A new primaquine analogue, tafenoquine (WR 238605), for prophylaxis against Plasmodium falciparum malaria.

We tested tafenoquine (WR 238605), a new long-acting 8-aminoquinoline, for its ability to prevent malaria in an area that is holoendemic for Plasmodium falciparum. In a double-blinded, placebo-controlled, randomized clinical trial in western Kenya, adult volunteers received a treatment course of 250 mg halofantrine per day for 3 days, to effect clearance of preexisting parasites. The volunteers were then assigned to 1 of 4 drug regimens: placebo throughout; 3 days of 400 mg (base) of tafenoquine per day, followed by placebo weekly; 3 days of 200 mg of tafenoquine per day, followed by 200 mg per week; and 3 days of 400 mg of tafenoquine per day, followed by 400 mg per week. Prophylaxis was continued for up to 13 weeks. Of the evaluable subjects (223 of 249 randomized subjects), volunteers who received 400 mg tafenoquine for only 3 days had a protective efficacy of 68% (95% confidence interval [CI], 53%-79%), as compared with placebo recipients; those who received 200 mg per day for 3 days followed by 200 mg per week had a protective efficacy of 86% (95% CI, 73%-93%); and those who received 400 mg for 3 days followed by 400 mg per week had a protective efficacy of 89% (95% CI, 77%-95%). A similar number of volunteers in the 4 treatment groups reported adverse events. Prophylactic regimens of 200 mg or 400 mg of tafenoquine, taken weekly for < or =13 weeks, are highly efficacious in preventing falciparum malaria and are well tolerated.

Efficacy and safety of atovaquone/proguanil as suppressive prophylaxis for Plasmodium falciparum malaria.

Currently recommended prophylactic regimens for Plasmodium falciparum malaria are associated with a high incidence of adverse events and/or suboptimal efficacy. In a double-blind, placebo-controlled, randomized clinical trial in western Kenya, adult volunteers received a treatment course of atovaquone/proguanil hydrochloride (250 mg/100 mg per tablet) to eliminate preexisting infection. Immediately thereafter, subjects were randomized to one of the three prophylactic regimens to receive one atovaquone/proguanil tablet daily (n = 68), two atovaquone/proguanil tablets daily (n = 65), or placebo (n = 65) for 10 weeks. The study endpoint for any subject was the development of parasitemia, evident on blood smear, during prophylaxis. Of the evaluable subjects, all in the low-dose (54 of 54) and high-dose (54 of 54) atovaquone/proguanil groups remained malaria-free during the 10-week prophylaxis period, in contrast to only 48% (26 of 54) in the placebo group (P < .001). Both atovaquone/proguanil prophylactic regimens were as well tolerated as placebo. Thus, atovaquone/proguanil appears to be highly efficacious and safe as prophylaxis for P. falciparum malaria.

Cross-reactive epitope among proteins in Plasmodium falciparum Maurer's clefts and primate leukocytes and platelets.

Erythrocytes infected with malaria parasites often contain membranous vesicles that are presumed to facilitate macromolecule traffic between the parasite and erythrocyte membranes. One such vesicle network, called Maurer's clefts, is expressed in Plasmodium falciparum-infected erythrocytes and contains a 50-kD polypeptide. Using a monoclonal antibody reactive with this polypeptide, we show that hepatic stages of P. falciparum express an epitope common to this blood-stage antigen. In addition, this epitope is cross-reactive with antigens expressed by primate leukocytes and platelets. Such epitopes may induce autoantibodies commonly seen in patients with malaria.

Co-localization of inducible-nitric oxide synthase and Plasmodium berghei in hepatocytes from rats immunized with irradiated sporozoites.

