Blocking malaria’s key virulence factors: scientists discovered antibodies that can disarm the disease

January 8, 2026 – by Santina Russo

Still today, the impact of malaria is staggering. According to the World Health Organization’s (WHO) latest World Malaria Report, an estimated 263 million people were infected and almost 600,000 died from the disease in 2023. The burden is heaviest in African countries, which account for 94 percent of all malaria cases worldwide. But the eastern mediterranean region is also affected and has suffered 57-percent increase in incidence since 2021 with 20 in 1000 people getting the disease. Young children are especially vulnerable, as they haven’t yet developed partial immunity, and pregnant women are at risk because the infection can cause maternal anaemia, low birth weight, and infant mortality.

In a recent study published in Nature, researchers including Monica Fernández-Quintero, now in Andrew Ward’s group at the Scripps Research Institute in La Jolla, uncovered a class of antibodies capable of blocking central virulence mechanisms of Plasmodium falciparum, the parasite responsible for the most severe form of malaria. Using the CSCS supercomputer "Piz Daint", Fernández-Quintero contributed to this study with an in-depth characterization of the antibody binding interface.

From initial to potent antibodies

Much of the disease’s damage occurs when infected red blood cells adhere to the endothelial walls of blood vessels. As the infected cells accumulate, the smallest vessels become blocked, reducing the supply of oxygen and nutrients and triggering inflammation. This clogging is mediated by the polymorphic erythrocyte membrane protein 1 (PfEMP1) of the Plasmodium parasite. In the process, it binds to receptors at the surface of endothelial cells, which form the inner lining of blood vessels.

Through extensive lab experiments and molecular dynamics simulations, Fernández-Quintero and her co-authors now identified two antibodies—isolated from two different patients—that consistently and potently inhibit the endothelial binding of five out of six PfEMP1 subclasses, thereby providing protection from the disease.

For their simulations at CSCS, the team used experimentally obtained structures of the two antibodies as the basis for modelling the antibody-antigen complexes. This term refers to antibodies bound to their target protein—in this case, a specific subdomain of the PfEMP1 protein known as CIDRα1. “This is the domain responsible for the endothelial binding that mediates the disease,” explained Fernández-Quintero. The scientists focused on the antibodies’ maturation: from the state before the infection, meaning before the immune system first encountered the virulent CIDRα1 domain, through their evolution into potent antibody against it.

How to double-check AI methods

When the team compared structures resulting from different AI-enhanced antibody structure prediction tools, they discovered that many of them contained inaccuracies and errors. They found clashes between atoms, or cis-amide bonds and incorrect stereochemistry—features that are theoretically possible but never found in proteins in nature. Overall, the team examined 137 antibody models which collectively contained more than 500 physical errors.

Cover Image: The Anopheles arabiensis mosquito is a known carrier of the pathogen causing malaria, the parasite Plasmodium falciparum.
(Image credit: Wikimedia Commons, CDC/James Gathany)