Computational Design of Vaccines Against Plasmodium falciparum Malaria

Abstract

Malaria caused by P. falciparum resulted in over 600,000 deaths per year in 2022, with the majority of deaths occurring in children under five years of age. Recently, two vaccines have been recommended by the WHO for use in children in areas of moderate-to-high malaria transmission activity in Africa. These vaccines, RTS,S/AS01 and R21/Matrix-M, offer vaccine efficacies of 50-70% at 1 year but there are concerns about antibody quality and durability. Here, in my dissertation, I describe multiple approaches to address issues with antibody quality elicited by CSP vaccination. First, I explore the use of a computationally designed self-assembling nanoparticle displaying a series of different CSP repeat epitopes to elicit antibodies towards novel epitopes not found in RTS,S or R21. Building on this, the next project utilizes CSP nanoparticles where the nanoparticle surface is coated in N-linked oligomannose glycans to take advantage of mannose-binding lectin complement pathways so that they traffic efficiently to the germinal center. These CSP glycosylated nanoparticles enhance early B cell responses and improve protection in mouse models of infection. Finally, I investigate the use of deep-learning based protein design methods for the display of protective epitopes from CSP such as CIS43 and L9, two monoclonal antibodies that have shown promise in the field as passive immunizations. These de novo immunogens exhibit specific binding to their target antibodies with little cross-reactivity to known non-protective antibodies and are able to activate and recruit inferred-germline CIS43 and L9 B cells to the germinal centers of recipient mice. These immunogens offer a new way to try and improve anti-CSP repeat B cell responses.