PMC

Sequence alignment of PvRBP family. (Left) Alignment of the sequence corresponding to the crystallized fragment of PvRBP2a with sequences of six other members of the PvRBP family highlights a conserved domain. The secondary structure elements of PvRBP2a are shown above the alignment. The bonded pairs of cysteine residues are highlighted in yellow and marked with green numbers. (Right) Two orthogonal views of the PvRBP2a structure, with conserved amino acid residues highlighted as magenta sticks. The conserved residues have a mostly hydrophobic character and are involved in packing of the helical core.

Crystal structure of the PvRBP2a erythrocyte-binding domain. (A) Three orthogonal views of the molecule. The overall fold of the protein is formed by 10 antiparallel α-helices and one short β-hairpin. Molecule is shown in ribbon representation colored in rainbow from blue at the N terminus until red at the C terminus. The cysteine residues forming disulphide bridges are shown as yellow sticks. (B) Sample of the electron density around the residue F181 forming a γ-turn. The 2F_obs − F_calc map is contoured at 2σ and shown as blue mesh. Protein is shown in ball and stick representation with carbon shown in green, nitrogen in blue, oxygen in red, and sulfur in yellow.

Asymmetric unit content analysis. (A) An overall view of the asymmetric unit. Molecule A includes 298 residues spanning 160–455 amino acids of PvRBP2a and a di-peptide GS fragment that was introduced as a cloning artifact. Molecule B spans residues 160–450 but is less well-defined, in which residues 392–398 could not be convincingly modeled, yielding a structure with a total of 284 residues. The buried interface area between two molecules present in the ASU is around 1,340 Ų as calculated using the PISA server. Molecules A and B are colored in blue and green, respectively, and are shown as ribbon and surface representation. (B) Superimposition of the two molecules present in the asymmetric unit. Molecules are shown as ribbons colored with the colors corresponding to the rmsd between two molecules. Dark blue shows good alignment, and higher deviations are in orange/yellow/red. Residues not used for alignment are colored white.

Structural differences between PvRBP2a (magenta) and PfRh5 (blue). The PDB accession code for PfRh5 is 4WAT. (A) Close view of the N-terminal part of the molecule. The β-hairpin is rotated by 180 degrees between two structures. The N termini of both molecules are labeled. The loop connecting strand β2 and helix α1 in PvRBP2a includes a classic γ-turn with F181 localized in its center. The loop in PvRBP2a is colored in yellow, with F181 shown as sticks. (B) Helix α3 in PvRBP2a is divided into two parts by a β-turn formed by amino acids Q273 to M276. Additionally, the helix α2a is shifted compared with the corresponding helix in PfRh5. The β-turn formed by amino acids Q273 to M276 is colored in yellow. The cysteine residues forming an overlapping disulphide bridge are shown as yellow sticks. (C) The disordered loop in PfRh5, including residues S257 to D294 and represented schematically as a dashed line. A corresponding place in PvRBP2a forms only a very short helix α2c and consists of residues P244 to H247.

PvRBP2a erythrocyte-binding characteristics. (A) Dot plots showing the enriched reticulocyte purification stained with both TO and CD71-PECy5 (Upper) or CD71-PECy5 alone (Lower). (B) Dot plots showing the binding of PvRBP2a to red blood cells (Upper). Binding was detected using an anti-PvRBP2a rabbit IgG antibody, followed by a secondary anti-rabbit Alexa 647 antibody. (Lower) A binding control where no protein was added before incubation with primary and secondary antibodies. (C) Dot plots showing the binding of PfRh4 to red blood cells (Upper). Binding was detected using an anti-PfRh4 rabbit IgG antibody, followed by a secondary anti-rabbit Alexa 647 antibody. (Lower) A binding control where no protein was added before incubation with primary and secondary antibodies. (D) Bar charts showing the percentage of binding of CD71-PECy5, PvRBP2a, and PfRh4 to mature erythrocyte (TO⁻) vs. reticulocyte (TO⁺) populations. Error bars represent SEM of three independent repeats. (E) PvRBP2a binding to erythrocytes is neuraminidase- and chymotrypsin-resistant but trypsin-sensitive whereas PfRh4 binding is unperturbed by neuraminidase treatment. A flow cytometry-based red blood cell-binding assay was performed using untreated erythrocytes (Un), neuraminidase (Nm), low trypsin (LT), high trypsin (HT), and chymotrypsin-treated erythrocytes (CHY). Low trypsin and high trypsin refer to treatments with 0.1 and 1.5 mg/mL enzyme, respectively.

