Structure-based design of a strain transcending AMA1-RON2L malaria vaccine

Abstract

Apical membrane antigen 1 (AMA1) is a key malaria vaccine candidate and target of neutralizing antibodies. AMA1 binds to a loop in rhoptry neck protein 2 (RON2L) to form the moving junction during parasite invasion of host cells, and this complex is conserved among apicomplexan parasites. AMA1-RON2L complex immunization achieves higher growth inhibitory activity than AMA1 alone and protects mice against Plasmodium yoelii challenge. Here, three single-component AMA1-RON2L immunogens were designed that retain the structure of the two-component AMA1-RON2L complex: one structure-based design (SBD1) and two insertion fusions. All immunogens elicited high antibody titers with potent growth inhibitory activity, yet these antibodies did not block RON2L binding to AMA1. The SBD1 immunogen induced significantly more potent strain-transcending neutralizing antibody responses against diverse strains of Plasmodium falciparum than AMA1 or AMA1-RON2L complex vaccination. This indicates that SBD1 directs neutralizing antibody responses to strain-transcending epitopes in AMA1 that are independent of RON2L binding. This work underscores the importance of neutralization mechanisms that are distinct from RON2 blockade. The stable single-component SBD1 immunogen elicits potent strain-transcending protection that may drive the development of next-generation vaccines for improved malaria and apicomplexan parasite control.

Fig. 1. Overview of the design of single-component immunogens.

a Structure of apo AMA1 DI-DII showing the domain II loop (DII loop) and DIf loop (gray cartoon) and the location of RON2L in the bound complex (orange cartoon). b A circularly permutated SBD1 immunogen was created by introducing a Gly/Ser linker between the original termini (not shown) and by removing the DII loop, which produced novel N- and C-termini at residues Lys386 and Thr357, respectively. Then, RON2L was fused to this new C-terminus without a linker. c, d Insertion fusion immunogens 2 and 3 were constructed by replacing the DIf loop of AMA1 DI-DII with RON2L and by removing the DII loop. Insertion fusion immunogen 3 retains Cys263 and its disulfide bridge (yellow). This figure was created using structures of apo AMA1 (PDB ID: 4r19) and the AMA1-RON2L complex (PDB ID: 3zwz). An arrow indicates the point of fusion. e Schematic illustrating the design processes for the three immunogens discussed in this article. Proteins are shown from the N- to C-terminus, and numbers indicate residue numbering based on the wild-type AMA1 and RON2 sequences.

Fig. 2. The yield and stability of single-component immunogens are higher than those of AMA1 DI-DII alone and the AMA1 DI-DII-RON2L complex.

a All three immunogens were expressed at higher levels than AMA1 DI-DII and eluted as monomers by size exclusion chromatography (SEC). The inset in (a) confirms the high purity of the immunogens through reducing SDS-polyacrylamide gel electrophoresis (PAGE). b Purification yield from three separate purifications. Bars represent the mean yield from three separate purifications. c Differential scanning fluorimetry indicated that three immunogens have higher thermostability than AMA1 DI-DII and the AMA1 DI-DII-RON2L complex. d T_m from five independent measurements. Bars represent the mean. Source data are provided as a Source data file.

Fig. 3. RON2L is bound to AMA1 in the designed immunogens, preventing accessibility to the RON2L binding site.

a Representative biolayer interferometry (BLI) traces used to quantitatively measure the binding of immunogens to IgNAR 14I-1 demonstrating inaccessibility of the epitope located in the RON2L binding pocket in the immunogens. Immunogens were two-fold serially diluted in the range of 200 nM to 3.125 nM. b IgNAR 14I-1 shows little or no binding to immunogens in three independent ELISAs. c Representative BLI traces used to measure the binding of immunogens to exogenous RON2L demonstrating that the binding site for exogenous RON2L is occupied by the fused RON2L in the designed immunogens. Immunogens were two-fold serially diluted in the range of 200 nM to 3.125 nM. d Exogenous RON2L does not bind to immunogens in three independent ELISAs. In b and d, bovine serum albumin (BSA) was used as a negative control. Source data are provided as a Source data file.

Fig. 4. Single-component immunogens have a very similar structure to the AMA1-RON2L complex, with the SBD1 immunogen possessing the greatest structural similarity.

a Crystal structures of the single-component SBD1 immunogen (light blue, blue) and insertion fusion immunogens 2 (light pink, magenta) and 3 (cyan, teal). The fused RON2L portion of the immunogen is shaded darker than the AMA1 portion. b Single-component immunogens superimposed on the AMA1-RON2L complex (PDB ID: 3zwz, white and orange). c A focused view of RON2L and the surrounding loops in single-component immunogens superimposed on the AMA1-RON2L complex (PDB ID: 3zwz). An arrow indicates local structural perturbations in the insertion fusion immunogens.

Fig. 5. Neutralizing antibody levels in rats immunized with single-component immunogens are similar to those of AMA1 DI-DII alone or the AMA1 DI-DII-RON2L complex.

a Immunization and blood draw scheme for rats. The figure was created with BioRender.com. b Serum IgG titers against AMA1 DI-DII from three independent biological replicates. The dashed line indicates the detection limit of the assay, and the bars represent the geometric mean titers (GMTs). c Serum antibody titers blocking the AMA1 DI-DII/RON2L interaction from two independent biological replicates depicted as described in (b). d In vitro GIA of purified IgG from individual rats from each group at day 63 was tested at 5.0 mg/ml against Plasmodium falciparum 3D7 blood stage in three independent assays. Bars represent the median. The dashed line indicates the median of the adjuvant only group. Statistical comparisons and P values for (b), (c), and (d) were obtained using a Kruskal‒Wallis analysis followed by Dunn’s test to correct for multiple comparisons of the AMA1 DI-DII, immunogens and adjuvant only groups with the AMA1 DI-DII-RON2L complex group. Source data are provided as a Source data file.

Fig. 6. SBD1 immunogen elicits significantly more potent strain-transcending antibodies than AMA1 DI-DII alone or the AMA1 DI-DII-RON2L complex.

In vitro GIA dilution series of pooled purified IgG from each group at day 63 against Plasmodium falciparum (a) 3D7 (b) FVO (c) Dd2. The data are plotted as the median with 95% CI and arise from three independent biological replicates for SBD1 immunogen, AMA1 DI-DII-RON2L Complex, insertion fusion immunogens 2 and 3, and two biological replicates for AMA1 DI-DII alone due to limited IgG for this group. Concentration (mg/ml) of pooled purified IgG required to demonstrate 50% inhibition (IC₅₀) against Plasmodium falciparum (d) 3D7 (e) FVO and (f) Dd2 were determined by interpolation after fitting data globally to a four-parameter dose-response curve. The bars represent the IC₅₀ (center) and 95% CI of a global fit of three independent biological replicates (two biological replicates for the AMA1 DI-DII group). Points represent IC₅₀ values for individual fits of each biological replicate. Statistical comparisons were made using a two-tailed extra sum-of-squares F-test (with Bonferroni correction for d). Source data are provided as a Source data file.