Summary of discussions on the biology of the malaria parasite.

Bulletin of the World Health Organization, 55 (2-3): 347-360 (1977) Summary of discussions on the biology of the malaria parasite PARASITE INVASION Erythrocytic stages Invasion of erythrocytes by malarial parasites is a rapid process that is completed in about 30 seconds. Although it is difficult with existing methods to distinguish the different stages of this continuous process, there is evidence for three distinct phases. These are (a) the initial attachment of the merozoite to the erythrocyte, (b) the invagination of the erythrocyte membrane and the formation of the parasitophorous vacuole, and (c) the resealing of the vacuole and the erythrocyte membrane. There is still a great deal to be learnt about the surface structure of merozoites and their interactions with erythrocytes before the invasion process can be fully understood and the potential for interference by immunological or chemotherapeutic means fully ex- ploited. Surface of the merozoite Electron micrographs indicate that the plasma membrane of the merozoite is covered by a " surface coat ". Some workers have suggested that this struc- ture is the result of interaction between serum components, e.g., immunoglobulin or albumin, and the merozoite plasma membrane and that the coat is present in the intraerythrocytic merozoite only if the erythrocyte itself is damaged. Amorphous aggrega- tions of material can occur on the merozoite surface but these should not be confused with the regularly arranged peg-like structures that have been described as being attached to the merozoite plasma mem- brane. These structures form at an early stage in merogamy and this has been taken to indicate that they do not originate from material discharged from the rhopteries or micronemes. The surface coat gives cytochemical reactions typi- cal of acidic glycoproteins and the electron-dense peg-like structures are lost after enzyme treatment with preparations containing proteases. It has been speculated that the carbohydrate residues are not exposed to the exterior since merozoites are not ag- glutinated by a number of lectins. However, charac- terization of structural details at this level probably must await the isolation and purification of the surface coat material. The shedding of the surface coat during invasion (see also page 350) might provide a method for its isolation, albeit possibly in a modified form. Immunology of the merozoite surface coat Attention has been drawn to the difficulties raised by the hypothesis that the merozoite surface contains not only determinants that express antigenic varia- tion, but also ligands that interact with receptors on erythrocytes of susceptible host species. It is evident from the chronicity of erythrocytic malaria in natu- ral hosts, and also from the strain specificity of protective immunity, that infected hosts are unable to recognize as antigenic a ligand for erythrocytes that is common to merozoites with different anti- genic specificities. The evidence for the hypothesis that antigenic variation is expressed by surface coat constituents is that variant-specific antibody inhibits invasion of erythrocytes. Analysis of merozoite in- hibitory antibodies in different hosts has shown that Macaca mulatta produces a monovariant response 3-4 weeks after infection, whereas M. fascicularis produces antibody that reacts with several variants. This response can be detected 7-14 days after infec- tion with a population of parasites assumed to be of a single variant. Before one can conclude that these are true cross-reacting antibodies however, the homo- geneity of the variant population in the infective inoculum has to be established. For example, dif- ferences in T cell helper activity in M. mulatta and M. fascicularis might account for a difference in the ability to produce antibody to minor variants in a mixed population of parasites. Unfortunately, the extent of heterogeneity present in different popula- tions of P. knowlesi variants is unknown, as is any difference in the T cell responsiveness of the two host species cited above. Needless to say, there is also no exact information concerning the distribution of merozoite surface antigens and other ligands. - 3473615 SUMMARY OF DISCUSSIONS Motility of the merozoite Merozoites of several genera in the sub-phylum Apicomplexa are known to be motile and Plasmo- dium spp. have a similar arrangement of subpellicu- lar microtubules. Movements of the conoid end of merozoites attached to erythrocytes and distortion of the merozoite at the point of entry into the parasitophorous vacuole have been reported. There is however, no detailed account of these manoeuvres by the parasite or of their significance in invasion. Attachment of merozoites to erythrocytes Merozoites can attach to erythrocytes initially by any part of their cell surface, but invasion occurs only if the merozoite is oriented so that the apical region of the merozoite is in contact with the erythrocyte surface. Detachment of the erythrocyte can occur at this stage if, for example, the erythro- cyte is not susceptible to invasion, or if the cell is haemolysed and only the erythrocyte membrane is present, i.e., it is a ghost. Invasion of other cells can take place subsequent to detachment, indicating that attachment itself does not alter the surface of the merozoite and prevent further attachment and inva- sion. Merozoites of P. knowlesi attach to erythro- cytes of species that are susceptible to infection suggesting that specific forces are more important than non-specific differences, for example, differ- ences in electrostatic forces between the merozoite and the erythrocyte surface. Indeed, electrostatic repulsion between the negatively charged surfaces of both the merozoite and the erythrocyte seems to have little effect, as adhesion and invasion is appar- ently unaffected by neuraminidase treatment of the erythrocyte-this treatment reducing its negative charge. It has been suggested that the surface of the parasite might be hydrophilic owing to its high lipid content and that this effect might be sufficiently strong to overcome the weak static repulsive charges between the host and parasite cell surfaces. This seems unlikely, however, as the merozoite plasma- lemma must expose a hydrophilic surface to the aqueous environment that surrounds it and it is in any case covered by the surface coat. Following attachment of the apical region of the merozoite to a suitable erythrocyte a considerable cohesive force appears to be generated, and there is visible morphological deformation of the host cell. This radiates from the point of parasite attachment. Even at this stage, however, the merozoite can detach and successfully invade another cell. The above observations have led to the hypothesis that species specificity of infection depends in part on the presence of specific receptors for merozoite ligands on the erythrocyte surface. Receptors for merozoites on erythrocytes Experimental evidence for specific interactions be- tween the merozoite and erythrocyte surfaces rests on the relative susceptibility to invasion of various types of erythrocyte as well as the blocking effects of enzyme treatments and specific antibodies or lectins. Nevertheless, a well defined receptor has not been identified or characterized. Merozoites of P. vivax and P. knowlesi have different receptor requirements since P. vivax cannot infect Old World monkeys susceptible to P. knowlesi. Both are capable of infecting Duffy positive human erythrocytes, but although they can attach to and deform Duffy negative erythrocytes they detach be- fore invasion occurs. Chymotrypsin treatment, which removes Fya and Fyb Duffy determinants, and treat- ment with antibody to Fya both partially block invasion of Duffy-positive erythrocytes by P. knowl- esi. Other evidence suggests, however, that none of the known Duffy-positive determinants (Fya, Fyb, Fy3) are required alone for invasion. To account for the refractoriness to P. vivax of Duffy-negative individuals of black African descent and for the insusceptibility to P. knowlesi of Duffy-negative ery- throcytes from non-black individuals, it may be supposed that other unknown determinants associ- ated with the Duffy locus have yet to be found. It is also not unlikely that more than a single receptor may be involved. Human erythrocytes of null type for some 10 different blood group systems (including Duffy) are, on the other hand, susceptible to invasion by P. fal- ciparum. Treatment of susceptible cells with chymo- trypsin did not block invasion whereas treatment with trypsin did (the opposite effect was found with cells suited to invasion by P. knowlesi). Some isolates of P. falciparum have a predilection for reticulocytes and young erythrocytes, as also occurs in certain species of rodent malaria parasite. Storage of ery- throcytes at 37°C, or for prolonged periods at 4°C, also decreases their susceptibility to infection. Dif- ferences in the properties of these different cell membranes might provide insight into the require- ments for merozoite binding and invasion. Further- more, it might be possible to relate differences in parasite virulence to such properties. 