Detailed analysis of the PvRBP2a structure. (A) The molecule is schematically divided into two subdomains of similar size that are related to each other through pseudo-twofold rotation symmetry. The alpha-helices are represented as cylinders and labeled. The N-terminal domain is colored in blue, and the C-terminal in red. (B) B-factors analysis. Molecule A is drawn in schematic putty representation. The tip of the molecule, including a disulphide bond formed between C299 and C303, is relatively flexible compared with the rest of the molecule. The protein regions with high temperature factors are shown as wide orange/red tubes. (C) Enlargement of the N-terminal part of the molecule. This fragment is in contact with both subdomains, forming an extensive network of hydrogen bond interactions. Several amino acid residues are also involved in the stacking interactions: like F181 with I224 and F167 with K269, for example. Interacting amino acid residues are shown in sticks and labeled. Residues belonging to the N-terminal part of the protein are underlined.

Plasmodium vivax RBP2a amino acid haplotypes and allele frequencies. High quality consensus sequences for the entire 7,464-bp pvrbp2a coding locus were obtained from 22 P. vivax clinical isolates from Papua New Guinea, and these sequences were aligned together with published data for five reference strains and four field isolates from Thailand. A total of 105 single nucleotide polymorphisms were identified, which encoded 81 polymorphic amino acids. All 31 P. vivax isolates surveyed had unique amino acid haplotypes demonstrating very high diversity of PvRBP2a. Of the 31 polymorphic sites with minor allele frequencies of more than 10%, 10 mapped to the corresponding region of the crystal structure at residues N186S, E277K, K285N/I, K289E, E304K, D306V, E351Q, P399S, K421M, and G438E.

Gene structure and expression of PvRBP2a. (A) Amplification of pvrbp2a confirms splicing near the start codon. Primers flanking the putative intron were used to amplify gDNA and cDNA, yielding amplicons consistent with the 356- and 177-base pair product lengths predicted by the gene structure in B. The no-reverse-transcriptase (−RT) control detected no gDNA contamination in cDNA. The molecular weight for the DNA standard is highlighted in base pair (bp). (B) Schematic of the gene structure for pvrbp2a. The pvrbp2a gene consists of two exons and one intron, with the coding sequence and the noncoding intron shown in capital and small italicized letters, respectively. ATG highlights the start methionine in the 5′ region whereas TAA represents the stop codon on the 3′ end. (C) Western blots of P. falciparum and P. vivax protein extracts probed with anti-PvRBP2a and anti-EBA175 antibodies. Molecular mass marker is shown on the left hand side in kDa.

Structural comparison between PvRBP2a and PfRh5. (A) Superimposition of PvRBP2a (magenta) and PfRh5 (blue) structures in ribbon representation. The molecules can be aligned with the calculated rmsd of atomic positions 3.4 Å. The cysteine residues forming disulfide bridges are shown as yellow sticks. The PDB ID code for PfRh5 is 4WAT. (B) Comparison of the electrostatic potential surfaces of two orthogonal views of PfRh5 (Upper) and PvRBP2a (Lower). The negatively charged area is clearly visible in the upper part of the PvRBP2a molecule. Electrostatic surface potentials were calculated using the program APBS with the nonlinear Poisson–Boltzmann equation and contoured at ±5 kT/e. Negatively and positively charged surface areas are colored red and blue, respectively. (C) Basigin-binding site in PfRh5 is superimposed with the corresponding region in PvRBP2a. Residues forming the negatively charged area in PvRBP2a are localized in helix α4 and include E304, D306, E309, E313, E317, and E321. This glutamate-rich region overlaps with the basigin-binding site in PfRh5 (shown as a white surface representation, with the residues interacting with basigin highlighted in teal). The PvRBP2a molecule is shown as ribbons colored in magenta, and amino acid residues bearing negative charge are shown as yellow/red sticks.

