Duffy-blood-group-negative human erythrocytes, FyFy, are resistant to invasion in vitro by Plasmodium knowlesi. The FyFy genotype is found predominantly in African and American blacks, who are the only groups completely resistant to infection by P. vivax. To determine if the FyFy genotype is the vivax resistance factor, we performed blood typing on 11 black and six white volunteers who had been exposed to the bites of P. vivax-infected mosquitoes. Only the five FyFy blacks were resistant to erythrocytic infection; the remaining six blacks and all whites had the Duffy-positive determinants (Fya or Fyb or both) and had contracted malaria. We conclude that Duffy determinants (Fya or Fyb or both) on the erythrocyte surface are required for invasion of erythrocytes by vivax merozoites.

Duffy blood group negative human erythrocytes (FyFy) are resistant to infection by Plasmodium knowlesi, a simian malaria that infects Duffy positive human erythrocytes. The P. knowlesi resistance factor, Duffy negative erythrocytes, occurs in high frequency in West Africa, where the people are resistant to vivax malaria. This suggests that Duffy blood group determinants (Fya or Fyb) may be erythrocyte receptors for P. vivax.

Human erythrocytes that lack the Duffy blood group antigens are refractory to infection by Plasmodium knowlesi malaria parasites in vitro and Plasmodium vivax in vivo, suggesting that these antigens are used as ligands for invasion of erythrocytes by merozoites (1-5). This specificity has been the prototype for understanding malaria parasite invasion oferythrocytes . The molecule that carries a Duffy determinant was characterized recently as a 35-45-kD glycoprotein (6), but no parasite molecule with specificity for Duffy has been described . A study with Plasmodium falciparum demonstrated that after release of merozoites from cultured schizont-infected erythrocytes, a 175-kD protein accumulates in the culture supernatant . This molecule binds to both Duffy-positive and Duffy-negative human erythrocytes that are invaded, but does not bind to neuraminidase-or trypsin-treated human erythrocytes that are refractory to invasion by P. falciparum (7). We used modifications of methods of Camus and Hadley (reference 7, Haynes, J. D., manuscript in preparation) to determine whether P. knowlesi uses a receptor analogous to that of P. falciparum but with the appropriate specificity for Duffy blood group antigen . Erythrocytes were incubated with the metabolically labeled proteins in culture supernatant, centrifuged through oil, and the bound proteins were eluted with salt. We now describe a 135-kD protein synthesized by P. knowlesi that bound to Duffy blood group-positive, but not to Duffy blood group-negative, human erythrocytes . It seems likely that the 135-kD Duffy binding protein is one of the receptors responsible for invasion ofP. knowlesi merozoites into Duffy-positive erythrocytes . Materials and MethodsParasites and Labeled Culture Supernatants. Schizont-infected rhesus erythrocytes (Malaysian H strain of P. knowlesi) were enriched and metabolically labeled at 1 .5 x 10'/ml with 100 gCi/ml ["Slmethionine in methionine-deficient RPMI-1640 culture medium (3, 8) with 1% rabbit serum. The parasites were cultured for 6-10 h to allow the maturation of schizonts, J. P.

The asexual erythrocytic cycle responsible for the clinical manifestations of malaria is perpetuated by merozoites which rupture out of infected erythrocytes and invade uninfected ones. Interaction between the merozoites and receptors on the erythrocyte surface initiate invasion (1-3). Chemical or immunologic interference between the merozoite and its receptor could block the infection at this point in the life cycle. Previously, we studied the characteristics of the erythroycte surface that are required for interaction and subsequent invasion by Plasmodium knowlesi merozoites (3, 4). Erythrocytes which lacked Duffy blood group determinants (Duffy negative erythrocytes) (3) or had been treated with chymotrypsin (4) were resistant to invasion. With the recent advances in in vitro cultivation ofPlasmodium falciparum (5, 6), it was possible to compare the erythrocyte surface determinants required for invasion by P. falciparum, the major malaria of man, and P. knowlesi, a monkey malaria that infects man. Materials and MethodsHuman erythrocytes (test cells) which either lacked a blood group determinant (Table I) or had been treated with various enzymes (Table II) were exposed to P. falciparum and P. knowlesi merozoites in vitro, and the percentage of test cells infected was determined. Chimpanzee erythrocytes infected with P. falciparum (Camp strain) and frozen in liquid nitrogen were thawed (7) on the day of the experiment. They contained approximately 50% infected erythroeytes and varied in viability from experiment to experiment. The infected cells were centrifuged in microhematocrit tubes. Cells from the upper half of the tubes, containing about 90% infected erythrocytes were mixed with test erythrocytes in a ratio of 1 to 25 and cultured as previously described (6). Erythrocytes were removed from culture 48 h later, spread in a thin film on a slide, stained with Giemsa, and the percent of erythrocytes infected with young trophozoites (ring forms) determined.Rhesus erythrocytes that are refractory to invasion by P. falciparum were included in each experiment (Table I). In parallel experiments, samples of test erythrocytes were cultured with P.knowlesi-infected cells (3). Since P. falciparum merozoites were released after 48 h in culture and P. knowlesi within 1 h, test erythrocytes were incubated at 35°C for 48 h before addition of P. knowlesi-infected cells.