Date Published: December 26, 2013
Publisher: Public Library of Science
Author(s): Sheetij Dutta, Lisa S. Dlugosz, Damien R. Drew, Xiopeng Ge, Diouf Ababacar, Yazmin I. Rovira, J. Kathleen Moch, Meng Shi, Carole A. Long, Michael Foley, James G. Beeson, Robin F. Anders, Kazutoyo Miura, J. David Haynes, Adrian H. Batchelor, Michael J. Blackman.
Malaria vaccine candidate Apical Membrane Antigen-1 (AMA1) induces protection, but only against parasite strains that are closely related to the vaccine. Overcoming the AMA1 diversity problem will require an understanding of the structural basis of cross-strain invasion inhibition. A vaccine containing four diverse allelic proteins 3D7, FVO, HB3 and W2mef (AMA1 Quadvax or QV) elicited polyclonal rabbit antibodies that similarly inhibited the invasion of four vaccine and 22 non-vaccine strains of P. falciparum. Comparing polyclonal anti-QV with antibodies against a strain-specific, monovalent, 3D7 AMA1 vaccine revealed that QV induced higher levels of broadly inhibitory antibodies which were associated with increased conserved face and domain-3 responses and reduced domain-2 response. Inhibitory monoclonal antibodies (mAb) raised against the QV reacted with a novel cross-reactive epitope at the rim of the hydrophobic trough on domain-1; this epitope mapped to the conserved face of AMA1 and it encompassed the 1e-loop. MAbs binding to the 1e-loop region (1B10, 4E8 and 4E11) were ∼10-fold more potent than previously characterized AMA1-inhibitory mAbs and a mode of action of these 1e-loop mAbs was the inhibition of AMA1 binding to its ligand RON2. Unlike the epitope of a previously characterized 3D7-specific mAb, 1F9, the 1e-loop inhibitory epitope was partially conserved across strains. Another novel mAb, 1E10, which bound to domain-3, was broadly inhibitory and it blocked the proteolytic processing of AMA1. By itself mAb 1E10 was weakly inhibitory but it synergized with a previously characterized, strain-transcending mAb, 4G2, which binds close to the hydrophobic trough on the conserved face and inhibits RON2 binding to AMA1. Novel inhibition susceptible regions and epitopes, identified here, can form the basis for improving the antigenic breadth and inhibitory response of AMA1 vaccines. Vaccination with a few diverse antigenic proteins could provide universal coverage by redirecting the immune response towards conserved epitopes.
Despite the availability of effective drugs mortality caused by malaria remains a global health and economic concern , and drug resistance to front-line antimalarials is increasing. A vaccine that either prevents the disease or can reduce the parasite burden is urgently needed to reduce mortality and morbidity of infants and young children living in many of the world’s poorest countries. Human anti-malarial antibodies can inhibit invasion and some studies suggest that growth inhibitory activity is associated with decreased risk of Plasmodium falciparum infection . Antigens that induce invasion inhibitory antibodies are therefore primary candidates for the development of a vaccine that targets parasite blood stages , . One such vaccine candidate is Apical Membrane Antigen-1 (AMA1) . Anti-AMA1 antibodies inhibit the invasion of merozoites into red blood cells  and this inhibitory activity correlates with protection in non-human primate malaria models , . Inhibitory AMA-1 antibodies are also acquired by humans exposed to P. falciparum infection , and antibodies to AMA1 are associated with protection from symptomatic malaria , . The positive selection of polymorphisms that map to the epitopes of inhibitory antibodies is further evidence that such antibodies have a protective role .
Given the failure of AMA1 vaccines to protect broadly, this study investigates the molecular basis of anti-AMA1 mediated cross-strain invasion inhibition. Using a mixture of antibodies against monovalent vaccines, Drew et al. have suggested that overlapping strain-specific antibodies can achieve broad inhibition . We show that vaccinating with a mixture of four divergent AMA1 allelic proteins (3D7, FVO, HB3, and W2mef) can outperform the GIA activity of a mixture of monovalent antisera (Fig. 1F) and whilst overlapping strain specific antibodies almost certainly contributed, QV vaccination achieved a kind of mixed-allele immunogenicity bonus, similar to that reported by mixing three alleles of the Duffy binding protein vaccine . The enhanced cross-strain inhibition by anti-QV was associated with no change in polymorphic face response, decrease in domain-2 response and increased response to two relatively non-polymorphic regions, the domain-3 and the conserved face. Reduction in domain-2 response was interesting because polymorphisms within domain-2 are important for antigenic escape in vitro,  and molecular analysis of a Phase 2b trial showed active selection at a domain-2 polymorphic site . Not surprisingly, all three domain-2 mAbs identified here were also strain-specific (Table 1). Our findings support the further development of multivalent AMA1 vaccines because broad antigenic coverage is achieved not only by expanded antigenic footprint of strain-specific antibodies but by inducing higher levels of antibodies against conserved inhibitory epitopes , .