasitaemia cW60 % (asynchronous), infected erythrocytes and are probably parasite associated, i.e., 1,2-diacylglycerols, unesterified fatty acids, and triacylglycerols (14, 29, 31, 34, 39, 40). Several other lipid types are relatively unchanged in amount or are decreased in infected erythrocytes, notably sphingomyelin, the neutral glycosyl cera- mides, the alkoxy forms of phosphatidylethanola- mine, and phosphatidylserine (when present in the erythrocyte) (30, 31, 34-38, 41). These phospho- lipids are believed to be associated principally with the plasma membrane of the erythrocyte. The lipid composition of the infected erythrocyte, therefore, is a reflection of the separate contributions of the erythrocyte and of the intracellular parasite (Table 1), with contributions also from the lipids of the plasma, which themselves show alterations during malaria infection. Accompanying the qualitative and quantitative changes in the complex lipids of the intact host cell- parasite system are changes in the amounts and proportions of the fatty acids. The most notable are a striking increase in octadecenoic fatty acids in the infected erythrocyte and decreases in polyun- saturated fatty acids, in particular a reduction in reticulocytosis c%a45 %. arachidonic acid (15, 29, 34, 39, 40, 43-47). These changes are also attributable to differences between the fatty acids of erythrocyte and parasite, and to changes in plasma fatty acids (Table 2). Analyses of infected erythrocytes are complicated by the fact that parasitaemia, reticulocytosis, and the degree of synchrony of infection vary with Plasmodium species and host species. Consequently, such analyses are rarely performed on homogeneous populations of parasitized cells. It should also be noted that all the lipid analyses of malarial parasites that have been performed to date have been com- promised by the association of the " free parasites " with some erythrocyte membrane and/or erythro- cyte-derived membrane. Immune haemolysis and saponin lysis both leave the parasite enclosed in the parasitophorous vacuole membrane, and saponin lysis also leaves the parasite associated with frag- ments of the erythrocyte plasma membrane (5, 35, 49). LIPIDS OF PLASMA (SERUM) Changes in the lipid composition of the plasma environment of the erythrocytes accompany malaria 240 LIPIDS AND THE MALARIAL PARASITE Table 3. Weight (mg/i 00 ml) and percentages (in parentheses) of the lipid classes and major lipid types of the plasma (serum) of normal and malarious animals Plasmodium lophurae (34) Plasmodium knowlesi (50) Plasmodium berghei (51) Lipid Normal Infected a Normal Infected b Normal Infected c Total lipids 533 688 451 780 380 350 Neutral lipids 292 (55) 433 (63) 281 (63) 530 (68) 215 (57) 243 (69) monoacylglycerols - - - - 11 (3) 22 (6) cholesterol 64 (12) 69 (10) 35 (8) 92 (12) 42 (12) 51 (15) diacylglycerols 1 (tr) 1 (tr) - - 7 (2) 1 8 (5) unesterified fatty acids 5 (1) 21 (3) 12 (3) 40 (5) 19 (5) 7 (2) triacylglycerols 117 (22) 200 (29) 50 (11) 267 (34) 45 (12) 99 (28) cholesterol esters 107 (20) 142 (21) 184 (41) 131 (17) 91 (24) 46 (13) Glycosphingolipids 1 1 (2) 14 (2) - - - - Phospholipids 229 (43) 241 (35) 170 (37) 250 (32) 164 (43) 107 (31) less polar lipids 5 (1) 2 (1) 7 (1) 6 (1) - - phosphatidylethanolamine 27 (5) 20 (3) 5 (1) 17 (2) 7 (2) 5 (1) phosphatidylserine 2 (< 1) 2 (< 1) - - - - phosphatidylinositol 5 (1 ) 6 (1 ) - - 7 (2) 7 (2) phosphatidylcholine 176 (33) 195 (29) 127 (28) 183 (23) 98 (26) 75 (21) sphingomyelin 1 1 (2) 14 (2) 19 (4) 29 (4) 16 (4) 9 (3) lysolipids 3 (< 1) 2 (< 1) 12 (3) 15 (2) 37 (9) 12 (4) a Parasitaemia C%a 90 %. b Parasitaemia Ca 60 %. c Parasitaemia c) 55 %. infections (Table 3). These also vary with the species of Plasmodium and with the host. Major lipids of the normal plasma (serum) of birds, monkeys, and rats are cholesterol and cholesterol esters, triacylglycerols, and phosphatidylcholine (33, 34, 50-52). Plasma (serum) triacylglycerols are elevated in the duck (P. lophurae) (34), the rhesus monkey (P. knowlesi) (46, 50), and the rat (P. ber- ghei) (51). Unesterified fatty acids also rise in the duck (34) and in the rhesus monkey (P. knowlesi) (57), but are reported to fall in the rat (51) and in the gibbon (53). Total plasma cholesterol and phospholipids are increased in the duck (34), the turkey (P. gallinaceum) (33), and the rhesus monkey infected with P. knowlesi (50). Total cholesterol, however, is decreased in rhesus monkeys infected with P. coatneyi (54), and in gibbons (55). An increase in the cholesterol/cholesterol esters ratio was seen in rhesus monkeys (P. knowlesi) (50, 52) and in rats (51). In malarious animals, alterations in the propor- tions of fatty acids in plasma lipids are similar to those observed for infected erythrocytes; however, Table 4. Unesterified fatty acids of the plasma (serum) of normal and malarious animals Total fatty acids (%) Fatt8y Duck plasma, Duck plasma, Monkey Monkeyacid normal (34) infected (34)a serum, serum, normal (57) infected (57) b 14:0 1 1 4 1 16 :0 22 23 28 35 16:1 3 3 3 5 18:0 12 11 14 11 18:1 21 30 22 31 18:2 8 7 20 12 18:3 1 > 1 - - 20:2 4 2 - - 20:3 4 3 - - 20:4 10 9 3 1 20:5 1 1 - - 22:5 3 2 - - 22:6 4 3 _ - a Plasmodium lophurae. b Plasmodium knowlesi. 