Development of malaria vaccines: Memorandum from a USAID/WHO meeting

Bulletin ofthe World Health Organization, 61 (1): 81-92 (1983) c. World Health Organization 1983 Development of malaria vaccines: Memorandum from a USAID/WHO meeting The fifth meeting of the Scientific Working Group on the Immunology of Malaria evaluated studies of the production and analysis of defined malarial antigens. Rapid pro- gress has been made in the study ofprotective antigens on the surface ofsporozoites and it is likely that a family of analogous polypeptides occurs in several species of Plasmodium. New assays have been developedfor the detection of these antigens andfor the detection of infected mosquitos. Exoerythrocytic stages of several parasite species can be cultivated in vitro, providing an assay system for antibody and allowing the characterization ofexoery- throcytic stage antigens. Progress has also been made in the identification of species- and stage-specific antigens of the asexual blood stages of rodent, simian, and human malaria parasites. In some instances, protective immunity has been shown to be directed against polypeptides (with a high relative molecular mass) synthesized at a late stage of schizont development. Messenger RNA (mRNA) species from P. knowlesi and P. yoelii have been successfully translated in vitro to give polypeptides with a high relative molecular mass (Mr). Monoclonal antibodies have been used to identify and to purify important parasite antigens and purified P. yoelii antigens induced protective immunity. Monoclonal anti- bodies reactive with merozoite surface antigens have been used, as well as S-antigens, to distinguish between different isolates of P. falciparum. Recombinant DNA technology is being applied to Plasmodium: differences were found between repetitive DNA sequences from the genome oftwo isolates of P. falciparum; the genesfor ribosomalRNA of P. falci- parum and P. yoelii, and sequences homologous to the actin gene were identified in frag- ments of Plasmodium DNA cloned in prokaryotic vectors; by means of hybrid selection, complementary DNA (cDNA) probes were used to purify mRNAs encoding proteins of P. knowlesi of up to 100 000 Mr. Since the inception, in 1976, of the research pro- gramme on the immunology of malaria, rapid advances have been made in identifying malaria anti- gens that elicit protective immune responses. Pro- tective immunity is directed principally at extracellu- lar forms of the parasite, namely the sporozoite, the merozoite, and the gamete, which come into contact with the immune system. These responses are directed against species- and stage-specific antigens on the sur- face of the extracellular forms. Major efforts are directed towards purifying the antigens concerned and exploring the possibility that they could be pro- duced by a combination of recombinant DNA tech- nology and protein chemistry. The fifth meeting of the Scientific Working Group on the Immunology of * This Memorandum is based on the contributions made by the signatories listed on pages 91-92 on the occasion of the fifth meeting of the Scientific Working Group on the Immunology of Malaria, held in Geneva in March 1982 under the auspices of the United States Agency for International Development and the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases. Requests for reprints should be addressed to Chief, Research and Technical Intelligence, Malaria Action Pro- gramme, World Health Organization, 1211 Geneva 27, Switzerland. A French translation of this article will appear in a later issue of the Bulletin. Malaria, held in Geneva on 8-10 March 1982, was devoted to recent advances in the analysis of malaria antigens and to techniques of current or potential application in this field, including macromolecular separation techniques, microanalytical methods, the analysis of protein structure, and recombinant DNA technology. Effective collaboration between para- sitologists, immunologists, chemists and molecular biologists is taking place and research on the immunology of malaria is currently in an extremely productive phase. THE IDENTIFICATION, STRUCTURE AND FUNCTION OF MALARIAL ANTIGENS Sporozoites of rodent, simian and human malarias Sporozoite antigens. Important progress has been made in the study of protective sporozoite antigens, using the Plasmodium berghei murine malaria model. The surface of fully differentiated sporozoites of P. berghei is covered uniformly by a major, 4262 -81- 82 MEMORANDUM species-specific immuno-dominant membrane pro- tein, Pb44, which migrates as a 44 kilodalton (kd) polypeptide (Mr: 44 000)" on sodium dodecyl sulfate- polyacrylamide gel electrophoresis) (SDS-PAGE). As little as 10 yg of purified monoclonal antibody (3D1 1) to Pb44 completely protects mice against sporozoite- induced infection. The infectivity of sporozoites is also abolished by preincubation in vitro with the monoclonal antibody or its monovalent Fab frag- ment. This treatment prevents the attachment to, and penetration of, mammalian target cells by P. berghei sporozoites in vitro (see below: exoerythrocytic stages of mammalian malaria parasites). Synthesis of the circumsporozoite (CS) membrane protein Pb44 is a major metabolic activity; during differentiation of sporozoites maturing in the sali- vary gland, as much as 20% of the metabolically labelled (35S-methionine) protein is immunoprecipi- tated by the monoclonal antibody (3D1 1). Three polypeptides share an epitope that is recognized by this antibody-including two polypeptides (Pb54 and Pb52) with higher relative molecular mass which are probably intracellular precursors of the surface anti- gen itself (Pb44). Sporozoites from oocysts do not contain these precursors but translation of mRNA from mature sporozoites in a cell-free system yields products closely resembling the precursors of the CS protein. Pulse-chase experiments in vitro indicate that Pb52 may be converted into Pb44. The precursor/product relation of Pb54 is uncertain; it may be processed at an earlier stage of sporozoite development in the mosquito although it is still present in exoerythrocytic forms after invasion of the mammalian target cell. Pulse-labelling in vivo in the mosquito salivary gland might give further infor- mation on this point. A family of analogous polypeptides with similar biosynthetic, physicochemical and antigenic prop- erties probably exists in other species of plasmodia. In P. knowlesi, the synthesis of CS protein and its intracellular precursors has been studied using nine monoclonal antibodies-five of these (or their Fab fragments) abolished infectivity of sporozoites by preincubation in vitro. These particular monoclonal antibodies immunoprecipitated a trypsin-sensitive (surface) protein of Mr 42 000 and two trypsin- insensitive (intracellular precursor) polypeptides bearing the same epitope, Pk52 and Pk5O. Five monoclonal antibodies against surface anti- gens of P. knowlesi sporozoites cross-reacted with sporozoite antigens of P. cynomolgi but immuno- precipitated polypeptides of different relative mol- ecular mass; by analogy, these correspond to a surface antigen, Pc49, and two precursor polypep- tides, Pc58 and Pc56. a The kilodalton is not an S.I. unit. The relative molecular mass (M,) should be used instead (same figure x 103). Monoclonal antibodies to the surface antigens of sporozoites of P.falciparum (Gambia), as well as a polyclonal antiserum from an immunized volunteer, immunoprecipitated