Both CD8+ T cells and IFN-gamma (IFN-gamma) are important components in the regulation of inducible-nitric oxide synthase (iNOS) which contribute to liver stage anti-malarial activity in rodents immunized with irradiated sporozoites. IFN-gamma, provided by malaria-specific CD8+ T cells, stimulates liver cells to produce nitric oxide (NO) for the destruction of infected hepatocytes or the parasite within these cells. To identify the cell source of iNOS in livers from Brown Norway rats challenged with Plasmodium berghei sporozoites, we probed tissue sections with antisera that recognize iNOS and the malarial exoerythrocytic stage parasite. Immunofluorescence analysis of parasitized livers demonstrate that 1) iNOS was found in infected hepatocytes, not Kupffer or endothelial cells; and 2) a higher proportion of infected hepatocytes express iNOS in immunized rats compared with naive animals after challenge. There was no immunoreactivity to the iNOS antisera in liver sections of immunized rats 15 h after sporozoite challenge, however, iNOS activity was present in 18% of the infected hepatocytes by 24 h and reached 81% by 31 h. In contrast, < 10% of the infected hepatocytes displayed iNOS activity in naive or immune animals 48 h after challenge. We also found a significant decrease in the ability of the immunized animals to express iNOS in response to sporozoite challenge by accelerating the removal of pre-existing irradiated-attenuated parasites from hepatocytes with the antimalarial drug, primaquine. Therefore, induction and maintenance of iNOS activity were dependent on intrahepatic persistence of the irradiated-attenuated parasite. These results suggest that liver-iNOS expression following sporozoite challenge is restricted to the infected hepatocyte and dependent on the presence of the irradiated-attenuated parasite in immune animals.

Plasmodium falciparum protein associated with the invasion junction contains a conserved oxidoreductase domain.

The merozoite cap protein-1 (MCP-1) of Plasmodium falciparum follows the distribution of the moving junction during invasion of erythrocytes. We have cloned the gene encoding this protein from a cDNA library using a monoclonal antibody. The protein lacks a signal sequence and has no predicted transmembrane domains; none of the antisera reacts with the surfaces of intact merozoites, indicating that the cap distribution is submembranous. MCP-1 is divided into three domains. The N-terminal domain includes a 52-amino-acid region that is highly conserved in a large family of bacterial and eukaryotic proteins. Based on the known functions of two proteins of this family and the pattern of amino acid conservation, it is predicted that this domain may possess oxido-reductase activity, since the active cysteine residue of this domain is invariant in all proteins of the family. The other two domains of MCP-1 are not found in any other members of this protein family and may reflect the specific function of MCP-1 in invasion. The middle domain is negatively charged and enriched in glutamate; the C-terminal domain is positively charged and enriched in lysine. By virtue of its positive charge, the C-terminal domain resembles domains in some cytoskeleton-associated proteins and may mediate the interaction of MCP-1 with cytoskeleton in Plasmodium.

Nitric oxide: cytokine-regulation of nitric oxide in host resistance to intracellular pathogens.

To discover how nitric oxide (NO) synthesis is controlled in different tissues as cells within these tissues combat intracellular pathogens, we examined three distinctively different experimental murine models designed for studying parasite-host interactions: macrophage killing of Leishmania major; nonspecific protection against tularemia (Francisella tularensis) by Mycobacterium bovis (BCG); and specific vaccine-induced protection against hepatic malaria with Plasmodium berghei. Each model parasite and host system provides information on the source and role of NO during infection and the factors that induce or inhibit its production. The in vitro assay for macrophage antimicrobial activity against L. major identified cytokines involved in regulating NO-mediated killing of this intracellular protozoan. L. major induced the production of two competing cytokines in infected macrophages: (1) the parasite activated the gene for tumor necrosis factor (TNF), and production of TNF protein was enhanced by the presence of interferon-gamma (IFN-gamma). TNF then acted as a autocrine signal to amplify IFN-gamma-induced production of NO; and (2) the parasite upregulated production of transforming growth factor-beta (TGF-beta), which blocked IFN-gamma-induced production of NO. Whether parasite-induced TNF (parasite destruction) or TGF-beta (parasite survival) prevailed depended upon the presence and quantity of IFN-gamma at the time of infection. The relationship between NO production in vivo and host resistance to infection was demonstrated with M. bovis (BCG).(ABSTRACT TRUNCATED AT 250 WORDS)

Induction of nitric oxide synthase protects against malaria in mice exposed to irradiated Plasmodium berghei infected mosquitoes: involvement of interferon gamma and CD8+ T cells.

Exposure of BALB/c mice to mosquitoes infected with irradiated Plasmodium berghei confers protective immunity against subsequent sporozoite challenge. Immunized mice challenged with viable sporozoites develop parasitemia when treated orally with substrate inhibitors of nitric oxide synthase (NOS). This suggests that the production of nitric oxide (NO) prevents the development of exoerythrocytic stages of malaria in liver. Liver tissue from immunized mice expressed maximal levels of mRNA for inducible NOS (iNOS) between 12 and 24 h after challenge with sporozoites. Intraperitoneal injection of neutralizing monoclonal antibody against interferon gamma (IFN-gamma) or in vivo depletion of CD8+ T cells, but not CD4+ T cells, at the time of challenge blocked expression of iNOS mRNA and ablated protection in immunized mice. These results show that both CD8+ T cells and IFN-gamma are important components in the regulation of iNOS in liver which contributes to the protective response of mice immunized with irradiated malaria sporozoites. IFN-gamma, likely provided by malaria-specific CD8+ T cells, induces liver cells, hepatocytes and/or Kupffer cells, to produce NO for the destruction of infected hepatocytes or the parasite within these cells.