241 242 G. G. HOLZ, JR the changes in absolute amounts are relatively small. The most interesting difference is an increase in octadecenoic fatty acids within the unesterified fatty acid fraction (34, 56, 57) (Table 4). OCTADECENOIC FATTY ACIDS AND CHOLESTEROL In 1948, Laser (47) reported that a haemolytic principle, which he had earlier observed in normal plasma (58), was elevated in the blood (erythrocytes and plasma) of P. knowlesi-infected rhesus monkeys. Chemical and physical studies identified the material as cis-vaccenic acid (18:1 (n-7)) accompanied by a small amount of oleic acid (18:1 (n-9)) (59, 60). Subsequent examination of the total fatty acids of the erythrocytes and of the unesterified fatty acids of the serum of P. knowlesi-infected rhesus monkeys by gas-liquid chromatography (29, 56, 57) suggested that the octadecenoic component was oleic acid but the presence of its positional isomer, cis-vaccenic acid, was not ruled out. The increases in octadecenoic fatty acids in the malarious monkeys were not limited to the infected erythrocytes and the plasma, but were seen as well in the lipids of the uninfected erythrocytes (15, 29). Recently, Laser et al. (39) demonstrated by improved gas-liquid chromato- graphic and degradative techniques that the increase in octadecenoic fatty acid in erythrocytes para- sitized by P. knowlesi was attributable to oleic acid, not to cis-vaccenic acid. Holz et al. (unpublished observations, 1976) have also studied the question of the nature and the roles of oleic and cis-vaccenic acids in malaria. Octa- decenoic fatty acids were isolated from P. lophurae- infected ducks by preparative thin-layer and gas- liquid chromatography, and their methyl esters ozonized. The resulting fragments of the esters were hydrogenated and their reduction products were then separated, characterized, and quantified by gas-liquid chromatography (61). It was found that oleic and cis-vaccenic acids coexisted in all the materials examined, with oleic as the major form. They were the only 18:1 positional isomers identified, and no trans isomers were encountered. The highest ratio of oleic to cis-vaccenic acid was in P. lophurae lipids (Table 5). Parasitization of the duck erythrocyte resulted in a tripling of the isomer ratio in the phospholipids. This change in the ratio, coupled with an absolute increase in the amount of phospholipids, resulted in an overall five-fold increase in the oleic acid content and a two-fold increase in cis-vaccenic acid. The increase in the octadecenoic fatty acid content of the infected Table 5. Proportions of octadecenoic fatty acid positional isomers (ratio of oleic acid to cis-vaccenic acid) in P. Iophurae and in the erythrocytes and plasma of normal and P. lophurae-infected ducks a Un- Neutral esterified Phospho- lipids fatty lipids acids Duck erythrocytes; normal b - _ 5 Duck erythrocytes; parasitized C 16 - 16 Plasmodium lophurae d 24 - 19 Duck plasma; normal 13 11 7 Duck plasma; infected 13 10 8 a Holz et al., unpublished observations, 1976. b Reticulocytosis c. 4 %. c Parasitaemia c'.90 %, schizonts co 80 %, reticulocytosis c'7%. d Freed by immune haemolysis. erythrocyte could not be attributed solely to the presence of the parasite lipids; it reflected changes in the erythrocyte membrane lipids as well, since alkoxy forms of phosphatidylethanolamine (alk-l-enyl-acyl and alkyl-acyl) isolated from infected erythrocytes were enriched with the octadecenoic fatty acids. Alkoxy molecular species of phosphatidylethanol- amine comprise c'.j4O% of the phosphatidylethanol- amine fraction of normal duck erythrocytes but are present in only trace amounts in P. lophurae phos- phatidylethanolamine. No change was seen in the oleic/cis-vaccenic ratio in the duck plasma total neutral lipids and phospholipids on infection, nor in the unesterified fatty acids, but absolute increases in these lipids resulted in two-fold increases in both isomers in the neutral lipids and phospholipids and 6-fold increases in both isomers in the unesterified fatty acids. Laser's emphasis on the haemolytic potential of the octadecenoic fatty acids in malaria infections (62, 63) seems to be supported by the studies on the P. lophurae-duck system. Direct tests of oleic and cis-vaccenic acids on the osmotic fragility of duck erythrocytes have