CS polypeptides (doublets) of Mr 67 000 and 58 000 from both homologous and heterologous (Thailand) sporozoites. Likewise, monoclonal antibodies to the CS protein of P. vivax immunoprecipitated polypeptide doublets, Pv5 1 and Pv45. The monoclonal antibodies to the CS proteins of P. vivax and P.falciparum sporozoites consider- ably decreased their infectivity for chimpanzees. That molecular relationships exist within this family of putatively analogous polypeptides is sup- ported by the results of peptide mapping after tryptic digestion of CS proteins. For example, peptide maps of Pk52 and Pk42 are very similar. Also, Pk52 and Pc58 showed some similarities although there was less between Pc58 and Pf67. The isoelectric point (pI) of the surface antigen is always less than that of the precursor. Considerable further progress can be anticipated from current work on the analysis of immunoprecipitated tryptic peptides bearing the species-specific epitopes of CS proteins. Antibodies to sporozoites-applications to im- munodiagnosis and epidemiology. A new immuno- radiometric test (the inhibition of idiotype-anti-idio- type interaction, or 4 i-assay) has been applied to the detection of the sporozoite CS protein (Pb44) by Potocnjak et al. (5). The principle of this assay is that antigen inhibits the interaction between two mono- clonal antibodies: the first monoclonal antibody is directed against the antigen and the second is directed against the idiotype of the first monoclonal. Both extracts of sporozoites of P. berghei and the purified protein Pb44 were shown to inhibit the idiotype-anti- idiotype interaction. Boiled extracts of salivary glands from individual mosquitos inhibited the inter- action (the epitope on Pb44 recognized by the first monoclonal is heat stable). In tests for species-speci- ficity, sporozoites of P. berghei were readily distin- guished from those of P. cynomolgi. The method is of general applicability and has the advantage that it measures the concentration of single epitopes (unlike most variants of sandwich or two-site immunoradio- metric assays). Furthermore, purified antigen is not required, and the two immunoglobulin reagents are homogeneous. Generating the second monoclonal antibody does not seem to be difficult and anti- idiotype antibodies which do not inhibit the reaction with antigen can be screened out by using the antigen inhibition binding assay itself. A simple assay employing monoclonal antibody to detect infected mosquitos has been developed for epidemiological purposes. The assay permits fairly rapid screening of large numbers of mosquitos, and the species of sporozoite can be determined readily. DEVELOPMENT OF MALARIA VACCINES The method employs dried or etherized individual mosquitos extracted in separate wells of a microtitre plate; after freezing and thawing, followed by heat treatment, sporozoite antigens are released and these adhere to the wells. After washing to remove extraneous material, the sporozoite antigen can be detected in a direct binding assay using labelled monoclonal antibody. The addition of species-speci- fic radiolabelled monoclonal antibodies to CS proteins distinguished P. berghei, P. cynomolgi, P. vivax and P.falciparum. The assay was effective at a minimum level of approximately 200 sporozoites. Exoerythrocytic stages of mammalian malaria parasites In 1981, the complete exoerythrocytic (EE) cycle of development of P. berghei was established in vitro (2). This significant advance allows study of the entry of sporozoites into target cells and the subsequent transformation into trophozoites, schizonts and EE merozoites. Within minutes after the introduction of sporozoites of P. berghei into cultures of W138 cells (a cell line from human embryonic lung), the elongated sporozoites attached to the cells, shed a "cloud" of CS protein (Pb44 antigen), and were engulfed within a prominent parasitophorous vacuole. Peroxidase-labelled antibody stained this vacuole strongly for Pb44. A bulge developed around the mid-region of intracellular sporozoites and the remainder of the sporozoite retracted into this area to form a small spherical trophozoite. During further development, complex cytoplasmic processes extended from the parasite or parasitophorous vacuole around the nucleus of the host cell. Irradiated sporozoites also attached to and entered cultured W138 cells and transformed into trophozoites but did not develop to schizonts. By contrast, sporozoites incubated with monoclonal antibody (or its Fab fragment) to Pb44 failed to attach and shed CS protein onto target cells; they also lost their infectivity for mice. This suggests that immune protection to sporozoites in mice is mediated by the binding of antibody to Pb44, thus preventing its interaction with target hepatic cells. The culture system provides an in vitro antibody assay: 16 Ag/ml anti-Pb44 inhibits invasion in vitro. Evidence from in vitro studies indicates that sporo- zoites infect the target cells of the host by first attaching to a receptor on the cell membrane and then penetrating within the cell by means of an active process involving movement between the parasite and host cell. Attachment requires a protein receptor, since treatment of W138 cells with proteolytic enzymes, such as chymotrypsin, reduced their suscep- tibility to infection whereas removal of sialic acid by neuraminidase had no effect. Penetration requires movement of host cell components, since inhibition of microtubule function by cytochalasin B permitted attachment and shedding of CS protein, but pre- vented penetration. Sporozoites heated at 42 °C were non-infective In sporozoite-induced P. berghei infection of mice, the parasites undergo only one cycle of development in the liver. It has also been found that in cultures of the EE stages of P. berghei, a single cycle of develop- ment occurred; the resulting EE merozoites produced erythrocyte infection in mice. EE merozoites of P. berghei still contain the large Mr precursor form of the Pb44 antigen. Attempts to cultivate in vitro the EE forms of simian and human malarias are in progress. Culti- vation of P. vivax is of particular importance because of the apparently crucial role of the liver stages in relapses. In cultures of monkey lung cells, sporo- zoites of P. knowlesi were observed to attach and release the CS protein. Sporozoites of P.falciparum interacted with cultures of W138 cells in vitro, released a "cloud" of CS protein, entered the cells, and underwent at least the initial stages of dedifferen- tiation. A monoclonal antibody to the CS protein of P.falciparum inhibited this interaction. So far, hepa- toma cell lines have proved not to be susceptible to sporozoite infection. From a practical point of view, it is important to determine whether the antigenic makeup of the EE merozoites of human malaria parasites allows them to escape from the immune response to asexual erythrocytic forms even for an initial round of erythrocyte invasion. That this may be so is suggested by the immunization study of Richards et al. (6). Blood stage antigens The complex array of antigens associated with asexual blood forms have been studied with a view to determining their stage-specificity and immunogeni- city. Pulse-labelling of asexual intraerythrocytic parasites indicates that most polypeptides are syn- thesized in a sequential manner during growth and differentiation. In several instances, it has been shown that protective immunity is