Binding of Plasmodium falciparum 175-kilodalton erythrocyte binding antigen and invasion of murine erythrocytes requires N-acetylneuraminic acid but not its O-acetylated form.

Sialic acid on human erythrocytes is involved in invasion by the human malaria parasite, Plasmodium falciparum. Mouse erythrocytes were used as a reagent to explore the question of whether erythrocyte sialic acid functions as a nonspecific negative charge or whether the sialic acid is a necessary structural part of the receptor for merozoites. Human erythrocytes contain N-acetylneuraminic acid (Neu5Ac), whereas mouse erythrocytes, which are also invaded by P. falciparum merozoites, contain 9-O-acetyl-N-acetylneuraminic acid (Neu5,9Ac2) and N-glycoloylneuraminic acid (Neu5Gc), in addition to Neu5Ac. We compared the effects of sialidase and influenza C virus esterase treatments of mouse erythrocytes on invasion and the binding of a 175-kDa P. falciparum protein (EBA-175), a sialic acid-dependent malaria ligand implicated in the invasion process. Sialidase-treated mouse erythrocytes were refractory to invasion by P. falciparum merozoites and failed to bind EBA-175. Influenza C virus esterase, which converts Neu5,9Ac2 to Neu5Ac, increased both invasion efficiency and EBA-175 binding to mouse erythrocytes. Thus, the parasite and EBA-175 discriminate between Neu5Ac and Neu5,9Ac2, that is, the C-9 acetyl group interferes with EBA-175 binding and invasion by P. falciparum merozoites. This indicates that sialic acid is part of a receptor for invasion.

A malaria invasion receptor, the 175-kilodalton erythrocyte binding antigen of Plasmodium falciparum recognizes the terminal Neu5Ac(alpha 2-3)Gal- sequences of glycophorin A.

Plasmodium falciparum malaria parasites invade human erythrocytes by means of a parasite receptor for erythrocytes, the 175-kD erythrocyte binding antigen (EBA-175). Similar to invasion efficiency, binding requires N-acetylneuraminic acid (Neu5Ac) on human erythrocytes, specifically the glycophorins. EBA-175 bound to erythrocytes with receptor-like specificity and was saturable. The specificity of EBA-175 binding was studied to determine if its binding is influenced either by simple electrostatic interaction with the negatively charged Neu5Ac (on the erythrocyte surface); or if Neu5Ac indirectly affected the conformation of an unknown ligand, or if Neu5Ac itself in specific linkage and carbohydrate composition was the primary ligand for EBA-175 as demonstrated for hemagglutinins of influenza viruses. Most Neu5Ac on human erythrocytes is linked to galactose by alpha 2-3 and alpha 2-6 linkages on glycophorin A. Soluble Neu5Ac by itself in solution did not competitively inhibit the binding of EBA-175 to erythrocytes, suggesting that linkage to an underlying sugar is required for binding in contrast to charge alone. Binding was competitively inhibited only by Neu5Ac(alpha 2-3)Gal-containing oligosaccharides. Similar oligosaccharides containing Neu5Ac(alpha 2-6)Gal-linkages had only slight inhibitory effects. Binding inhibition assays with modified sialic acids and other saccharides confirmed that oligosaccharide composition and linkage were primary factors for efficient binding. EBA-175 bound tightly enough to glycophorin A that the complex could be precipitated with an anti-glycophorin A monoclonal antibody. Selective cleavage of O-linked tetrasaccharides clustered at the NH2 terminus of glycophorin A markedly reduced binding in inhibition studies. We conclude that the Neu5Ac(a2,3)-Gal- determinant on O-linked tetrasaccharides of glycophorin A appear to be the preferential erythrocyte ligand for EBA-175.

Sequestrin, a CD36 recognition protein on Plasmodium falciparum malaria-infected erythrocytes identified by anti-idiotype antibodies.