shown that they are some- what more sensitive to these fatty acids than are human and horse erythrocytes, and that cis-vaccenic acid is approximately twice as active as oleic acid. In any consideration of the haemolytic effects on erythrocytes of increases in environmental and intra- membrane concentrations of octadecenoic fatty acids, however, one must recognize that haemolysis (defined as haemoglobin release) is a catastrophic endpoint, and that more subtle influences on erythro- LIPIDS AND THE MALARIAL PARASITE cyte structure and function are exerted before lysis. Such prehaemolytic changes include many that are demonstrated by parasitized and unparasitized erythrocytes in a malaria infection, for example, increased osmotic fragility (64), increased passive permeability (65), and reduced adenosinetriphos- phatase-dependent transport of sodium ions (66). The octadecenoic fatty acids may cause such changes in the barrier properties of the erythrocyte plasma membrane by their esterification to membrane phospholipids, displacing other fatty acids and altering the fluidity of the inner and/or outer leaflets of the phospholipid bilayer by changing the degree of unsaturation of their fatty acids (10). Such modifi- cations can, in turn, vary the interactions of mem- brane proteins with bilayer phospholipids, leading to conformational changes influencing protein func- tions (67). Octadecenoic fatty acids may also influence membrane fluidity by acting directly on hormonal regulatory mechanisms. It has been shown that cis- vaccenic acid and oleic acid are among the most active fatty acids in enhancing the catecholamine- activated adenylate cyclase (EC 4.6.1.1) of turkey erythrocytes (68). Adenylate cyclase-dependent pro- tein kinases are common in plasma membranes. Their phosphorylating activity can modify membrane proteins interacting with bilayer phospholipids, with consequent changes in membrane fluidity (69) and barrier properties (70). Reduction in the membrane cholesterol/phospho- lipid ratio can also be invoked to account for the modified membrane properties of parasitized ery- throcytes. Seed & Kreier (33) calculated the choleste- rol content of the turkey erythrocyte before and after infection with P. gallinaceum (corrected for the cho- lesterol of the parasite) and found a sizeable decrease. Loss of membrane cholesterol has been reported to increase osmotic fragility and passive permeability and to reduce active transport (71, 72). Cholesterol is asymmetrically distributed in the membrane lipid bilayer, being concentrated more in the outer leaflet (73) where it readily exchanges with plasma chol- esterol (26). The rise in plasma cholesterol esters in malaria, documented earlier, probably results from plasma lecithin-cholesterol acyltransferase (EC 2.3.1.26) catalysed esterification of the cholesterol lost from erythrocytes (74). LIPID METABOLISM: ERYTHROCYTE AND PARASITE Any consideration of the lipid metabolism of the malarial parasite must begin with an examination of the host erythrocyte and its plasma milieu, since it is their lipid composition and lipid metabolism that govern the availability of lipids and lipid precursors to the parasite. The erythrocytes that have received the most attention are the mature, non-nucleated rhesus erythrocyte (P. knowlesi), the immature, non-nucleated rat erythrocyte (reticu- locyte) (P. berghei), and the mature, nucleated duck erythrocyte (P. lophurae). Their lipids and those of their plasma environments are shown in Tables 1-4. Lipids of the mature mammalian erythrocyte are thought to be located exclusively in the plasma membrane (26). Those of the rodent reticulocyte are found as well in mitochondria and in sparsely distributed rough endoplasmic reticulum, if present (75); consequently, the lipid content is considerably higher than in mature erythrocytes (cholesterol c.2.5-fold, phospholipids c-i4-fold) (76). In mature avian erythrocytes it is also higher than in mature mammalian erythrocytes, since the nuclear envelope is a double-membraned structure with a total surface area roughly equivalent to that of the plasma membrane (77, 78). The lipid metabolism of each of the erythrocyte types also has special characteristics. The rhesus erythrocyte, as all mature mammalian erythrocytes, is incapable of de novo fatty acid (14, 26), cholesterol (12), and phospholipid (79) biosynthesis. Such erythrocytes exchange lipids with the plasma in reactions shown diagrammatically in Fig. 1. In this manner, plasma lipids become accessible to the malarial parasite. In contrast, rat reticulocytes probably retain some capacity for lipid synthesis as well as for lipid exchange. Finally, mature, P L A S M A |E R Y T H R C Y T E CHOLESTEROL < , CHOLESTEROL PHOSPHATIDYLCHOLINE HOSPHATIDYLCHOLINE CoA LYSOPHOSPHATIDYLCHOLINE * LYSOPHOSPHATIDYLCHOLINE CHOLESTEROL ESTERS I GLYCEROPHOSPHORYL- GLYCEROPHOSPHORYL- CHOLINE CHOLINE + + FATTY ACIDS 4-4- -- -* FATTY ACIDS SPHINGOMYELIN 4 . SPHINGOMYELIN Fig. 1. Exchanges of lipids between plasma and mam- malian erythrocyte (after van Deenen & de Gier, 26). 