directed against protein antigens synthesized at a late stage of the para- site's cell cycle; in particular, these are polypeptides of high relative molecular mass synthesized during the differentiation of schizonts and merozoites. Studies on the cellular location, species specificity, biosyn- thesis, processing and degradation, as well as on the purification and composition of these antigens are outlined below. Monoclonal antibody reagents have made possible many of the studies, including tests of the capacity of purified antigens to induce protective immunity. Minor differences in the degree of corres- pondence between some of the polypeptides identi- 83 MEMORANDUM fied in different studies are likely to be resolved by further proteochemical comparisons. Rodent malarias. The murine malaria parasite P. chabaudi provides a valuable model for the study of stage-specific antigens, since the infection is syn- chronous. Polypeptide synthesis has been studied in in vitro cultures of P. chabaudi and stage-specific polypeptide synthesis was demonstrated in both membrane-enriched and total lysate fractions of the parasites. Pulse-labelling with 35S-methionine indi- cated that the schizonts made a prominent protein of Mr 250 000 (i.e., it migrated above the spectrin bands on SDS-PAGE). There were several other stage- specific components, including histidine-rich proteins of Mr 80 000 and 30 000. The 250 000 Mr protein was accessible at the surface of metabolically labelled, intact, infected erythrocytes incubated with anti- serum from repeatedly infected mice. A similar but species-specific protein was demonstrated in reciprocal experiments with P. berghei. Metabolic labelling of the M2 250 000 protein with H-glucos- amine was not inhibited by tunicamycin, so the evi- dence for glycosylation is still circumstantial. Mono- clonal antibodies produced against P. chabaudi recognized an antigen of Mr 250 000 that co-migrated on SDS gels with the late stage-specific polypeptide. Immunofluorescence suggests that this antigen is on the surface of merozoites. Metabolically labelled parasites that were recovered after undergoing red cell invasion in vivo, retained many labelled polypeptides but not the Mr 250 000 protein. Two of eleven mono- clonal antibodies, directed against the Mr 250 000 component, passively protected infected mice and acted synergistically to cause delayed patency and suppression (but also protraction) of parasitaemia in a dose-response related manner. An important recent advance has been the induc- tion of protective immunity in mice using single purified parasite antigens. Using monoclonal anti- bodies, five antigens of P. yoelii were immunoprecipi- tated (the estimated Mr are 235 000, 230 000, 140 000, 66 000 and 54 000). Preliminary results indicate that the Mr 66 000 antigen is a soluble protein that is neither stage-specific nor protective. By contrast, both antigens with the highest relative molecular mass are protective and have been purified. The Mr 235 000 antigen (i.e., it migrates between spectrin bands) is probably located within the paired organelles of merozoites and may be secreted during invasion of host cells. From extracts of 5 x 1011 schizont-infected erythrocytes, 400 jig of protein were obtained after two cycles of affinity chromatography (see reference I for details). Serum from mice immunized with the antigen immunoprecipitated the same polypeptides as the monoclonal antibodies used in the purification procedure. Challenge infections in mice immunized with this antigen were mild, with the parasitaemia restricted to reticulocytes. The protein had a pl of 5.8-6.8, did not stain with para-aminosalicylic acid (PAS), and did not cross-react with P.falciparum. The Mr 230 000 antigen of P.yoelii is present (probably in a processed form) on the surface of free merozoites and also on residual bodies. It is a major antigen of P.yoelii (1600 ug protein were recovered from 5 x 10" schizont-infected red cells). The pl of the protein is 5.4-5.6 and again there is no evidence for glycosylation. Labelled antigen is processed in vivo into a series of fragments (Mr 230 000, 197 000, 160 000, 150 000, 90 000, 50 000). Mice immunized with the Mr 230 000 antigen, and fragments that co- select with it, strongly resisted challenge infection. Few parasites were seen either in reticulocytes or mature red cells. Presumably, interaction between the merozoite and the host cell was blocked. It is not known if the fragments with lower relative molecular mass are protective per se. This aspect and the chemical relationships between the fragments are being studied. Antiserum raised in mice against the Mr 230 000 protein cross-reacts weakly with P.falci- parum schizonts (indirect fluorescent antibody test), but does not immunoprecipitate polypeptides of P.falciparum. There is evidence that auto-immune responses may influence protective immunity in malaria. In an experimental model in which mice were immunized with rat red blood cells, the mice developed antibody to two determinants, one of which is rat specific and the other is shared by rat and mouse red cells. The anti-mouse red cell response was specifically sup- pressed on further injection of rat red cells. The adop- tive transfer of this suppressive cell population (spleen cells) conferred a degree of protection to recipient mice challenged with P. chabaudi. Further studies of the mechanisms involved are in progress. Simian malarias. Antigens which may prove to be important in protective immunity have been identi- fied in P. knowlesi merozoites. Monoclonal anti- bodies to P. knowlesi were raised by immunizing AO rats with merozoites (WI variant) and fusing the spleen cells with Y3 rat myeloma cells. Thirteen categories of immunofluorescent staining patterns were found with methanol-fixed schizonts and mero- zoites. Two monoclonals, purified from ascitic fluids, inhibited the multiplication of parasites in vitro. Both of these antibodies produced a diffuse fluorescence over the merozoite surface, but neither antibody cross-reacted with P.falciparum. The monoclonals did not damage parasites inside intact erythrocytes. No synergistic combinations of other, noninhibitory monoclonals were found. Both of the inhibitory anti- bodies immunoprecipitated a metabolically labelled (35S-methionine) polypeptide of Mr 66 000. Competi- 84 DEVELOPMENT OF MALARIA VACCINES tive binding studies showed that the two monoclonals reacted with the same or closely adjacent epitopes on the Mr 66 000 antigen, suggesting that there may be only one determinant on each antigen molecule. The two monoclonal antibodies seemed identical except that one inhibited parasite invasion of red cells more effectively than the other. Eight main polypeptides ranging downwards in size from Mr 150 000 have been demonstrated on the surface of intact merozoites of P. knowlesi by lacto- peroxidase-catalyzed radio-iodination (3). A mono- clonal antibody to one of these components (Mr 75 000) also reacts with a large polypeptide of approximately Mr 230 000 that migrates between the spectrin bands on SDS-PAGE. Both the Mr 75 000 and 230 000 polypeptides can be labelled metaboli- cally with 35S-methionine but surface labelling with 1251 reveals only the Mr 75 000 protein. Antibody to the Mr 75 000 protein reacts with the whole surface of merozoites and with residual bodies, but not red cells. A band corresponding to the Mr 230 000 protein stains with Coomassie Blue and