Naturally Acquired Binding-Inhibitory Antibodies to Plasmodium vivax Duffy Binding Protein in Pregnant Women Are Associated with Higher Birth Weight in a Multicenter Study

Pilar Requena, Myriam Arévalo-Herrera, Michela Menegon, Flor E. Martínez-Espinosa, Norma Padilla, Camila Bôtto-Menezes, Adriana Malheiro, Dhiraj Hans, Maria Eugenia Castellanos, Leanne Robinson, Paula Samol, Swati Kochar, Sanjay K. Kochar, Dhanpat K. Kochar, Meghna Desai, Sergi Sanz, Llorenç Quintó, Alfredo Mayor, Stephen Rogerson, Ivo Mueller, Carlo Severini, Hernando A. del Portillo, Azucena Bardají, Chetan C. Chitnis, Clara Menéndez, Carlota Dobaño

A vaccine to eliminate malaria would need a multi-stage and multi-species composition to achieve robust protection, but the lack of knowledge about antigen targets and mechanisms of protection precludes the development of fully efficacious malaria vaccines, especially for Plasmodium vivax (Pv). Pregnant women constitute a risk population who would greatly benefit from a vaccine preventing the adverse events of Plasmodium infection during gestation. We hypothesized that functional immune responses against putative targets of naturally acquired immunity to malaria and vaccine candidates will be associated with protection against malaria infection and/or poor outcomes during pregnancy. We measured (i) IgG responses to a large panel of Pv and Plasmodium falciparum (Pf) antigens, (ii) the capacity of anti-Pv ligand Duffy binding protein (PvDBP) antibodies to inhibit binding to Duffy antigen, and (iii) cellular immune responses to two Pv antigens, in a subset of 1,056 pregnant women from Brazil, Colombia, Guatemala, India, and Papua New Guinea (PNG). There were significant intraspecies and interspecies correlations for most antibody responses (e.g., PfMSP119 versus PfAMA1, Spearman's rho = 0.81). Women from PNG and Colombia had the highest levels of IgG overall. Submicroscopic infections seemed sufficient to boost antibody responses in Guatemala but not antigen-specific cellular responses in PNG. Brazil had the highest percentage of Duffy binding inhibition (p-values versus Colombia: 0.040; Guatemala: 0.047; India: 0.003, and PNG: 0.153) despite having low anti-PvDBP IgG levels. Almost all antibodies had a positive association with present infection, and coinfection with the other species increased this association. Anti-PvDBP, anti-PfMSP1, and anti-PfAMA1 IgG levels at recruitment were positively associated with infection at delivery (p-values: 0.010, 0.003, and 0.023, respectively), suggesting that they are markers of malaria exposure. Peripheral blood mononuclear cells from Pv-infected women presented fewer CD8(+)IFN-γ(+) T cells and secreted more G-CSF and IL-4 independently of the stimulus used in vitro. Functional anti-PvDBP levels at recruitment had a positive association with birth weight (difference per doubling antibody levels: 45 g, p-value: 0.046). Thus, naturally acquired binding-inhibitory antibodies to PvDBP might confer protection against poor outcomes of Pv malaria in pregnancy.