243 G. G. HOLZ, JR nucleated avian erythrocytes can apparently carry out de novo phospholipid synthesis (80) but not de novo fatty acid synthesis (81). Their phospholipid exchange reactions with the plasma are not identical to those of mammals; for example, turnover of erythrocyte plasma membrane lysolecithin is low (80). Also, their nuclear membrane-associated acyl- CoA synthetase (EC 6.2.1.3) assists in providing a much more active incorporation of fatty acids into membrane phospholipids than does the comparable plasma membrane enzyme system (81). The lipid composition of the parasite is regulated by the lipids and lipid precursors of the intra- erythrocytic environment and by the parasite's metabolic capacities. There are numerous qualitative and quantitative differences between the lipids of malarial parasites and those of mature and immature mammalian and avian erythrocytes. Among the three species of Plasmodium cited in Tables 1 and 2, unesterified fatty acids, triacylglycerols, polyglycerol phosphatides, 1,2-diacylglycerols, diacyl phospha- tidylethanolamine, and phosphatidylinositol make up a larger proportion and cholesterol, phosphatidyl- serine, and sphingomyelin a smaller proportion of the total lipids than is the case in the various host erythrocytes. Also, the fatty acids acylating these lipids are less polyunsaturated in the parasites, and oleic acid is a particularly notable fatty acid. Attempts have been made to describe the mechan- isms by which parasite lipids are formed, and also to define the contributions of the erythrocyte and the plasma as sources of lipids and lipid precursors. The most complete studies have been those of Rock (14, 15, 79) and Trigg (12) on the P. knoWlesi-rhesus erythrocyte system (Fig. 2). Parasites in erythrocytes, and parasites (60-80% schizonts) freed by saponin lysis from a mixture of purified, infected (30-50%) and uninfected erythrocytes, but still enclosed in CHOLINE * PHOSPHATIDYLCHOLINE PHDSPHATIDYLETHANOLAMINJE -. A NDLANIRE P 0 PHOSPHATIDYLINOSITOL - ACEITATE 80o POLYGLYCEROLPHOSPHATIDE |o: GLUCOSE / PHIOSPHSATIDIC ACID PALMITATE CLYCERYL) (ACYL) STEARATE \ 1.2-DIACYLCLYCEROL OLEATE 2 UNESTERIFIED FATTY ACID 20S TRIACYLGLYCEROL CHOLESTEROL - CHOLESTEROL Fig. 2. Lipid biosynthesis in P. knowlesi (12, 14, 15, 79). the parasitophorous vacuole membrane and con- taminated (c 15%) by fragments of erythrocyte membrane, incorporated 33P-orthophosphate into phospholipids (79). The free parasite was less active in this respect than the intracellular form. In short- term incubations, as much as 95% of the isotopic phosphorus taken up was found in membrane phosphatidylcholine, phosphatidylethanolamine and phosphatidylinositol, and the remainder in poly- glycerol phosphatide and phosphatidic acid. The highest specific activities occurred in the latter three phospholipids. In contrast, 33P-orthophosphate in- corporation into erythrocyte membrane phospho- lipids was 80- to 100-fold less than that into the parasite, and only one phospholipid (phosphatidic acid) was labelled. The rates of incorporation and the patterns of labelling in these experiments indicate a rapid turnover of phosphorus in the phospholipids of the parasite. They may be interpreted as support for de novo phospholipid synthesis. The labelling of only phosphatidic acid in the erythrocyte is seen in all such experiments with mature mammalian erythrocytes and is generally not considered evidence for de novo synthesis (26). Substrate and inhibitor experiments on normal and infected erythrocytes and on free parasites also indicated that 33P-ortho- phosphate incorporation was dependent on glycolysis. Exposure of the same preparations of normal and infected rhesus erythrocytes, and free P. knowlesi, to 14C-labelled choline, ethanolamine, glucose, glycerol, and acetate resulted in incorporation of radioactivity into parasite lipids (the intracellular parasite was 30-500% more active than the free parasite), but not into erythrocyte lipids (14). 