Functional and Immunologic Mapping of Domains of the Reticulocyte-Binding Protein Plasmodium vivax PvRBP2a

Abstract

We previously described a novel Plasmodium vivax invasion mechanism into human reticulocytes via the PvRBP2a-CD98 receptor-ligand pair. Using linear epitope mapping, we assessed the PvRBP2a epitopes involved in CD98 binding and recognized by antibodies from patients who were infected. We identified 2 epitope clusters mediating PvRBP2a-CD98 interaction. Cluster B (PvRBP2a431-448, TAALKEKGKLLANLYNKL) was the target of antibody responses in humans infected by P vivax. Peptides from each cluster were able to prevent live parasite invasion of human reticulocytes. These results provide new insights for development of a malaria blood-stage vaccine against P vivax.

Keywords: Plasmodium vivax; CD98; PVRBP2a; invasion; reticulocytes.

Figure 1.

Identification and structural mapping of CD98-binding linear epitopes of PvRBP2a. A, The minimal CD98-binding fragment (as determined in Malleret et al [6]) is shown on a map of PvRBP2a, with putative signal peptide (SP), transmembrane domain (TM), and nucleotide-binding domain (NBD) labeled. A peptide library of 94 biotinylated 18-mers with 10-mer overlap was constructed from the minimal CD98-binding fragment, PvRBP2a_23-767. B, Left: the PvRBP2a-derived peptide library was tested for binding to reticulocytes via flow cytometry. Thiazole orange–positive reticulocytes were stained with biotinylated peptides and detected with streptavidin-APC. Data are presented as mean (SD) of 2 independent replicates. Right: candidate-binding peptides were further tested to see whether their binding could be blocked by anti-CD98 antibody, showing a dependence on CD98 as the cognate ligand. Reticulocytes were blocked with the polyclonal rabbit anti-CD98 antibody (KE020) or a rabbit isotype control antibody for 15 minutes before staining with biotinylated peptide candidates and detection with streptavidin-APC. Reticulocytes were distinguished from normocytes via thiazole orange staining. Data are presented as median (range). C, Peptides showing immunoreactivity or CD98-binding activity are highlighted in the published 3-dimensional structure of PvRBP2a (PDB 4Z8N). PepLib_21, blue; PepLib_25, medium blue; PepLib_36, cornflower blue; PepLib_39 and PepLib_40, forest green; PepLib_41, green; PepLib_49, cyan; PepLib_52, red. D, Candidate-binding peptides or vehicle control (VC) was tested for direct binding to CD98 via biolayer interferometry. Biotinylated peptides (2 µM) were loaded onto streptavidin-coated sensors, which were immersed in a solution of 200nM CD98 to assess binding.

Figure 2.

Identification of antigenic and functional linear epitopes of PvRBP2a. A, Peptides were pooled into groups of 5 and coated on streptavidin plates. Pooled sera were added from donors infected with Plasmodium vivax. Following washing, binding of peptides to donor sera was quantified with an anti-human horseradish peroxidase secondary antibody. The baseline was determined to be the mean + 3 SD of the reactivity of the seronegative sample pool to all peptide pools, and peptide pools showing a greater optical density (OD) value were deemed immunoreactive. B, The antigenicity of the antigenic peptide hit, PepLib_52, was confirmed by plasma from 20 donors who were seropositive and 22 who were seronegative. C, The effect of PvRBP2a-derived linear epitope peptides on P vivax invasion was examined on 5 P vivax clinical isolates. Selected CD98-binding PvRBP2a peptide candidates that were in cluster A (PepLib_25, PepLib_49) or cluster B/immunodominant (PepLib_41, PepLib_52), as well as negative control peptides (PepLib_50, PepLib_56) and positive control monoclonal antibodies (anti-Duffy 2C3, anti-CD98 HBJ127), were tested for their ability to inhibit P vivax invasion into reticulocytes. One well without addition of peptide or antibody served as a control for measurement of baseline reinvasion level. *P < .05 by 2-tailed t test with Shapiro-Wilk normality test.