O-GlcNAc modification of the anti-malarial vaccine candidate PfAMA1: in silico-defined structural changes and potential to generate a better vaccine.

The complex life cycle of plasmodial parasites makes the selection of a single subunit protein a less than optimal strategy to generate an efficient vaccinal protection against malaria. Moreover, the full protection afforded by malarial proteins carried by intact parasites implies that immune responses against different antigens expressed in different phases of the cycle are required, but also suggests that native malarial antigens are presented to the host immune system in a manner that recombinant proteins do not achieve. The malarial apical membrane antigen 1 (AMA1) represents a suitable vaccine candidate because AMA1 is expressed on sporozoites and merozoites and allows them to invade hepatocytes and erythrocytes, respectively. Anti-AMA1 antibodies and cytotoxic T-cells are therefore expected to interfere both with the primary invasion of hepatocytes by sporozoites and with the later propagation of merozoites in erythrocytes, and thus efficiently counteract parasite development in its human host. AMA1 bears potential glycosylation sites and the human erythrocytic O-linked N-acetylglucosamine transferase (OGT) could glycosylate AMA1 through combinatorial metabolism. This hypothesis was tested in silico by developing binding models of AMA1 with human OGT complexed with UDP-GlcNc, and followed by the binding of O-GlcNAc with the hydroxyl group of AMA1 serine and threonine residues. Our results suggests that AMA1 shows potential for glycosylation at Thr517 and Ser498 and that O-GlcNAc AMA1 may constitute a conformationally more appropriate antigen for developing a protective anti-malarial immune response.