it can also be labelled with 3H-glucosamine (the Mr 75 000 protein is not). To test the hypothesis that processing of the Mr 230 000 polypeptide yields the Mr 75 000 protein, syn- chronized schizonts at the 2 to 4 nuclei stage were pulse chased with 35S-methionine. As the schizonts matured in the first 2-3 hours of the experiment, there was no evidence that label was chased from the Mr 230 000 polypeptide to the Mr 75 000 surface protein. In another approach, mice were immunized with the Mr 150 000 trypsin-sensitive surface protein of merozoites to try to avoid the parasite's major surface immunogen (Mr 75 000/230 000) that might be liable to antigenic variation or other modulations. The Mr 150 000 protein was extracted after SDS-PAGE and incorporated into liposomes composed of deoxy- cholate and a mixture of lipids based on those of the parasite. The mouse antiserum was found to be monospecific to the Mr 150 000 protein and attempts will be made to obtain a monoclonal antibody. It was also reported that a monoclonal antibody (or its Fab fragment) to the band 3 protein of the erythro- cyte membrane blocked invasion by P. knowlesi. The relationship between the red cell membrane receptor for the parasite, band 3, and the Duffy blood group antigens still has to be clarified. A comparison was made of proteins synthesized by P. knowlesi and P.falciparum, either by intra- erythrocytic parasites or by translation of mRNAs in vitro. To minimize contamination of parasites with leukocytes or platelets, infected rhesus or squirrel monkeys were irradiated twice with 3 Gy (300 rads) in the course of the infection. Purified schizonts con- tained 0.01%Olo leukocytes. Messenger RNAs from P. knowlesi corresponding to parasite proteins of Mr 140 000, 102 000, 74 000 and 40 000 were translated in large amounts in lysates of rabbit reticulocytes. These parasite components, and an additional protein (Mr 230 000), were also synthesized by intraerythro- cytic parasites. It is possible that the mRNA encoding the Mr 230 000 protein has secondary structures (loops) which inhibit translation in vitro. Alter- natively, the polysomes may have disaggregated pre- maturely in the translation system. Subcellular fractionation of parasitized erythro- cytes showed that parasite proteins of high relative molecular mass were also associated with the host red cell membrane. Sera from immunized monkeys reacted somewhat variably with all the components with high relative molecular mass, but antibody to the Mr 74 000 protein correlated positively with protec- tion. The immune sera also immunoprecipitated the Mr 240 000, 102 000 and 74 000 components from the translation mixture. Peptide homology was shown for the Mr 74 000 protein, produced either by the parasite or translated in vitro. Vaccination of monkeys with the Mr 74 000 protein in Freund's adjuvant resulted in self-cure after chal- lenge with I04 parasites (but one animal became severely anaemic and had to be transfused). Serum collected from monkeys during the immunization procedure reacted only with the Mr 74 000 and 40 000 proteins at first, but later also with the Mr 140 000, 230 000 and other components. Homology between these proteins is being studied by peptide mapping. Performic acid treatment of the Mr 230 000 compo- nent yielded a product of Mr 74 000. It was also observed that the Mr 74 000 protein gave a sharp band in extracts from parasites but a diffuse band in ex- tracts of host cell membranes. The P.falciparum system gave poor levels of mRNA translation in vitro but a component of Mr 180 000 was found. Antibodies from squirrel mon- keys immune to P.falciparum cross-reacted with the Mr 140 000, 120 000, and 74 000 proteins of P. knowlesi translated in vitro. Plasmodium falciparum antigens. The antigens of P.falciparum asexual blood forms have been ana- lysed by SDS-PAGE and a comparison was made of antigens reacting with hyperimmune human sera and with sera from non-immune patients recovering from a first infection, in order to identify antigens of poten- tial importance in protection. Synchronized cultures of schizonts of P.falciparum produce characteristic antigens of Mr 200 000, 160 000, 140 000, 105 000, 82 000, 55 000 and 41 000. Antigens of Mr 200 000, 82 000 and 41 000 are accessible on the surface of schizont-infected red cells purified on Percoll- gradients; however, adsorption from culture fluids cannot be excluded, since proteins of Mr 200 000, 140 000, 82 000 and 50 000 are released during the invasion cycle. Serum from immune and nonimmune 85 MEMORANDUM adults preferentially immunoprecipitate different metabolically labelled antigens. Hyperimmune serum recognized proteins of Mr 140 000, 55 000 and 41 000. Monoclonal antibodies raised against three antigens of schizonts and merozoites (Mr 140 000, 82 000 and 41 000) inhibited parasite multiplication in vitro. Three isolates of P.falciparum from differ- ent geographical areas were all inhibited, suggesting that the epitopes are common rather than variant specific. An antibody against the Mr 200 000 protein was not inhibitory. The possible relationships between the Mr 140 000, 82 000 and 41 000 proteins may be resolved by peptide mapping. Immune Aotus monkey serum agglutinated meta- bolically labelled merozoites released from cultures of synchronous metrizamide purified schizonts. Lysis of agglutinates with neutral Vo Triton X-100 yielded a post-lysis supernatant and a post-lysis pellet. Certain merozoite- and schizont-specific antigens (Mr 155 000, 80 000, 75 000, 40 000 and 34 000) were concentrated in the pellet fraction of agglutinated merozoites and were less abundant or absent in the post-lysis pellet and the post-lysis supernatant of merozoites from a control culture containing normal Aotus monkey serum. These antigens, plus a Mr 200 000 antigen, could be solubilized by pH 2.7 Triton X-100 and reprecipitated with Pansorbin. Although the Mr 80 000, 75 000, 40 000 and 34 000 proteins could be detected in fluid from the control culture, the spent culture media were not depleted of free membranous vesicles which may carry these components. Four monoclonal antibodies which react with the Mr 200 000 protein on the periphery of merozoites did not inhibit multiplication in vitro. Merozoites were isolated by either free-flow electrophoresis or by a sieving procedure less than one hour after spontaneous release from schizonts. The merozoites had excellent morphological integrity and were of high purity (acetyl cholinesterase was used to measure contamination with erythrocyte mem- branes). Extracts from metabolically labelled mero- zoites (methionine, isoleucine and glucosamine) con- tained a protein of high Mr (approximately 200 000) that migrated in the region of spectrin on SDS- PAGE. This protein was not labelled with 3H-glucosamine and was not labelled by lacto- peroxidase-catalysed radioiodination of the surface of free merozoites; surface labelling revealed compo- nents of only Mr 140 000, 80/90 000, 50 000 and 40 000. The polypeptide of high Mr (200 000) is there- fore likely to be intracellular. To fix merozoites for surface labelling with immuno-reagents, merozoites or late-stage schizonts should be washed free of serum proteins and fixed for the briefest possible time (5-10 minutes) at room tem- perature. This brief fixation period includes suspend- ing the cells in 0.0750Go glutaraldehyde (free of reac- tive groups) and centrifugation to remove the fixing solution. The