Protective immunity against severe malaria is associated with antibodies to a conserved repertoire of PfEMP1 variants

Extreme diversity of the major surface antigen and virulence determinant of the malaria parasite Plasmodium falciparum, Erythrocyte Membrane Protein-1 (PfEMP1), poses a major barrier to identifying targets of protective immunity. To overcome this problem, we developed a PfEMP1 protein microarray containing 456 DBL domains, which was used to characterize the immunome of a cohort of semi-immune children and to identify variants associated with protective immune responses. Children with high mean antibody levels to DBL group 2 had a 26-36% reduced risk of uncomplicated (clinical) malaria, however only 8 diverse DBL variants were weakly associated with protection from clinical malaria and had low predictive accuracy. On the other hand, children with high mean antibodies to DBL groups 1 and 2 (which are markers for pathogenic \"Type A\" PfEMP1) and elevated antibodies to 85 (18.6%) of individual DBL variants had a 70 -100% reduced risk of severe malaria. Of the top 20 predictive variants for severe disease protection, 17 were strongly associated with protection (86 - 100% reduction in risk of severe malaria) and had high predictive accuracy for severe disease risk. Many variants were conserved and had highly correlated antibody responses, including the three highest-ranking variants, which were linked to EPCR-binding CIDR domains. The results suggest that while immunity to uncomplicated malaria is characterised by antibodies to a diverse repertoire of PfEMP1, immunity to severe malaria requires antibodies to a limited subset of antigenically conserved variants. These findings provide new insights into antimalarial immunity and potential biomarkers for tracking disease risk.