14C-choline labelled only phosphatidylcholine and 14C-ethanolamine only phosphatidylethanolamine in both parasite prepara- tions. More than 80% of the carbon-14 of glucose and glycerol incorporated was found in the phos- phatidylcholine, phosphatidylinositol, phosphati- dylethanolamine (sum 97%), and phosphatidic acid and polyglycerol phosphatides (sum 3%) of intracellular and free parasites. The remainder was in the diacylglycerols, triacylglycerols, and un- esterified fatty acids. Analysis of the distribution of carbon-14 in the glyceryl and acyl fractions of the phospholipids showed that in intracellular and free parasites, over 95% of the carbon-14 of glucose and glycerol incorporated was located in the glyceryl moiety (except for 15% of the carbon-14 of glucose in the acyl groups of free parasite phospholipids). In contrast, 14C-acetate labelled only the phospho- lipids of the intracellular parasites (only phos- 244 LIPIDS AND THE MALARIAL PARASITE phatidylethanolamine in the free parasites) and 85% of the carbon-14 of acetate was found in the acyl groups. Incorporation was relatively poor and was enhanced by ATP and CoA. These experiments demonstrated that P. knowlesi was able to: (a) introduce nitrogenous bases into phosphoglycerides; (b) metabolize glucose and glycerol to form the glyceryl backbone of glycerides and phosphoglycerides; and (c) utilize the carbon of glucose and acetate in the formation of the acyl groups of phosphoglycerides. De novo fatty acid synthesis was not suggested; rather, the exogenous acetate, or acetate generated endogenously from glucose, seemed to be used to a limited extent to elongate extant fatty acyl chains. When 14C-labelled palmitic, stearic, and oleic acids (bound to bovine serum albumin) were incu- bated with normal and infected rhesus erythrocytes and with free P. knowlesi, the fatty acids were incorporated into the lipids (80-90% into the phos- pholipids) of the erythrocytes, the intracellular parasites, and the free parasites (15). The free para- sites were 10-15 times more active in this respect than the erythrocytes. Fatty acid degradation and recycling of carbon-14 were ruled out by the finding that all of the radioactivity incorporated by erythro- cytes and parasites was present in the fatty acyl chains of the complex lipids, and not in their glyceryl backbones. The possibility of chain elongation of the labelled fatty acids was recognized but not tested. Of the 80% of fatty acid carbon-14 incorporated into the phospholipids, 98 % was distributed among phosphatidylcholine, phosphatidylethanolamine, and phosphatidylinositol (PC-PE>PI). The remaining 20% was found among diacylglycerols (not found in erythrocyte lipids), triacylglycerols, and fatty acids. The rapid and relatively extensive incorporation of fatty acids by P. knowlesi (free parasites incorporated 40% of available 14C-labelled palmitic acid into cell lipids in 2 h) and the lack of evidence for de novo fatty acid synthesis suggests that the parasite is normally dependent on and makes extensive use of host fatty acids. This interpretation is supported by the apparent requirement of P. knowlesi for a fatty acid, satisfied by stearic acid, for in vitro cultivation (82, 83). The modest use of extracellular fatty acids by the rhesus erythrocytes (incorporation of 2-5 % of available 14C-palmitic acid in 2 h) is in harmony with the experience of others with mature mammalian erythrocytes. Plasma fatty acids are known to be activated to acyl-CoA thioesters and then trans- esterified to lysophospholipids formed by phospho- lipase action in the erythrocyte membrane or present as a result of exchanges with plasma lysophospho- lipids (Fig. 1). In experiments similar to those of Rock, Trigg (12) examined the potential sources of the cholesterol of P. knowlesi. Normal and infected rhesus erythro- cytes were exposed to 14C-labelled acetate, meva- lonate, and cholesterol in an in vitro culture system. Infected erythrocytes were then separated from other formed elements and lysed by immune serum. The cholesterol fractions of the freed parasites (enclosed in the parasitophorous vacuole membrane but separated from any erythrocyte plasma mem- brane), and of the control normal erythrocytes, were labelled only by the 14C-cholesterol. It would appear, therefore, that as in the case of fatty acids, P. know- lesi is incapable of de novo sterol biosynthesis and parasitizes the host cell cholesterol supply. Again, supporting this finding is the observation that cholesterol is required for in vitro cultivation of P. knowlesi (83). Mature mammalian erythrocytes also fail to synthesize cholesterol and exchange it with plasma cholesterol (Fig. 1). Other Plasmodium-erythrocyte systems that have been examined include P. fallax-turkey (11), P. lophurae-duck (84, 85), P. berghei-rat (13), and P. berghei-mouse (86). The lipid metabolism of these dual cell systems is more difficult to study because of the presence of nuclear membrane- associated lipid metabolism in mature avian erythro- cytes (81) and cytoplasmic membrane-associated lipid metabolism in rodent reticulocytes. There has been only one attempt, to date, to compare isolated nuclei of avian erythrocytes with intact malarial parasites for their ability to incorporate