348 BIOLOGY OF THE MALARIA PARASITE Formation of the parasitophorous vacuole The mechanism of formation of the parasitopho- rous vacuole is unknown but structural evidence suggests that initially material released from the rhoptery-microneme complex in Plasmodium and other apicomplexans may be the cause of the invagi- nation of the erythrocyte membrane that finally results in the formation of the parasitophorous vacuole. However, there are still basic questions to be answered about the organization of these orga- nelles and their contents before this hypothesis can be confirmed. Although the rhopteries have been shown cytochemically to contain protein, they do not appear to contain proteases. Intraerythrocytic stages of P. lophurae contain granules that have been shown to contain a histidine-rich (73 %) protein. Indirect experimental evidence has also indicated that the protein may also be a component of the polar organelles of the merozoites of this species. A similar protein has been found in merozoites of P. falciparum. The highly polar nature of the pro- tein, its possible location, and its effect on erythro- cytes, suggest that it may play a role in the penetra- tion of the erythrocyte by the merozoite, but the evidence so far is circumstantial. At present, the isolation of these organelles is proving technically difficult and improved methods of subcellular frac- tionation of sufficiently large parasite preparations are required in order to be able to determine the chemical nature and function of these organelles. Freeze-fracture experiments indicate that the for- mation of the parasitophorous vacuole is initially caused by the inward buckling of the erythrocyte plasma membrane without any alteration in the density and arrangement of the intramembraneous particles. This indicates that initially at least, gross changes that would be detectable at the electron microscopical level, do not occur at the time of invagination. Research on erythrocyte membranes has suggested that the maintenance of cell shape is controlled by a macromolecular complex of spectrin and an actin- like protein associated with the plasma membrane. The mode of action of these structural proteins is not known but some workers report that the ATP/Ca++ ratio is important in controlling the form and de- formability of the cell by the phosphorylation of spectrin. Although there is much evidence in favour of this hypothesis, other workers postulate that this mechanism is necessary only for maintaining the discocytic form of the erythrocyte and that minor alterations in the cell membrane, such as endocyto- sis, may occur by other mechanisms. For instance, it has been suggested that the unequal expansion of the two lipid leaflets of the membrane bilayer could cause inwards and outwards bending. Most amphi- pathic compounds that crenate erythrocytes are anionic whereas it is suggested that cationic amphi- pathic compounds distribute preferentially into the cytoplasmic half of the bilayer, causing that half to expand relative to the exterior half and thereby induce invagination. Unfortunately, the relation, if one exists, between current work on the deformabil- ity of erythrocytes and formation of the parasito- phorous vacuole during parasite invasion by the action of the rhoptery-microneme complex is un- known at the present time. In this context, it is of interest that rhesus monkey erythrocytes pretreated with colchicine and vinblastine, at concentrations that inhibit microtubule formation and function, are less susceptible to invasion than are untreated ery- throcytes. Although this may infer that invasion is inhibited by the interaction of the drug and the structural proteins of the erythrocyte membrane, one cannot eliminate an effect on the parasite because of the close apposition of host and parasite membranes during adhesion and invasion. However, although electron microscope studies do not detect marked changes in the invaginated membrane during the initial stages of invasion, they do show that the final parasitophorous vacuole is lined with a highly modi- fied membrane in comparison with the normal ery- throcyte plasmalemma. The characteristics of this membrane are discussed later (see also page 352). Resealing It is assumed that once the parasitophorous va- cuole is formed the erythrocyte membrane reseals, since although cytoplasmic clefts have been described in infected erythrocytes, no connexion between the exterior of the cell and the parasitophorous vacuole has been described in electron micrographs. Prelimi- nary attempts to detect such a connexion using chemical probes that are large enough not to pass across but only between membranes, have produced contradictory evidence. Lactoperoxidase labelling of intact infected erythrocytes indicates that there is an absence of label in the parasitophorous vacuole suggesting either that the erythrocyte membrane has resealed, or that the channel, if it exists, is smaller than this molecule. However, some results with lanthanum nitrate-labelling of infected cells showed that the label could pass into the vacuole but this observation was not confirmed in other experiments. 349 SUMMARY OF DISCUSSIONS The mechanism of removal of the merozoite sur- face coat during invasion by the parasite is a further unsolved problem. While the parasite is invading, the coat forms a seal between the merozoite surface and the rim of the parasitophorous vacuole. At this stage the coat does not appear to be lost but accumulates at the posterior surface of the merozoite, extending from the erythrocyte. It has been suggested that this is similar to the " capping " observed in other eukaryotic cells, and results from the flow of lipids and other membrane components to the pos- terior end of the merozoite. The coat is finally liberated at the end of invasion and as the coat fila- ments appear to be normally connected to the inner membranes of the merozoite this may be under enzymatic control. Mechanisms of host cell invasion by other stages of Plasmodium The mechanism of invasion of host cells by ooki- netes and sporozoites and the details of the host cell-parasite interactions in the sporogonic and exo- erythrocytic stages remain to be clarified. The dem- onstration of a surface coat on sporozoites and a prominent surface deposition upon interaction with immune serum, make it possible to characterize the parasite surface components. However, the role of this surface coat and that of the rhoptery-microneme complex found at the anterior end of both sporozoites and ookinetes in the mechanism of cell invasion, will remain circumstantial until a system can be devel- oped to study the interaction of these stages and the host cells in vitro (see also pages 414-415). MEMBRANE CHARACTERIZATION The importance of the characterization of both erythrocyte and parasite membranes has been out- lined in the previous section on parasite invasion and was a recurrent theme during this Workshop. The ultimate objective of research on membrane charac- terization is toward an understanding of both the chemical and the antigenic composition of the var- ious membranes of the infected cell, as well as their physiological roles in parasite development and their interactions with the host system. Such research is essential to permit a directed attack on the disease by chemotherapeutic and immunological methods. However, in order to gain an understanding of the physiological and immunological roles of the com- ponent membranes of the infected cell, it is necessary to develop methods for their efficient purification and isolation from one another in quantities suffi- cient for subsequent analysis. This discussion con- centrates solely on the membranes of the erythrocytic stages, as systems are not yet available by which such studies could be made on other stages of the parasite life cycle. Definition of the problems Four types of membrane are of particular interest in the infected erythrocyte; the erythrocyte plasma membrane, the parasitophorous vacuole membrane, the parasite plasma membrane, and the internal parasite membranes. These membranes function in supplying the parasite with nutrients, in the interac- tions with the host's immune system, in the develop- ment of the internal structure and concomitant compartmentalization of intermediary metabolism of the parasite, and in the transmission of the various " homeostatic " signals from the external environment to the parasite. As yet, little characterization and purification of membranes has been achieved, presumably for the following interrelated reasons: 1. Suitable methods for the sequential " stripping off" of the various membranes has either not been applied or not been developed. 