Polymorphism and natural selection on the genes encoding PvRBP2a. (A) Sliding window analysis showing nucleotide diversity (π) values in pvrbp2a for the 31 sequences analyzed. A window size of 100 bp and a step size of 3 bp were used. (B) Sliding window calculation of Tajima’s D statistic was performed for the PNG population (n = 22). A window size of 100 and a step size of 3 were used. The gray box in A and B refers to a highly polymorphic region within pvrbp2a that is under balancing selection. (C) Three views of the surface representation of the crystallized fragment of PvRBP2a, with polymorphic residues highlighted in green. The polymorphic residues cluster around three main sites. The first one is localized on the tip of the molecule in the region overlapping with the basigin-binding region in PfRh5 and includes residues E304K, D306V, and P399S. The second region is localized on the side of the molecule within helix α3b and includes E277K, K285N/I, and K289E. The third region is localized in the cleft between helices α1 and α7 and includes N186S and K421M. There are two isolated single polymorphic sites, including E351Q and G438E.

Identification of amino acid residues in PvRBP2a important for erythrocyte binding. (A) Far UV CD spectrum for the WT PvRBP2a_160–1000 (WT) compared with the spectra of various structural mutants, including PvRBP2a_E/DmutK, PvRBP2a_{C299S, C303S}, PvRBP2a_Δ160–460, and PvRBP2a_{P180A, F181A, Y182A}, as labeled. Only the last triple-alanine mutant exhibits a different CD spectrum compared with the WT protein. (B) Far UV CD spectrum for the WT PvRBP2a_160–1000 (WT) and various SNP mutants, including PvRBP2a_{E277K, K285N, K289E}, PvRBP2a_{E304K, D306V}, PvRBP2a_{E304K, D306V, P399S}, PvRBP2a_G438E, PvRBP2a_E351Q, PvRBP2a_{N186S, K421M}, and PvRBP2a_{E277K, K285I, K289E}, as labeled. All presented spectra superimpose very well, suggesting that the introduced mutations did not alter the proper folding of the protein. (C) SDS/PAGE gel of purified PvRBP2a constructs loaded, as labeled. Two micrograms of each construct were loaded onto a NuPAGE gradient gel (4–12%) under reducing conditions and stained with Coomassie Brilliant Blue. Molecular mass marker indicated in kDa. (D) Deletion and mutational analyses of the erythrocyte-binding domain within PvRBP2a. Recombinant PvRBP2a proteins were tested for their capability to bind erythrocytes in a flow cytometry-based assay. % RBC binding, the percentage of erythrocytes with bound PvRBP2a protein determined by normalizing the number of erythrocytes exhibiting a positive Alexa Fluor 488 signal that is above the background (which is the Alexa Fluor 488 signal of erythrocytes without protein added) on the total number of erythrocytes.

AUC and SAXS data analysis for PvRBP2a. (A) Sedimentation velocity analytical ultracentrifugation for PvRBP2a_160–1000 at 0.4, 1.0, and 1.7 mg/mL (green, red, and black, respectively) is consistent with a monomeric protein. The measured sedimentation coefficient 3.5 S corresponds to an apparent molecular mass of 93.6 kDa with the theoretical calculated molecular mass as 98.4 kDa. No formation of higher order oligomers was detected. (B) Experimental scattering profile (black squares) overlaid with the calculated scattering pattern (red line) of a representative ab initio model generated using the program DAMMIF. The data are presented as the natural logarithm of the intensity vs. q. (Inset) The Guinier plot was linear, which is consistent with an absence of detectable aggregates in the sample. (C) The interatomic distance distribution function, P(r), of PvRBP2a_160–1000. The curve was calculated by indirect Fourier transform using the program GNOM. (D) Kratky plot analysis of the SAXS data, suggesting that the protein molecule is fairly rigid. (E) Two orthogonal views of the crystal structure of PvRBP2a_160–455 docked into an averaged and filtered ab initio SAXS envelope of PvRBP2a_160–1000 derived from 20 independent DAMMIF calculations. The PvRBP2a_160–1000 construct adopts an elongated boomerang-like shape. The crystal structure was fitted into one of its extremities by rigid-body docking. The fragment of PvRBP2a not present in the crystal structure forms a long tail after the C terminus.