a lipid precursor. Trager (16) showed that duck erythrocyte nuclei were unable to carry out a CoA-dependent incorporation of 14C-choline into lipids that was readily observed with extracellular P. lophurae. In some cases, investigators have incubated erythrocytes in vitro with isotopically-labelled fatty acids in the presence of leucocytes and platelets, without appropriate consideration of the active lipid metabolism of the extraneous cells. Reports of the incorporation of 14C-acetate into the lipids (mainly the phospholipids) of normal and infected avian erythrocytes and their parasites (11, 84, 85) are probably attributable to this oversight. In other cases in which leucocytes and platelets were separated from erythrocytes after incubation with isotopically- labelled fatty acids, and before extraction of erythro- 245 G. G. HOLZ, JR cyte lipids (84, 85), there was opportunity for exchange of labelled lipids of the leucocytes and platelets with unlabelled lipids of the erythrocytes during the long incubations used (3-24 h). Also, in vivo exposures of plasmodia have been attempted in which isotopically-labelled lipid pre- cursors were infused into malarious hosts (11, 13). In such experiments, host tissue metabolism of the infused materials must be taken into account and one is uncertain of the precise nature of the labelled materials that ultimately reached the intracellular parasites. Finally, malarial parasites themselves may differ in the ways in which they generate lipid precursors and in the ways in which they activate exogenous lipid precursors. Avian plasmodia have typical protozoan mitochondria with tubular christae. Mammalian plasmodia contain instead various membranous structures (with associated cytochrome oxidase) that may be considered " acristate " mito- chondria (3, 25). Also, avian plasmodia have a complete Krebs cycle while mammalian plasmodia do not (87). Given the equivocal nature of the results of experiments on the incorporation of lipid precursors into the lipids of avian and rodent malarial parasites, it is surprising that these studies have generally supported those done with P. knowlesi. For example, P. fallax-infected turkey erythrocytes introduced 14C-labelled palmitic, stearic, and oleic acids and P. berghei-infected mouse erythrocytes introduced carbon-14 from glucose into phospholipids in vitro (11, 86). Also, continuous perfusion of 3H-glucose and 14C-oleic acid into rats infected with P. berghei resulted in labelling of parasite phospholipids with both 3H and 14C; phospholipid fatty acids contained 14C but not 3H. Tritium appeared to be restricted to the glyceryl moiety of the phospholipids (13). PARASITE LIPID SOURCES Our knowledge of the lipids and lipid metabolism of the plasma, the parasitized erythrocyte, and the parasite provide a basis for speculation on the sources of parasite lipids in vivo. In the case of the mature mammalian erythrocyte, the plasma membrane exchanges phospholipids, lysophospholipids, and cholesterol with the plasma, and phospholipids are also formed by the activation of plasma unesterified fatty acids to acyl-CoA thioesters and their trans- esterification to membrane lysophospholipids. Ex- changes of these phospholipids and lysophospho- lipids among the erythrocyte plasma membrane, the parasitophorous vacuole membrane, and the parasite plasma membrane might then be mediated by phospholipid exchange proteins (88). Net transfer of fatty acids to the parasite would also have to occur since the parasite has a fatty acid requirement for in vitro culture (82) and is incapable of de novo fatty acid synthesis (14). Of note in this connexion is the fact that free fatty acids can be activated and incorporated into phospholipids by parasites (in parasitophorous vacuoles) freed from infected erythrocytes (15). Other phospholipid precursors (for example, nitrogenous bases, inositol, glycerol, phosphorus) may also be transported to the parasite from the plasma and from the erythrocyte, since free parasites incorporate exogenous choline, ethanola- mine, glycerol, and phosphorus (14, 16, 79). The activation and incorporation reactions require CoA and ATP (16, 79), also obtained by the parasite from the host cell (89, 90, 91). Exchanges of cholesterol between the erythrocyte and the plasma, and among the erythrocyte, para- sitophorous vacuole, and parasite membranes must also occur. Net transfer is inferred from the sterol requirement for culture in vitro (83) and from the absence of sterol biosynthesis (12). Finally, parasites feed by pinocytosis and phago- cytosis on erythrocyte cytoplasm and, in so doing, ingest large amounts of their own plasma membrane and the membrane of the parasitophorous vacuole. Presumably, the lipids of the endocytosed cytoplasm and membranes