2. Markers such as enzymes, antigens, and co- valent labels specific for the various membranes have either not been utilized or not been developed. 3. The various membranes may change their com- position during the successive stages of development of the parasite within the host cell. The two most accessible membranes are the ery- throcyte plasma membrane and that of the merozoite. Erythrocyte plasma membranes, after purification of infected cells from other blood cell components, could be released by osmotic or detergent lysis or by physical methods, such as nitrogen cavitation. How- ever, these preparations might be contaminated by parasitophorous membranes, especially at times in parasite development when the two membranes are closely associated. The merozoite plasma membrane could be prepared from merozoites released in vitro, but the quantities would be small and probably only suitable for electron microscopy and immunological study. Also, the small quantities and the short viability of the cells may present major difficulties for biochemical experimentation. Larger quantities of parasite plasma membrane could, in theory, prob- ably be prepared from infected erythrocytes, but techniques that remove both erythrocyte and parasi- 350 BIOLOGY OF THE MALARIA PARASITE tophorous membranes from infected cells involve complications related to the instability of parasites and also to differences in the developmental stage of the parasite. The parasitophorous vacuole mem- brane, while in many ways the most interesting component membrane, appears to be the most diffi- cult to obtain because of the difficulty in separating it from other membrane fractions, and because no suitable markers have yet been developed for this membrane. Biological properties of membranes As yet little is known of the biological characteris- tics of the membranes of the parasite and of infected erythrocytes but research has been started in some areas. Transport studies Although it is well known that the infected ery- throcyte can take up glucose and amino acids from the external environment, the methods of transport of these substrates have been studied only with avian malarial parasites. Transport of substrates or drugs into parasitized cells is a complex event since it involves a multicompartmental system. However, it does appear that in the systems studied so far, the parasite changes the permeability of the erythrocyte, both to glucose and to certain amino acids, and that the enhanced rate of entry is due to an increased diffusion rather than to changes in carrier mediated transport. Changes in permeability of the erythrocyte membrane have also been detected using non-meta- bolizable sugars. Uninfected erythrocytes from normal mice and mice infected with P. berghei are not permeable to L-glucose, but this sugar readily enters infected cells by diffusion and not by facilitated transport. It is not clear whether this permeability is related to the stage of development of the parasite within the host cell, as the P. berghei infections are highly asynchronous. But in synchronous P. lophurae infections, the efflux of another non-metabolizable sugar, 3-0-methylglucose was faster in erythrocytes containing larger parasites. Permeability changes in cation transport have also been observed in malarial infections. Monkey erythrocytes infected with P. knowlesi show an increase in intracellular concen- trations of Na+ and a decrease in K+ compared with erythrocytes from normal monkeys. Significant in- creases in Na+ concentrations have also been de- tected in chimpanzee erythrocytes infected with P. falciparum, rhesus monkey erythrocytes infected with P. coatneyi, and hamster erythrocytes infected with P. berghei, but the decreases in K+ levels were not as pronounced in the P. knowlesi infections. Such permeability changes would be important in the functioning of the membrane located ATPases. Uninfected erythrocytes from monkeys infected with P. knowlesi also showed such changes, this finding being correlated with the increased osmotic fragility observed in normal erythrocytes from infected mon- keys. Differences between infected and uninfected cells have been detected in the accumulation of antimala- rial drugs. In vitro, uninfected erythrocytes take up little chloroquine; however, in erythrocytes infected with P. berghei CS (chloroquine-sensitive) strain, there is a great hyperbolic accumulation of the drug if glucose is present as substrate. When glucose is omitted the accumulation is decreased. Erythrocytes infected with P. berghei CR (chloroquine-resistant) strain, on the other hand show a sigmoid accumula- tion curve in the presence of glucose. Upon treat- ment with pronase, erythrocytes infected with CR parasites behave similarly to those infected with CS parasites with respect to drug accumulation. There- fore, chloroquine sensitivity may be due to parasite modification of the erythrocyte surface (perhaps by proteases). The accumulation seems to be a two-step process in which the parasite perhaps modifies the erythrocyte surface first before energy is needed in the form of glucose for stimulation of uptake. Alternatively, it has been shown that infected ery- throcytes do not accumulate chloroquine in the cold but that protease-treated cells do. Thus the energy derived from glucose may be necessary only for making the receptor available and not for the actual uptake. These findings are consistent with the hypo- thesis of an erythrocyte membrane binding site for chloroquine, the accessibility of which is somehow controlled by the parasite. It appears that the para- site induces permeability changes in the plasma membrane of both infected and uninfected erythro- cytes. The mechanism of this induction of permeabil- ity change-in the former by making the host cell anaerobic (see also page 354) and in the latter, by the production of either toxins or an immune response in the host-is unknown. In addition, new transport systems may be required by the infected cell. For instance, if lactate is not removed from cells on glucose pulsing then the internal pH will decrease. Normal erythrocytes do not produce much lactate and the buffering capacity of these cells is adequate but there is a much greater production of lactate in 351 SUMMARY OF DISCUSSIONS infected cells and this may necessitate a transport system for its removal. It has also been suggested that the parasite membrane has lost many of the active transport systems regulating the passage of molecules and has become freely permeable to all kinds of molecules. Present experiments support this hypo- thesis. However, since most of the transport studies, e.g., of amino acids and of nucleosides, have been performed with " free " parasites still surrounded by the erythrocyte membrane, these studies are probably measuring, at least in part, the properties of the erythrocyte membrane and not those of the parasite. Studies are required on parasites that are completely free of erythrocyte plasma membranes. Antigens Both immunological and electron microscope stud- ies indicate that new antigens are expressed on the surface of infected cells. Erythrocytes