cells are then incubated in buffer con- taining 1% glycine for 30 mins and washed several times to remove the glutaraldehyde. All reagents are Millipore-filtered beforehand. The histidine-rich protein of P. lophurae has pre- viously been reported to elicit protective immunity in immunized ducks. In recent studies, comparison has been made of the histidine-rich granules and the knobs that appear on the surface of erythrocytes infected with P.falciparum. The knob protein (KP) was found to contain approximately 10%/o histidine. It is synthesized early in the parasite's cell cycle and is located on the cytosol side of the erythrocyte mem- brane (it cannot be iodinated from outside). KP may, however, be shed by vesiculation of the infected red cell's membrane. Antibody to the histidine-rich pro- tein (HRP) of P. Iophurae immunoprecipitated the KP, and rabbit antisera, raised with a membrane- enriched fraction of P.falciparum, immunoprecipi- tated HRP. Both HRP and KP are insoluble in Nonidet 40 (NP-40) but soluble in sodium dodecyl sulfate (SDS). High Mr aggregates of HRP and com- ponents of Mr 80 000 and 40 000, as well as degra- dation products, were resolved by SDS-PAGE. Both HRP and KP gave a band migrating at Mr 80 000, but the rabbit antiserum to P.falciparum membranes immunoprecipitated the Mr 40 000 band only. Sera from some individuals infected with P.falciparum in Africa or Viet Nam reacted with HRP but not with the Mr 80 000 protein. Antibody to HRP will be used to purify KP by means of affinity chromatography and it will be important to determine whether it induces protective immunity. Studies of the species and strain specificity of P.falciparum antigens have been made using mono- clonal antibodies raised to parasite isolates of different geographic origins. Monoclonal antibodies to the FVO (Viet Nam) or FCB (Columbia) strains of P.falciparum reacted with structures within indivi- dual merozoites, with the surface coat of merozoites, and with the red cell side of the parasitophorous vacuole membrane. The antibodies did not react with other species of plasmodia except a few which reacted with P. lophurae. Most of the antibodies reacted with all the strains of P.falciparum tested, but some did not react with a strain from Africa (Uganda) or with one from Honduras, or both. The strain-specificity of these monoclonal antibodies was related to their reac- tivity with a surface antigen on free merozoites as well as with the parasitophorous vacuole membrane. Studies on induced malaria in man suggest that P.falciparum is composed of races or strains with distinct characteristics. S-antigens associated with the asexual phase behave like strain-specific antigens in the following respects. Clones of parasites (recog- nized by isoenzyme type) produce characteristic indi- 86 DEVELOPMENT OF MALARIA VACCINES vidual S-antigens. The S-antigen serotypes of clones remain constant during continuous cultivation in vitro and after transfer to Aotus monkeys. Mixtures of parasites that produce different S-antigens yield predictable mixtures of S-antigens when grown in red cells from a single source. Panels of patients' sera contain a wide variety of S-antigens and mixtures of 2 or 3 antigens are common in individual patients. Individuals usually present with different S-antigens during clinical attacks in endemic areas; however, the same S-antigen can sometimes recur in different attacks. In immune adults, the spectrum of antibodies to S-antigens changes irregularly owing to reinfection presumably. Antibodies to S-antigens from West Africa are readily found in endemic areas ranging from Brazil to Papua New Guinea. Heterogeneity between isolates has also been shown by others working with isozymes, proteins, or monoclonal antibodies. This heterogeneity should be borne in mind when seeking to explain immunity to the asexual forms in terms of the structure and function of particular macromolecules. Gametes Antibodies to gametes confer transmission- blocking immunity by preventing fertilization within the mosquito stomach. The antigens involved in this form of immunity are being investigated, since a transmission-blocking vaccine could play an impor- tant role in malaria control. Antigamete antibodies, taken up with a blood meal, act inside the mosquito midgut to neutralize gametes after they emerge from red cells; antibodies to microgametes block fertilization and antibodies to ookinetes prevent zygote transformation. In the presence of complement there is lysis; without complement, agglutination occurs. Rabbit antisera were raised to purified micro- and macrogametes and zygotes of P. gallinaceum. Surface antigens of the purified stages were radiolabelled with 125I by the lactoperoxidase method. About 15 labelled protein bands ranging from Mr 230 000 to 40 000 were extracted from the surface of both macrogametes and zygotes. Microgametes gave only three bands (Mr 230 000, 210 000 and 55 000); only the Mr 210 000 protein was present on micro- and not on macrogametes. Antibody to microgametes blocked transmission. A major band from ookinetes (Mr 30 000) might be the target for the transmission block. Monoclonal antibodies (IgG) to surface antigens of gametes and zygotes, even those shown to react with the Mr 230 000 and 55 000 iodinated proteins, failed to block transmission. Several monoclonals of IgM isotype did not immunoprecipitate radioiodinated material but one reduced the infectivity to mosquitos by 9901o. It is conceivable that IgM antibody of low avidity might fail to retain its target antigen during solubilization in the first stage of extraction in the immunoprecipitation procedure. Bands labelled with 3H-glucosamine were not found. In two instances, mixtures of ineffective antibodies of IgG and IgM isotypes acted synergistically to suppress infectivity. The target antigens of transmission-blocking immu- nity remain to be identified. REVIEW OF SOME RECENT ADVANCES IN PROTEIN CHEMISTRY AND RECOMBINANT DNA TECHNOLOGY A section of the meeting was devoted to a review of recent technical developments which are of current or potential relevance to the analysis and production of malaria antigens. Cell and macromolecule separation techniques The separation of parasitized cells from other cells and from extraneous material is a prerequisite for many of the studies described earlier in this report. In addition, many parasite antigens are macromol- ecules. Recent developments in the field of cell and macromolecule separation were therefore reviewed. Cell separation may be carried out by density gradient centrifugation. This method involves a physical interaction depending on a general property of the whole cell, in this case its density. The density of any particular type of cell correlates with its degree of differentation, position in the cell cycle, and osmo- tic environment. Cells may also be separated by affinity chromatography. This is based on a biospecific interaction, involving specific chemically distinct sites on the cell surface, such as the binding to a ligand via a cell surface receptor. Separation in this case depends on a functional property of the cell. Three recent developments for purifying macro- molecules are noteworthy: (i) chromatofocusing gives a high resolution according to the isoelectric point; (ii) affinity chromatography has the ability to concentrate a minor component from a large volume of starting material. The use of monoclonal anti- bodies as ligands