then become available for further parasite lipid biosynthesis (40). UMft LIPIDES ET PARASITE DU PALUDISME L'endocytose des merozoTtes marque le debut du developpement de Plasmodium dans une vacuole limitee par une membrane d'origine erythrocytaire dans laquelle la distribution asym6trique des lipides et proteines presente une orientation inversee par rapport A la membrane plasmique du parasite. La croissance de la membrane, liee A celle du parasite et A son alimentation par phagocytose et par pinocytose, peut s'accompagner 246 LIPIDS AND THE MALARIAL PARASITE 247 d'une reorientation de cette distribution. L'extension, de la surface de la membrane du parasite et dans certains cas de 1'erythrocyte, va de pair avec la croissance et la segmentation du parasite. Le developpement de l'ensemble membraneux de l'erythrocyte infecte provoque une augmentation rapide de la teneur en lipides, mais la composition des lipides du parasite differe de celle de 1'erythrocyte A maints egards: teneur plus elevee en diacylphosphatidylethanolamine, phosphatidylinositol, polyglycerol phosphatides, diacylglycerols, acides gras non esterifies, triacylglycerols, et acides gras hexadeca- nolque et octaddc6nolque; une teneur plus faible en sphingomyeline, phosphatidylserine, alcoxy phosphatidy- lethanolamine, cholesterol et acides gras polyinsatures. La croissance de la membrane liMe aux processus d'ali- mentation, de croissance et de reproduction, s'accom- pagne d'un metabolisme lipidique actif. Plasmodium est incapable d'operer A nouveau la biosynthese des acides gras et du cholesterol; cependant, il peut fabriquer ses propres glycerides et phosphoglycerides avec les acides gras, les bases azotees, les alcools, I'ATP et le CoA fournis par l'h6te, et peut produire la fraction glyceryle pendant la glycolyse. Le cholesterol provient de l'h6te, mais on ignore tout des origines des sphingo- lipides. Le metabolisme des lipides du parasite peut etre en rapport avec des modifications de la teneur de la membrane plasmique erythrocytaire en acides gras octade&cnolque et en cholesterol, responsables A leur tour des variations de permeabilite et de fragilite. REFERENCES 1. TRAGER, W. Transactions of the Royal Society of Tropical Medicine and Hygiene, 50: 419-420 (1956). 2. LADDA, R. L. ET AL. Journal of parasitology, 55: 633-644 (1969). 3. RUDZINSKA, M. International review of cytology, 25: 161-199 (1969). 4. TRAGER, W. ET AL. Proceedings of the Helmintho- logical Society of Washington, 39: 220-230 (1972). 5. LANGRETH, S. G. & TRAGER, W. Journal of proto- zoology, 20: 606-613 (1973). 6. DVORAK, J. ET AL. Science, 187: 748-750 (1975). 7. BANNISTER, L. H. ET AL. Parasitology, 1: 483-491 (1975). 8. GINN, F. L. ET AL. Science, 164: 843-845 (1969). 9. BRETSCHER, M. S. & RAFF, M. C. Nature, 258: 43- 49 (1975). 10. EMMELOT, P. & VAN HOEVEN, R. P. Chemistry and physics of lipids, 14: 236-246 (1975). 11. GUTIERREZ, J. American journal of tropical medicine and hygiene, 15: 818-822 (1966). 12. TRIGG, P. I. Annals of tropical medicine and para- sitology, 62: 481-487 (1968). 13. CENEDELLA, R. J. ET AL. Military medicine, 134: 1045-1055 (1969). 14. ROCK, R. C. Comparative biochemistry and phy- siology, 40B: 657-669 (1971). 15. ROCK, R. C. Comparative biochemistry and phy- siology, 40B: 893-906 (1971). 16. TRAGER, W. Journal of protozoology, 18: 392-399 (1971). 17. TRAGER, W. & JENSEN, J. B. Science, 193: 673-675 (1976). 18. HAYNES, J. D. ET AL. Nature, 263: 767-769 (1976). 19. STERLING, C. R. ET AL. Proceedings of the Hel- minthological Society of Washington, 39: 109-129 (1972). 20. LANGRETH, S. G. In: de Puytorac, P. & Grain, J., ed. Progress in protozoology. Proceedings of the Fourth International Congress on Protozoology, September 1973, Clermont-Ferrand, Universite de Clermont, 1973, p. 242. 21. TRAGER, W. Science, 183: 269-273 (1974). 22. RENOOLJ, W. ET AL. European journal of biochemistry, 61: 53-58 (1976). 23. ROSEMAN, M. ET AL. Biochemistry, 14: 4826-4830 (1975). 24. POZANSKY, M. & LANGE, Y. Nature, 259: 420-421 (1976). 25. AIKAWA, M. & STERLING, C. Intracellular parasitic protozoa. New York, Academic Press, 1974. 26. VAN DEENEN, L. L. M. & DE GIER, J. In: Surgenor, D. M. ed. The red cell. New York, Academic Press, 1974, vol. 1, pp. 147-211. 27. MORRISON, D. B. & JESKEY, H. A. Journal of the National Malaria Society, 7: 259-264 (1948). 28. BALL, E. G. ET AL. Journal of biological chemistry, 175: 547-571 (1948). 29. ANGUS, M. G. N. ET AL. Annals of tropical medicine andparasitology, 65: 429-439 (1971). 30. LAWRENCE, C. W. & CENEDELLA, R. J. Experimental parasitology, 26: 181-186 (1969). 31. RAO, K. N. ET AL. Experimental parasitology, 27: 22-27 (1970). 32. LEWERT, R. M. Journal of infectious diseases, 91: 125-144 (1952). 