infected with P. knowlesi-will agglutinate in immune serum when the parasite has grown to the late trophozoite stage and when the parasite has grossly distorted the host erythrocyte. Knob-like protrusions are found on the surface of the erythrocyte plasma membrane of infected cells and in erythrocytes in- fected with P. chabaudi and P. yoelii these have been shown by immunofluorescence techniques to react with immune serum. In addition, certain mouse ery- throcyte antigens have been shown to be lacking from the erythrocyte and from the parasitophorous vac- uole membranes of infected cells. However, nothing is known of the structure and characterization of these antigens at the molecular level. Enzymes Alterations in the activities of some membrane- associated enzymes have been detected in malarial infected erythrocytes. The specific activity of acetyl- cholinesterase (EC 3.1.1.7), which is probably located on the outer surface of the erythrocyte membrane, and of adenylate cyclase (EC 4.6.1.1), some activity of which is located probably on the inside of the membrane, are elevated in mouse erythrocytes in- fected with P. chabaudi. The physiological significance of these changes is not known but it is of interest to note that increases in acetylthiocholinesterase activi- ties are observed in human erythrocytes as they change from the discoid to the spheroid shape. The specific activity of the ATPase of the erythro- cyte plasma membrane appears to be unchanged upon invasion of the erythrocyte by the parasite but histochemical localization studies with both ATPase and NADH oxidase have indicated that the parasito- phorous vacuole membrane differs markedly from that of the erythrocyte. ATPase activity is found on the inside and NADH oxidase on the outside of the erythrocyte ghost plasma membrane and the situa- tion is similar in the parasitophorous vacuole mem- brane. If the vacuolar membrane is merely an inva- ginated erythrocyte membrane then the positions of the enzymes should have been reversed, but spectrin appears to be present at the outside of the parasito- phorous vacuole as one might expect from the invagination of the erythrocyte membrane. This has led to the hypothesis that there is a change in the polarity of the membrane of the parasitophorous vacuole that may be caused by the presence of the parasite. Such changes in protein polarity of the membrane of the vacuole would probably require lipid changes. This may be possible during the growth of the parasite within the vacuole but it is difficult to conceive how this could occur in the short interval (a few minutes) in which changes in ATPase and NADH oxidase activities are seen to occur. Freeze-fracture results indicate that just after the merozoite has completed invasion of the erythrocyte there is a depletion of intramembraneous particles from both faces (parasite/erythrocyte) of the vacuolar membrane. These observations might be explained by the proposal that merozoite rhoptery material may, in aqueous surroundings, form lamellae, per- haps analogous to liposomes, that could be incor- porated into the erythrocyte surface. Structural properties of membranes Electron microscope studies have shown that in P. knowlesi the erythrocyte plasma membrane is extremely dense and thickened in schizont-infected cells. In addition, there appears to be a 20% reduc- tion in the number of intramembraneous particles on the P-fracture face of the erythrocyte plasma mem- brane in schizont-infected cells with rearrangement and clumping of the remaining particles. This clump- ing could result from an alteration or degradation in the spectrin-actin network of the erythrocyte mem- brane or could be the effect of amphiphilic com- pounds or lyso-phospholipids such as lysolecithin. Such spectrin degradation is believed to occur in erythrocytes infected by P. chabaudi and P. berghei but not in those infected by P. knowlesi. Particle aggregation cannot have been a fixation artefact as unfixed samples of normal and infected erythrocytes were used throughout the freeze-fracture studies. Purely morphological criteria are inadequate for 352 BIOLOGY OF THE MALARIA PARASITE efficient membrane characterization. These studies, although important, need to be linked if possible to biochemical characterization of the membranes be- fore a proper understanding of the structure and function of the membranes can be obtained. There is an urgent need for specific marker en- zymes for the identification of the various membrane fractions of the infected cell. Techniques that have been successfully used in other systems are now being applied to studies on the malarial parasite and it appears that some markers for the identification of the erythrocyte plasma membrane of infected cells have been recognized. It appears that both sialic acid and the glycoproteins present in the erythrocyte plasma membrane are not found in either the parasi- tophorous vacuolar membrane or the parasite mem- brane of infected erythrocytes. In P. knowlesi, an iodinatable component, a glyco- protein with a molecular weight of 125 000-130 000, is found in the erythrocyte plasma membranes of schizont-infected cells but not in uninfected erythro- cyte membranes. It seems unlikely that this protein arises from the proteolytic action on an existing membrane protein as its molecular weight is higher than other glycoproteins present and this has led to the hypothesis that this new protein is inserted by the parasite into the erythrocyte membrane. It would be of great interest to know when this protein appears during the growth of the parasite within the host cell and if this new component is antigenic. Examples of protein insertion into membranes are known to occur. For instance, the transplantation antigen of SV40 virus, which has a molecular weight of about 55 000, is glycosylated by the host cell and inserted into the host membrane. A glycosylating system is probably absent in mature erythrocytes but it could be present in the metabolically active parasite. An analogy can be drawn with schisto- somes in which there is apparently a movement of glycoprotein from the interior of the parasite to the surface membrane. It is perhaps possible in malarial infected erythrocytes that, if the vacuolar membrane is closely apposed to the erythrocyte plasma mem- brane (as occurs in the later stages of parasite development), such (glyco)protein insertion could occur via the vacuolar membrane to the exterior erythrocyte plasma membrane. Electron microsopi- cal observations could be interpreted as indicating that protein insertion occurs also in the parasito- phorous vacuole membrane, since, although this is devoid of intramembraneous particles at invasion, there is a gradual increase in total particle number as the parasite grows to maturity. However, these observations could be a result of individual compo- nents of the transmembraneous particles being expressed as separate entities, although the size of the components argues against this idea, and these results should be confirmed by biochemical studies. Periodic acid-Schiff's staining of membrane pro- teins isolated from P. knowlesi-, P. berghei-, and P. chabaudi-infected erythrocytes, and galactose oxi- dase-labelling of P. knowlesi-infected erythrocytes, indicate that some glycoproteins are also lost from the erythrocyte plasma membrane. The loss of a protein from a membrane could be due to proteolytic activity and in the case of the galactose oxidase method, the enzyme itself, unless very pure, usually has proteolytic activity. However, in these studies, if the loss of glycoprotein were due to proteolytic activity then it is more likely to have been a result of parasite proteolytic activity since the same enzyme preparations were used to label both infected and uninfected cells. Loss of glycoprotein from the ery- throcyte plasma membrane has also been indicated by some studies where the number of concanavalin A binding sites on the surface of erythrocytes in- fected by P. knowlesi was reduced. However, in contrast, other results indicate an increase in con- canavalin A binding sites in infected cells. This anomaly requires to be solved. Present results indicate therefore that the major changes between membranes of infected and unin- fected cells seem to be in the glycoprotein pattern. They also indicate that membrane markers are