in affinity chromatography is now boosting interest in this technique, which allows iso- lation of significant amount of antigens from complex mixtures; (iii) fast protein liquid chrom- atography, performed on a new range of separation particles, gives high resolution separation of proteins and macromolecules in 20-60 minutes. Such rapid separation facilitates the optimization of separation schemes. Microtechniques In many instances, studies of malaria antigens are impeded by the extremely limited quantities of ma- 87 MEMORANDUM terial available for analysis. For this reason, micro- techniques developed for protein chemistry were reviewed. Spot analysis can yield a lot of information from small amounts of material by the ingenious applica- tion of staining techniques (protein stains, fluoro- chromes, labelled lectins, etc.) and densitometry. Microdisc gel electrophoresis in homogeneous or gradient gels, in capillaries of 10-1l volume have been used to determine relative molecular masses, isoelec- tric points and the kinetics of isoenzyme/substrate reactions. Slab gels on a microscale (3.0 cm2) permit both single and double dimension separations of proteins, peptides or nucleic acids and can be quantitated photometrically at the microscopic level. Details of these techniques are given by Neuhoff (4) who is preparing a further publication on this subject; see also reference 7. Analysis ofprotein structure When suitable antigens for malaria vaccines have been identified, they will have to be produced in large quantities. Whether this is eventually done by chemi- cal synthesis or by recombinant DNA technology, it will be necessary first to analyse the amino acid structure of the antigens concerned. Current methodology for protein sequencing was therefore discussed. The identification of a protein by its amino acid sequence is a powerful tool for comparative purposes or for identifying regions of special interest. Improve- ments have been made in automated methods of se- quencing and of identifying the amino acid residues by high pressure liquid chromatography. Significant achievements-concerning, in particular, membrane antigens present in small amounts on cell surfaces- have been made using microsequencing, following biosynthetic radiolabelling of amino acids. Sequences from a native functional protein sometimes provide information not available from a synthesized protein because of post-translation modifications (see below: expression of cloned genes in host cells). The initial strategy of analysis for a large protein includes separation of peptide chains, determination of relative molecular masses and N-terminals, proteo- lytic cleavage at high yield, and sequencing and deter- mination of primary structure; this strategy has been successfully applied to the analysis of the structure of fibrinogen. Knowledge of at least partial sequences of protein is extremely useful, if not indispensable, for rapid progress in the isolation or the identification of the genes coding for it. The synthesis of antigenic peptides Structural information is essential as a starting point for the synthetic approach. Such information is not yet available for any malarial antigen. However, experiments employing certain other antigens indicate what can be achieved. Synthetic peptide antigens capable of eliciting specific anti-protein immune responses have been demonstrated with the "loop" region of lysozyme, the N-terminal region of carcinoembryonic antigen, and a fragment of the coat protein of coliphage MS-2. A conjugate of this last- mentioned peptide with a synthetic carrier was attached to a synthetic adjuvant (N-acetyl muramyl dipeptide) to yield a completely synthetic molecule that induced a high neutralizing response. Anti-toxic immunization against diphtheria toxin has also been achieved with a synthetic tetradecapeptide linked to a protein carrier. Other examples where part of an anti- genic molecule has been synthesized and used to elicit protective immune responses are hepatitis B (virus envelope) and influenza (haemagglutinin). Synthesis of polypeptides can only be achieved with prior knowledge of the amino acid sequence and, in general, known sequences of approximately 20 amino acids are required. It is possible that genetic engineering may provide larger chains. Epitopes of interest could be identified by monoclonal antibodies. The application of recombinant DNA techniques Since it is possible that malaria antigens for use in vaccines may be produced by means of recombinant DNA technology, the meeting included a general review of the basic principles and techniques involved and discussion of the current status of research on malaria antigens in this field. Basic principles. The basic steps of genetic engin- eering are technically simple. Fragments of DNA are made and joined to the DNA of a vector, allowing them to be biologically replicated in bacterial or eukaryotic cells. Ligation of the DNA molecules is accomplished through extended cohesive ends ex- posed after cleavage with a restriction endonuclease; alternatively, a short "tail" of poly A, for example, can be added to one molecule and a complementary tail of poly T to the other (homo-polymeric tails); otherwise, artificial linkers containing restriction sites can be attached to the foreign fragments to allow insertion into the vector. Once joined, the DNA mol- ecules are introduced into bacterial or eukaryotic cells by transformation or transfection to give derivatives which propagate the new recombinants in a stable manner. There are several types of vector to choose from: plasmids that are small, independently replicating units, commonly found in bacteria; and phages such as those that can be adapted for use as vectors. In general, it is possible to insert more DNA (up to 16 kilobases) into phages than into plasmids. Cosmids combine features of phages and plasmids; they are 88 DEVELOPMENT OF MALARIA VACCINES designed to carry larger inserted fragments (40-50 kilobases). Cloning of genomic genes and of synthetic genes. Fragments of DNA, from a foreign genome may be inserted directly into an appropriate vector. This allows the direct analysis of large fragments (up to 50 kilobases) or the study of the regulation of gene expression. If Plasmodium genes are discontinuous (containing introns like many other eukaryotic genes), such clones will not be expressed in bacteria. To solve this problem, RNA transcripts (mRNA) from the Plasmodium are converted to cDNA in vitro and inserted into the vector (usually a plasmid). The mRNA can be a single species with one protein product, or a mixture. This is the method of choice for the expression and production of eukaryotic pro- teins in bacterial cells. Molecular cloning ofspecific genes. The bottleneck in most cloning strategies is the identification of clones of interest. There are four main approaches: (a) the clone can be sequenced; (b) immobilized DNA from the clone can be used selectively to hybridize complementary mRNA that is then eluted and identified by its translation product in vitro (though elaborate, this is a good procedure for detecting rare clones); (c) determination of a partial amino acid sequence from the protein of interest allows chemical synthesis of a short sequence-specific nucleic acid probe (see below: special problems in DNA technology) that can be used as a primer of cDNA synthesis or directly as a radioactive probe for screening bacterial plasmids; (d) direct