The CD36 molecule expressed by human endothelial cells is a receptor for the adhesion of erythrocytes infected with the human malaria parasite Plasmodium falciparum. A CD36-specific monoclonal antibody, OKM8, inhibits the adhesion of malaria-infected erythrocytes (IRBC) to purified CD36 and cells expressing CD36. Monospecific polyclonal anti-idiotype (anti-Id) antibodies, raised against monoclonal antibody OKM8, expressed determinants molecularly mimicking the CD36 binding domain for the adhesion of IRBC. Purified rabbit anti-Id antibodies reacted with the surface of IRBC by immunofluorescence, directly supported the adhesion of wild-type P. falciparum malaria isolates, and inhibited IRBC cytoadherence to melanoma cells. An approximately 270-kDa protein was immunoprecipitated by the anti-Id antibodies from surface-labeled and metabolically labeled IRBC and was competitively inhibited by soluble CD36. These results support the hypothesis that CD36 is a receptor and the approximately 270-kDa protein, sequestrin, is a complementary ligand involved in the adhesion of IRBC to host-cell endothelium. Sequestrin is a candidate malaria vaccine antigen, and anti-Id antibodies that recognize this molecule may be useful for passive immunotherapy of cerebral and severe P. falciparum malaria.

Primary structure of the 175K Plasmodium falciparum erythrocyte binding antigen and identification of a peptide which elicits antibodies that inhibit malaria merozoite invasion.

The Plasmodium falciparum gene encoding erythrocyte binding antigen-175 (EBA-175), a putative receptor for red cell invasion (Camus, D., and T. J. Hadley. 1985. Science (Wash. DC). 230:553-556.), has been isolated and characterized. DNA sequencing demonstrated a single open reading frame encoding a translation product of 1,435 amino acid residues. Peptides corresponding to regions on the deduced amino acid sequence predicted to be B cell epitopes were assessed for immunogenicity. Immunization of mice and rabbits with EBA-peptide 4, a synthetic peptide encompassing amino acid residues 1,062-1,103, produced antibodies that recognized P. falciparum merozoites in an indirect fluorescent antibody assay. When compared to sera from rabbits immunized with the same adjuvant and carrier protein, sera from rabbits immunized with EBA-peptide 4 inhibited merozoite invasion of erythrocytes in vitro by 80% at a 1:5 dilution. Furthermore, these sera inhibited the binding of purified, authentic EBA-175 to erythrocytes, suggesting that their activity in inhibiting merozoite invasion of erythrocytes is mediated by blocking the binding of EBA-175 to erythrocytes. Since the nucleotide sequence of EBA-peptide 4 is conserved among seven strains of P. falciparum from throughout the world (Sim, B. K. L. 1990. Mol. Biochem. Parasitol. 41:293-296.), these data identify a region of the protein that should be a focus of vaccine development efforts.

A 60-kDa Plasmodium falciparum protein at the moving junction formed between merozoite and erythrocyte during invasion.

Invasion of erythrocytes by malaria merozoites requires the formation of a junction of attachment between erythrocyte and merozoite membranes. The attachment junction initially forms at the apical region of the merozoite. It then moves around to the posterior of the merozoite as invasion proceeds. A monoclonal antibody against a 60-kDa merozoite protein (termed MCP-1 for merozoite capping protein 1) of Plasmodium falciparum reacts in an immunofluorescence pattern resembling the moving junction. By two-color immunofluorescence, MCP-1 was located at the attachment site formed between the merozoite apical region and erythrocyte. During invasion, MCP-1 separated and migrated around merozoites at the orifice of the parasitophorous vacuole. In newly-invaded erythrocytes, MCP-1 persisted at the pole of the young parasite nearest the erythrocyte membrane, suggesting its anterior-to-posterior movement. MCP-1 exhibited no variability in molecular mass among the FCR-3, Camp and 7G8 strains of P. falciparum, and the epitope was invariant in the P. falciparum strains studied. We conclude that MCP-1 may participate in merozoite invasion of erythrocytes by facilitating attachment or movement of the junction along the parasite cytoskeletal network.

Receptor-like specificity of a Plasmodium knowlesi malarial protein that binds to Duffy antigen ligands on erythrocytes.