avail- able for the identification of the erythrocyte plasma membrane but that specific markers for vacuole and parasite membranes are required before fractionation and identification of the various membrane compo- nents of the infected cell can be achieved. PARASITE METABOLISM General comments The basic reason for studying the biochemistry of the malarial parasite is to obtain a thorough under- standing of the physiological mechanisms of the parasite and its relationship to the host cell. Such an understanding has direct relevance to the develop- ment of antimalarial drugs and of techniques for the cultivation of the parasite in vitro, and to an under- standing of the immunological reactions of the host and parasite at the molecular level. Most of the biochemical studies on the malarial parasite have been restricted to the intraerythrocytic stages since 353 SUMMARY OF DISCUSSIONS these are the only stages that can be prepared in sufficient quantity for such studies. Information on the metabolism of the sporogonic and exoerythrocy- tic stages can, as yet, be inferred only from cyto- chemical and electron microscope studies. Metabolic studies on these stages await the development of methods for the preparation and fractionation of parasites in sufficiently large quantities for bio- chemical analysis. It is perhaps a disadvantage that only a limited number of Plasmodium species, namely P. knowlesi, P. lophurae, and P. berghei and the other malarial parasites of rodents have been studied. The reasons for this are practical, as these parasites generally produce a virulent infection that allows parasites to be prepared in large quantities. However, these species may not be the best models for human parasites such as P. falciparum and P. vivax. Never- theless, although there are obvious differences be- tween parasites, the same basic pathways are prob- ably common to all parasites and there are probably fewer differences than similarities even between the avian and mammalian parasites. This may not be too surprising since the erythrocytic parasites are developing in similar environments although in dif- ferent hosts. The fact that the malarial parasite does develop in a specialized environment, i.e., the ery- throcyte, may mean that there are deviations from the standard pathways of eukaryotic metabolism that may be related to this specialized environment and the parasite's evolutionary status. In order to assess rationally the metabolic differ- ences and similarities between species and even strains of species, reports of experiments should describe clearly the characteristics of the strain be- cause each strain may differ genetically and may also vary in different environments. For instance, some strains of P. berghei are found in reticulocytes, whereas others preferentially invade mature erythro- cytes. This may be a result of different metabolic characteristics between the strains. In addition, me- tabolic changes may occur during the cell cycle of the parasite and these must also be considered. Carbohydrates It is accepted that in all species the major energy substrate of the erythrocytic stages is glucose, which is metabolized glycolytically to lactate. However, there appear to be minor differences in the glycolytic rates and the further metabolism of glucose between different species. Originally it was thought that there were marked differences between avian and mamma- lian malarial parasites as the erythrocytic stages of avian parasites contained a cristate mitochondrion and therefore presumably metabolized glucose com- pletely via the citric acid cycle. However, it has not been proved conclusively that there is a functional citric acid cycle in the intraerythrocytic stages of avian plasmodia, and is it not certain whether these stages of any malarial parasite utilize oxygen. Although erythrocytic stages of malarial parasites growing in vitro need low oxygen concentrations, there is no evidence as yet concerning the role of oxygen in parasite metabolism. The possibility that malarial parasites have some form of electron trans- port chain is generally considered unlikely as none of the constituents of such a chain have been reported with any certainty. For instance, the observation that mouse blood cells infected with Friend leukaemia virus change the structure of their CoQ in a similar way to the changes attributable to malaria parasites casts doubt on the reports of the synthesis of CoQ by the parasite itself. However, very recently, prelimi- nary experiments with preparations of P. knowlesi and P. inui indicated peaks of cytochrome activity at 540 and 615 nm, but there was no evidence of cytochromes a and b. An iron-sulfur protein was also detected. These results await further confirma- tion before the role of cytochromes in plasmodial metabolism can be ascertained. It appears possible that the oxygen-carrying capa- city of blood might be reduced by the parasite both by the destruction of haemoglobin and by alterations in the oxygen-carrying capacity of haemoglobin it- self. A reduction in the intracellular pH and in the amount of 2,3-diphosphoglycerate in infected cells would reduce the affinity of haemoglobin for oxygen. A previous report has indicated that 2,3-diphospho- glycerate concentrations are reduced in erythrocytes infected by P. berghei which might result in making the host cell " anaerobic" and in turn affecting the permeability of the cell membrane (see also page 351). There have been many investigations of the pen- tose-phosphate pathway of the erythrocytic stages of malaria parasites since it was suggested that a gene- tically determined deficiency of glucose-6-phosphate dehydrogenase (G-6-PDH) (EC 1.1.1.49) afforded some degree of protection against P. falciparum infections in man. The activity of this pathway in malarial infected cells is relatively low but there appears to be some variation between the species studied. The parasite itself probably does not possess 354 BIOLOGY OF THE MALARIA PARASITE a fully functioning pathway in spite of the presence in infected cells and in "free" parasites of an isoenzyme of 6-phosphogluconate dehydrogenase (6- PGDH) (EC 1.1.1.43) which is attributable to the parasite. It seems most likely that the parasite utilizes the pathway of the host erythrocyte. The function of the pentose-phosphate pathway in ery- throcytes is believed to be linked to membrane stabilization. Obviously the demand on this function would be greater when most membrane is present, i.e., at the schizont stage, and it is at this stage when most recycling via the pentose-phosphate pathway occurs in infected erythrocytes. It has also been suggested that the pentose-phosphate pathway could provide ribose for nucleic acid synthesis. This is possible but it has been argued that insufficient quantities of pentose sugars would be generated for this to occur. It is difficult to make a rational assessment of the contribution of pentose-phosphate pathway activity to nucleic acid synthesis without quantitative data. However, in some systems with extensive pentose-phosphate activity, pentose sugars have been shown to be generated by a reverse pathway and not by the action of G-6-PDH and 6- PDGH. The origin of the pentose sugars involved in nucleic acid synthesis is not known but if these do not arise from the activity of pentose-phosphate pathway, host nucleosides could be the source. There is no evidence suggesting the major role of other substrates, for example ribose and glycerol, in energy metabolism. It has been suggested that the homeostatic mechanism of the host ensures a con- stant supply of glucose and therefore the metabolism of other sugars appears less profitable. The depen- dence of the parasite on large amounts of glucose may indicate an avenue for chemotherapeutic attack if differences exist between the enzymes of carbo- hydrate metabolism in the host and the parasite. Sugar analogues, such as arabinosyl-adenine, which probably act finally on protein synthesis but whose action depends initially on its phosphorylation by hexokinase (EC 2.7.1.1), may prove to be antimala- rial in action. However, the use of inhibitors of glucose utilization in combination with other drugs would not be useful because of the practical difficul- ties involved in testing. Lipid metabolism The erythrocytic stages of the malarial parasite appear to be incapable