identification of expressed protein either by a sensitive assay system such as radioimmunoassay or enzyme-linked im- munosorbent assay (such a system requires that a large number of clones be screened), or by a biological assay such as complementation with a selectable product (this requires compatibility in the genetic system of the host and of the donor DNA). Expression of cloned genes in host cells. A gene cloned in a bacterial plasmid can be expressed if the signals regulating RNA and protein synthesis are present and functional. The product of transcription and translation of a foreign gene inserted within a plasmid gene is a hybrid or fused protein. Partial expression of a foreign gene may yield a product that has lost its primary biological activity (e.g., in the case of an enzyme), yet retains an immunogenic portion. Elucidation of the nucleotide and hence the amino acid sequence may make the organic synthesis of anti- genic peptides possible. Alternatively, the physical relationship between the vector and inserted genes may be manipulated to obtain "perfect construction" that permits more faithful transcription of the foreign gene product. Expression of cloned genes has been achieved in eukaryotic cells. This is important when a glyco- sylated protein product is the objective, or to study the expression of DNA fragments that have already been identified. A selection procedure using trans- formation of thymidine kinase deficient (TK-) mouse cells has been devised. Yeast may be unsuitable for the general cloning of eukaryotic genes but progress is being made in the development of new eukaryotic host cell systems and new vectors. Special problems in DNA technology. Chemical synthesis of DNA is another way to provide radio- active probes for screening clones of interest. When the amino acid sequence of the gene product, or a part of it, is known, a short oligonucleotide sequence of these codons is synthesized, labelled and used as a probe. The commonest strategy for the synthesis of short oligonucleotides is to build the oligonucleotide on a solid support such as polyacrylamide. Protected nucleotides are attached to the support. The protec- tive group is then removed and a new mono-, di- or tri- nucleotide presented to the system. Step-wise addition of this kind extends the oligomer and oligo- nucleotides can be generated up to 17 bases long. With the use of high pressure liquid chromatography the synthesis is fast (2 days) and has high yields, but it is commercially very expensive. Yet larger primers can be made by joining the 17 base oligomers in tandem using a bridging molecule and DNA ligase. Discussion on molecular cloning of genes from Plasmodium Present achievements. Two major approaches are being adopted to study the genes of Plasmodium. The first is to take fragments of the genomic DNA and ligate them into a prokaryotic vector-a phage or plasmid, which permits the bacteria to replicate the fragment. The collections of fragments that have thus been obtained have been screened for specific sequen- ces and genes. The following cloned fragments have been found using 32P-labelled probes of known sequence: repetitive DNA sequences from the genome of P.falciparum (differences were found between two isolates); the genes for ribosomal RNA of P.falcipa- rum and P. yoelii; sequences homologous to the actin gene of Dictyostelium and the chicken. An intensive search for cloned Plasmodium genes that are able to express and complement mutations in E. coli for five enzymes (thymidylate synthetase, dihydroorotate dehydrogenase, aspartate carbamylase, phospho- enolpyruvate carboxylase, and dihydrofolate reductase-the target for pyrimethamine) has not yielded a positive result. This strongly suggests that functional expression of the genomic DNA of Plas- modium encounters problems in the bacterial 89 MEMORANDUM cytoplasm. Perhaps it cannot be transcribed or translated. Alternatively, protein products may not be processed or may be broken down in the bacte- rium. The proportion of G and C nucleotides in the total DNA of Plasmodium is abnormally low (approxi- mately 1807, compared with approximately 500%o in many other organisms). Recent work suggests that the previously reported base composition of P. knowlesi DNA, (G + C, approximately 370/o), is likely to be due to host cell contamination and that another com- ponent with a G + C content of approximately 19%7o reflects the true value. Various possibilities were raised to explain the low G + C content of plasmodial DNA such as non-coding sequences with a high A + T content, a modified genetic code, enrichment of plas- modial proteins with amino acids such as proline and glycine; and it was suggested that perhaps it will be difficult for Plasmodium genes to be expressed in other organisms. Nevertheless, there is a possibility that the part of Plasmodium DNA that codes for proteins has a higher G + C content than the average. This is certainly true for the recombinant RNA (rRNA) genes which, although they do not code for a protein, contain 400o G + C. In the second approach, mRNA has been extracted from mature asexual erythrocytic stages of P. know- lesi and P. falciparum and made into cDNA. The cDNA has been cloned in plasmids to create libraries of fragments corresponding to genes expressed in the erythrocytic stages of the Plasmodium. Screening procedures are required to find clones containing the whole or part of a gene of interest. A promising approach is to divide the total mRNA into single frac- tions (for example, by electrophoretic fractionation on a methyl mercury gel). mRNAs of different size correspond to precursor proteins of comparable size and can themselves be used as probes by converting them to cDNA labelled with 32p. Such probes will hybridize to clones coding for proteins in a certain size range. Amongst these, it is possible to seek a clone coding for a specified polypeptide by means of a technique called hybrid selection: each DNA clone is fixed to paper and total mRNA is allowed to hybridize to it. Only the homologous message will stick, while the rest is washed off leaving a messenger species whose protein product can be identified in a cell-free translation system. Polypeptides coresponding to mRNA species of Mr 100 000 and 74 000 have been translated. This approach allows a cloned gene or gene fragment to be detected although it cannot be expressed. Genes expressed in the sporozoite stage are now being investigated. Although limited by very small quantities of parasites, total mRNA has been trans- lated from infected salivary glands, and the high relative molecular mass precursor of the CS protein of P. knowlesi sporozoites has been detected by im- munoprecipitation with monoclonal antibodies (see above: sporozoite antigens). Current problems in the genetic manipulation of Plasmodium. An important surface protein of many Plasmodium spp. is a very large protein of up to Mr 250 000. In order to recognize clones containing all or part of this gene, it is necessary to make the cell-free translation system (from rabbit reticulocytes) synthesize big proteins. Many investigators, however, find it hard to make proteins larger than Mr 150 000. A partial solution to this problem is to supplement the translation systems with transfer RNA (tRNA) from Plasmodium or other sources. This interesting observation would seem to indicate that the coding pattern of Plasmodium may be unusual. Perhaps because its