A 135-kD parasite protein, a minor component of the Plasmodium knowlesi malaria radiolabeled proteins released into culture supernatant at the time of merozoite release and reinvasion, specifically bound to human erythrocytes that are invaded and carry a Duffy blood group determinant (Fya or Fyb), but did not bind to human erythrocytes that are not invaded and do not carry a Duffy determinant (FyFy). Specific anti-Duffy antibodies blocked the binding of the 135-kD protein to erythrocytes carrying that specific Duffy determinant. Purified 135-kD protein bound specifically to the 35-45-kD Duffy glycoprotein on a blot of electrophoretically separated membrane proteins from Fya and Fyb erythrocytes but not from FyFy erythrocytes. Binding of the 135-kD protein was consistently greater to Fyb than to Fya both on the blot and on intact erythrocytes. The 135-kD protein also bound to rhesus erythrocytes that are Fyb and are invaded, but not to rabbit or guinea pig erythrocytes that are Duffy-negative and are not invaded. Cleavage of the Duffy determinant by pretreating Fyb human erythrocytes with chymotrypsin greatly reduced both invasion and binding of the 135-kD protein, whereas pretreating Fyb erythrocytes with trypsin had little effect on the Duffy antigen, the 135-kD protein binding, or on invasion. However, instances of invasion of other enzyme-treated erythrocytes that are Duffy-negative and do not bind the 135-kD protein suggest that alternative pathways for invasion do exist.

Falciparum malaria parasites invade erythrocytes that lack glycophorin A and B (MkMk). Strain differences indicate receptor heterogeneity and two pathways for invasion.

To determine the ligands on erythrocytes for invasion by Plasmodium falciparum, we tested invasion into MkMk erythrocytes that lack glycophorins A and B and enzyme-treated erythrocytes by parasites that differ in their requirement for erythrocyte sialic acid. The 7G8 strain invaded MkMk erythrocytes and neuraminidase-treated normal erythrocytes with greater than 50% the efficiency of normal erythrocytes. In contrast, the Camp strain invaded MkMk erythrocytes at 20% of control and neuraminidase-treated normal erythrocytes at only 1.8% of control. Invasion of MkMk erythrocytes by 7G8 parasites was unaffected by treatment with neuraminidase but was markedly reduced by treatment with trypsin. In contrast, invasion of MkMk cells by Camp parasites was markedly reduced by neuraminidase but was unaffected by trypsin. We conclude that the 7G8 and Camp strains differ in ligand requirements for invasion and that 7G8 requires a trypsin sensitive ligand distinct from glycophorins A and B.

Invasion of mouse erythrocytes by the human malaria parasite, Plasmodium falciparum.

Plasmodium falciparum malaria merozoites require erythrocyte sialic acid for optimal invasion of human erythrocytes. Since mouse erythrocytes have the form of sialic acid found on human erythrocytes (N-acetyl neuraminic acid), mouse erythrocytes were tested for invasion in vitro. The Camp and 7G8 strains of P. falciparum invaded mouse erythrocytes at 17-45% of the invasion rate of human erythrocytes. Newly invaded mouse erythrocytes morphologically resembled parasitized human erythrocytes as shown on Giemsa-stained blood films and by electron microscopy. The rim of parasitized mouse erythrocytes contained the P. falciparum 155-kD protein, which is on the rim of ring-infected human erythrocytes. Camp but not 7G8 invaded rat erythrocytes, indicating receptor heterogeneity. These data suggest that it may be possible to adapt the asexual erythrocytic stage of P. falciparum to rodents. The development of a rodent model of P. falciparum malaria could facilitate vaccine development.

Vaccination-induced variation in the 140 kD merozoite surface antigen of Plasmodium knowlesi malaria.

Immunity to 143/140 kD schizont antigens of a monkey malaria, Plasmodium knowlesi, provides partial protection to lethal malaria infection in rhesus monkeys challenged with uncloned parasites. To determine the capacity of a cloned parasite to generate variants of the 143/140 kD antigens, immunized monkeys were challenged with a clone of P. knowlesi. Parasites recovered 8 d after inoculation with a cloned parasite retained the 143/140 kD antigens. Parasites recovered 30 d after challenge had undergone changes in the 143/140 kD antigens. Antibodies that block erythrocyte invasion in vitro of the inoculum parasites did not inhibit invasion of erythrocytes by two isolates recovered from the immunized monkeys. An isolate from one monkey recovered on day 30 contained clones expressing new 76/72 kD antigens reactive with rabbit antiserum against the 143/140 kD proteins, and other clones expressing no antigens crossreactive with antisera against the 143/140 kD proteins. An isolate from another monkey obtained 59 d after challenge expressed new antigens of 160/155, 115/113, and 87/85 kD. Using monoclonal antibodies, we found that epitopes were lost from the variant proteins, but we were unable to determine whether new epitopes had appeared. We conclude that clones of P. knowlesi can rapidly vary antigenic determinants on the 143/140 kD proteins in animals immunized with these antigens.