of de novo biosynthesis of fatty acids. However, they can make glycerides and phosphoglycerides from the fatty acids, nitrogenous bases, alcohols, and CoA obtained from the host and can degenerate the glycerol moiety during gly- colysis. Many lipids, notably cholesterol, sphingo- myelin, lysolecithin, and fatty acids, of the plasma exchange with those of the erythrocyte membrane and the parasite can presumably take advantage of this dynamic situation to obtain lipids from the erythrocyte membrane. Published evidence also suggests that the malarial parasite is also incapable of de novo synthesis of sterols. However, recent preliminary work with P. knowlesi suggests that there is biosynthesis of cholesterol from acetate at the early stages of devel- opment and that this ceases later at the trophozoite and schizont stages. Although the experimental data is incomplete, it seems probable that the design of the early experiments may not have detected this change with time; this illustrates that the metabolic changes occurring during the life cycle of parasites should be further investigated. It appears odd that cholesterol synthesis should cease at the schizont stage when an increase in membrane material is most marked, but present evidence suggests that there is no de novo synthesis of sterols at this stage of development. There are many differences between the lipid com- position of malarial parasites and that of the host erythrocyte, and this indicates that lipid analyses may possibly be used as an indicator of the purity of parasite preparations and also may indicate some areas for chemotherapeutic attack. Phosphatidyl inositol is present in all plasmodia even though erythrocytes vary markedly in their content of this lipid. 1,2-Diacylglycerols, which promote fusion of membranes, are found in the parasite but not in the host erythrocyte. Also parasites preferentially incor- porate the 18:1 fatty acids, oleic acid and cis- vaccenic. The levels of these two acids, both of which have haemolytic properties, are raised in both the membrane of the infected cell and the blood plasma of infected animals. This has led to the suggestion that these fatty acids are involved in the permeability and fragility changes observed in ery- throcytes from infected hosts. There are several mechanisms by which this could occur: by their esterification to membrane phospholipids, by dis- placing other fatty acids, or by changing the fluidity of the phospholipid bilayer either by changing the degree of unsaturation of their fatty acids or by acting directly upon hormonal regulatory mecha- nisms. 355 SUMMARY OF DISCUSSIONS Nucleic acid metabolism The intraerythrocytic stages of rodent malarial parasites use exogenous purines for nucleic acid synthesis but must add the ribose and phosphate moieties to the preformed purine. This is accom- plished by the enzymes of the " salvage pathway " to form adenosine monophosphate from adenine or adenosine. Data obtained from studies of incorpora- tion of radioactive precursors into intact cells in- fected with P. chabaudi and from kinetic experiments on purified enzyme preparations suggest biosynthetic pathways in the parasite that are generally similar to those of the host cell metabolism. However, details of the regulation both of the uptake of nucleosides and bases into the infected cell, and of the metabolic routes involved are still lacking. Plasmodia are ca- pable of providing the nucleotides that they need for the synthesis of their nucleic acids, mainly by stimu- lating reactions that also exist in the host cell. However, minor differences in the sequence and regulation of some reactions may exist. How far such minor differences also exist between different species of malarial parasite is unknown since enzy- matic studies have been performed only with rodent malarias. Exogenous pyrimidines that are incorpor- ated into the nucleic acids of white cells are not used by the erythrocytic stages of malarial parasites. Initially it was suggested that the erythrocyte mem- brane was impermeable to pyrimidines but this now appears to be unlikely. Consequently it has been further suggested that the parasite membrane is impermeable to these precursors but this has not been proved conclusively. P. chabaudi lacks thymi- dine kinase (EC 2.7.1.75) and as a result is unable to phosphorylate thymidine. Whether the parasite lacks a pyrimidine " salvage pathway" or is impermeable to the precursors is not known but the result would be similar and the parasites would fail to incorporate pyrimidines into nucleic acids. It appears therefore that malarial parasites synthesize pyrimidines de novo. The rodent species appear to have at least some of the enzymes necessary to synthesize their own pyrimidines, although all the enzymatic steps have yet to be described. Although the descriptions of some of the enzymes could be attributed to host- cell contamination, thymidylate synthetase from P. chabaudi infected cells is of parasite origin. The activity of this enzyme rises sharply during the growth of the parasite from ring to mature stages, reaching a peak just before the nuclear division. Most studies on the incorporation of purine pre- cursors suggest that DNA is synthesized only during the earlier stages of the cell cycle, almost ceasing during schizogony. But there is other conflicting evidence; for example, the pyrimidine precursor, orotic acid, is incorporated into DNA and RNA mainly during the schizont development. However, since all these methods measured simply incorpora- tion of precursors into DNA, and not net synthesis, they do not give equivocable answers. Nevetheless, there is some evidence from direct biochemical esti- mations of nucleic acid contents of the various developmental stages of P. knowlesi that the majority of DNA is synthesized during growth from the ring to the late trophozoite stage and that little DNA is synthesized during schizogony. These results agree with those on periodicity in the synthesis of thymi- dylate synthase (EC 2.1.1.45) mentioned above. These studies might appear to conflict with the known schizonticidal action of tetrahydrofolate de- hydrogenase (EC 1.5.1.3) inhibitors, e.g., pyrimetha- mine, which are assumed to act by inhibiting DNA synthesis. One possible explanation could be the use of folate cofactors in some reaction other than that of thymidylate biosynthesis but there is no firm evidence as yet for this hypothesis (see also page 358). Further work on the periodicity of DNA synthesis and the action of pyrimethamine is required before this apparent anomaly can be resolved. Protein synthesis Since the free amino acids present in the erythro- cyte are presumably not present in sufficient quantity to serve as a source for plasmodial protein synthesis and the de novo biosynthesis of amino acids is severely restricted, haemoglobin remains the most abundant reservoir of amino acids for the growth of malarial parasites. Malarial parasites appear to use proteolytic enzymes to break down the host's hae- moglobin within a food vacuole, liberating amino acids and forming the malarial pigment haemozoin. The physicochemical properties of haemozoin indi- cate its equivalence to haemin. However, no reliable characterization of parasite proteases is available since the reported results could be attributable to the contamination of the preparations with host-cell components. Cathepsin D is not present in the parasite, but as chloroquine is both antimalarial and lysosomotropic and inhibits cathepsin B1 in macro- phages, it is possible that cathepsin B1 might be present in the parasite. Chloroquine is concentrated in the food vacuole where haemoglobin digestion occurs. Thus studies on proteases responsible for the digestion of haemoglobin might help in understand- 356 BIOLOGY OF THE MALARIA PARASITE ing the mode of action of the drug and add confir- mation to the hypothesis that the drug acts by preventing haemoglobin digestion. The digestion of haemoglobin by a lysosomal mechanism is suggested by the observation that acid phosphatase is present in the food vacuole, although some electron microscopical studies have led to the suggestion that the malarial parasite does not possess lysosomes. This may be misleading as all these studies were morphological and cytochemical and there is no morphological/cytochemical assay for defining a