33. SEED, T. M. & KREIER, J. P. Proceedings of the Hel- minthological Society of Washington, 39: 387-411 (1972). 34. BEACH, D. H. ET AL. Journal ofparasitology, 63: 62- 75 (1977). 35. ROCK, R. C. ET AL. Comparative biochemistry and physiology, 38B: 425-437 (1971). 36. DE ZEEUW, R. A. ET AL. Proceedings of the Hel- minthological Society of Washington, 39: 412-418 (1972). 248 G. G. HOLZ, JR 37. MCCLEAN, S. ET AL. Analytica chimica acta, 82: 175-185 (1976). 38. BECKWITH, R. ET AL. Experimental parasitology, 37: 164-172 (1975). 39. LASER, H. ET AL. Parasitology, 71: 167-181 (1975). 40. CENEDELLA, R. J. ET AL. Experimental parasitology, 24: 130-136 (1969). 41. MCLAUGHLIN, J. & MCGHEE, B. Life sciences, 11: 397404 (1972). 42. NELSON, G. J. Journal of lipid research, 8: 374-379. 43. MORRISON, D. B. & JESKEY, H. A. Federation proceedings, 6: 279 (1947). 44. WALLACE, W. R. ET AL. American journal of tropical medicine and hygiene, 14: 715-718 (1965). 45. WALLACE, W. R. American journal of tropical medicine and hygiene, 15: 811-813 (1966). 46. GINGER, C. D. Transactions of the Royal Society of Tropical Medicine and Hygiene, 61: 2-3 (1967). 47. LASER, H. Nature, 161: 560 (1948). 48. WALKER, B. L. & YURKOWSKI, M. Biochemical journal, 103: 218-224 (1967). 49. AIKAWA, M. & COOK, R. T. Experimental parasito- logy, 31: 67-74 (1972). 50. ANGUS, M. G. N. ET AL. Annals of tropical medicine and parasitology, 65: 135-154 (1971). 51. RAo, K. N. ET AL. Indian journal ofmedical research, 57: 921-924 (1969). 52. MCCLEAN, S. ET AL. Analytica chimica acta, 86: 255-261 (1976). 53. LANGER, B. W., JR ET AL. Annals of tropical medicine and parasitology, 61: 384-385 (1965). 54. DESOWITZ, R. S. ET AL. Annals of tropical medicine and parasitology, 61: 365-374 (1965). 55. MILLER, L. H. ET AL. Annals of tropical medicine and parasitology, 61: 375-383 (1967). 56. ANGUS, M. G. N. ET AL. Transactions of the Royal Society of Tropical Medicine and Hygiene, 61: 4 (1967). 57. ANGUS, M. G. N. ET AL. Annals of tropical medicine andparasitology, 65: 155-167 (1971). 58. LASER, H. & FRIEDMANN, E. Nature, 156: 507 (1945). 59. MORTON, I. D. & TODD, A. R. Biochemical journal, 47: 327-330 (1950). 60. GODDARD, E. D. & ALEXANDER, A. E. Biochemical journal, 47: 331-335 (1950). 61. GELLERMAN, J. L. & SCHLENK, H. Journal ofproto- zoology, 12: 178-189 (1965). 62. LASER, H. Journal ofphysiology, 110: 338-355 (1950). 63. LASER, H. ET AL. Parasitology, 71: 167-181 (1975). 64. FOGEL, B. J. ET AL. American journal of tropical medicine and hygiene, 15: 269-275 (1966). 65. NEAME, K. D. & HOMEWOOD, C. A. International journal ofparasitology, 5: 537-540 (1975). 66. DUNN, M. J. Journal of clinical investigation, 48: 674-684 (1969). 67. SOLOMONSON, L. P. ET AL. Biochemistry, 15: 892- 897 (1976). 68. ORLY, J. & SCHRAMM, M. Proceedings ofthe National Academy of Sciences of the USA, 72: 3433-3437 (1975). 69. KURY, P. G. & MCCONNELL, H. M. Biochemistry, 14: 2798-2803 (1975). 70. RUDOLPH, S. A. & GREENGARD, P. Journal of biological chemistry, 249: 5684-5687 (1974). 71. BRUCKDORFER, K. R. ET AL. Biochimica et biophysica acta, 183: 334-345 (1969). 72. KIMELBERG, H. K. & PAPAHADJOPOULOS, D. Journal of biological chemistry, 249: 1071-1080 (1974). 73. FISHER, K. A. Proceedings of the National Academy of Sciences of the USA, 73: 173-177 (1976). 74. GLOMSET, J. A. Journal of lipid research, 9: 155-167 (1968). 75. HALLINAN, T. ET AL. Blood, 20: 547-556 (1962). 76. SHATTIL, S. J. & COOPER, R. A. Journal of laboratory and clinical medicine, 79: 215-227 (1972). 77. ZENTGRAF, H. ET AL. Journal of biological chemistry, 246: 2986-2995 (1971). 78. KLEINIG, H. ET AL. Journal of biological chemistry, 246: 2996-3000 (1971). 79. ROCK, R. C. ET AL. Comparative biochemistry and physiology, 40B: 543-561 (1971). 80. SOULA, G. ET AL. European journal of biochemistry, 30: 93-99 (1972). 81. SADLER, J. & FRANKE, W. W. Biochimica et bio- physica acta, 311: 205-213 (1973). 82. SIDDIQUI, W. A. ET AL. Science, 156: 1623-1625 (1967). 83. TRIGG, P. I. Transactions of the Royal Society of Tropical Medicine and Hygiene, 62: 3 (1968). 84. BRUNDAGE, W. G. ET AL. American journal of tropical medicine and hygiene, 18: 657-661 (1969). 85. HARDY, C. L. S. ET AL. Experimental parasitology, 37: 193-204 (1975). 86. CENEDELLA, R. J. American journal of tropical medicine and hygiene, 17: 680-684 (1968). 87. OELSHLEGEL, F. J., JR & BREWER, G. J. In: Surgenor, D. M., ed. The red cell. New York, Academic Press, 1975, vol. 2, pp. 1263-1302. 88. DAWSON, R. M. C. Sub-cellular biochemistry, 2: 69-89 (1973). 89. TRAGER, W. & BROHN, F. H. Proceedings of the National Academy of Sciences of the USA, 72: 1834- 1837 (1975). 90. BROHN, F. H. & TRAGER, W. Proceedings of the National Academy of Sciences of the USA, 72: 2456- 2458 (1975). 91. TRAGER, W. Journal of protozoology, 14: 110-114 (1967).