Invasion of erythrocytes by Plasmodium falciparum malaria parasites: evidence for receptor heterogeneity and two receptors.

Plasmodium falciparum malaria parasites with different capabilities of invading sialic acid-deficient erythrocytes were identified. Thai-2 parasites cultured in Tn erythrocytes invaded neuraminidase-treated and Tn erythrocytes twice as efficiently as Thai-2 parasites cultured in normal erythrocytes and seven to ten times more efficiently than a cloned line of Camp parasites cultured in normal erythrocytes. All three parasite lines required sialic acid for optimal invasion, but Thai-2 parasites cultured in Tn erythrocytes invaded neuraminidase-treated erythrocytes with 45% efficiency whereas Camp parasites invaded neuraminidase-treated erythrocytes with less than 10% efficiency. P falciparum malaria parasites probably possess two receptors: one that binds to a sialic acid-dependent ligand and another that binds to a sialic acid-independent ligand. Parasites may differ in the quantity or affinity of their receptors for the sialic acid-independent ligand.

A feline thymocyte antigen defined by a monoclonal antibody (FT2) identifies a subpopulation of non-helper cells capable of specific cytotoxicity.

A monoclonal antibody, FT2 (IgG1 kappa) prepared against cat thymocytes, was found to be reactive with an antigenic determinant expressed by approximately 76% of thymocytes, 15% of blood mononuclear cells, 14% of splenocytes, and 1% of bone marrow cells. The FT2-reactive determinant was not expressed on B cells, macrophages, granulocytes, or erythrocytes. Both FT2+ and FT2- populations of peripheral blood mononuclear cells were capable of proliferative responses to the T cell mitogens Con A and PHA. When splenocytes were sensitized to the lymphoblastoid cell line, 79p90, cytotoxic T cells were found in the FT2+ population and were absent from the FT2- population. Conversely, the FT2- population contained the helper T cell activity required for pokeweed mitogen-induced B cell differentiation. Under nonreducing conditions, the FT2 antigen had an apparent m.w. of 71,000. When reduced, subunits of 31,000 and 38,000 apparent m.w. were observed. The data suggest that the FT2 antibody identifies the feline analog of the human T8/Leu-2, murine Ly-2 molecules expressed by cytotoxic/suppressor T cells.

Immunization of monkeys with a 140 kilodalton merozoite surface protein of Plasmodium knowlesi malaria: appearance of alternate forms of this protein.

The merozoite is the invasive stage of the malaria parasite which is released by rupture of the schizont-infected erythrocyte. A monoclonal antibody against a 140 kilodalton (kDa) merozoite surface antigen of Plasmodium knowlesi was used to characterize and to purify this antigen. It was shown by pulse-chase metabolic labeling of mature schizonts that the 140 kDa merozoite antigen was the processed product of a 143 kDa schizont component, and that processing occurred at the time of erythrocyte rupture. Antiserum, prepared by immunizing a rabbit with the 143/140 kDa antigen purified by immunoaffinity chromatography with the monoclonal antibody, strongly inhibited invasion of erythrocytes in vitro; Fab fragments prepared from purified rabbit IgG were inactive at blocking invasion, suggesting that agglutination of merozoites was the mechanism of invasion inhibition. The purified 143/140 kDa antigen was used in Freund's adjuvant to immunize four rhesus monkeys. Two of the immunized animals developed fulminating infections on challenge with 10(4) schizonts, as did the three control animals. The remaining two immunized animals controlled their infections and developed chronic low-grade parasitemias. The animals which were partially protected were those that had developed anti-143/140 kDa antibodies capable of blocking invasion in vitro. Parasites were isolated from the chronic stage of infection (V5 population) and were compared with the original parasite population used for challenge (P population). Inhibition of invasion, immunofluorescence, and immunoprecipitation with anti-143/140 kDa monoclonal antibody, with immune rabbit, and with monkey sera showed that the 143/140 kDa surface antigen had been replaced by multiple cross-reacting alternate antigenic forms of the molecule in the V population. Thus, specific immune response directed against a purified merozoite surface antigen resulted in the replacement of this antigen by variant or mutant forms.