lysosome. Acid phosphatase has also been detected in the endoplasmic reticulum from which it is presumably transferred directly to the food vacu- oles. It is unlikely, however, that the enzymes are free within the cytoplasm, as in other systems pro- teolytic enzymes are packaged in the Golgi bodies into granules which then fuse with the lysosomal membrane. It would probably be difficult to detect such fusion of the granules with the electron micro- scope. Clearly the fractionation of parasites, free from erythrocyte membrane debris, using reliable enzyme markers is required before the detailed mechanism of haemoglobin digestion can be under- stood. Although it appears that the protein synthesizing systems in plasmodia are typically eukaryotic there is preliminary evidence suggesting that there are slight differences from this typical pattern in P. berghei, P. knowlesi, and other plasmodia. The sedimentation coefficient and base composition of rRNA from P. knowlesi are typically protozoan. These observa- tions indicate that the ribosomes are of parasite origin and that the suggestion that the large rRNA species of malarial ribosomes is provided by the host seems unlikely. rRNA precursors appear to be simi- lar to those of typical eukaryotes but the method of processing rRNA is unknown. These molecules are normally processed in the nucleolus but P. knowlesi, like other malarial species from mammalian hosts, does not appear to possess a nucleolus. The RNA polymerase activity of extracts of " free " parasites of P. knowlesi appears not to be typical of eukaryotes as judged by their sensitivity to a-amanitin, but this preliminary observation requires to be confirmed, preferably on enzymes extracted from nuclear prep- arations of malarial parasites, before any definite conclusion can be made. So far only monomeric ribosomes have been isolated from malarial parasites, but this does not exclude polysome formation since either these may have broken down to the monomer during extrac- tion, or even not been present in the schizont stage from which the ribosomes were isolated. Perhaps earlier stages of the life cycle might be more suitable for polysome extraction. Difficulties have been experienced in attempts to isolate mRNA and initiation factors from parasites, but amino-tRNA synthetases isolated from P. ber- ghei appear to differ from the highly specific enzymes from typical eukaryotes. The parasite enzymes ap- pear to be non-specific as they will aminoacylate tRNA from various sources such as E. coli, liver, and yeast. In spite of these examples of minor differences between the protein synthesizing systems of parasites and other eukaryotes it has proved difficult so far to inhibit parasite protein synthesis selectively. For instance, although tetracycline inhibits the growth of P. berghei, its effect is unlikely to be parasite specific, as the drug also affects protein synthesis in reticulo- cytes. Protein synthesis not only has importance in chemotherapy but also has important implications for the immunology of malaria and the production of parasite antigens. It has been shown that some proteins synthesized in vitro are antigenic. These studies on the synthesis of proteins, particularly of antigenic proteins, in relation to the parasite cell cycle and the genetic control of antigenic variability should be encouraged. Cofactors and vitamins Relatively few cofactors appear to be essential for the intracellular development of erythrocytic stages of malarial parasites. These are 4-aminobenzoic acid (which is presumably required for the synthesis of folates) biotin, and pantothenate. This evidence is based on the growth of the parasite in vitamin deficient hosts and on studies of growth in vitro. Malaria parasites differ from their hosts in that they synthesize folate cofactors de novo, like other micro- organisms. Consequently the pathway of folate bio- synthesis has been the site of chemotherapeutic attack, e.g., by the action of a sulfa drug that inhibits the malarial dihydropteroate synthase (EC 2.5.1.1.5), or by the action of pyrimethamine and other anti- folate drugs that bind to the tetrahydrofolate dehy- drogenase. Inhibition of dihydropteroate synthase and tetrahydrofolate dehydrogenase must result in a depletion of folate cofactors, but the final mode of action of these drugs is unknown. Evidence has already been presented which might argue against an effect on nucleic acid synthesis and which may 357 SUMMARY OF DISCUSSIONS suggest an action on another as yet unknown path- way for folate cofactors (see also page 356). Recently it has been reported that N5-methyltetrahydrofolate may be utilized for methionine biosynthesis. If this is so it would negate the hypothesis that the thymidyl- ate pathway was the only folate cofactor function in malaria and might help unravel some of the appa- rently contradictory evidence on the mode of action of pyrimethamine. The importance of this pathway to the utilization of folate cofactors should be determined since the parasite is able to incorporate this amino acid from exogenous sources and it is required for optimum growth of the parasite in vitro. Dihydrofolate biosynthesis in malaria may be inhibited by compounds other than analogues of pyrimethamine, as one compound that inhibits the tetrahydrofolate dehydrogenase appears to be a pteridine analogue. The use of pteridine analogues as antimalarial drugs is a possibility but to date suit- able compounds are not available. In strains resistant to pyrimethamine the mech- anism of resistance appears to be the production of an enzyme with a decreased affinity to the drug. These studies were made with P. berghei and it is possible that the mechanism of resistance to this drug may differ in P. falciparum since the level of resistance in the experimental malarias is several-fold greater than in P. falciparum. It is important, therefore, to study the mechanism of resistance to pyrimethamine either in the field or in vitro. No evidence is available on the mode of resistance to sulfonamides as no studies have been reported on the properties of dihydro- pteroate synthase in sulfonamide-resistant strains of malaria. Resistance of sulfonamides is often accom- panied by resistance to pyrimethamine, which may suggest either the unlikely event of two mutations at the same time, or the loss of the pathway of folate biosynthesis and the utilization of preformed folates by the parasites. The role of biotin in parasite metabolism is un- known but the metabolic fate of pantothenate has been elucidated, at least in the avian parasite P. lo- phurae. Pantothenate is not used directly by the parasite but rather as coenzyme A (CoA) synthesized by the host cell. None of the enzymes of CoA biosynthesis are found in the parasite but they are present in the erythrocyte. These studies indicate a major metabolic lesion in the metabolism of the parasite but how far such lesions exist in other parasites is not known. It would be of interest to discover whether other analogous biosynthetic lesions exist in the erythrocytic stages of malarial parasites. The possible utilization of the pentose- phosphate pathway of the host by the parasite has already been discussed (see also page 354) and there are several other cases of enzyme deficiencies in human erythrocytes that might result in resistance to infection by malaria parasites. For example, pyri- doxine kinase has a lower activity in the erythrocytes of Africans than in those of Caucasians and ,B- thalassaemia patients have a slow rate of conversion of pyridoxine to pyridoxal phosphate in their ery- throcytes. It seems possible that studies using ery- throcytes with known enzyme deficiencies will yield information regarding possible metabolic lesions in malarial parasites. The recent continuous cultivation of P. falciparum make these studies a practical possibility. PARASITE PREPARATION AND ISOLATION The difficulty of interpreting experimental data obtained from parasite preparations that might have been or were contaminated with host components or other materials, or which may have lost components of the parasite, was a recurrent theme of discussion and contention at every session of these Workshops. To minimize this problem, methods of parasite preparation that are judged suitable for particular kinds of investigation, as well as criteria for evaluat- ing