DNA has an exceptionally low proportion of G + C in the overall base composition, the parasite commonly uses codons for which there is very little tRNA present in the translation extracts. It was pointed out that success in translating large polypep- tides (Mr 230 000 and 235 000) of P. yoelii may be partly due to the abundance of their messengers. It was stressed that the Mr 230 000 band in P. knowlesi is also a major protein as judged by the intensity of protein staining. Monoclonal antibodies may not be able to detect a protein made in vitro because generation of the epi- tope requires modification of the primary polypep- tide chain. This problem might be resolved by mixing monoclonal antibodies directed against the same antigen or by raising a polyclonal antiserum against native antigen first purified by the monoclonal antibody. This approach has been used to purify large amounts of the two protective antigens (Mr 230 000 and 235 000) of P. yoelii. To use this technique on a preparative scale, the optimum binding and elution conditions should be defined for each monoclonal antibody. It is also essential to optimize the elution conditions (pH or concentration of chaotropic ion) so as to avoid denaturation of the parasite protein and loss of its immunogenicity. It was suggested that leakage of mouse IgG from immunosorbent columns might be minimized by careful washing after coupling of antibody, or the use of an epoxy-link to the substrate to obviate hydrolysis during elution. CONCLUSIONS Important progress has been made recently on the identification of protective malaria antigens. It can be expected that the application of monoclonal anti- bodies to antigen purification techniques will lead to the analysis of the fine structure of several stage- specific protective antigens. Recombinant DNA tech- 90 DEVELOPMENT OF MALARIA VACCINES nology will make it possible to analyse the structure, organization, and expression of Plasmodium genes. Many of the studies reported at this meeting focused on antibody-dependent effector mechanisms which are related to antigens on the surface of extra- cellular forms of the malaria parasite. It should be emphasized that various other protective effector mechanisms are likely to operate as well. It will be important to foster further research into the basic biology of the parasite, since this is likely to produce new leads relevant to the orientation and the prospects of malaria vaccine development. * * R. Arnon, Immunochemical Department, The Weiz- man Institute of Science, Rehovot, Israel S. Bartlett, Pharmacia Fine Chemicals AB, Upp- sala, Sweden R. L. Beaudoin, US Navy Malaria Vaccine Develop- ment Program, Naval Medical Research Institute, National Naval Medical Center, Bethesda, MD, USA G. Campbell, Division of Tropical and Geographic Medicine, University of New Mexico School of Medicine, Albuquerque, NM, USA R. Carter, Malaria Section, Laboratory of Parasitic Diseases, National Institute of Allergy and Infec- tious Diseases, National Institute of Health, Bethesda, MD, USA S. Cohen, Department of Chemical Pathology, Guy's Hospital Medical School, London, England W. E. Collins, Bureau of Tropical Diseases, Depart- ment of Health and Human Services, Centers for Disease Control, Atlanta, GA, USA G. A. M. Cross, Department of Immunochemistry, Wellcome Research Laboratories, Beckenham, Kent, England J. A. Deans, Department of Chemical Pathology, Guy's Hospital Medical School, London, England C. L. Diggs, Department of Immunology, Walter Reed Army Institute of Research, Walter Reed Army Medical Center, Washington, DC, USA M. Edge, ICI Pharmaceutical Division, Mereside, Macclesfield, Cheshire, England B. Enders, Parasitology Division, Behringwerke AG, Marburg, Federal Republic of Germany R. Freeman, Department of Immunochemistry, Well- come Research Laboratories, Beckenham, Kent, England A. Henschen-Edman, Max-Planck Institute of Bio- chemistry, Munich, Federal Republic of Germany M. R. Hollingdale, Biomedical Research Institute, Rockville, MD, USA A. Kilejian, Laboratory of Parasitology, The Rocke- feller University, New York, NY, USA F. M. Kourilsky, Centre d'Immunologie de Mar- seille-Luminy, Marseilles, France S. Langreth, Uniformed Services University of the Health Sciences, School of Medicine, Bethesda, MD, USA B. Mach, Department of Microbiology, University of Geneva, Geneva, Switzerland I. A. McGregor, Department of Tropical Medicine, Liverpool School of Tropical Medicine, Liver- pool, England L. H. Miller, Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Dis- eases, National Institutes of Health, Bethesda, MD, USA P. Muller, Max Planck Institute for Biochemistry, Munich, Federal Republic of Germany V. Neuhof, Max Planck Institute for Experimental Medicine, Gottingen, Federal Republic of Germany C. Newbold, Division of Parasitology, National Institute for Medical Research, London, England R. S. Nussenzweig, Division of Parasitology, Department of Preventive Medicine, School of Medicine, New York University Medical Center, New York, NY, USA V. Nussenzweig, New York University Medical Center, New York, NY, USA F. Perler, Biolabs, Beverly, MA, USA P. Perlman, Department of Immunology, Univer- sity of Stockholm, The Wenner-Gren Institute, Stockholm, Sweden L. Perrin, Division of Haematology, Hopital Cantonal, University of Geneva, Geneva, Switzer- land R. T. Reese, Scripps Clinic and Research Foun- dation, La Jolla, CA, USA P. K. Russell, Walter Reed Army Institute of Research, Walter Reed Army Medical Center, Washington, DC, USA J. Scaife, Department of Molecular Biology, Uni- versity of Edinburgh, Edinburgh, Scotland R. Schmidt-Ullrich, Tuft's New England Medical Center, Boston, MA, USA A. Vaidya, Department of Microbiology and Im- munology, The Hahnemann Medical College, Philadelphia, PA, USA 91 92 MEMORANDUM R. J. M. Wilson, MRC National Institute of Medical Research, Mill Hill, London, England J. Erickson, Bureau of Science and Technology, Office of Health, United States International Development Cooperation Agency, Agency for International Development, Washington, DC, USA WHO Secretariat P.-H. Lambert, Immunology, World Health Organ- ization, Geneva, Switzerland L. Martinez, Malaria Action Programme, World Health Organization, Geneva, Switzerland G. Torrigiani, Immunology, World Health Organ- ization, Geneva, Switzerland P. I. Trigg, Malaria Action Programme, World Health Organization, Geneva, Switzerland W. H. Wernsdorfer, Research and Technical Intelligence, Malaria Action Programme, World Health Organization, Geneva, Switzerland Observers D. C. Comb, Biolabs, Beverly, MA, USA H. Heidrich, Max-Planck Institute of Biochemistry, Munich, Federal Republic of Germany ACKNOWLEDGEMENT The assistance of Dr. R. J. M. Wilson in the preparation of this report is gratefully acknowledged. REFERENCES 1. HOLDER, A. A. & FREEMAN, R. R., Nature (London), 294: 361-364 (1981). 2. HOLLINGDALE, M. R. ET AL. Science, 213: 1021-1022 (1981). 3. JOHNSON, J. G. ET AL. Journal of protozoology, 28: 160-164 (1981). 4. NEUHOFF, V. Micromethods in molecular biology. Berlin, Heidelberg & New York, Springer-Verlag, 1973. 5. POTOCNJAK, P. ET AL. Science, 215: 1637-1639 (1982). 6. RICHARDS, W. H. G. ET AL. Parasitology, 74: 191-198 (1977). 7. UNDP/WORLD BANK/WHO SPECIAL PROGRAMME FOR RESEARCH AND TRAINING IN TROPICAL DISEASES. Application of biochemical micromethods for the investigation of tropical disease pathogens. Geneva, World Health Organization, 1982.