existing methods, clearly need to be developed and applied in a rigorous way. Probably no single method can be recommended to produce " clean " parasites for study. Nevertheless, it is obviously desirable to develop procedures that give high yields of intact and metabolically active parasites free of other cells, cell debris, etc. Development of such techniques and their strict application is a prerequi- site for meaningful future work on many studies outlined in this report. Methods of separating parasitized erythrocytes from contaminating materials Several methods are currently available for the separation of parasitized erythrocytes from host cells, but they give variable yields with different degrees of contamination. (a) Simple differential centrifugation tends to give poor yields but has the advantage that preparations are chemically uncontaminated. (b) Gravity filtration through buffer-equilibrated columns of filter paper or cellulose powder provides an efficient method of reducing leucocyte contamina- tion in blood infected with Plasmodium berghei and 358 BIOLOGY OF THE MALARIA PARASITE P. knowlesi., but overloading of such columns should be avoided. (c) Various methods that remove platelets from blood are available and they should be applied during parasite preparation. (d) Density gradients improve parasite yields by concentrating the infected cells in bands of similar density but they introduce foreign materials into the preparation. Dextran gradients are not efficient and Ficoll/hypaque gradients give only partial purifica- tion. A simple sedimentation method with " Plasma- gel" enriches yields of schizonts from cultures of P. falciparum. The different densities of reticulocytes and older red cells need to be taken into account in such fractionation procedures. Methods of releasing intraerythrocytic parasites None of the methods for freeing parasites from erythrocytes by cell lysis are entirely satisfactory but immune lysis, continuous flow sonication, and nitro- gen cavitation seem to be most promising as regards obtaining " free " parasites (" free " according to morphological criteria). The metabolic activity of " free " parasites prepared by saponin, ammonium chloride, or hypotonic lysis, or by immune lysis has been evaluated by the uptake of radioactive precur- sors. Only the saponin and immune lysed parasites, which are frequently contaminated with erythrocyte membranes, were metabolically active and it was concluded that " good " morphology need not imply biochemical integrity or activity. The metabolism of parasites freed by continuous flow sonication has yet to be evaluated, but French pressure cell lysis does not produce viable parasites. Immune lysis. Some 50% of P. lophurae parasites present in erythrocytes can be prepared free of host erythrocyte tissue by immune lysis with a high-titre antiserum. This method requires that the incubation in immune serum be performed in a favourable growth medium, that pipetting, if used, is gentle and that low-speed centrifugation is used to separate agglutinated cells from the non-agglutinated para- sites. Parasites produced by this method are metabo- lically active and can be grown to the schizont stage in vitro. Continuous flow sonication. About 80% of the erythrocytes are broken by this procedure but fur- ther purification results in the loss of some free parasites. Sonication appears to shatter schizont- stage parasites releasing merozoites without a sur- rounding parasitophorous vacuole. Nitrogen cavitation. There have been two reports of the use of nitrogen cavitation for the release of " free " parasites and it appears that this method is capable of the sequential removal of membranes from infected cells under carefully controlled condi- tions. The method appears to be promising but needs further evaluation. Natural release. Sonication, as well as all the other lytic procedures, releases whatever form of parasite is present in the erythrocyte. The use of culture sys- tems for the natural release of merozoites seems an extremely promising alternative but additional efforts are needed to investigate this method further and to make it reproducible. Unfortunately, free merozoites rapidly degenerate in conventional cul- ture media and so there is an urgent need to develop and improve harvesting conditions for these stages (see also page 412). It was concluded that there should be an effort to standardize conditions used, or at least to standard- ize the reporting of conditions used, by different laboratories for the preparation of " free " parasites. There also appears to be a lack of standard bio- chemical markers for the quantitive detection of host cell contaminants, although qualitative monitoring of preparations can be made by the use of low-power electron microscopy. Isolation of stages from the invertebrate Contamination is a major problem in sporozoite preparation. This problem can be reduced but not eliminated by a procedure in which the salivary glands are separated from the mosquito thorax and the sporozoites are subsequently separated by density gradient centrifugation. The density separation meth- ods employed include the use of biphasic systems with homologous serum and Hypaque rather than albumin and Hypaque. Isolation of exoerythrocytic stages The isolation of the exoerythrocytic stages from mammalian hosts is not a practical possibility until methods are available for the cultivation of these stages in vitro in large quantities. STRAIN DIFFERENTIATION IN MALARIA PARASITES It is fundamental to any programme designed to control pathogenic organisms that the precise iden- 359 SUMMARY OF DISCUSSIONS tity of the pathogen can be established. This is particularly true in relation to the development of vaccines and epidemiological problems such as the spread of drug resistance. In malaria, the term " strain " is generally used to describe parasites that possess their own distinctive characteristics. Characteristics used to differentiate strains have included morphological features, relapse patterns, drug response, infectivity to various vector species, and immunological differences. A clear un- derstanding of the composition of a " strain ", how- ever, depends on knowledge of: (1) the basic genetic organization of the parasite; and (2) the manner in which the genetic factors involved are inherited and dispersed in the parasite population. Recently, biochemical techniques have been used to examine genetic differences between isolates of both rodent and human malarial species. The two methods principally used have been: (1) enzyme electrophoresis, which reveals differences in the pro- ducts of individual gene loci, and (2) DNA charac- teristics, which reveal differences in the total genetic information of the organisms. So far these methods have been applied mainly to the Plasmodium species infecting rodents. These investigations, together with genetic studies involv- ing hybridization and cloning, have shown that: (1) the blood forms of the parasite are haploid; (2) there is a considerable degree of random mating among parasites of a given subspecies, thus permit- ting the production of numerous gene combinations; and (3) parasite populations that are reproductively isolated from one another can be identified. However, the basic genetics of the malarial parasite still needs much more investigation. Enzyme variants and DNA characteristics are considered the most reliable markers for strain dif- ferentiation. For the full characterization of a given parasite, however, it is important that other charac- teristics such as antigenic diversity and drug response should be studied. An example of the importance of examining a variety of characteristics is given by the distinguishing features of the rodent species P. ber- ghei and P. yoelii. These two species possess similar blood form morphology, yet differ in their innate response to chloroquine and in their enzyme and DNA characteristics. Enzyme analysis shows that the two species are almost certainly reproductively isolated. It will thus be of considerable interest to examine the enzyme forms of drug-resistant and drug-sensitive P. falciparum from various regions of the world to determine whether similarly genetically distinct forms exist in the parasite population. This approach should be extended further among the malarial species infecting man. 360